| 00:00.56 | c4757p | DHL |
| 00:01.08 | c4757p | they were scheduled today, but customs seems to have taken a bit longer than usual |
| 00:01.28 | xzcvczx | thats what happens when you build missile control boards |
| 00:01.33 | c4757p | hah |
| 00:02.04 | *** part/#kicad ridesau7_ (~quassel@pool-74-110-107-3.nrflva.fios.verizon.net) |
| 00:02.48 | c4757p | hm, I do find it annoying that there are options like "no_std" that have to be set in the /source/ rather than at build time |
| 00:03.01 | c4757p | I'd much rather put that with the link options in the makefile than have to write that at the top of /every source file/ |
| 00:03.44 | xzcvczx | but surely you would want std all the time |
| 00:03.50 | xzcvczx | and only very very rarely not |
| 00:03.55 | c4757p | sure |
| 00:04.16 | c4757p | but it's still more of a link option than a source option |
| 00:05.51 | *** join/#kicad AndyBotwin (~Gustavo@unaffiliated/andybotwin) |
| 00:05.55 | c4757p | (also, being targeted at systems-level stuff, non-stdlib builds are going to be rather more common with Rust than with some other languages) |
| 00:11.53 | c4757p | "The same year, work shifted from the initial compiler (written in OCaml) to the self-hosting compiler" ...that explains some of the syntax quirks :) |
| 00:13.30 | *** join/#kicad intart (~intart@148.103.127.48) |
| 00:18.03 | Plugh | c4757p: I've always thought of it as a British expression and they drink a lot of tea over there. :) |
| 00:18.08 | c4757p | heh |
| 00:26.50 | xzcvczx | for every litre of tea consumed across britain c4 seems to consume a gallon of coffee |
| 00:27.06 | xzcvczx | thats an imperial gallon not the stupid us gallon |
| 00:27.30 | c4757p | I really don't drink that much coffee. I actually drink a lot of tea |
| 00:27.46 | xzcvczx | o_O |
| 00:28.14 | c4757p | usually I have two cups of coffee in the morning. I've started having more coffee during the day at work, but that's because there's a big lazybox at the end of the hall that magically makes me a cup of coffee in an instant |
| 00:28.17 | c4757p | >_> |
| 00:28.59 | xzcvczx | haha |
| 00:29.27 | xzcvczx | most of those sort of machines make pretty terrible drinks though |
| 00:29.32 | c4757p | yeah |
| 00:29.47 | c4757p | but if I'm drinking it while I work, I'm not really paying much attention to what it tastes like anyway ;) |
| 00:29.53 | xzcvczx | then again you are american so you probably just add 5 tablespoons of high-fructose corn syrup to your coffee |
| 00:30.00 | c4757p | nah, black |
| 00:30.26 | xzcvczx | errrr the introduction or lack there of of sweetners does not a not black coffee make |
| 00:30.41 | c4757p | might be regional, around here 'black' usually means totally plain |
| 00:30.52 | xzcvczx | oh |
| 00:31.11 | xzcvczx | (i don't drink tea or coffee) but i think over here black implies no milk only |
| 00:31.23 | c4757p | that makes a bit more sense :P |
| 00:33.35 | xzcvczx | that i don't drink tea/coffee/rat piss or that black only means no milk? |
| 00:33.40 | c4757p | the latter |
| 00:33.54 | c4757p | tea is delicious, I think you're wrong :D |
| 00:34.32 | xzcvczx | you may think but i know :) |
| 00:35.16 | c4757p | stupid people rarely realize that they are ^-^ |
| 00:35.47 | xzcvczx | stupid people rarely realise that they are wrong? |
| 00:35.52 | c4757p | that they are stupid :) |
| 00:35.53 | xzcvczx | well thats sort of implied |
| 00:36.13 | c4757p | facepalms at Wayne, tries to respond gently |
| 00:37.15 | xzcvczx | translation code is a bitch..... it would be nice to use the new pretty wxtranslations api but that doesn't support stupid fucked up numbering systems |
| 00:37.21 | xzcvczx | not stupid fucked up american date codes |
| 00:40.05 | *** join/#kicad bugzc_ (~1@unaffiliated/bugzc) |
| 00:40.38 | c4757p | bah |
| 00:40.51 | c4757p | I need to finish bringing up this power supply card before I get the motherboard assembled |
| 00:40.58 | c4757p | instead of working on writing Rust bindings :) |
| 00:41.07 | c4757p | moves to the workbench, warms up the scope |
| 00:41.19 | xzcvczx | "bringing up this power supply card"? |
| 00:41.38 | c4757p | need to finish the firmware and testing for the power supply card :P |
| 00:42.00 | xzcvczx | you yanks have weird ass sayings |
| 00:42.28 | c4757p | I'm pretty sure "bringing up" a board is more EE lingo than an Americanism |
| 00:44.28 | c4757p | ooh |
| 00:44.35 | c4757p | anyone using zsh |
| 00:44.39 | c4757p | zstyle ':completion:*' matcher-list '' 'r:|?=** m:{a-z\-}={A-Z\_}' |
| 00:44.42 | xzcvczx | never heard of it |
| 00:44.45 | c4757p | ^fuzzy matching on tabcomplete |
| 00:44.54 | c4757p | (pcw tabcompletes powercard-fw) |
| 00:45.09 | xzcvczx | lol |
| 00:45.14 | c4757p | that's cool :) |
| 00:45.51 | c4757p | 'mc' completes main.c in one go rather than "main." and requiring me to clarify .c, not .o |
| 00:45.53 | c4757p | :D |
| 00:47.32 | xzcvczx | well if you built into a different directory that would not be an issue |
| 00:47.39 | c4757p | shrugs |
| 00:47.43 | c4757p | simple makefile build |
| 00:47.50 | c4757p | besides, that's hardly the only use case :P |
| 00:47.54 | xzcvczx | that doesn't stop you |
| 00:48.16 | xzcvczx | you could even avoid the .o's with a small enough project |
| 00:48.31 | xzcvczx | i mean c'mon if build time is < 10secs then ain't shit gonna be saved from keeping objects around |
| 00:48.39 | c4757p | so? |
| 00:49.02 | xzcvczx | so i don't think your sister is the only hell-beast on this planet |
| 00:55.41 | *** join/#kicad eszett (~Eszett@dslb-084-059-074-203.084.059.pools.vodafone-ip.de) |
| 00:56.15 | c4757p | hmm |
| 00:56.16 | eszett | Hi kicad users! Can someone give me (or link me) a JTAG/SWD 10 pin header footprint, please? |
| 00:56.26 | c4757p | I think I should dedicate Friday night to cleaning my bench, before assembling the mobo on Saturday |
| 00:56.38 | c4757p | a specific footprint for that? I just use a ten-pin header |
| 00:57.13 | xzcvczx | or shrouded for pedantic people |
| 00:57.14 | eszett | <PROTECTED> |
| 00:57.23 | xzcvczx | yeah |
| 00:57.26 | c4757p | well, I don't use the standard lib, so |
| 00:57.36 | c4757p | I don't know if it has the right size in it |
| 00:57.44 | c4757p | they're usually 50mil headers, not 100mil |
| 00:57.46 | xzcvczx | meh 0.1" is hard to fuck up |
| 00:57.54 | eszett | ye and it is not that easy to browse |
| 00:57.54 | c4757p | ...^^ |
| 00:58.01 | xzcvczx | oh you mean 0.05in? |
| 00:58.27 | wpwrak | hmm, i wonder what the logic is behind "rescuing" things from the cache. typical situation: i edit a component in the library, save it, update the schematics, save and commit. then, on the next run, eeschema helpfully offers to "rescue" the component. i'm a bit at a loss why it even thinks anything needs rescuing in this situation. blind actionism, or am i missing some deeper meaning ? |
| 00:58.49 | xzcvczx | wpwrak: are you sure you didn't change the schematic symbol in the mean time? |
| 00:58.55 | eszett | xzcvczx i mean this one https://cdn.sparkfun.com//assets/parts/1/1/7/4/08506-02.jpg |
| 00:59.19 | xzcvczx | shrouded 0.1" in the standard libs should be fine |
| 00:59.37 | wpwrak | xzcvczx: yes, i changed it, with the library editor. saved. then updated things in the schematics. all in the same session. |
| 00:59.59 | xzcvczx | oh shrug |
| 01:01.17 | c4757p | reattaches a few more knobs he forgot to his 485, sees how cracked and chipped they are, sighs |
| 01:01.26 | c4757p | I wonder if I can 3D-print some new knobs for it like I did the buttons |
| 01:01.51 | xzcvczx | well if you don't treat your equipment like shit it won't break |
| 01:01.52 | c4757p | work even has a 3D printer - certainly not the sort I'd want for the final things, I'd send off for those, but decent for prototyping? |
| 01:02.03 | c4757p | dude it's 40 years old :P |
| 01:02.25 | xzcvczx | so are you but you don't see me replacing all your broken parts with 3d printed shit |
| 01:03.43 | c4757p | oh, that's interesting, I never noticed that the "B intensity" knob is relative to the "A intensity" knob |
| 01:03.45 | c4757p | that's very, very nice :) |
| 01:08.25 | *** join/#kicad dumle29 (~quassel@64.137.168.54) |
| 01:09.59 | *** join/#kicad cosmobird (~firebird@125.16.18.198) |
| 01:10.39 | *** join/#kicad CarlPoirier (~CarlPoiri@modemcable019.249-59-74.mc.videotron.ca) |
| 01:20.15 | *** join/#kicad crazy_im1 (~mj@a89-182-21-147.net-htp.de) |
| 01:20.58 | *** join/#kicad mozzarella (~sam@unaffiliated/sam113101) |
| 01:22.36 | cirilo | ok .. Smoke Test #1 passed - FTDI gizmo is recognized. Now I wish I had a DSO (or even non-storage) so I could check other parts of the circuit. |
| 01:23.24 | cirilo | but I feel like a royal goose ... just bought a USB micro-B cable to test the device I designed, and found out just now that I'd misplaced my programmer's micro-B cable so I haven't got a functional programmer. :( |
| 01:24.54 | xzcvczx | you must be the one person in 2016 who doesn't have 5million micro-usb cables sitting around your house |
| 01:26.03 | cirilo | wut can I say - I'm an ultra-conservative Amish |
| 01:26.29 | *** join/#kicad f11f12c (~f11f12@p5B2956D0.dip0.t-ipconnect.de) |
| 01:26.32 | cirilo | I even have a Dumb Phone |
| 01:26.43 | xzcvczx | apparently so...... hows that goldrush prospecting treating ya |
| 01:27.17 | cirilo | I got no need for gold - I got my barn and chickens |
| 01:27.59 | xzcvczx | and desert..... you can't forget the desert |
| 01:32.23 | mroszko | opengl3 dropped matrix stack |
| 01:32.27 | mroszko | explains my suffering |
| 01:32.28 | mroszko | QQ |
| 01:34.39 | mroszko | i see since 3d viewer uses glm |
| 01:34.43 | mroszko | its forwards compatible! |
| 01:34.44 | mroszko | lol |
| 01:41.19 | *** join/#kicad Novus (~Novus@2605:e000:628b:1400:9d45:9ad3:6423:5d0a) |
| 01:42.42 | inductiveload | maybe farnell and friends need to get on the amazon dash button bandwagon |
| 01:42.52 | inductiveload | run out of mico USB cable, press the button! |
| 01:45.40 | mroszko | just buy 50 from monoprice |
| 01:45.43 | mroszko | at 50 cents a piece |
| 01:45.47 | xzcvczx | or just don't be cirilo |
| 01:45.59 | xzcvczx | every other human has ample micro-usb cables these days |
| 01:46.00 | mroszko | i bought 300 usb type Bs from monoprice at work |
| 01:46.02 | mroszko | life is good |
| 01:46.07 | xzcvczx | i havem ore than i can feasibly count or find |
| 01:46.16 | mroszko | lol |
| 01:47.47 | inductiveload | i bought some super cheap spiral usb micro cables on ebay |
| 01:48.22 | inductiveload | i have no earthly clue how they work, the wires are only slightly thicker than hairs, and there is no sheilding at all |
| 01:48.26 | xzcvczx | just be cautious with where you source your usb-c cables...... they are getting too expensive so people have found dodgy ass ways to cut costs |
| 01:48.34 | mroszko | lol |
| 01:48.40 | mroszko | i had to explain at work |
| 01:48.48 | mroszko | how the $30mil military charger we are working on |
| 01:48.54 | mroszko | has to be idiot proofed on usb type c |
| 01:48.55 | cirilo | mroszko, is OGL3 a problem in other kicad code? |
| 01:48.59 | mroszko | "BUT WHY, ITS ONLY 5V?" |
| 01:49.13 | xzcvczx | mroszko: show them the usb-killer |
| 01:49.13 | mroszko | cirilo, nah im doing my own thing |
| 01:49.13 | inductiveload | mroszko, that's when you crack out the $50/set keyed lemos |
| 01:49.33 | mroszko | one engineer keeps begging me to find a 5V micro... |
| 01:49.39 | mroszko | because he doesnt want to design for 3.3V |
| 01:49.39 | mroszko | ugh |
| 01:49.49 | mroszko | doesnt want 8051s |
| 01:49.51 | inductiveload | is it that hard? |
| 01:50.05 | inductiveload | well that's reasonable, at least, i hate 8051s |
| 01:50.09 | mroszko | if you want one capable of running an RTOS and have > 50kb of ram yea |
| 01:50.10 | mroszko | lol |
| 01:50.27 | inductiveload | no i mean design for 3.3 |
| 01:50.50 | inductiveload | isn't it easier these days? |
| 01:50.57 | xzcvczx | if you are interfacing with 5v parts it can be a unnecessary complication |
| 01:51.21 | inductiveload | well i guess if loads of other stuff doesn't also go to 3.3v |
| 01:51.49 | mroszko | "BUT I DONT WANT TO RECALCULATE ALL THE DIVIDERS WE HAVE FROM A 20 YEAR OLD DESIGN" -- the only excuse |
| 01:51.53 | xzcvczx | 3v3 is also nice in usb as you can ldo it with out really having any issues |
| 01:52.55 | inductiveload | mroszko, i'm doing that right now, it is a pain ;-) |
| 01:53.02 | mroszko | pfft |
| 01:53.10 | mroszko | thats why in the last 20 years |
| 01:53.14 | mroszko | we have better calculators! |
| 01:53.23 | mroszko | the TI 89 is a thing now! |
| 01:53.57 | inductiveload | but i avoid a bank of level shifters next to the cpld, so probably worth it |
| 01:54.04 | *** join/#kicad CarlPoirier (~CarlPoiri@modemcable019.249-59-74.mc.videotron.ca) |
| 01:56.39 | inductiveload | mroszko, the ti89 has been a thing for 18 years |
| 01:57.39 | decimad | <3 wxMaxima |
| 01:58.15 | inductiveload | "Because the TI-Nspire lacks a QWERTY keyboard, it is acceptable for use on the PSAT, SAT,[1] SAT II, ACT,[2] AP, and IB Exams." |
| 01:58.29 | inductiveload | whaaaaa |
| 01:58.40 | decimad | I would enter their market with a QWERTZ design |
| 01:58.55 | inductiveload | that calculator has an ABCDE keyboard |
| 01:59.42 | inductiveload | that is a retarded rule and an inevitable workaround |
| 02:01.50 | xzcvczx | just use a dvorak keyboard |
| 02:02.45 | inductiveload | i don't think touch typing on a calculator is very effective anyway |
| 02:03.18 | inductiveload | ABCDEF keyborads do piss me off when used on screen keyboards, like at the ralway station ticket machines |
| 02:03.42 | inductiveload | also when the confirmation code is 10 random numbers and letters |
| 02:03.50 | xzcvczx | people said the same about t9 yet many people were quite well versed in touch typing on them prior to smartphones |
| 02:05.23 | inductiveload | so your hunting and pecking and trying to remember 45GHE4DFGD |
| 02:06.06 | inductiveload | i'd much rather get a code like three random words |
| 02:07.36 | inductiveload | i never really got the thing with calculators, a 10 pound scientific casio one did me all though school and uni |
| 02:08.55 | inductiveload | and even then, i'm sure i spent longer reading the intsturctons for that that I even did using things like matrix inverison and the stats mode |
| 02:09.11 | xzcvczx | was it the casio with the green back & white front? |
| 02:10.10 | inductiveload | no it was more like this: https://d29h7ql7qnxkqx.cloudfront.net/pix/flashwiz/C07s.jpg |
| 02:11.24 | xzcvczx | we had https://upload.wikimedia.org/wikipedia/en/thumb/9/9d/Casio_CFX9850GC.jpg/800px-Casio_CFX9850GC.jpg at school |
| 02:11.48 | inductiveload | ugh i hate the rounded look ones |
| 02:12.04 | decimad | when I visited the US I had some trouble with the ti 89 that I had with me ;) |
| 02:12.16 | decimad | that was before 9/11 even! |
| 02:12.36 | xzcvczx | inductiveload: people at school also made a fake clear memory application on them |
| 02:12.49 | xzcvczx | so that you could pretend you had cleared their memory before a test or exam |
| 02:13.09 | inductiveload | i can't remember if graphing calculators were allowed in exams |
| 02:13.29 | xzcvczx | we weren't prior to 5th form (15years old) i think |
| 02:14.54 | inductiveload | plus you get points for working, so at best it's after the fact confirmation that you got it right |
| 02:16.13 | inductiveload | and at uni, exams weren't ridiculously maths-heavy, they saved the worst of it for assignments, where you could use a computer |
| 02:16.23 | xzcvczx | inductiveload: our "statistics" "exam" had like 2 sides of an a4 paper of just explaing/justifying some stupid line of best fit graph that we were pretty much spoon fed the explanation for and were expected to memorise it and write it out in the exam |
| 02:17.16 | inductiveload | oh god, reminds me of physics exams |
| 02:17.30 | inductiveload | they were so formulaic |
| 02:18.21 | inductiveload | "when you get to the question with a log graph of something, they'll ask you for sometihng, you find the gradient and ... and ... andthen you write ..." |
| 02:18.49 | xzcvczx | lol formulaic got me through more than one exam at university.... i remember one well, i had only studied for a couple of days, wrote down all the formulas over i think 4 or 5 sides of a4, converted that to 2 sides of a4, was sitting outside prior to the exam jsut running over the formulas, got into exam as soon as i could i wrote the damn things down and voila |
| 02:19.18 | inductiveload | for a-level physics you didn't even need that, just handle on how big numbers should be |
| 02:20.06 | inductiveload | if the answer came out as 4 billion kg, you should have divided, not multiplied |
| 02:20.16 | xzcvczx | haha yeah until you get a question about the power usage in the first few microseconds of turning on an incandescent bulb |
| 02:21.30 | inductiveload | i'm not even sure if inductance was even in the physics syllabus |
| 02:22.39 | *** join/#kicad Zhenn (~who@199.241.146.179) |
| 02:22.44 | inductiveload | you certainly didn't have to be able to integrate or differentiate, though you might have to find the area under a line graph |
| 02:23.20 | *** join/#kicad jjrh (~jjrh@2607:f0b0:8:801c:98a8:2254:f971:e3f4) |
| 02:23.55 | Zhenn | heya. I've got a board with ground/vcc fill zones. is there a way to run the autorouter (and then re-fill zones with "B") that doesn't result in stranded areas in either plane? (not all gnd or vcc fill zones are interconnected) |
| 02:25.26 | xzcvczx | yeah don't use autorouter |
| 02:25.55 | xzcvczx | but no there is no way to not fill in islands |
| 02:25.57 | xzcvczx | unfortuantely |
| 02:26.05 | Bird|otherbox | Zhenn: don't bother with the autorouter. manual routing is more fun ;) |
| 02:26.47 | Zhenn | is there a way to manually make the bridges by specifying a "no trace" zone that the autorouter will avoid? |
| 02:34.14 | xzcvczx | are you meaning freerouter? |
| 02:34.53 | Zhenn | yup yup,. |
| 02:35.42 | xzcvczx | meh i have no idea as i don't use freerouter, and its likely that noone in here does |
| 02:35.55 | xzcvczx | autorouters are not well regarded |
| 02:40.07 | Zhenn | I've never worked in this field at all. just watching the autorouter makes me realize I've virtually no chance at doing it myself. thanks for the heads up though, I've decided to manually route the islands and auto-fill. |
| 02:42.33 | xzcvczx | expecting an autorouter to remove the need for knowledge is a dangerous concept and seldom workable |
| 02:43.39 | eszett | I think routing is the fun part of pcb design. Its a challenge for me better than sudoku. |
| 02:44.20 | eszett | How to connect everthing with the least amount of vias? How to connect the most elegant way? How to connect with low interference? How to connect in an aesthetical manner? |
| 02:44.29 | eszett | That is the challenge.. |
| 02:45.59 | xzcvczx | well the latter one there is easy.... just do a c4 and hide all the really ugly shit on an inner layer |
| 02:46.39 | eszett | inner layer: additional costs. I forgot to mention budged considerations :-) |
| 02:51.08 | inductiveload | Zhenn, emulating the autorouter is not a very good way to attempt routing |
| 02:51.46 | inductiveload | you know a lot more about how your board works than any autorouter does |
| 02:52.09 | *** join/#kicad ayjay_t (~ajp_@24-151-28-56.dhcp.nwtn.ct.charter.com) |
| 02:58.38 | Zhenn | I'm starting to see that =) |
| 02:58.44 | Zhenn | the autorouter made some terrible choices. |
| 03:00.58 | *** join/#kicad Wetmelon (~wetmelon@host-73-169.ncasgro.asheville.nc.us.clients.pavlovmedia.com) |
| 03:02.16 | cirilo | now this is frustrating - as far as I can tell my wiring is right, but I'm getting no response when interrogating the MCU with the programmer. :( |
| 03:02.49 | xzcvczx | cirilo: haha probably dieless chips :P |
| 03:03.18 | cirilo | xzcvczx, could be - just legs and ceramic block |
| 03:03.45 | cirilo | oh - and a resistor so the #RESET pin reads 3,3V when I check |
| 03:03.48 | xzcvczx | sparkfun got shafted by that |
| 03:04.10 | xzcvczx | not the resistor thing |
| 03:04.14 | cirilo | did sparkfun say where they sourced the chips? |
| 03:04.20 | xzcvczx | china :) |
| 03:04.33 | cirilo | well, China's an awfully big place |
| 03:05.11 | xzcvczx | they were out of stock on some chipset that noone had, so they found a "supplier" in china, bought a reel or something proceeded to assemble a bunchashit then wondered why all their tests failed |
| 03:05.24 | cirilo | heh |
| 03:05.40 | cirilo | did they name the supplier? |
| 03:05.51 | Zhenn | for the ti shift registers / bcd units can I just throw the same 5v/gnd lines at them? |
| 03:06.22 | xzcvczx | https://www.sparkfun.com/news/350 |
| 03:06.37 | cirilo | anyway, I suspect the components are good; might be a bad wire on my programmer - wish I had a DSO to check it all. :( |
| 03:15.47 | eszett | xzcvczx wait, if i get this news article right, there wasnt any die inside, but just a chunk of copper? ? |
| 03:16.38 | eszett | what a daring fake. |
| 03:16.45 | xzcvczx | nice and cheap though |
| 03:17.07 | cirilo | + higher profit margin than fakes that rebadge other products |
| 03:17.12 | eszett | ye |
| 03:17.20 | inductiveload | i wonder if the factory making those is actually making fakes |
| 03:17.52 | cirilo | oo .. nice storm in the Pacific with wind gusting to 370kph |
| 03:17.55 | inductiveload | or if they're making blanks for testing reflow or something and someone else is marking them and selling them |
| 03:18.11 | inductiveload | (or both) |
| 03:18.23 | xzcvczx | eszett: also if you can flog enough of them quick enough that you have many orders while previous are still in the post then as soon as one company figures out its a sham you just choose a new name and a new chip and try again |
| 03:19.16 | inductiveload | xzcvczx, that's where your industry 4.0 line comes in, it automatically lasers the right numbers onto the chip as the orders come in |
| 03:20.44 | eszett | yes,but obviously turns out profitable only when selling huge volumes. And buyers like sparkfun wont do the same mistake again |
| 03:21.31 | xzcvczx | thats what makes it hard when you keep changing trading names |
| 03:27.11 | *** join/#kicad kyranf (~kyranf@S0106bcd165673658.ok.shawcable.net) |
| 03:32.44 | *** join/#kicad cosmobird (~firebird@125.16.18.198) |
| 03:38.43 | *** join/#kicad C47 (~d3bug@189.242.29.24) |
| 03:38.53 | cirilo | yes .. found spare cable .. now the device is talking to the programmer. :) |
| 03:40.29 | cirilo | hmm .. packaging of the assembled boards looks suspicious though .. clear bubblewrap. I guess I'd better pray that it happens to be a clear antistatic bubble wrap |
| 03:43.02 | *** join/#kicad bugzc (~1@unaffiliated/bugzc) |
| 03:43.34 | xzcvczx | lol |
| 03:43.44 | xzcvczx | anti-static is too expensive |
| 03:57.57 | inductiveload | what evil troll would make clear anti static bubble wrap? |
| 04:01.27 | *** join/#kicad cosmobird (~firebird@125.16.18.198) |
| 04:01.52 | *** join/#kicad DocScrutinizer05 (~saturn@openmoko/engineers/joerg) |
| 04:09.03 | Zhenn | stupid question, do I need resistors for ic's if they're wired in parallel @ 5v & gnd ? |
| 04:17.14 | *** join/#kicad Zhenn (~who@199.241.146.179) |
| 04:17.32 | cirilo | what's wired in parallel? |
| 04:20.43 | Zhenn | cirilo: a bunch of ic's. eight of them. two central headers, 5v/gnd. |
| 04:21.25 | Zhenn | I'll be honest, I really don't know what I'm doing here. |
| 04:23.52 | c4757p | resistors for ICs? |
| 04:23.59 | Zhenn | I understand that an LED requires a resistor based upon it's mA rating, and that's placed between it and ground. is the same true for IC's? |
| 04:24.04 | c4757p | oh |
| 04:24.07 | c4757p | no. |
| 04:24.24 | c4757p | the vast majority of ICs are voltage-input devices, so they get connected directly to the power supply |
| 04:24.43 | c4757p | LEDs are current-input - they try to clamp the voltage across them to a certain level, and then you use the difference between that and the supply to set the voltage |
| 04:24.46 | c4757p | er |
| 04:24.48 | c4757p | to set the current* |
| 04:25.01 | Zhenn | oh whoa. |
| 04:25.14 | Zhenn | kk. |
| 04:26.16 | Zhenn | 7-segment (cathode) displays require resistors then, yeah? |
| 04:26.55 | cirilo | yes, unless you use a constant current driver to light them |
| 04:27.57 | cirilo | 1 resistor per cathode - if you put the resistor on the common anode then the lighting will be very uneven and some segments may never switch on while others are on |
| 04:28.08 | Zhenn | I'm using the 4511 IC's, afaik they're just converting binary to the segments? |
| 04:28.21 | *** join/#kicad daey_ (~Flutterba@unaffiliated/day) |
| 04:28.34 | Zhenn | kk, that was the next question. each segment has to have it's own resistor before the common gnd. |
| 04:28.53 | Zhenn | wait, no, that doesn't make sense. =p |
| 04:29.06 | Zhenn | 1 resistor per display you mean? |
| 04:29.44 | *** join/#kicad gzamboni (~gzamboni@gzamboni.zael.fr) |
| 04:30.55 | c4757p | one per segment |
| 04:31.32 | Zhenn | that's if you're lighting multiples at a time though, isn't it? |
| 04:31.39 | c4757p | yeah, but aren't you? :P |
| 04:31.41 | cirilo | oh - you have common cathode? Then 1 resistor per anode :) |
| 04:31.49 | Zhenn | the 4511 (afaik) lights the segments independently (in rapid succession) |
| 04:32.01 | cirilo | + make sure your IC can actually supply the current that you want in each segment |
| 04:32.05 | Zhenn | so that'd just be the one resistor? |
| 04:32.08 | c4757p | the 4511? I don't think it does. |
| 04:32.12 | Zhenn | oshit |
| 04:32.23 | c4757p | it's too simple a chip for that. it's /ancient/ |
| 04:32.53 | Zhenn | is there an ic that does this? |
| 04:33.07 | c4757p | there are all sorts of very sophisticated LED driver ICs, yes |
| 04:33.18 | c4757p | but if you want to use the 4511, you can, just put a resistor on each segment ;) |
| 04:33.39 | Zhenn | that'd be messy, but that's one resistor between the outputs of the 4511 and each segment? |
| 04:34.00 | cirilo | keep in mind that LEDs will appear dimmer if they are switched on and off at a high frequency |
| 04:34.57 | cirilo | also, I typically run LEDs at <=80% of their max. current; the lifetime of the LED generally suffers badly if you apply >90% |
| 04:35.55 | Zhenn | "peak forward 80, continuous forward 20, reccomended 12" |
| 04:36.03 | cirilo | looking at real devices, I see an opto switch where the LED is rated at 50mA but the recommended operating current is 20mA |
| 04:36.03 | Zhenn | (mA) |
| 04:36.32 | c4757p | I'm not sure it's so messy to have one resistor per segment |
| 04:36.38 | cirilo | haha - see - the LED display manufacturers use an even lower current than I typically do. :) |
| 04:36.40 | c4757p | that's just normal for driving multiplexed displays |
| 04:37.06 | Zhenn | I'll be soldering this by hand, so it just seems like a lot to me I suppose. |
| 04:37.22 | cirilo | (though having said that, there are some types of LEDs that I pulse with currents way above the rated maximum) |
| 04:37.45 | c4757p | maximum for a diode is typically average, not peak :P |
| 04:37.48 | Zhenn | there's also a scale problem that I don't know how to see/calculate. it's uh, three shift registers, six bcd's (4511), six 7-segments. |
| 04:38.24 | cirilo | well, even with "peak" there is quite a bit of leeway - unless the peak was already determined with nanosecond pulses. :) |
| 04:38.26 | Zhenn | how would I understand the total draw in mA? |
| 04:38.38 | Zhenn | (and how does that even work?) |
| 04:39.05 | cirilo | total draw of what - of the entire assembly? |
| 04:39.10 | Zhenn | yup yup. |
| 04:39.32 | Zhenn | has never in his life messed with circuits. |
| 04:39.37 | cirilo | easy - current used by chips + current used assuming all LEDs are on |
| 04:40.11 | cirilo | asking questions isn't the right way to learn - need to read first, otherwise your questions won't make much sense |
| 04:40.13 | Zhenn | is it really that easy? let's say 1A in, and that total comes to .5A - does that mean there's .5A of headroom remaining? |
| 04:40.42 | cirilo | you mean if the power supply can provide 1A? |
| 04:40.47 | Zhenn | yup. |
| 04:40.50 | cirilo | how are you regulating the power? |
| 04:41.36 | cirilo | if your total projected max. consumption is ,5A then you still have a spare ,5A if your supply is rated at 1A continuous |
| 04:41.49 | cirilo | (not counting other things like derating in a hot environment etc) |
| 04:41.50 | Zhenn | should be regulated out of the Pi? 5V ~1.5A |
| 04:42.23 | cirilo | I don't know anything about the RPIs (even though I have 3 of them in a drawer next to me) |
| 04:42.51 | cirilo | is that on the USB connector or is it a special power connector? |
| 04:43.37 | Zhenn | it's a gpio header. |
| 04:43.48 | Zhenn | (which will feed to the board with the IC+display arrangement) |
| 04:44.40 | cirilo | and how is that voltage regulated by the Pi? |
| 04:48.01 | *** join/#kicad sandeepkr (~sandeepkr@103.49.155.114) |
| 04:48.02 | *** join/#kicad kuldeep (~kuldeep@unaffiliated/kuldeepdhaka) |
| 04:51.03 | Zhenn | couldn't tell you, there's not much information on the subject. |
| 04:51.14 | cirilo | what model PI? |
| 04:51.58 | Zhenn | zero? |
| 04:53.45 | Zhenn | it looks like the ld1117 would drop the 5v down to 3.3v, and I could expect anywhere from 200 to 500mA. |
| 04:54.43 | benishor_ | hi all, is there such a thing like eagle to kicad conversion tool or an eagle importer component in kicad? |
| 04:54.47 | *** join/#kicad eduardas_m (~eduardas_@213.197.143.19) |
| 04:55.31 | benishor_ | I was just htinking that would ease adoption |
| 05:07.49 | *** join/#kicad Sujith (~Sujith@59.91.45.26) |
| 05:15.02 | cyborg_ar | cirilo, i found a small buck in kicad2mcad |
| 05:15.08 | cyborg_ar | bug* |
| 05:15.38 | Zhenn | cirilo, c4757p: it looks like I'd have to solder 42 resistors (one to each 7segment anode) if I was going for the simultaneous light method. that would look insane in person. is there a way to bank a bunch of them together or should I just go with one of those IC's that alternates between segments, and use six resistors on the common cathodes instead? |
| 05:16.11 | c4757p | well, personally I'd buck up and solder the 42 resistors |
| 05:16.21 | c4757p | but you could multiplex them :P |
| 05:19.19 | Zhenn | multiplex is the term for alternating segments? |
| 05:19.49 | *** join/#kicad mgottschlag (~quassel@reactos/tester/phoenix64) |
| 05:19.51 | c4757p | generally not - alternating entire digits |
| 05:20.14 | Zhenn | lol. I was struggling to define as much =/ |
| 05:20.23 | Zhenn | screw it, resistor grid it is (it's a hobby project anyway) |
| 05:20.53 | Zhenn | thank you, by the way. |
| 05:21.03 | c4757p | np :P |
| 05:22.27 | cirilo | cyborg_ar, how small is the bug? |
| 05:22.54 | cirilo | Zhenn, you can look for "resistor network" packages |
| 05:23.19 | cirilo | Zhenn, there are 2 types - individual resistors and resistors with 1 common point |
| 05:23.51 | Zhenn | https://i.imgur.com/2ZYd5Bv.png this is the idea, yeah? just put a resistor on each anode (the schematic wrongly says CA, wheras the unit is CC ) |
| 05:24.03 | cyborg_ar | it is when SMD pads have an offset |
| 05:24.30 | cyborg_ar | you get a pad that has somethinbg like (drill (offset 1 2)) which makes your parser choke because there is no X dimension |
| 05:24.32 | c4757p | yes, that's the idea |
| 05:24.34 | cirilo | ah - oops - nope - can't use respack with a common pin because the resistor is between LED and driver IC .. d'oh |
| 05:24.51 | cirilo | but at least a RP IC makes the wiring neater |
| 05:25.02 | *** join/#kicad cosmobird (~firebird@125.16.18.198) |
| 05:25.03 | Zhenn | =) |
| 05:25.33 | cirilo | cyborg_ar, can you email me a simple PCB with that example (even just 1 square PCB with a 1-pin footprint) |
| 05:25.44 | cyborg_ar | okay |
| 05:25.48 | Zhenn | yup yup, looks like the weird multiple resistor line (individual but placed vertically) is the way to go |
| 05:27.09 | Zhenn | "isolated resistor network" how awesome. |
| 05:31.10 | cyborg_ar | cirilo, https://gist.githubusercontent.com/iromero91/2eb332925b1b08bda7e46ed98b98a6a2/raw/a30530a47d1f434134453c96044386f8dea480fd/bad.kicad_pcb |
| 05:34.20 | cyborg_ar | i patched it out by commenting out that check |
| 05:35.12 | cyborg_ar | it is really annoying when stuff of no consequence cause a parser to fail, it should ignore errors instead of failing |
| 05:40.10 | wpwrak | cyborg_ar: another one is if you're trying to load a 4.1-saved board with 4.0, which failed because it doesn't recognize the default parameters for differential pairs 4.1 saved (even if there are no diff pairs at all) |
| 05:41.14 | c4757p | uuuurgh "4.1" |
| 05:41.35 | cyborg_ar | i had an idea on how to fix that but it's complicated and i don't have the time nor skill to implement it properly |
| 05:43.07 | *** join/#kicad rohitksingh (~rohit@119.82.107.167) |
| 05:43.31 | wpwrak | c4757p: you really don't like the bleeding edge, do you ? :) |
| 05:43.48 | c4757p | I don't like that some people are calling it "4.1" :P |
| 05:44.27 | wpwrak | c4757p: that's what it calls itself :) |
| 05:44.43 | c4757p | well, that's what the PPA dude has taken to calling it |
| 05:45.46 | cirilo | cyborg_ar, thanks, I'll try to look at it later or tomorrow |
| 05:46.00 | Zhenn | whats the difference between normal dip+x and the "longpads" version? |
| 05:46.22 | inductiveload | Zhenn, one has longer pads |
| 05:46.29 | c4757p | lol |
| 05:46.32 | Zhenn | =p |
| 05:46.33 | inductiveload | longer pads are easier to hand solder |
| 05:46.59 | inductiveload | and easier to visually see if the joint looks good |
| 05:47.05 | *** join/#kicad berjan (~chatzilla@ip54538619.speed.planet.nl) |
| 05:47.23 | cirilo | I'm surprised DIPs hadn't been phased out yet. :) |
| 05:47.37 | cirilo | I hadn't used one in a design in over 10 years |
| 05:48.22 | Zhenn | I'm guessing the through-hole off-the-shelf DIPs are longpad |
| 05:48.39 | inductiveload | Zhenn, the DIP itself is the same |
| 05:49.03 | inductiveload | you can fit a DIP to the round pad or long pad footprint |
| 05:49.37 | Zhenn | offs, I'm dense. |
| 05:49.46 | Zhenn | it's literally the pads. whether they're circular or ovular. |
| 05:49.47 | Zhenn | =p |
| 05:50.16 | inductiveload | if you haven't done a lot of circuit assemblt before, go for long pad if you have space |
| 05:51.08 | inductiveload | cirilo, i picked up a board rammed with about 40 of the wide body DIPs in sockets the other day - you know how they used to be, all laid out end to end in rows |
| 05:51.50 | inductiveload | the sound and feel of the sockets flexing arond the chip leads brought back a lot of memories |
| 05:57.40 | *** join/#kicad plaes (~plaes@unaffiliated/amd) |
| 06:06.27 | cirilo | inductiveload, yeah - I remember a vector processor board like that - was 18x30 inches. They sure creaked a lot :) |
| 06:06.58 | inductiveload | and all for what was probably about 200 lines of HDL ;-) |
| 06:07.17 | cirilo | amazing how these days a 2048-core vector processor will fit in your pocket |
| 06:09.36 | cyborg_ar | cirilo, i've used dip on an industrial board pretty recently, but mostly because it was wire wrap |
| 06:12.05 | cyborg_ar | inductiveload, lol the times you could fix boards by reseating the damn things |
| 06:12.58 | inductiveload | well you can also sometimes 'reflow' suspiciious chips today |
| 06:14.35 | cyborg_ar | hah |
| 06:14.39 | cyborg_ar | true that |
| 06:14.55 | cyborg_ar | i liked how louis rossman debunked part of that |
| 06:15.16 | cyborg_ar | turns out that most "reflow" cures are just reforming tantalum capacitors with heat |
| 06:15.17 | inductiveload | 'reflow' meaning wave a heat gun around becuase oyur all out of better ideasa |
| 06:15.37 | cyborg_ar | so they behave well for a little bit |
| 06:16.10 | inductiveload | good to know |
| 06:16.26 | inductiveload | so the solution then is to replace the tantalum cap? |
| 06:16.55 | cyborg_ar | yep |
| 06:17.00 | inductiveload | must be why the packages are yellow, so you can zoom in on the problem right away |
| 06:17.08 | cyborg_ar | not really |
| 06:17.37 | cyborg_ar | tantalum caps are just a common example. but there are many components that might return back to spec by just heating or cooling them |
| 06:17.50 | cyborg_ar | that's the whole logic behind repairmen using freezer sprays |
| 06:17.51 | inductiveload | or so when they'e been on fire, you can see the difference clearly ;-) |
| 06:18.30 | cyborg_ar | thing is, if you heat up the whole board, then you're out of luck trying to find out which cap was it that failed |
| 06:18.52 | *** join/#kicad maiden (~maiden@c83-255-72-208.bredband.comhem.se) |
| 06:19.03 | inductiveload | yeah, i wasn't serious that you just go nuts with the heat gun ;-) |
| 06:20.04 | cyborg_ar | most people do though |
| 06:20.13 | cyborg_ar | they just toss the whole board in the oven |
| 06:20.48 | inductiveload | i'd be more scared of scrwing it up more |
| 06:20.57 | *** join/#kicad stelarcf_ (~stelarcf@178.138.34.53) |
| 06:21.17 | inductiveload | plus any troubleshooting you've already done is now suspect |
| 06:21.48 | cyborg_ar | yup |
| 06:22.08 | inductiveload | if one chip has some wierd behaviour and everything else looks good, a quick heat gun to the soldering is about as far as I'd go for a reflow cure |
| 06:22.53 | inductiveload | even then, you'd normally see the problem under a microscope first unless it's some part where you can't properly see all the pads |
| 06:23.12 | inductiveload | and if it's a bga, you're just screwed ;-) |
| 06:29.50 | trygvis | in pcbnew, if I make a component value invisible .. how do I make it visible again? |
| 06:36.07 | *** join/#kicad sandeepkr_ (~sandeepkr@111.235.65.5) |
| 06:41.22 | *** join/#kicad sandeepkr_ (~sandeepkr@111.235.65.5) |
| 06:44.54 | *** join/#kicad kuldeep (~kuldeep@unaffiliated/kuldeepdhaka) |
| 06:48.20 | *** join/#kicad kuldeep (~kuldeep@unaffiliated/kuldeepdhaka) |
| 06:49.15 | cirilo | why not just hit them with a hammer? hit it hard enough and you get a replacement board - all fixed :) |
| 06:49.38 | *** join/#kicad pointhi21 (~pointhi21@dyn166055.wlan.jku.at) |
| 06:49.40 | cirilo | trygvis, hover and edit the component |
| 06:50.06 | cirilo | one of the menu items gives you the "edit value" UI |
| 06:56.34 | *** join/#kicad chepaz (~chepaz@dsl60b7235.sdtnet.de) |
| 06:58.41 | *** join/#kicad berjan (~chatzilla@2001:67c:2564:a156:e90f:d52e:c15d:17b7) |
| 06:59.43 | *** join/#kicad sandos (~jhn@1071ob.scansafe.net) |
| 07:04.51 | *** join/#kicad alphaferret (~user@54034F08.catv.pool.telekom.hu) |
| 07:08.15 | nickoe | xzcvczx: it was simply because swig was not installed. |
| 07:08.28 | *** part/#kicad sandos (~jhn@1071ob.scansafe.net) |
| 07:08.59 | inductiveload | email from recruiter: Good Evening First Middle Surname, Ever thought about re-locating to be the âbestâ city in the world? |
| 07:09.41 | inductiveload | (they did put the full name in) |
| 07:10.41 | xzcvczx | and what is the best city in the world? beirut? |
| 07:10.47 | xzcvczx | nickoe: ah ok, added |
| 07:10.50 | inductiveload | london, apparently |
| 07:10.59 | xzcvczx | s/best/most expensive/ |
| 07:11.09 | inductiveload | eh, london's not that bad |
| 07:11.21 | xzcvczx | it sorta is |
| 07:11.25 | inductiveload | i've been there for 4 years which makes the email even better |
| 07:11.27 | inductiveload | it |
| 07:11.44 | inductiveload | is pretty shit, but not _that_ bad |
| 07:12.17 | inductiveload | house prices are pretty silly |
| 07:12.50 | xzcvczx | tbh house prices are pretty nuts here |
| 07:13.29 | inductiveload | public transport costs are expensive, but if you compare on a ticket by ticket basis, they're actually better than lots of other places |
| 07:14.36 | xzcvczx | we stayed in canary wharf area i think its called for a month in 2012 |
| 07:14.50 | inductiveload | well that's going to be $$$ |
| 07:15.09 | xzcvczx | lol, well we didn't pay nothing as we stayed with an old colleague of my fathers |
| 07:15.22 | xzcvczx | woops 2002 not 2012 |
| 07:15.33 | inductiveload | well it was probably a bit better back then |
| 07:15.41 | xzcvczx | and someone in the same building had a nice ass lotus elise in electric blue |
| 07:15.41 | inductiveload | i was going to ask if tat was the olympics time |
| 07:15.49 | xzcvczx | commonwealth games |
| 07:15.55 | xzcvczx | but that was up in manchester |
| 07:16.03 | inductiveload | these days, that's probably the shittest car in that place ;-) |
| 07:16.17 | xzcvczx | haha wouldn't surprise me |
| 07:17.57 | inductiveload | i had a friend who lived in his mum's 'spare' appartment in st katharine docks |
| 07:18.07 | *** join/#kicad orsonmmz (~orsonmmz@orson.net.pl) |
| 07:18.09 | inductiveload | and that _was_ around 2012 |
| 07:18.36 | xzcvczx | lol |
| 07:18.36 | inductiveload | but the docklands is a like the rest of london, just a few streets over and it turns into a shithole |
| 07:19.16 | xzcvczx | what area do you live in? |
| 07:19.36 | inductiveload | way out of central, can't afford it |
| 07:20.08 | xzcvczx | lol you just too embarrased to admit you live in milton keynes huh |
| 07:20.20 | inductiveload | hah, no, inside the m25 |
| 07:20.35 | xzcvczx | ah ok |
| 07:20.55 | xzcvczx | are your signs in miles or km? |
| 07:20.59 | inductiveload | london's pretty horrible, but on the scale of cities in the world it's not too bad, but defintely won't live there forever |
| 07:21.00 | *** join/#kicad adamwwolf (~quassel@monique.wayneandlayne.com) |
| 07:21.01 | inductiveload | miles |
| 07:21.08 | xzcvczx | ewww |
| 07:21.17 | xzcvczx | i thought it was based on my memory of '02 |
| 07:21.51 | inductiveload | yeah, well if the EU had made us chage, all the brexit types would have revolted and voted to leave years ago |
| 07:21.56 | xzcvczx | i remember my parents stopping to ask what the speed limit was and being told as long as you are doing like < 90 they typically leave you alone |
| 07:22.13 | inductiveload | depends where |
| 07:22.22 | xzcvczx | aren't you mainly metric over there already though? |
| 07:22.38 | inductiveload | yeah exept road signs and beer |
| 07:22.48 | xzcvczx | on the highway between your major international airport and cambridge |
| 07:23.15 | xzcvczx | heathrow |
| 07:23.19 | inductiveload | well on the motorway, if everyone else is doing it, you'll be fine |
| 07:23.45 | inductiveload | but there're lots of average speed zones now, where they clock you in and out of a long stretch |
| 07:24.21 | xzcvczx | yeah i remember top gear saying in those you go 200mph for most them then just before you get to the 2nd one you stop for a coffee |
| 07:24.59 | inductiveload | so double the cost of the petrol, plus 10 quid for a shit coffee and stale croissant and you arrive at the same time |
| 07:25.23 | xzcvczx | nah man.... drive a tesla.... no fueld :P |
| 07:25.24 | xzcvczx | -d |
| 07:25.28 | inductiveload | clarksons a millionaire and the bbc/amaxon pays for hte petrol and foods on expenses, so what does he care |
| 07:25.38 | xzcvczx | ffs it was a joke dude |
| 07:25.43 | inductiveload | :-p |
| 07:26.09 | xzcvczx | did you ever go to a top gear filiming? |
| 07:26.16 | inductiveload | no |
| 07:26.27 | inductiveload | i'm not really into cars |
| 07:26.31 | xzcvczx | ah k |
| 07:26.38 | inductiveload | or jeremy clarkson ;-) |
| 07:27.06 | xzcvczx | can you just put in perspective how long it would take to get from point a to point a on the m25? |
| 07:27.16 | inductiveload | all the way around? |
| 07:27.19 | xzcvczx | yeah |
| 07:27.24 | inductiveload | when? |
| 07:27.29 | xzcvczx | middle of the night |
| 07:27.36 | xzcvczx | say 2:35am |
| 07:28.16 | inductiveload | as long as there aren't any major roadworks or spped limits, you'd probably be able to do 70 all the way |
| 07:29.01 | inductiveload | so a couple of hours |
| 07:29.05 | xzcvczx | ah ok |
| 07:29.14 | xzcvczx | and only 70 :O |
| 07:29.34 | xzcvczx | meh i guess thats 110km/hr |
| 07:29.48 | xzcvczx | our max speed on any nz road is only 100 :S |
| 07:30.02 | xzcvczx | although my mother lost her licence for 143km/hr |
| 07:30.20 | inductiveload | well would you go faster than that? you can't ogle the sheep if you go to fast |
| 07:30.28 | *** join/#kicad pointhi21 (~pointhi21@dyn166055.wlan.jku.at) |
| 07:30.39 | xzcvczx | quite hard to see the sheep at night |
| 07:30.56 | inductiveload | more nuclear power is what you need then |
| 07:31.06 | xzcvczx | haha more/any |
| 07:31.48 | inductiveload | is NZ a nuclear-free zone? sounds like a place that would be |
| 07:32.09 | xzcvczx | it is indeed |
| 07:32.24 | xzcvczx | pretty much since the french blew up the greenpeace boat |
| 07:32.31 | xzcvczx | that was docked in nz |
| 07:32.47 | xzcvczx | strained relationship to the us pretty badly |
| 07:33.17 | archivist | nuclear free zone, a joke, Derby city council declared themselves nuke free, except RR has a plant making the reactors for submarines! |
| 07:34.01 | inductiveload | south yorkshire is the same - apparently, swann morton has a gamma steriliser |
| 07:34.17 | xzcvczx | well we allow nuclear for medical reasons but thats it |
| 07:34.22 | xzcvczx | aka radiotherapy |
| 07:34.28 | archivist | and all the smoke detectors, and all the radioctive rocks |
| 07:34.39 | xzcvczx | and big rocks of uranium for universities to show off in 1st year physics classes |
| 07:34.49 | inductiveload | and bananas |
| 07:34.54 | xzcvczx | archivist: most smoke detectors are moving away from radio these days |
| 07:35.00 | archivist | I know |
| 07:35.41 | archivist | I used to work for .gov testing all the radiation monitors in case the bomb dropped |
| 07:36.17 | archivist | now that location is a B&Q supply depot :) |
| 07:37.19 | inductiveload | that's what they told you, anyway ;-) |
| 07:37.54 | inductiveload | but if you go into the kitchen section and pull on the right handle.... |
| 07:38.21 | xzcvczx | lol |
| 07:38.49 | archivist | it supplies the B&Q customer sites |
| 07:38.52 | xzcvczx | https://bugs.launchpad.net/kicad/+bug/1623364 <--- there we go c4 |
| 07:38.56 | xzcvczx | c4757p: ^^ |
| 07:40.06 | *** join/#kicad xorly (~xorly@ip-89-176-10-118.net.upcbroadband.cz) |
| 07:40.08 | archivist | inductiveload, fun part was working there during the firemen's strike 1977, we had the green goddess fire engines in store |
| 07:42.26 | inductiveload | i always wondered my they couldn't just use thre normal fire engines, but staff them with the army, but I just found out they do now do that |
| 07:43.04 | inductiveload | when the tube is on strike, they bring out the old green buses, reminds me of the green goddess fire engines |
| 07:43.49 | archivist | used to travel on the greenline......1960's |
| 07:44.01 | inductiveload | and then you realise that the new buses really are _much_ quieter and cleaner |
| 07:44.19 | *** join/#kicad rdmitry (~Dmitry@95-37-249-206.dynamic.mts-nn.ru) |
| 07:44.19 | trygvis | what is a good footprint to use with pogo pins? |
| 07:44.29 | xzcvczx | a big enough hole? |
| 07:45.00 | inductiveload | trygvis, ^^ that's what I used before, just a round pad |
| 07:45.06 | archivist | large enough to aim at |
| 07:45.17 | trygvis | I was hoping for something similar to the 1x03 pin header kind |
| 07:45.32 | inductiveload | you mean to mount the pin, or to poke with the pin? |
| 07:45.48 | archivist | depends how good your jig will be |
| 07:45.53 | trygvis | I'll poke the board with the programmer that has the pogo pins |
| 07:46.19 | trygvis | the jig will most likely be my hand so not very good |
| 07:47.11 | inductiveload | well if your pads have plated holes, the pins will self centre? |
| 07:47.55 | trygvis | sure, I can just use the 01x03 pin header footprint |
| 07:47.59 | inductiveload | but much less pressure around the rim of a hole than at a point, so the ribbed-head pins might be better |
| 07:48.12 | archivist | I had holes and used a bit of header and leaned it to make contact for programming by hand |
| 07:48.34 | trygvis | but I was hoping for a footprint with a 3d-model without the pins |
| 07:48.41 | trygvis | so just holes I guess |
| 07:48.51 | inductiveload | well then just remove the model? |
| 07:49.09 | inductiveload | since there won't actually be anything mounted on that footprint? |
| 07:49.38 | trygvis | that'll work |
| 07:50.34 | *** join/#kicad metacollin (~anonymous@c-174-56-101-29.hsd1.nm.comcast.net) |
| 07:51.51 | xzcvczx | tries to decide what to get for dinner |
| 07:51.59 | inductiveload | lamb |
| 07:52.07 | xzcvczx | naw not a big fan of lamb |
| 07:52.14 | inductiveload | just lean outside and scoop one up |
| 07:52.27 | metacollin | Just timed a from scratch build (no cached files etc.) of KiCad on my new workstation. Start to finish 2m41.218s :D. |
| 07:52.57 | xzcvczx | bullshit |
| 07:53.06 | metacollin | I have 28 cores |
| 07:53.08 | xzcvczx | something was cached there |
| 07:53.15 | xzcvczx | i don't care i still call bullshit |
| 07:53.31 | metacollin | Nothing was cashed =P |
| 07:53.40 | xzcvczx | you could have 500 cores but kicad isn't that parallelable |
| 07:53.49 | metacollin | er thats beginning with make |
| 07:53.55 | metacollin | not counting cmake config time etc. |
| 07:54.00 | inductiveload | xzcvczx, it's not bad |
| 07:54.02 | metacollin | er, building it is |
| 07:54.10 | inductiveload | becuase it builds cpp by cpp |
| 07:54.22 | metacollin | in fact, the entire compile, I had 100% utilization across all 56 logical cores |
| 07:54.28 | inductiveload | linking takes time, but you can still link one target per core |
| 07:54.41 | metacollin | I just ran make -j56 |
| 07:54.51 | metacollin | clang did the rest =P |
| 07:54.52 | xzcvczx | kicad is also dependency hell |
| 07:55.03 | metacollin | oh I mean I already had the dependencies |
| 07:55.30 | xzcvczx | nah i mean file a needs b c d e f g h i j k l file b needs m n o p q r s t u v w x y z |
| 07:55.37 | xzcvczx | aka not parallelable |
| 07:56.22 | xzcvczx | but if you insist its only 2 mins, then nuke your build folder, then run cmake and time make |
| 07:56.27 | inductiveload | xzcvczx, doesn't matter for building, that's the linkers problem |
| 07:56.28 | xzcvczx | and pastebin it and i will believe you |
| 07:56.32 | metacollin | I did time make |
| 07:56.42 | inductiveload | xzcvczx, also clear ccache |
| 07:56.46 | metacollin | real2m41.218s |
| 07:56.47 | metacollin | user105m8.167s |
| 07:56.47 | metacollin | sys7m23.575s |
| 07:56.52 | xzcvczx | clang doesn't ccache |
| 07:56.59 | metacollin | I don't even have ccache installed |
| 07:57.13 | metacollin | But yeah, i'll rm -rf build dir |
| 07:57.34 | inductiveload | gcc doesn't ccache either, ccache gcc's ;-) |
| 07:57.55 | inductiveload | does clange do its own caching? |
| 07:58.50 | metacollin | not as far as I know |
| 07:59.18 | xzcvczx | nor i |
| 07:59.29 | xzcvczx | inductiveload: it never gets any faster for me |
| 07:59.50 | xzcvczx | metacollin: ssd? |
| 08:00.01 | *** join/#kicad ricardocrudo (~ricardocr@185.19.197.52) |
| 08:02.26 | metacollin | Oh yeah, I do have an SSD though I don't think the build is particularly IO limited |
| 08:02.29 | metacollin | but I could be wrong |
| 08:02.48 | xzcvczx | every little bit helps |
| 08:02.55 | metacollin | here one sec |
| 08:03.09 | xzcvczx | esspecially as i am pretty sure that spotlight would fuck you over pretty bad if you had hdd |
| 08:03.15 | inductiveload | i guess you probably have enough ram to get all your headers 'n' crap into ram |
| 08:03.25 | inductiveload | and keep them there |
| 08:03.31 | metacollin | 64GB of ECC ram so yeah |
| 08:03.40 | inductiveload | i take it this is a mac pro? |
| 08:04.00 | metacollin | naw, its a super micro dual xeon v4 rig I built |
| 08:04.22 | xzcvczx | lol apple really shouldn't have given up rackmount pros |
| 08:04.24 | inductiveload | oh, sorry, I thought you were on a mac |
| 08:04.29 | inductiveload | they didn't |
| 08:04.30 | metacollin | Well, I am using OS X |
| 08:04.35 | xzcvczx | thief |
| 08:04.44 | xzcvczx | hackintosh or virtual machine? |
| 08:04.45 | metacollin | No, I have a legal copy. |
| 08:04.49 | inductiveload | http://www.sonnettech.com/product/images/rackmacpro_2modulesmacpros.png |
| 08:05.09 | xzcvczx | metacollin: doesn't mean you are allowed to install it on non-apple hardware |
| 08:05.31 | inductiveload | i love sonnet, they're like a whole company dedicated to working around problems caused by apple design choices |
| 08:05.38 | xzcvczx | inductiveload: that looks suspiciously like a 4u when weren't the old mac pros 1u? |
| 08:06.34 | inductiveload | what the old silver ones? |
| 08:07.04 | xzcvczx | yeah |
| 08:07.30 | inductiveload | they were huge? |
| 08:08.00 | xzcvczx | oh i am thinking of xserve |
| 08:09.52 | inductiveload | is it a really bad idea to put componnets on the other side of a board from a power supply filter coil? |
| 08:10.10 | metacollin | xzcvczx: I am allowed to. There are a dozen legal precedents in the US where EULAs with clauses like those found in OS X's EULA were found to be invalid and unenforceable. A EULA must be valid to be agreed to. |
| 08:10.18 | xzcvczx | inductiveload: how intolerant are they to noise |
| 08:10.44 | xzcvczx | metacollin: yet did apple not succesfully sue some hackintosh company into obscurity? |
| 08:10.58 | inductiveload | they're the fuses and zeners and stuff on the "clean" side of the filter |
| 08:10.59 | metacollin | No, there was a company making mac clones |
| 08:11.07 | metacollin | Which is very different |
| 08:11.20 | metacollin | they're selling something, I'm just installing software on my computer |
| 08:13.22 | metacollin | Plus I've given apple no small sum of money (at least $10,000) over the last 15 years or so (last several machines were real macs, including a 2008 mac pro). And the reason I built this was because there simply wasn't a recent mac pro with enough power available. There is no mac pro with v4 or even v3 xeons, so I made my own. *shrug*. |
| 08:14.04 | inductiveload | mac pros are wierd |
| 08:14.14 | metacollin | They keep ignoring them recently |
| 08:14.35 | inductiveload | they ram them full of top-flight stuff, then let them rot for years |
| 08:14.35 | xzcvczx | metacollin: tbh i haven't heard great things about hackintosh |
| 08:14.36 | metacollin | Oh you mean the form factor? |
| 08:14.49 | inductiveload | form factor too |
| 08:14.55 | inductiveload | especially the GPU |
| 08:15.01 | metacollin | what's hackintosh? is that a company? |
| 08:15.04 | xzcvczx | inductiveload: thats happening with all their non-mobile hardware these days |
| 08:15.07 | inductiveload | the one thing you'd want to be able to upgrade |
| 08:15.14 | xzcvczx | metacollin: nah osx on non-apple hardware |
| 08:15.25 | metacollin | i know I'm just trolling =P. Sorry heh |
| 08:15.31 | xzcvczx | oh meh |
| 08:15.35 | metacollin | It works great as long as you know what you're doing |
| 08:15.40 | xzcvczx | too early in the morning for all the green under the bridge shit :P |
| 08:15.44 | metacollin | which most people in the hackintosh community don't |
| 08:15.51 | xzcvczx | ah that makes sense |
| 08:16.17 | xzcvczx | metacollin: are you running sierra on anything yet? |
| 08:16.21 | inductiveload | metacollin, well they _were_ mac users, can't expect too much |
| 08:18.28 | metacollin | xzcvczx, 10.11.6 but going to upgrade to sierra tonight probably |
| 08:19.25 | *** join/#kicad tumnao (b91f64ed@gateway/web/freenode/ip.185.31.100.237) |
| 08:19.47 | metacollin | I patched super micro's bios for my mobo so I can run OS X without a patched kernel too, which makes updating so much easier heh |
| 08:19.58 | metacollin | (and added the apple logo to the boot screen heh) |
| 08:20.03 | xzcvczx | metacollin: can you ping me with what version of python its running (and if py3 is installed by default)? and also whether any of the shit thats come up as deprecated has been nuked that kicad needs |
| 08:20.28 | xzcvczx | wow nice, thats probably much more reliable than kernel patches |
| 08:22.55 | xzcvczx | bbs getting food |
| 08:26.11 | metacollin | yeah its pretty much fool proof. Basically OS X tries to write to MSR registers (specific configuration registers on intel CPUs) that most BIOSes lock. Writing to a locked MSR causes the CPU instashit itself and you reboot immediately. So you just make the BIOS not lock the right registers and you're good. Way more robust than trying to make the kernel not try to write to it, since you probably have to do a new patch for each m |
| 08:26.12 | metacollin | of the kernel |
| 08:29.08 | metacollin | oh, no py3 is not default on 10.11. Or did you mean 10.12> |
| 08:29.22 | metacollin | oh 10.12 probably |
| 08:33.30 | metacollin | lol |
| 08:33.53 | *** join/#kicad cirilo (~cirilo@CPE-120-144-53-157.lnse5.win.bigpond.net.au) |
| 08:34.09 | metacollin | ok |
| 08:34.14 | metacollin | I made a gif for xzcvczx |
| 08:35.17 | *** join/#kicad Thaolia (~thaolia@80.90.49.230) |
| 08:37.47 | cirilo | sometimes I just loathe clion |
| 08:38.12 | cirilo | this java shit is taking close to 100% on all 8 processor cores |
| 08:38.53 | cirilo | I might have to resort to running clion on a VM with only 2 cores |
| 08:40.06 | cirilo | cyborg_ar, I fixed the kicad2step code; now it skips any (drill) specification in an SMD pad |
| 08:40.32 | cirilo | it must be one of kicad's strange things - specifying a drill location for a pad with no drill. |
| 08:45.22 | metacollin | IDEs are not worth it IMO. I use a good editor (like Atom or SublimeText3) which can do all the stuff an IDE does (well, the ones that actually matter) but without being heavy weight java bloatware and you only pay a little bit more in setup time, but you can build using some scripts etc. |
| 08:46.00 | xzcvczx | metacollin: is finder not encrypted or some shit on osx? |
| 08:46.14 | xzcvczx | that use to be rumour not sure about practicality though |
| 08:46.31 | metacollin | yeah, actually most everything except what you need to boot up enough to decrypt the rest is encrypted |
| 08:47.20 | xzcvczx | ah ok |
| 08:48.52 | xzcvczx | wow apple could reaaly take some hints from gopro on how to avoid product leaks |
| 08:49.13 | metacollin | i mean, the password to the encrypted parts is ourhardworkbythesewordsguardedpleasedontsteal(c)AppleComputerInc |
| 08:49.26 | metacollin | because it gets the password from the SMC |
| 08:49.31 | metacollin | so some dude just sniffed the bus |
| 08:49.43 | metacollin | they honestly didn't try *that* hard lol |
| 08:49.46 | metacollin | apple I mean |
| 08:50.26 | xzcvczx | oh lol |
| 08:50.46 | xzcvczx | wait for them to use the tpm :P |
| 08:50.53 | inductiveload | yeah, but do they really care? |
| 08:51.01 | metacollin | Thats the thing, I don't think they care |
| 08:51.21 | metacollin | They've never ever seemed to care and the hackintosh community has been thriving for years |
| 08:51.25 | xzcvczx | software aint really their breadwinner |
| 08:51.29 | inductiveload | as soon as anyone make any sort of money, they'll shut it down |
| 08:51.37 | metacollin | yeah, definitely |
| 08:51.49 | metacollin | But they don't care about hobbyist/hacker types playing around |
| 08:51.52 | inductiveload | but if it's some people hacking and fiddling on their own, me |
| 08:51.53 | inductiveload | meh |
| 08:52.16 | metacollin | exactly. They even have easter eggs in the OS specifically for us to find |
| 08:52.26 | inductiveload | and those types probably also buy macbooks and iphones too, so if you do slam them, they'll mon online and then leave |
| 08:52.26 | metacollin | like there is a poem, let me find it |
| 08:52.42 | xzcvczx | tbh if you want a nice server their shit just dont cut it |
| 08:52.55 | metacollin | Oh yeah OS X Server is not very good |
| 08:53.21 | xzcvczx | i more meant hardware |
| 08:53.55 | xzcvczx | osx server is just a buncha mem leaks though |
| 08:53.57 | metacollin | they don't make any servers anymore |
| 08:54.02 | metacollin | i thought |
| 08:54.26 | xzcvczx | well mac pro/mac mini such as they are |
| 08:56.12 | metacollin | The mac pros used to be awesome |
| 08:56.39 | inductiveload | everything about them felt expensive |
| 08:56.39 | metacollin | the latest ones would be ok if it was still the year they were released, and they were about half to 1/3rd the price |
| 08:57.03 | inductiveload | the new ones have a silly form factor too |
| 08:57.25 | xzcvczx | i actually wonder whether something big is coming as all of their non-mobile hardware is old as shit currently |
| 08:57.26 | metacollin | but like, my prior computer was a mac pro 2008. That thing was a steal, Apple got their Xeons from intel at some really good price or something, because the cost of the CPUs retail was more than the entire mac pro (which had those same CPUs installed). |
| 08:57.49 | xzcvczx | as in bigger than just a slight update in hardware |
| 08:58.13 | metacollin | xzcvczx: I'm not holding my breath. They've been kind of neglecting the mac pro line for a while now |
| 08:58.28 | xzcvczx | its not just mac pro though |
| 08:58.31 | metacollin | oh |
| 08:59.10 | metacollin | oh woah holy crap |
| 08:59.14 | xzcvczx | pretty much the entire mac range i think is 2014 |
| 08:59.18 | xzcvczx | except the macbook |
| 08:59.25 | xzcvczx | aka the single usb-c thing |
| 08:59.25 | metacollin | you're right, basically their entire hardware lineup is due for a new release |
| 08:59.29 | metacollin | http://buyersguide.macrumors.com/#Mac |
| 09:00.39 | xzcvczx | i am always right :) |
| 09:01.27 | xzcvczx | i still wonder whether they are going to move away from x86_64 |
| 09:02.32 | metacollin | like to ARM? |
| 09:02.36 | xzcvczx | yeah |
| 09:03.35 | metacollin | they really need to hire someone whose job is just to tell johnny ive no sometimes |
| 09:03.51 | metacollin | like when he wants to get rid of headphone jacks, that person can step in and be like, "no" |
| 09:04.02 | metacollin | or he wants to move everything to arm |
| 09:05.01 | xzcvczx | wth is johnny ive? |
| 09:05.16 | metacollin | https://en.wikipedia.org/wiki/Jonathan_Ive |
| 09:05.45 | metacollin | he is more or less responsible for the look and design of all of apple's hardware products, starting with the bondi blue imac |
| 09:05.45 | cirilo | I've the crackpot hailed as a genius by Jobs etc |
| 09:06.03 | metacollin | yeah he has a real chub for minimalism |
| 09:06.20 | xzcvczx | lol they claim hes the designer of all those products.... yeah that ain't gonna happen..... hes todays picasso, he ain't designed all those |
| 09:06.46 | archivist | pencil drawing is probably all he manages |
| 09:06.54 | cirilo | nah, standard "I take the credit for what my team did because I'm so wonderful" type of manager |
| 09:07.23 | xzcvczx | cirilo: indeed, its like many of the old school painters, if they had painted all that their name was on they wouldn't have had time to sleep or eat |
| 09:07.37 | xzcvczx | and other arts |
| 09:07.49 | cirilo | yup |
| 09:08.47 | cirilo | truth is the most he likely does is shout at the actual artists and tell them to make some trivial change. then tell everyone he designed the whole lot and that the sun is shining brightly from his ass that day |
| 09:08.55 | metacollin | i mean, watch the documentary objectified sometime. It's fantastic, btw. Made me realize why design should be appreciated and valued, and how important good design really is. And its just fascinating. But there are some parts on Ive |
| 09:09.15 | *** join/#kicad firewalker (~firewalke@46.246.152.93.dsl.dyn.forthnet.gr) |
| 09:09.27 | metacollin | and I don't think the degree his hand played in design of apple products is really that exaggerated |
| 09:09.51 | cirilo | if you've got enough money there's no shortage of stooges who will hail you as a genius |
| 09:11.23 | metacollin | No no, I mean, he could talk about the design of the internal assembly and screw holes of a mac book pro being made at the time on camera, from memory, in a way that very much convinced me he largely did the cad work for the casing |
| 09:11.44 | metacollin | I mean, I am not saying he's a genious, I'm just saying I don't think he's just a figurehead |
| 09:11.47 | cirilo | meh - probably stuff he remembers from a design review |
| 09:12.37 | xzcvczx | metacollin: yeah but all the brains are more in the coming up with concept stage, and he at most would be involved in maybe giving direction to the concept designers, and parts of the developement stage |
| 09:13.12 | xzcvczx | the external designer sure as shit ain't the same person as the internal designer |
| 09:14.11 | xzcvczx | oooooo sdf has been pushed |
| 09:14.16 | xzcvczx | congrats decimad |
| 09:17.18 | metacollin | well and fixtures |
| 09:17.19 | metacollin | https://youtu.be/nUHROAtyGIg?t=3m10s |
| 09:17.24 | metacollin | oh nice |
| 09:20.04 | metacollin | i also suspect he's a massive douche in private but thats just a suspicion heh |
| 09:20.12 | xzcvczx | lol |
| 09:20.16 | xzcvczx | hes french of course he is |
| 09:20.27 | xzcvczx | well he has a french sounding name at least :P |
| 09:20.30 | metacollin | is JP around? ;) |
| 09:20.35 | xzcvczx | lol |
| 09:20.36 | metacollin | He's british |
| 09:20.40 | metacollin | and a knight =P |
| 09:20.49 | xzcvczx | ives i am pretty sure is not a british name |
| 09:20.55 | xzcvczx | or ive |
| 09:21.05 | metacollin | ooh you mean like his lineage |
| 09:21.18 | archivist | StIves is a uk village/town |
| 09:21.24 | xzcvczx | wow can't even make a simple damn joke no more |
| 09:22.07 | xzcvczx | metacollin: and meh knight don't mean shit..... my cousin was Dr Sir Sir |
| 09:22.16 | xzcvczx | Dr being honorary iirc |
| 09:23.29 | cyborg_ar | cirilo, great! |
| 09:23.35 | metacollin | oh, the tone was like, "they make anyone a knight now" but didn't transfer over IRC |
| 09:23.49 | metacollin | Er we're on the same page heh |
| 09:24.26 | metacollin | honorary just means in no ways but name |
| 09:24.42 | cyborg_ar | if only apple did good design |
| 09:24.56 | cyborg_ar | like laptops that dont overheat |
| 09:24.58 | xzcvczx | metacollin: the phd was honorary |
| 09:25.05 | xzcvczx | or maybe not its hard to tell |
| 09:25.14 | xzcvczx | phds get given out for the most random ass shit these days though |
| 09:25.43 | xzcvczx | cyborg_ar: meh i barely have the fan come on in my air |
| 09:25.50 | xzcvczx | mainly only when encoding video |
| 09:26.03 | xzcvczx | or having 80 tabs open in safari |
| 09:26.25 | cyborg_ar | still, nobody is as bad as HP |
| 09:26.36 | metacollin | there are no honorary Ph.Ds |
| 09:26.40 | xzcvczx | hahahahaha hahahaha.... apparently my cousin shot him self in the foot in 1953 :S |
| 09:26.40 | metacollin | there are honorary doctorates |
| 09:26.53 | xzcvczx | pedantic bastard : |
| 09:26.54 | xzcvczx | P |
| 09:26.55 | metacollin | but they aren't allowed to put Ph.D after their name and its not considered one |
| 09:26.56 | xzcvczx | :P |
| 09:27.12 | metacollin | No no, I mean, its like that so you know if it says Ph.D, its legit |
| 09:28.30 | metacollin | And you must write and successfully defend a thesis, which is usually 100-300 pages long. If you are able to defend it against peer review and its felt that the contribution is significant enough, it gets published and you get your Ph.D |
| 09:28.39 | xzcvczx | but yeah he was knighted twice so was pretty nuts |
| 09:28.43 | metacollin | lol wut |
| 09:29.02 | metacollin | thats like losing your virginity twice |
| 09:29.24 | xzcvczx | metacollin: yeah don't get me started on contribution being significant..... my father's phd ain't worth the paper its printed on with his thesis |
| 09:29.52 | cyborg_ar | lol |
| 09:30.22 | metacollin | well, its safe to say personal bias isn't influencing that statement lol |
| 09:30.39 | cyborg_ar | well if they only graduated earth shattering geniuses it would look pretty discouraging |
| 09:30.41 | metacollin | but yeah, I imagine that's true |
| 09:31.10 | metacollin | Well, contribution can be just finding out how not to do something |
| 09:31.16 | xzcvczx | metacollin: i am just trying to find the title of his thessis now |
| 09:31.16 | cyborg_ar | in anything you'll find people that got there they are by just drifting/getting by |
| 09:32.41 | xzcvczx | metacollin: Perceptions of accounting and accountants: an investigation into how and why these perceptions were forme |
| 09:32.42 | metacollin | yeah, there are people like that =/. I think some of it might be that it might be kind of hard to tell this person who just spent the last few motnhs sleeping 5 hours a day and doing this huge amount of work that it sucks to much try again |
| 09:32.46 | cyborg_ar | a bigger problem is the bias against publishing null/negative results |
| 09:32.51 | metacollin | even if it really does suck too much |
| 09:33.21 | metacollin | since the professors are also the advisors to the candidates etc. |
| 09:33.22 | metacollin | I dunno nvm |
| 09:33.51 | xzcvczx | metacollin: abstract: There is growing concern that a widespread perception of accounting and accountants is discouraging individuals with the required skills and capabilities from entering the accounting profession. At the same time, an inappropriate perception may be attracting those who lack the required skills and |
| 09:34.06 | metacollin | LOL |
| 09:34.15 | metacollin | what field was this? |
| 09:34.17 | xzcvczx | so yeah don't talk to me about significant contribution |
| 09:34.18 | xzcvczx | accounting |
| 09:34.29 | archivist | giggles |
| 09:35.06 | metacollin | is there a link anywhere to a PDF? I need to send this to my experimental physicsit friend (he just got his phd) |
| 09:35.14 | cirilo | xzcvczx, I always thought Ive was an old English name - kind of like Higham |
| 09:35.15 | xzcvczx | http://aut.researchgateway.ac.nz/handle/10292/830 |
| 09:35.30 | metacollin | thanks :D |
| 09:35.32 | cirilo | so old it had gone out of fashion centuries ago |
| 09:35.35 | xzcvczx | cirilo: meh stop fucking up my chi |
| 09:35.56 | xzcvczx | metacollin: theres a pdf link on that page if they are to read 1.182MB of drivel |
| 09:36.03 | xzcvczx | and waffling |
| 09:36.18 | cyborg_ar | i dunno |
| 09:36.23 | cyborg_ar | it sounds interesting |
| 09:36.28 | metacollin | yeah |
| 09:36.33 | metacollin | its 230 pages |
| 09:36.35 | metacollin | lol |
| 09:36.44 | metacollin | I mean |
| 09:36.51 | xzcvczx | hes an accountant... what do you expect |
| 09:36.56 | cirilo | wtf .. clion still doing ${RANDOM_SHIT} and eating my CPU time |
| 09:36.57 | *** join/#kicad NoGodDamnIdea (~King@188.241.239.53) |
| 09:36.57 | *** join/#kicad NoGodDamnIdea (~King@unaffiliated/nogoddamnidea) |
| 09:36.58 | metacollin | he wasn't lazy, thats for sure |
| 09:37.44 | xzcvczx | cirilo: its java... what do you expect it to do instead? |
| 09:37.49 | metacollin | its probably just running a bunch of __asm__ nop loops |
| 09:37.56 | metacollin | that's how java rolls |
| 09:38.06 | metacollin | java just wants to watch the world burn |
| 09:38.18 | cyborg_ar | god the guy really padded it |
| 09:38.19 | xzcvczx | metacollin: and damn that paper/thesis was harder to find than i thought it would be |
| 09:38.22 | xzcvczx | stupid aut website |
| 09:38.23 | cyborg_ar | so much white space |
| 09:38.34 | xzcvczx | cyborg_ar: i will tell him you think so |
| 09:38.35 | archivist | java and boost both need to be removed from the gene pool |
| 09:38.46 | metacollin | I mean, most stuff is behind a paywall so its actually awesome you could get it at all |
| 09:38.49 | cyborg_ar | it could have been the style guide |
| 09:38.57 | cyborg_ar | but this text should fit in 100 pages |
| 09:38.58 | cyborg_ar | lol |
| 09:39.07 | xzcvczx | metacollin: paywalls are not nealy as popular direct from universities over here |
| 09:39.11 | metacollin | or maybe we should rewrite boost, but in java |
| 09:39.12 | cyborg_ar | double margin, double spacing |
| 09:39.14 | metacollin | KIDDING |
| 09:39.22 | metacollin | oh well good on NZ then |
| 09:39.26 | metacollin | that's how it should be |
| 09:39.48 | xzcvczx | metacollin: you know in southpark how the devil could invade after so many swear words or whatever it was..... same thing happens if you rewrite boost in java |
| 09:39.55 | xzcvczx | devil/satan |
| 09:39.58 | metacollin | lol |
| 09:40.30 | xzcvczx | meh tbh if it was behind a paywall i would have just emailed him and told him to send it to me :) |
| 09:40.31 | cyborg_ar | damn boost in java |
| 09:40.40 | xzcvczx | if i cared enough |
| 09:40.42 | metacollin | isn't that the worst thing you've heard all week? |
| 09:40.43 | cyborg_ar | it's like rewriting autotools in perl |
| 09:40.50 | metacollin | HAHAHAHAHAH |
| 09:40.59 | xzcvczx | metacollin: may i ask what your friends phd was on? |
| 09:41.04 | xzcvczx | and don't say experimental physics |
| 09:41.18 | metacollin | High energy gamma ray sources |
| 09:41.25 | xzcvczx | cyborg_ar: i dunno.... m4 vs perl is debatable |
| 09:41.29 | cyborg_ar | mmm gamma rays |
| 09:41.32 | xzcvczx | metacollin: cool |
| 09:41.34 | metacollin | he was also the sysadmin for the entire http://www.hawc-observatory.org/ |
| 09:41.59 | cyborg_ar | xzcvczx, well to be entirely compatible you need to write a M4 interpreter in perl |
| 09:41.59 | xzcvczx | metacollin: although high energy gamma ray sources would probably be a bitch to research here |
| 09:42.09 | xzcvczx | fuck compatibility |
| 09:42.14 | cyborg_ar | lol |
| 09:42.21 | xzcvczx | compatibility is for people living in the past..... quit living in the past |
| 09:42.36 | cirilo | clion died - java shit chewed up 4G and wants more |
| 09:42.40 | metacollin | yeah, the actual observatory is in mexico |
| 09:42.48 | metacollin | at high latitude |
| 09:42.56 | cirilo | simply unusable - will probly ditch 'em and go back to a retarded text editor |
| 09:43.11 | xzcvczx | metacollin: meh i am more referring to the whole nuclear-free shit over here, purposely importing high gamma ray sources probably wouldn't be easy |
| 09:43.24 | xzcvczx | cirilo: you could try vscode |
| 09:43.26 | xzcvczx | or atom |
| 09:43.34 | cyborg_ar | or eclipse even |
| 09:43.42 | xzcvczx | or not |
| 09:44.35 | metacollin | and the sensors are giant vats of water. There are photo multiplier tubes inside that can detect cherenkov radiation which is the electromagnetic version of a 'sonic boom', when something moves faster than the speed of light in a medium (speed of light in water is like 90% that of vacuum I think?). So it works well for detecting the huge particle showers of near light-speed protons and crap a gamma ray creates when it hits the at |
| 09:44.58 | metacollin | then they actually back calculate the position/energy/etc of the gamma ray from the shower of particles and their trajactories |
| 09:45.08 | metacollin | its pretty cool stuff |
| 09:46.30 | cyborg_ar | ooooh jp wants to remove old code :D |
| 09:47.15 | metacollin | xzcvczx: uh these are gamma rays that come from elsewhere in the galaxy, or near by galaxies |
| 09:47.28 | metacollin | you don't need to import them, they're importing themselves all the time |
| 09:47.29 | xzcvczx | oh duh |
| 09:47.44 | xzcvczx | i assumed high-energy would be local shit relatively speaking |
| 09:48.19 | metacollin | naw, high energy like, our current understanding of physics can't explain how a gamma ray of such high energy can even come into being |
| 09:48.57 | metacollin | which is why they're observing and measuring them, to advance our understanding of the universe to use wank phrasing |
| 09:49.49 | xzcvczx | lol so what did he come up with? NFI? :P |
| 09:49.54 | *** join/#kicad Novus (~Novus@2605:e000:628b:1400:df2:8679:73d0:e704) |
| 09:50.08 | cyborg_ar | it's god's farts |
| 09:50.16 | metacollin | I actually proof red his thesis, and let me tell you |
| 09:50.19 | metacollin | I have no fucking idea |
| 09:50.25 | metacollin | way over my head lol |
| 09:50.25 | xzcvczx | haha |
| 09:50.37 | xzcvczx | awww and i had respect for you :P |
| 09:51.08 | metacollin | I mean, the maths got hard :( |
| 09:51.13 | xzcvczx | orsonmmz: btw thanks |
| 09:51.37 | metacollin | oh ok |
| 09:51.47 | xzcvczx | metacollin: let me guess.... it had more than 2 operators? :P |
| 09:51.59 | orsonmmz | xzcvczx: you're welcome, whatever I've done ;) |
| 09:52.17 | metacollin | he figured out how to back calculate the gamma rays from particle showers for TEV gamma rays emitted from active galactic nuclei |
| 09:52.23 | metacollin | lemme see if Ican just find the paper |
| 09:52.34 | xzcvczx | orsonmmz: seeing as no one else had i was thanking you for merging the sdf stuff :) |
| 09:52.46 | cyborg_ar | and the undo |
| 09:53.15 | xzcvczx | metacollin: he should have just set up his equipment under cheyenne mountain next to the stargate |
| 09:53.16 | cyborg_ar | lots of work for a no-op on the user's side |
| 09:53.25 | xzcvczx | wasn't undo tom's? |
| 09:53.26 | metacollin | lol naw thats for neutrinos |
| 09:53.33 | cyborg_ar | oh undo was tom right |
| 09:53.44 | cyborg_ar | i keep confusing them, sorry |
| 09:53.51 | xzcvczx | everyone does :) |
| 09:54.05 | cyborg_ar | wait |
| 09:54.18 | cyborg_ar | no it was orson |
| 09:54.40 | xzcvczx | metacollin: there are plenty of space gates that would be able to get other locations of gamma rays |
| 09:54.53 | xzcvczx | cyborg_ar: my bad |
| 09:55.26 | cyborg_ar | :) |
| 09:55.39 | xzcvczx | wow mr suminski has been busy |
| 09:55.48 | xzcvczx | i don't even have all those nutty letters on my keyboard |
| 09:56.02 | xzcvczx | well the 'n' anyways |
| 09:56.59 | xzcvczx | orsonmmz: btw did you ever proceed with your osx building? |
| 09:57.40 | orsonmmz | xzcvczx: the real thx should go to decimad, that's his hard work |
| 09:57.47 | orsonmmz | xzcvczx: nope, I've lost all my incentives |
| 09:57.53 | xzcvczx | ok |
| 09:57.58 | orsonmmz | are there any particular problems with osx now? |
| 09:58.23 | metacollin | https://metacollin.com/thesis.pdf |
| 09:58.38 | metacollin | I'm in the awknowledgements :D. But its pretty serious heh |
| 09:58.42 | metacollin | er the thesis |
| 09:58.56 | metacollin | oh wait |
| 09:58.56 | xzcvczx | just the broke/lack of nightly builds.... some python thing i have introduced which i have no damn clue about in my builds, and the bugs on the tracker (many of which are wx bugs) |
| 09:58.58 | metacollin | thats not the thesis |
| 09:59.00 | metacollin | thats 4 pages |
| 09:59.06 | xzcvczx | lol |
| 09:59.15 | xzcvczx | shortest phd thesis in history |
| 09:59.16 | cyborg_ar | that's a short thesis |
| 09:59.42 | xzcvczx | hmmmm he shoould really change his middle initial to m |
| 09:59.46 | xzcvczx | Im Wisher |
| 10:01.29 | xzcvczx | orsonmmz: with the nightly issue being know |
| 10:01.30 | xzcvczx | n |
| 10:01.44 | xzcvczx | but after adam gets that shit fixed up we should get it sorted pretty quick |
| 10:01.56 | xzcvczx | although next stage is to get shit signe |
| 10:01.56 | xzcvczx | d |
| 10:03.56 | xzcvczx | c4757p: i wonder whether 1041066 will make your board any faster |
| 10:04.19 | metacollin | https://metacollin.com/Thesis.pdf |
| 10:04.20 | metacollin | there |
| 10:04.26 | cirilo | fucking java: 755% of CPU time |
| 10:04.31 | xzcvczx | lol |
| 10:04.37 | xzcvczx | quad core? |
| 10:04.48 | metacollin | not anymore, java is using all of it |
| 10:04.48 | cirilo | yup |
| 10:05.00 | xzcvczx | metacollin: was this generated out of latex? |
| 10:05.06 | metacollin | I have no idea |
| 10:05.07 | cirilo | and 6GB .. WTF .. seriously |
| 10:05.27 | metacollin | Well, the math was |
| 10:05.33 | metacollin | the rest was matplotlib I think |
| 10:05.37 | *** join/#kicad maiden (~maiden@c83-255-72-208.bredband.comhem.se) |
| 10:05.38 | xzcvczx | pdfTeX-1.40.11 |
| 10:05.50 | xzcvczx | LaTeX with hyperref package |
| 10:05.56 | cirilo | and I thought 16GB would be enough - even SolidWorks doesn't puke so much with that .. but crap JAVA code is special |
| 10:06.18 | xzcvczx | cirilo: whats your problem.... you had a whole 45% free for other tasks |
| 10:08.40 | cirilo | whoop - up to 7GB now |
| 10:09.03 | xzcvczx | cirilo: half the fault is probably kicads fucked up tree |
| 10:09.07 | cirilo | xzcvczx, well computer is *somewhat* responsive with that 45% .. earlier I was getting more like 0.5% |
| 10:09.10 | xzcvczx | and its insistence on <> everywhere |
| 10:09.51 | cirilo | well CLion was working fine with kicad until a clion update late last year - been shit since and following instructions to 'fix' it only delays the shit |
| 10:10.19 | cirilo | + it's also shit when I work on other code which is un-kicad-like |
| 10:10.39 | xzcvczx | oh ok |
| 10:11.08 | xzcvczx | metacollin: ok i officially hate your firend....... "DISCARD THIS PAGE" is almost as bad as "THIS PAGE IS MEANT TO BE BLANK" |
| 10:11.28 | metacollin | oh its tghe rough draft |
| 10:11.29 | cirilo | fucking piece of shit .. finished indexing and I closed clion and started again .. and it's doing the indexing again... fucking genius |
| 10:11.31 | metacollin | not the final version |
| 10:11.45 | xzcvczx | ah ok |
| 10:11.47 | metacollin | I can't find the actual published one anywhere, paywalls =/ |
| 10:11.58 | metacollin | the discard this page was a note to himself heh |
| 10:12.15 | metacollin | oh yeah, I forgot |
| 10:12.21 | *** part/#kicad thearduinoguy (sid138303@gateway/web/irccloud.com/x-thnkdhstvzxzeyzp) |
| 10:12.30 | metacollin | xzcvczx: https://metacollin.com/compile.mp4 |
| 10:13.26 | xzcvczx | ewwww wtf seriouslly...... "-DUSE_IMAGES_IN_MENU=ON" are you insane? |
| 10:13.41 | *** join/#kicad cosmobird (~firebird@125.16.18.198) |
| 10:13.58 | xzcvczx | and boo no scripting |
| 10:14.07 | xzcvczx | and why do you bother to specify c compiler and cxx compiler |
| 10:15.08 | xzcvczx | goddamned warnings |
| 10:16.15 | metacollin | I love the images in the menus =P |
| 10:16.22 | cyborg_ar | xzcvczx, the "Page purposely blank" has a purpose, say that the page was intended to be blank |
| 10:16.34 | cyborg_ar | otherwise if you see a blank page you could think it is a printing error |
| 10:16.38 | metacollin | xzcvczx: because I have gcc installed |
| 10:16.45 | xzcvczx | cyborg_ar: oh the irony of a blank page not being blank though :P |
| 10:16.46 | metacollin | and cmake will find it and use it over clang |
| 10:16.54 | xzcvczx | oh? |
| 10:16.56 | metacollin | you should probably add those lines to your build script =P |
| 10:16.56 | xzcvczx | fucking clang |
| 10:16.59 | xzcvczx | why do you have gcc? |
| 10:17.06 | metacollin | fortran |
| 10:17.07 | xzcvczx | metacollin: don't worry i do |
| 10:17.08 | xzcvczx | :P |
| 10:17.10 | xzcvczx | ah |
| 10:17.17 | metacollin | its not my choice |
| 10:17.19 | metacollin | *by |
| 10:17.30 | xzcvczx | what do you use fortran for? |
| 10:17.45 | xzcvczx | metacollin: is there a reason for no scripting? |
| 10:17.56 | xzcvczx | do you use wx3.1? |
| 10:18.20 | xzcvczx | metacollin: can you try a -j29 when you get bored? |
| 10:18.40 | metacollin | -j29 takes about 40% longer |
| 10:18.51 | xzcvczx | hmmm i should try -8 |
| 10:18.56 | xzcvczx | -j8 |
| 10:19.10 | metacollin | I find using the number of logical cores tends to work best |
| 10:19.14 | xzcvczx | i assume you have 14 w/ ht? |
| 10:19.25 | metacollin | I have 28 physical cores, 56 logical |
| 10:19.38 | metacollin | I have dual Xeon E5-2690v4s |
| 10:19.54 | xzcvczx | have you tried -j57? as i thought general rule of thumb was logical+1? |
| 10:20.20 | xzcvczx | aka isn't it 1 control job + local build cores? |
| 10:20.29 | xzcvczx | logical* |
| 10:20.32 | metacollin | Well, like I said earlier, I have 100% CPU utilization with all 28 cores at their max all-core turbo (3.1GHz) the entire compile |
| 10:20.39 | metacollin | so I think more jobs would just add overhead |
| 10:21.07 | metacollin | I mean I am happy, kicad took like 40 minutes to build on my computer I was using just a bit ago heh |
| 10:21.12 | xzcvczx | i guess when logical+1 was the rule people didn't have such insane numbers of cores |
| 10:22.09 | metacollin | Well, 29 jobs doesn't actually max out my cores, hyper threading or not |
| 10:22.23 | metacollin | another thing to keep in mind is those rules of thumb are really old |
| 10:22.50 | xzcvczx | i guess when logical+1 was the rule people didn't have such insane numbers of cores |
| 10:22.51 | metacollin | back when multithreading and utilization of multicore cpus was sort of crappy (at least compared to now) |
| 10:23.27 | metacollin | software has actually gotten MUCH better at using those extra cores and in a relatively short amount of time. But honestly, it probably depends on the thing being built too |
| 10:23.43 | metacollin | I know brew always builts stuff with physical cores * 2 |
| 10:24.58 | metacollin | Oh, I use fortran for BLAS stuff which I in turn use for torch |
| 10:25.04 | metacollin | er, just uh, deep learning stuff |
| 10:25.18 | metacollin | I don't really use it, I just have to compile it heh |
| 10:25.23 | metacollin | and its gfortran or bust |
| 10:25.24 | xzcvczx | oh i think opencv is almost as abd |
| 10:25.27 | xzcvczx | bad |
| 10:25.48 | xzcvczx | although opencv is now bottled at least |
| 10:25.49 | metacollin | yeah all that stuff can't get fortran's nuts out of their mouths |
| 10:25.54 | metacollin | oh wait |
| 10:26.27 | xzcvczx | metacollin: speaking of do you have any clue as to why if there is a bottled package that brew would insist on compiling shit? |
| 10:26.32 | metacollin | oh torch isn't in brew. Er, its all stuff that is under rapid development like kicad so I rebuild it often |
| 10:26.41 | xzcvczx | nick was trying to install oce and it insisted on building it |
| 10:26.55 | *** join/#kicad crazy_imp (~mj@a89-182-21-147.net-htp.de) |
| 10:26.55 | *** join/#kicad crazy_imp (~mj@unaffiliated/crazy-imp/x-8704303) |
| 10:27.10 | metacollin | xzcvczx: bottles are only available for the default build flags, so if you use any flags it will have to build from source. |
| 10:27.24 | xzcvczx | nickoe: you weren't/didn |
| 10:27.31 | xzcvczx | 't specify any flags did you |
| 10:27.37 | metacollin | oh mm |
| 10:27.39 | xzcvczx | metacollin: i don't believe he was |
| 10:28.58 | metacollin | did he have any of the dependencies already built? |
| 10:29.32 | metacollin | hmm it installs from the bottle for me |
| 10:29.46 | xzcvczx | metacollin: it did for me as well, thats why i was lost as hell |
| 10:30.01 | metacollin | oh well then its just nick's fault |
| 10:30.18 | metacollin | he fucked something up, but its ok |
| 10:30.20 | metacollin | jk |
| 10:30.25 | xzcvczx | i am more inclined to blame c4 |
| 10:30.31 | xzcvczx | as everythings always his fault |
| 10:30.34 | metacollin | lol |
| 10:30.39 | xzcvczx | and he owns the mac |
| 10:31.10 | xzcvczx | metacollin: hmmm i wonder whether the fucked prefix could affect stuff based on hard coded paths to stuff |
| 10:31.19 | *** join/#kicad tumnao (b91f64ed@gateway/web/freenode/ip.185.31.100.237) |
| 10:31.58 | metacollin | what do you mean? |
| 10:32.07 | metacollin | er what is the prefix |
| 10:32.47 | xzcvczx | its inside of ~ for nick |
| 10:34.19 | metacollin | brew is supposed to be fully portable but who knows |
| 10:34.43 | xzcvczx | oce also has some really fucked up ideas of how to do stuff |
| 10:34.56 | metacollin | I sometimes have brew randomly build things for no reason, I think it might happen when a dependency gets out of sync with the bottle or something |
| 10:35.17 | metacollin | the bottles are rebuilt by the brew built bot, which is a mac mini in a closet somewhere |
| 10:35.21 | metacollin | not joking |
| 10:35.33 | metacollin | so bottles aren't always super responsive |
| 10:35.39 | xzcvczx | my mac mini is in a closet...... it seems scared to come out of the closet |
| 10:36.07 | xzcvczx | i call it tom sometimes |
| 10:36.57 | metacollin | what does OCE do thats fucked up? Genuinely curious, I've never actually used it |
| 10:37.21 | xzcvczx | c4757p: ~7secs per hide layer |
| 10:56.03 | metacollin | brb gonna upgrade to sierra |
| 10:56.09 | metacollin | if I'm not back in an hour, i'm already dead |
| 10:56.41 | xzcvczx | you don't have another machine to talk through your experience? |
| 10:56.57 | xzcvczx | you using the GM? |
| 11:02.30 | *** join/#kicad LeoNerd (~leo@genoa.dictatorshipcake.co.uk) |
| 11:13.17 | *** join/#kicad stelarcf (~stelarcf@92.81.115.213) |
| 11:30.24 | *** join/#kicad shaun413 (uid121475@gateway/web/irccloud.com/x-kuhotfrniocebhbw) |
| 11:37.12 | nickoe | xzcvczx: does your script also package it? |
| 11:37.22 | xzcvczx | not yet |
| 11:37.41 | xzcvczx | by package i assume you mean dmg it? |
| 11:39.37 | *** join/#kicad decimad (~deci@ip5f5be505.dynamic.kabel-deutschland.de) |
| 11:39.44 | nickoe | xzcvczx: i just did brew install science/whatever/oce |
| 11:41.54 | *** join/#kicad stelarcf (~stelarcf@92.81.115.213) |
| 11:42.21 | xzcvczx | c4757p: don't upgrade xcode |
| 11:43.09 | xzcvczx | fucking apple |
| 11:43.33 | xzcvczx | if you are going to release ide+1 to os+0 then make sure ide+1 comes with os+0 sdk |
| 11:43.38 | nickoe | Why would he? |
| 11:43.58 | xzcvczx | because mac app store is now offering xcode 8 as an update |
| 11:44.45 | nickoe | Did he install it via app store? |
| 11:44.50 | xzcvczx | yes |
| 11:47.24 | nickoe | Insane. Anyway, when it queried me, I said it should not auto update. |
| 11:48.19 | xzcvczx | why is installing it via mas insane? |
| 11:48.50 | nickoe | Because it could auyoupdate crap :P |
| 11:49.16 | xzcvczx | as can the app itself |
| 11:52.28 | nickoe | I wonder if that works when not being sogned in to the app store |
| 11:53.43 | *** join/#kicad tumnao (b91f64ed@gateway/web/freenode/ip.185.31.100.237) |
| 11:55.51 | xzcvczx | nickoe: open system preferences and select the app store pane |
| 11:56.06 | xzcvczx | and see if "Install app updates" or "Install OS X updates" is selected |
| 12:01.18 | *** join/#kicad stelarcf (~stelarcf@92.81.115.213) |
| 12:03.52 | *** join/#kicad cyborg_ar (~cyborg_ar@c-98-215-226-11.hsd1.il.comcast.net) |
| 12:04.46 | xzcvczx | hahahahah "Excuse my ignorance but cat blah.zip.weird-ending > blah.zip |
| 12:04.49 | xzcvczx | " |
| 12:09.43 | decimad | drc should really work on squared distances right? And I see a few spots where the rounding is not done correctly |
| 12:11.25 | xzcvczx | metacollin: where are yuo |
| 12:11.32 | xzcvczx | decimad: congrats mi amigo |
| 12:12.47 | cyborg_ar | squared distances? |
| 12:12.58 | decimad | Thanks, I'm glad if that can help keeping off people complaining. |
| 12:13.23 | decimad | cyborg_ar: as to avoid the square roots and then calculating with floats |
| 12:13.28 | cyborg_ar | holy shit tom's patch |
| 12:13.42 | cyborg_ar | instant layer hide/show |
| 12:13.50 | cyborg_ar | wow |
| 12:13.51 | xzcvczx | cyborg_ar: not on c4's shitty mobo |
| 12:14.03 | xzcvczx | its about 7secs there |
| 12:14.09 | cyborg_ar | it's not magic |
| 12:14.31 | cyborg_ar | he is just doing a slow operation instead of a slow operation followed by a very slow operation |
| 12:14.38 | xzcvczx | oh ffs |
| 12:14.42 | xzcvczx | nope maybe not |
| 12:14.51 | xzcvczx | i might still be on steinbergers latest |
| 12:14.52 | xzcvczx | dammit |
| 12:14.57 | xzcvczx | must have pulled at just hte wrong time |
| 12:14.59 | xzcvczx | rebuild time |
| 12:15.13 | decimad | orsonmmz: I obviously forgot cleaning up the helpers/gal/ directory ... everything I did there was scrapped in the end, I'm sorry -.- |
| 12:15.25 | orsonmmz | decimad: not a big problem |
| 12:15.41 | cyborg_ar | haha |
| 12:15.53 | cyborg_ar | yeah it should actually be fast |
| 12:15.54 | orsonmmz | the overall gain is large enough to cover the mess :P |
| 12:15.57 | xzcvczx | any other osx users awake? |
| 12:16.06 | cyborg_ar | the slow operation is that it recaches everything, which seems to be cpu intensive |
| 12:16.12 | cyborg_ar | anyway it is a game changer |
| 12:16.18 | cyborg_ar | now i can actually hide layers |
| 12:16.25 | decimad | xzcvczx I'm no steinberger ;) |
| 12:16.39 | xzcvczx | steak burger then? |
| 12:16.47 | cyborg_ar | stained burger |
| 12:17.25 | *** join/#kicad sandeepkr (~sandeepkr@103.49.155.114) |
| 12:17.42 | decimad | are trace widths always a multiple of 2? |
| 12:17.44 | cyborg_ar | xzcvczx, i just love the fact that this patch was submitted a couple days after c4757p was done with his board |
| 12:17.48 | xzcvczx | poor decimad |
| 12:17.54 | cyborg_ar | decimad, they shouldnt need to |
| 12:17.57 | xzcvczx | cyborg_ar: i say that c4 deserved it |
| 12:18.15 | xzcvczx | decimad: well whats not a multiple of 2? |
| 12:18.25 | xzcvczx | unless you go down to 1nm |
| 12:19.01 | decimad | int d = KiROUND( dist ) - ((w1 + w2) / 2); |
| 12:19.28 | decimad | I'd argue that's wrong |
| 12:20.03 | decimad | w1 and w2 being integers |
| 12:20.23 | cyborg_ar | decimad, it is a known issue that the drc rounding behavior is wonky, and perhaps subtly incorrect |
| 12:20.32 | cyborg_ar | we know it disagrees with the PNS drc calculations |
| 12:20.42 | cyborg_ar | so the PNS ends up violating drc sometimes |
| 12:20.47 | xzcvczx | there is even a bug that a+b+c == limit is rejected |
| 12:21.16 | cyborg_ar | some real numerical analysis needs to be done |
| 12:22.31 | cyborg_ar | decimad, there you trade overflow for underflow |
| 12:22.34 | decimad | Well, these could be fixed... but wouldn't it make more sense to get the datastructures the same (line, line-strip) etc. so both could use the same source? |
| 12:22.45 | cyborg_ar | you could multiply dist by two, but that could cause overflow |
| 12:23.12 | decimad | cyborg_ar: yes, need to use int64 there of course |
| 12:23.16 | cyborg_ar | the coder there chose underflow to be more acceptable |
| 12:23.32 | cyborg_ar | aren't ints 64 bits? |
| 12:23.53 | xzcvczx | not on 32-bit |
| 12:24.06 | cyborg_ar | ahh |
| 12:24.10 | cyborg_ar | it is 32 bits |
| 12:24.28 | cyborg_ar | goddamn inconsistent C type system |
| 12:24.32 | xzcvczx | haha i see c4 screaming at tom for not fixing that hide/show issue a month or 2 ago |
| 12:24.34 | cyborg_ar | well, not inconsistent |
| 12:24.37 | cyborg_ar | loosely defined |
| 12:25.01 | cyborg_ar | he could have fixed it himself, it was a one line patch |
| 12:25.14 | xzcvczx | that would require competence |
| 12:25.22 | xzcvczx | and a fair idea of whether that would break shit |
| 12:25.41 | xzcvczx | i wonder why that was put in there |
| 12:26.03 | cyborg_ar | anyway that's why i almost always use stdint.h types in C |
| 12:26.18 | cyborg_ar | makes the code easier to read when doing tricky number stuff |
| 12:26.35 | decimad | I usually tend to use size_t and define headers with int32, int32, uint32 etc. |
| 12:26.46 | xzcvczx | least msvc finally got a fucking clue and added stdint.h to c |
| 12:26.59 | cyborg_ar | well, size_t is a must when dealing with posix, or memory |
| 12:27.16 | cyborg_ar | for the rest, stdint.h, no need to reinvent the wheel |
| 12:27.47 | decimad | They define it like uint32_t right? |
| 12:28.08 | cyborg_ar | size_t when you deal with sizes or pointer-like integers, stdint.h for the rest |
| 12:28.09 | xzcvczx | stdint.h yes |
| 12:28.10 | cyborg_ar | yes |
| 12:28.19 | decimad | I don't see what the _t brings to the table ;) |
| 12:28.23 | xzcvczx | type |
| 12:28.28 | cyborg_ar | consistency |
| 12:28.33 | xzcvczx | yet posix disallows anyone else to use _t |
| 12:28.43 | xzcvczx | its "reserved" |
| 12:28.51 | cyborg_ar | _t is for *standard* types |
| 12:28.54 | cyborg_ar | not your own types |
| 12:29.24 | cyborg_ar | doesnt stop me from using _t though |
| 12:29.33 | xzcvczx | nor me |
| 12:29.34 | cyborg_ar | so far the posix police hasn't caught me |
| 12:29.35 | decimad | I can see that it's super consistent and all, but I still think it would be better without |
| 12:30.08 | xzcvczx | decimad: thats because you are as delusional as mroszko and think that microsoft can do no wrong |
| 12:30.15 | cyborg_ar | why? that's how everyone writes it, and i can rely on stdint.h implementing the types somewhat correctly |
| 12:30.41 | cyborg_ar | instead of some homegrown header that might not deal with every compiler's quirks |
| 12:30.50 | decimad | because of all the _t's cluttering the source and bringing nothing to the table than being consistent of course ,) |
| 12:31.18 | decimad | using uint32 = uint32_t is an option of course |
| 12:31.49 | cyborg_ar | but uint32 looks like a function |
| 12:32.16 | decimad | It looks to be like "uint32" |
| 12:32.23 | decimad | why does it look like a function? |
| 12:33.21 | cyborg_ar | it's all lowercase |
| 12:33.25 | *** join/#kicad Shuggsy_ (~Shuggsy@2600:1005:b049:8449:bc01:5978:72f7:31d9) |
| 12:33.33 | cyborg_ar | anyway this is silly |
| 12:33.33 | xzcvczx | so is uint32_t :P |
| 12:33.51 | cyborg_ar | _ is not lowercase :P |
| 12:33.58 | cyborg_ar | i'm being silly |
| 12:34.01 | xzcvczx | decimad: if you really care #define uint32_t uint32 |
| 12:34.07 | cyborg_ar | hell no |
| 12:34.07 | xzcvczx | but you will be hated for it |
| 12:34.15 | cyborg_ar | typedef it at worst |
| 12:34.17 | decimad | I prefer the using declaration up there |
| 12:34.49 | cyborg_ar | anyway i find it silly to reject a mostly harmless convention like that |
| 12:35.20 | cyborg_ar | i'm actually kind of sure they added the _t to ensure it wouldnt collide with the mass of homegrown headers defining uint32 |
| 12:35.51 | xzcvczx | lol standards don't give a shit about what downstream shit they trod on |
| 12:35.57 | decimad | they do |
| 12:36.31 | decimad | Anyways, I'd just add a header that redefined all these... or put it into common.h or something |
| 12:38.06 | decimad | But I'm still thinking about PNS and BOARD using the same datastructures, so DRC can be shared and not use pointers to point arrays and stuff |
| 12:38.16 | orsonmmz | decimad: ++ |
| 12:38.32 | orsonmmz | geometry lib ftw |
| 12:38.40 | orsonmmz | +keeping everything in r-trees |
| 12:39.13 | decimad | Yeah, though I'd argue at least switching to boost::rtree then |
| 12:39.22 | xzcvczx | but b-streams are so much hipper |
| 12:39.36 | decimad | Or was that evaluated and didn't make it? |
| 12:39.59 | xzcvczx | woops trees not streams |
| 12:40.01 | orsonmmz | no idea, r-tree was in place when I came to the project |
| 12:40.12 | xzcvczx | and not more boost :S |
| 12:40.13 | *** join/#kicad NoFate17 (~NoFate@c49-177-221-69.eburwd6.vic.optusnet.com.au) |
| 12:41.02 | cyborg_ar | it would be nice if there was a single geometry library in all of kicad |
| 12:41.14 | cyborg_ar | a single type for a point, a single type for a vector |
| 12:41.28 | cyborg_ar | there must be like 5 different vector types |
| 12:41.54 | decimad | xzcvczx: b-tree? |
| 12:41.57 | xzcvczx | cyborg_ar: just enslave cirilo and make him write it for his freedom |
| 12:42.17 | xzcvczx | decimad: binary tree |
| 12:42.36 | decimad | xzcvczx: how does that help to sort stuff spatially? |
| 12:42.46 | cyborg_ar | use BSP |
| 12:42.50 | cyborg_ar | :P |
| 12:43.09 | decimad | I think R-Trees are perfect for what pcbnew does |
| 12:43.41 | cyborg_ar | yes |
| 12:44.08 | cyborg_ar | using the right data structure is 80% of the battle |
| 12:44.21 | cyborg_ar | having code that uses it effectively is the other 80% |
| 12:44.29 | decimad | hehe |
| 12:44.53 | xzcvczx | cyborg_ar: and the other 40%? |
| 12:45.09 | cyborg_ar | is getting the damn thing to compile on osx |
| 12:45.25 | xzcvczx | s/osx/windows/ |
| 12:46.08 | cyborg_ar | windows is always special |
| 12:47.30 | decimad | Well, I could try implementing a geometry library, but these kind of libraries usually get a flamewar starting about taste |
| 12:47.50 | xzcvczx | maybe the next update to windows will make it posix compatible |
| 12:47.51 | cyborg_ar | hmm i'm really not familiar with R-trees, could the VBOs be kept there too, so there's less gpu IO churn every frame? |
| 12:49.05 | decimad | cyborg_ar: you could store vbos for the bounding-rects, yeah... |
| 12:49.35 | cyborg_ar | because most of the board does not change |
| 12:49.54 | cyborg_ar | you could just update the vbo when the tree is updated |
| 12:50.32 | cyborg_ar | it might not make it much faster, there is caching in the current implementation |
| 12:50.58 | decimad | that adds another layer of complexity though... maybe boards are not complex enough to justify that |
| 12:56.46 | *** join/#kicad ziongate (~spritz@2001:41d0:8:935d::) |
| 12:59.13 | cyborg_ar | decimad, http://i.imgur.com/JntUx0e.jpg kicad layering |
| 13:01.11 | decimad | :) |
| 13:02.29 | *** join/#kicad Wetmelon (~wetmelon@host-73-169.ncasgro.asheville.nc.us.clients.pavlovmedia.com) |
| 13:09.16 | cyborg_ar | and yeah, there is a reason why there are 5 vector types around kicad |
| 13:09.32 | decimad | shoot |
| 13:10.44 | cyborg_ar | <decimad> these kind of libraries usually get a flamewar starting about taste |
| 13:12.54 | *** join/#kicad mozzarella (~sam@unaffiliated/sam113101) |
| 13:12.57 | cyborg_ar | is there a way in c++ to declare a class that is compatible with another one (you can pass it to a function that takes the latter as value) without linking/compiling in headers from the other object's library? |
| 13:14.05 | *** join/#kicad Thor-Arne (~email@34.79-161-66.customer.lyse.net) |
| 13:15.25 | decimad | I'd guess no. What's the use case |
| 13:15.57 | decimad | at call site you need to know the arguments memory layout/size |
| 13:17.35 | cyborg_ar | lame |
| 13:17.46 | cyborg_ar | so c++ is not object oriented yet |
| 13:18.22 | xzcvczx | cyborg_ar: class blah; |
| 13:18.29 | cyborg_ar | making a point class that can interoperate with wx, but doesnt need to compile with wx |
| 13:19.05 | decimad | well, if the memory layout is the same, you could do reinterpret_cast hacks |
| 13:19.15 | cyborg_ar | yeah |
| 13:19.34 | decimad | but strongly typed rules out doing that implicitly |
| 13:19.40 | cyborg_ar | another option is to conditionally compile casting functions if wx is present |
| 13:20.29 | cyborg_ar | it could probably just be converted in the UI code |
| 13:20.49 | cyborg_ar | it just feels wrong to have wx types all over the internal objects |
| 13:20.51 | decimad | you can however make templated code that does not rely on the specific point types being passed, if you provide member accessors for that type |
| 13:21.15 | cyborg_ar | hmm templated sounds better |
| 13:22.41 | decimad | I'd write "my" geometry library like that... Are there documents describing what the problems with boost::geometry were? |
| 13:22.55 | GyrosGeier | how about using separate point classes for UI coordinates and model coordinates? |
| 13:22.58 | cyborg_ar | it has to do with the thing before the :: |
| 13:23.28 | cyborg_ar | they like breaking shit between versions |
| 13:23.43 | cyborg_ar | i think that was the point when kicad didnt need to keep bundling boost anymore |
| 13:23.53 | cyborg_ar | when boost::geometry was removed |
| 13:25.37 | cyborg_ar | GyrosGeier, that would make sense |
| 13:27.21 | GyrosGeier | so we can't do that, naturally |
| 13:27.44 | decimad | I think there's no argue about not using wxPoint in the internals |
| 13:28.26 | cyborg_ar | that way you can also make proper python wrapper for those types |
| 13:28.34 | cyborg_ar | wrappers* |
| 13:29.15 | decimad | xzcvczx: are you guys looking for additional kicad workforce? :) |
| 13:29.21 | decimad | orsonmmz even |
| 13:29.46 | orsonmmz | decimad: sure! |
| 13:30.18 | orsonmmz | decimad: do you have any candidates? |
| 13:30.30 | decimad | :( |
| 13:31.18 | orsonmmz | em, are you talking about yourself? |
| 13:31.28 | orsonmmz | I assumed you already count as KiCad workforce :) |
| 13:32.34 | cyborg_ar | decimad, i just checked my ML archives |
| 13:32.43 | cyborg_ar | it was boost::polygon that was the sour one |
| 13:33.36 | *** join/#kicad alphaferret (~user@54034F08.catv.pool.telekom.hu) |
| 13:34.25 | decimad | orsonmmz: I rather thought about sending an application (maybe after _promising_ a huge donation) to get kicad straightened out ;) |
| 13:35.01 | decimad | cyborg_ar: Oh right, I was mixing that up |
| 13:36.05 | cyborg_ar | it would be nice if kicad applied for GSoC next year |
| 13:37.28 | cyborg_ar | there's a shortage of people deeply familiar with the codebase to act as mentors though |
| 13:37.42 | GyrosGeier | ow |
| 13:37.56 | GyrosGeier | these people are calling for a "biohacking meatup" |
| 13:38.08 | cyborg_ar | haha |
| 13:38.10 | cyborg_ar | so punny |
| 13:38.21 | GyrosGeier | yes |
| 13:38.32 | GyrosGeier | they actually want to grow meat in a petri dish |
| 13:39.12 | alphaferret | must i have an electrolytic cap in the negative feedback loop of audio op amp |
| 13:39.32 | archivist | no |
| 13:40.15 | archivist | GyrosGeier, been done I thought |
| 13:40.43 | cyborg_ar | alphaferret, you need at least 4 |
| 13:41.07 | archivist | back to back and arse about face |
| 13:41.19 | alphaferret | i'm having a low iq day be nice |
| 13:41.47 | cyborg_ar | okay |
| 13:41.59 | cyborg_ar | schematic plz |
| 13:44.01 | GyrosGeier | archivist, yes, but not in a hobbyist context yet |
| 13:47.57 | cyborg_ar | alphaferret, where do you want to add a capacitor? |
| 13:48.13 | alphaferret | at 1 and 5 at the switch |
| 13:48.27 | cyborg_ar | also do consider the topology i gave you, it gives you real continuous gain with no pops on shorting switches |
| 13:48.49 | alphaferret | ok |
| 13:49.05 | cyborg_ar | because it works like a wirewound pot |
| 13:49.20 | cyborg_ar | anyway on the cap |
| 13:49.31 | cyborg_ar | between 1 and what? |
| 13:49.36 | alphaferret | series |
| 13:49.46 | cyborg_ar | in series with the resitors? |
| 13:49.59 | alphaferret | yes |
| 13:50.07 | alphaferret | if 1 was the cap |
| 13:50.16 | cyborg_ar | that would turn your amp into a low pass |
| 13:50.24 | alphaferret | yes |
| 13:50.50 | cyborg_ar | wait no |
| 13:50.54 | alphaferret | i need a high-ish value, worst case is the 1k43 which necessitates something like 33uf |
| 13:51.15 | cyborg_ar | right there it would be a high pass, as higher frequencies see a smaller impedance there |
| 13:51.39 | cyborg_ar | i'm slightly confused, so dont mind me |
| 13:51.49 | cyborg_ar | why do you want to do that filtering? |
| 13:52.05 | alphaferret | i don't care about filtering it's to remove dc in the loop |
| 13:52.20 | cyborg_ar | what DC? |
| 13:52.40 | cyborg_ar | that leg of the circuit is just there to form a resistor divider with R9 |
| 13:52.52 | alphaferret | that's my question, because the original schematic i got had a 220uf there plus the output cap |
| 13:52.54 | cyborg_ar | it needs to be dc coupled with the output |
| 13:52.57 | alphaferret | i don't understand why it's necessary |
| 13:53.04 | cyborg_ar | link? |
| 13:54.06 | alphaferret | http://uploadpie.com/nYZ72 |
| 13:54.17 | alphaferret | that circuit is using a battery though |
| 13:54.48 | alphaferret | virtual ground rail |
| 13:55.10 | *** join/#kicad Plugh (~kcozens@CPE68b6fc4e50b3-CM68b6fc4e50b0.cpe.net.cable.rogers.com) |
| 13:58.22 | trygvis | is there a way to get the DRC to *not* fill my zones? |
| 14:02.28 | *** join/#kicad berjan (~chatzilla@2001:67c:2564:a156:e90f:d52e:c15d:17b7) |
| 14:02.30 | cyborg_ar | dang, 1 millifarad |
| 14:03.34 | alphaferret | he's a little ocd with low noise, i'm not using that guys strategy |
| 14:03.45 | alphaferret | and i'm not sure his low noise theories are even correct |
| 14:04.08 | cyborg_ar | i see no sense there |
| 14:04.27 | cyborg_ar | a 1 millifarad cap is pretty much a piece of wire at all audio frequencies |
| 14:04.33 | cyborg_ar | you might as well use a piece of wire |
| 14:04.49 | alphaferret | the docs say 220uF - 1mF |
| 14:05.04 | alphaferret | here's what someone else said if i leave out that cap... |
| 14:05.13 | alphaferret | Now the DC offset at the output will also divide. Therefore attenuating the DC feedback applied to the inverting input. Whereas with C7 you'll have full DC feedback. |
| 14:05.29 | alphaferret | C7 being the cap in question |
| 14:05.45 | cyborg_ar | hm i see |
| 14:06.03 | cyborg_ar | ahh now i see what he's doing |
| 14:06.10 | cyborg_ar | he's removing the DC offset |
| 14:06.15 | alphaferret | i think i said that |
| 14:06.52 | alphaferret | the high value is just to keep filter values out of audio |
| 14:06.52 | cyborg_ar | how high is your gain |
| 14:07.03 | alphaferret | gain of 1, 4 or 8 |
| 14:07.07 | cyborg_ar | ok |
| 14:07.18 | cyborg_ar | what is the input offset voltage? |
| 14:07.20 | alphaferret | no 2,4,8 |
| 14:07.27 | alphaferret | no idea |
| 14:07.35 | alphaferret | rails are bipolar so none? |
| 14:07.39 | alphaferret | or near zero |
| 14:07.42 | cyborg_ar | read the datasheet |
| 14:07.46 | cyborg_ar | input offset voltage |
| 14:07.59 | cyborg_ar | multiplied by gain, is your output offset voltage |
| 14:08.07 | cyborg_ar | with a pure dc feedback |
| 14:08.17 | cyborg_ar | in your case it should be just a few mV |
| 14:08.27 | cyborg_ar | so i would say dont bother with the capacitor |
| 14:08.36 | cyborg_ar | now if your gain was crazy like 1000 |
| 14:08.39 | cyborg_ar | then yes |
| 14:08.40 | alphaferret | +-3mv max |
| 14:08.57 | cyborg_ar | ok you are going to have 25mV at the output worst case |
| 14:09.01 | alphaferret | yes that's what i was thinking, since it's low gain it probably isn't necessary, but i aint smart enough to make the call |
| 14:09.02 | cyborg_ar | that's bugger all |
| 14:09.22 | cyborg_ar | if the gain was 1000 though, it would be 3 volts |
| 14:09.26 | alphaferret | plus there's the 10uf output cap |
| 14:09.27 | cyborg_ar | that would be bad |
| 14:09.30 | alphaferret | right |
| 14:09.39 | *** join/#kicad phinxy (~tehhhd@unaffiliated/phinxy) |
| 14:09.57 | cyborg_ar | yeah the only reason you would need to get rid of offset there is to avoid taking your amplifier out of the linear region |
| 14:10.30 | cyborg_ar | but 35mv in a 30v rail is not much |
| 14:10.38 | cyborg_ar | it's probably more than 35mv |
| 14:10.50 | cyborg_ar | you need to count the imbalance of offset currents |
| 14:10.51 | alphaferret | if i /only/ cared about keeping dc out of the next stage, would the 10uf be fine |
| 14:11.06 | cyborg_ar | and the fact that the impedance of the feedback changes with your gain |
| 14:11.19 | cyborg_ar | more than good enough |
| 14:11.33 | cyborg_ar | no dc is going through that cap |
| 14:11.51 | alphaferret | and the 20r following the 10u is in series with minimum 7k so i don't need a high value for the filter to be out of audio |
| 14:12.13 | alphaferret | more like 10-20k |
| 14:12.25 | cyborg_ar | yeah that's good enough |
| 14:13.41 | cyborg_ar | anyway it took me a bit to realize what was going on in the original, it was getting rid of the gain on the input offset |
| 14:14.06 | cyborg_ar | because the gain is so huge |
| 14:14.07 | alphaferret | i think you had a typo with your resistor topology, you said ...yours would be click free with mbb |
| 14:14.23 | alphaferret | right plus that one was battery powered |
| 14:14.35 | cyborg_ar | i dont think battery powered changes anything |
| 14:14.35 | alphaferret | biased at about 4.5v |
| 14:15.10 | *** join/#kicad pointhi21 (~pointhi21@dyn166055.wlan.jku.at) |
| 14:15.37 | cyborg_ar | yes i said it would be click free in make before break, some clicking with break before make |
| 14:15.47 | cyborg_ar | yours would always click a bit |
| 14:15.55 | alphaferret | oh ok, but not much |
| 14:16.41 | alphaferret | the other comment guy made about leaving out that 220u was this...You have gain select switch on the low side. You'll have awful clicks. If this was a potentiometer you'd have awful scratches. |
| 14:17.26 | alphaferret | hmm let me look at the other opamp i was considering and check dc offset |
| 14:17.27 | *** join/#kicad ricardocrudo (~ricardocr@185.19.197.52) |
| 14:18.47 | alphaferret | +-.7mv |
| 14:19.02 | alphaferret | nominal .1mv |
| 14:20.56 | *** join/#kicad niki (~niki@0x3e2c87f8.mobile.telia.dk) |
| 14:25.13 | *** join/#kicad riq_ (~riq_@c-24-6-115-254.hsd1.ca.comcast.net) |
| 14:35.17 | *** join/#kicad pointhi21 (~pointhi21@dyn166055.wlan.jku.at) |
| 14:49.08 | *** join/#kicad pazul (~pazul@ptr-2hj4tbi3h8zz81s2j7xokbjug.ip6.access.telenet.be) |
| 15:03.45 | *** join/#kicad Murr (~Murr@84-73-204-128.dclient.hispeed.ch) |
| 15:03.59 | *** join/#kicad pointhi (~pointhi@heim-032-115.raab-heim.uni-linz.ac.at) |
| 15:09.37 | *** join/#kicad pointhi1 (~pointhi@heim-032-115.raab-heim.uni-linz.ac.at) |
| 15:15.54 | *** join/#kicad ayjay_t (~ajp_@24-151-28-56.dhcp.nwtn.ct.charter.com) |
| 15:27.32 | *** join/#kicad Devil (~King@188.241.239.53) |
| 15:30.42 | *** join/#kicad NoGodDamnIdea (~King@unaffiliated/nogoddamnidea) |
| 15:40.35 | decimad | nickoe: While you can absolutely do that approximation, it has two drawbacks: 1) the approximation will have to be fixed unless you want to risk breaking existing boards on refinement. 2) doing the push operation on routing will become a nightmare |
| 15:42.04 | decimad | existing boards' DRC state I mean |
| 15:43.09 | *** join/#kicad berjan (~chatzilla@2001:67c:2564:a156:cc9c:3b14:fd97:d18f) |
| 15:44.18 | cyborg_ar | push with arcs aready sounds annoying |
| 15:44.36 | sandeepkr | What are your views on DAVE's suggestion https://www.youtube.com/watch?v=5wrSXCBdalc |
| 15:45.18 | cyborg_ar | it would be nice to treat meanders as a solid unit that cant be spread around, and arcs not being able to be shoved at all, as they would be typically critical traces you lay by hand anyway |
| 15:49.43 | c4757p | whoa, decimad patch already |
| 15:49.46 | c4757p | sweet! |
| 15:50.03 | c4757p | why do y'all have to keep making things better /after/ I spend a couple months working on a huge board |
| 15:50.08 | c4757p | that's just mean |
| 15:51.00 | decimad | we actually have second channel where we all organize our efforts to take place right after you needed them |
| 15:51.06 | c4757p | hehe |
| 15:52.14 | decimad | maybe one should drop a line at the originating kicad.info thread |
| 15:52.22 | decimad | I don't have an account tho |
| 15:52.27 | *** join/#kicad MacInLew (~textual@2001:1948:212:6010:f577:2ecf:f8f8:1ecd) |
| 15:52.29 | c4757p | I visit that site like once a month :P |
| 15:56.59 | decimad | well, the last thing in that thread is the guy proposing a 1024² map per character, I guess that should not be the last word |
| 15:57.50 | decimad | Although actually many boards would have the 6+gb vram now... it would be visually pleasing |
| 15:58.13 | Yatekii | guys, how can I add a wrl file to a lib with relative paths? |
| 15:58.22 | Yatekii | when I export the model all standard parts are included but mine |
| 15:58.26 | Yatekii | it's so annoying |
| 16:02.33 | Yatekii | is there a variable that contains the path to the current footprint I can set or anything? |
| 16:03.41 | *** join/#kicad Eli2| (~Eli2@pdpc/supporter/professional/eli2) |
| 16:21.46 | *** join/#kicad weatherhead (~somebody@vpn-public.physics.ox.ac.uk) |
| 16:23.37 | *** join/#kicad iglo (~user@dyndsl-037-138-248-091.ewe-ip-backbone.de) |
| 16:32.44 | decimad | @OSHW: Considering the gear looks amazingly like a sun as well, I'd maybe favour an encoding that ends up as a yellow/golden sun rise if the project is completely free ;) |
| 16:34.13 | c4757p | hm, I never noticed that |
| 16:36.29 | decimad | c4757p is it my drug use maybe? :P |
| 16:37.19 | c4757p | LOL |
| 16:40.58 | *** join/#kicad ricardocrudo (~ricardocr@185.19.197.52) |
| 16:42.29 | *** join/#kicad berjan (~chatzilla@ip54538619.speed.planet.nl) |
| 16:43.09 | decimad | Hrmmm, I'm wondering why chars like V are clipped when being inner characters |
| 16:44.05 | nickoe | decimad: existing boards? |
| 16:44.35 | decimad | nickoe: boards that have been designed vs. a certain implementation of approximation |
| 16:45.26 | nickoe | decimad: well, yes of course the approximation settings needs to be stored with the object explicitly for backward compatibility reasons. |
| 16:45.36 | nickoe | I don't see a problem with that. |
| 16:46.16 | decimad | I think that leaves a bit of elegance to be desired |
| 16:46.46 | decimad | ie. an implementation detail bubbles up into the file format |
| 16:56.49 | nickoe | decimad: what is the problem with that? |
| 16:57.07 | cyborg_ar | lol, for my next job i need to fit an adc, an arm micro, supporting components and connectors in two boards the size of a quarter |
| 16:57.23 | decimad | nickoe: it adds complexity that will never be able to be stripped again |
| 16:57.40 | nickoe | decimad: why should it be stripped? |
| 16:58.01 | decimad | because someone comes up with a clever exact representation or something |
| 16:58.20 | cyborg_ar | http://uploadpie.com/CJTQi |
| 16:58.33 | nickoe | well, they it should still approximate in the never version of kicad, FOR BACKWARD COMPATIBILITY. |
| 16:58.33 | cyborg_ar | imo bezier makes no sense for copper |
| 16:58.52 | decimad | see how you will have to keep that stuff in for eternity then? |
| 16:59.05 | decimad | and it will need to be touched for every refactor |
| 16:59.12 | *** join/#kicad ferdna (~ferdna@cpe-69-23-129-73.elp.res.rr.com) |
| 16:59.34 | cyborg_ar | kinda like how you have to store filled zones |
| 16:59.34 | alphaferret | cyborg_ar, i like that job description |
| 16:59.37 | decimad | so I'm playing the force on the conservative side now ;) |
| 16:59.37 | nickoe | decimad: you can simply bump the board format if you in the unlikely case would do exact DRC for example. |
| 16:59.52 | cyborg_ar | that's a huge space hog in board files |
| 17:00.24 | nickoe | cyborg_ar: what is a huge space hog? |
| 17:00.28 | cyborg_ar | but you can't avoid it much because it's very possible that zones won't be filled the same in the future |
| 17:00.34 | cyborg_ar | filled zones |
| 17:00.36 | alphaferret | there's a company who makes connectors for human nerve tissue, forgot their name |
| 17:00.49 | cyborg_ar | kicad stores every single vertex of the filled polygons |
| 17:01.10 | cyborg_ar | on complex boards it's a huge dump of (xy ) objects |
| 17:01.35 | *** join/#kicad Shuggsy (~Shuggsy@c-73-7-71-211.hsd1.ga.comcast.net) |
| 17:02.19 | cyborg_ar | it serves two functions, it accelerates file opening, and preserves the exact board that was saved in the event polygon filing behavior changes subtly in a future release |
| 17:02.22 | *** join/#kicad drod (~ldm@ip-95-221-235-169.bb.netbynet.ru) |
| 17:02.40 | cyborg_ar | the second one is a bit dubious, but when you need it you need it |
| 17:02.41 | *** join/#kicad bvernoux (~Ben@chl26-1-88-183-104-56.fbx.proxad.net) |
| 17:03.49 | nickoe | cyborg_ar: how would you store it otherwise? |
| 17:04.41 | *** join/#kicad ferdna (~ferdna@cpe-69-23-129-73.elp.res.rr.com) |
| 17:04.42 | decimad | Imho the filling actually differs there, since it can be done "exactly" if working with lines and arcs |
| 17:04.49 | *** join/#kicad C47 (~d3bug@189.242.29.24) |
| 17:04.55 | cyborg_ar | nickoe, not |
| 17:04.58 | cyborg_ar | just the edges |
| 17:05.04 | cyborg_ar | not the actual baked in filling |
| 17:05.23 | cyborg_ar | it's a huge waste of space and a source of churn for vcs |
| 17:05.39 | *** join/#kicad sftp (~sftp@unaffiliated/sftp) |
| 17:06.03 | cyborg_ar | there was some thought put into it though, it gets saved at the end of the file |
| 17:06.08 | nickoe | mmm, JP keeps reffering to bugs in another way that was discussied on the list ... c4757p |
| 17:06.09 | cyborg_ar | so the churn is not that bad |
| 17:06.46 | cyborg_ar | nickoe, the problem with JP is that he uses the -m switch to set the commit message |
| 17:06.53 | cyborg_ar | so everything is squashed in one line |
| 17:07.19 | cyborg_ar | perhaps he doesn't have an editor set up properly |
| 17:07.24 | nickoe | so he does git commit fixes -m "foo" ? |
| 17:07.33 | cyborg_ar | yes |
| 17:07.36 | nickoe | git commit fixes 4354325-m "foo" ? |
| 17:07.42 | cyborg_ar | his commit messages are always a single line |
| 17:07.47 | cyborg_ar | no |
| 17:07.55 | nickoe | Maybe the git hook could be modified to support that? |
| 17:08.01 | cyborg_ar | git commit -a -m "Fixes 32323 blablablabla" |
| 17:08.28 | cyborg_ar | i'm 89.4% sure that's the exact command he uses |
| 17:08.54 | nickoe | how are you supposed to use it normally? |
| 17:09.05 | cyborg_ar | it would be nice to set him up with a commit hook that reformats his messages |
| 17:09.23 | cyborg_ar | git add file1 file2... ; git commit |
| 17:09.32 | nickoe | but the fixes part! |
| 17:09.41 | nickoe | isn't that sort of a "command" ? |
| 17:09.45 | cyborg_ar | you type it into your text editor |
| 17:09.51 | cyborg_ar | at the end of the commit message |
| 17:09.56 | nickoe | and not part of the commit message |
| 17:10.06 | cyborg_ar | it is part of the commit message |
| 17:10.14 | cyborg_ar | when you run git commit a text editor pops up |
| 17:10.47 | *** join/#kicad OneSploit_ (~OneSploit@c-569672d5.022-26-73746f3.cust.bredbandsbolaget.se) |
| 17:10.50 | cyborg_ar | you write a summary of at most 54 characters, hit enter twice, type a detailed description if any, leave another blank line and type Fixes: 34243 |
| 17:11.01 | cyborg_ar | save and close |
| 17:11.03 | *** join/#kicad OneSploit (~OneSploit@unaffiliated/onesploit) |
| 17:11.04 | cyborg_ar | git will commit |
| 17:11.10 | OneSploit | Hi |
| 17:11.14 | cyborg_ar | that is how you work the git |
| 17:11.27 | nickoe | https://lists.launchpad.net/kicad-developers/msg25728.html |
| 17:11.44 | OneSploit | When I create a component and I set units not interchangable, why on earth erases rectangle on other units and makes it the same size in all units? |
| 17:11.54 | nickoe | cyborg_ar: I know how to work git. I don't know how to use the hook FFS |
| 17:12.18 | cyborg_ar | you put Fixes: <bug no> at the end of the commit message |
| 17:12.27 | nickoe | as you see there the bug id is not mentioned in the commit, but added with the fixes "command" |
| 17:12.30 | c4757p | I need to dig up that commit hook |
| 17:12.35 | c4757p | anyone have it handy? I couldn't find it quickly |
| 17:12.42 | c4757p | I've been typing the messages out by hand >_< |
| 17:12.45 | cyborg_ar | on it's own line |
| 17:12.48 | nickoe | c4757p: it is in the source |
| 17:13.02 | cyborg_ar | you could but it is the same from git's perspective |
| 17:13.03 | c4757p | where? |
| 17:13.12 | nickoe | c4757p: helpers/git |
| 17:13.14 | cyborg_ar | it might be an alias or something |
| 17:13.18 | c4757p | agg, thanks |
| 17:13.21 | c4757p | er |
| 17:13.23 | c4757p | ahh |
| 17:13.46 | cyborg_ar | ahh |
| 17:13.50 | cyborg_ar | it is an alias |
| 17:13.52 | cyborg_ar | not a hook |
| 17:13.55 | cyborg_ar | as i said |
| 17:14.05 | c4757p | hm, it's a little bit rough. I might tidy it up and submit an improvement |
| 17:14.05 | nickoe | terminology |
| 17:14.13 | cyborg_ar | the hook wouldnt care how you make the line |
| 17:14.29 | cyborg_ar | you can just type it manually in a text editor |
| 17:14.39 | nickoe | cyborg_ar: but can't the alias ammend the commit? |
| 17:14.56 | nickoe | cyborg_ar: we know we can write the commit message ourselves. That is not what we are talking about. |
| 17:15.06 | cyborg_ar | the way that alias works |
| 17:15.09 | nickoe | all I am talking about is using the alias |
| 17:15.14 | cyborg_ar | it loads a template when the editor pops up |
| 17:15.25 | cyborg_ar | if you use the -m switch you defeat it |
| 17:15.42 | cyborg_ar | hmm |
| 17:15.48 | cyborg_ar | now that i look at it better |
| 17:15.57 | cyborg_ar | maybe jp does use the editor |
| 17:15.58 | nickoe | c4757p: The email says git config --add include.path "$(pwd)/helpers/git/fixes_alias" |
| 17:16.09 | cyborg_ar | but he just keeps writing instead of leaving that premade line at the end |
| 17:16.24 | c4757p | nickoe: yeah, I know :P |
| 17:16.58 | cyborg_ar | nope, the hook indeed amends |
| 17:17.03 | cyborg_ar | err |
| 17:17.05 | cyborg_ar | alias |
| 17:17.08 | cyborg_ar | i got confused now |
| 17:17.10 | cyborg_ar | blergh |
| 17:17.48 | nickoe | cyborg_ar: what do you mean by keeps writing? |
| 17:17.53 | c4757p | I think I'd rather a hook, that looks for "Fixes: ...", and 1) verifies it's a valid bug number, and 2) amends to add the URL |
| 17:17.54 | cyborg_ar | nevermind |
| 17:18.10 | c4757p | I'd rather type out the "Fixes: ..." in the message, I'm used to that format |
| 17:18.14 | nickoe | is it another hook that orson has set up that actually adds the hyperlink when it sees fixes in the commit msg? |
| 17:18.22 | c4757p | ...what |
| 17:18.33 | nickoe | look at https://github.com/KiCad/kicad-source-mirror/commit/89a9b9177f72a446d7453e7ff5501bb16d10419e |
| 17:18.42 | nickoe | it has more than one line in the commit desc |
| 17:18.57 | c4757p | I have no idea. there's no hook in helpers/git |
| 17:19.01 | cyborg_ar | nickoe, it does not |
| 17:19.03 | nickoe | If we assume JP used the -m switch, he would only have made a oneline desc |
| 17:19.04 | c4757p | the alias adds both, but it puts them at the end |
| 17:19.13 | cyborg_ar | nickoe, it is one line |
| 17:19.15 | c4757p | I wonder if JP is just doing this manually or some shit |
| 17:19.19 | nickoe | c4757p: there is an alias in there. |
| 17:19.22 | c4757p | .....yes. |
| 17:19.28 | c4757p | that's why I"m talking about "the alias", yes. |
| 17:19.29 | cyborg_ar | jp is most definitely typing it out |
| 17:19.43 | cyborg_ar | he writes git commit -a -m |
| 17:19.44 | c4757p | the difference between hook and alias is important here. |
| 17:19.47 | nickoe | c4757p: why do you mention there is not hook in helpers/git? |
| 17:19.51 | c4757p | because there isn't |
| 17:19.52 | c4757p | there's an alias. |
| 17:19.58 | cyborg_ar | it's an alias not a hook |
| 17:20.03 | c4757p | <c4757p> the difference between hook and alias is important here. |
| 17:20.03 | nickoe | yeah, I know. I am talking about a hook now |
| 17:20.06 | c4757p | ...yes |
| 17:20.08 | c4757p | but there's no hook there. |
| 17:20.14 | c4757p | :P |
| 17:20.19 | nickoe | JP's commit is a multiline commit. |
| 17:20.28 | cyborg_ar | it's like the difference between a car and a food processor |
| 17:20.28 | c4757p | ¯\_(ã)_/¯ |
| 17:20.31 | nickoe | so it could not only be made with the -m swithc |
| 17:20.34 | cyborg_ar | nickoe, it is not |
| 17:20.41 | c4757p | (â¯Â°â¡Â°ï¼â¯ï¸µ â»ââ» |
| 17:20.43 | nickoe | cyborg_ar: so how is it made? |
| 17:21.10 | c4757p | ask him |
| 17:21.18 | nickoe | cyborg_ar: has a theory |
| 17:21.43 | cyborg_ar | ask him |
| 17:21.53 | cyborg_ar | i think he might answer |
| 17:21.59 | cyborg_ar | worst case he will answer in french |
| 17:22.15 | nickoe | cyborg_ar: but you said he made it with the -m switch |
| 17:22.23 | c4757p | (â¯Â°â¡Â°ï¼â¯ï¸µ â»ââ» |
| 17:22.31 | c4757p | ...I need a macro for that too |
| 17:22.36 | cyborg_ar | hahaha |
| 17:22.49 | cyborg_ar | c4757p, how do you type that out? |
| 17:22.55 | c4757p | paste |
| 17:23.41 | the_wench | ¯\_(ã)_/¯ |
| 17:23.41 | c4757p | though now I just added an alias for it, so I can just type /table :D |
| 17:23.55 | c4757p | already have /shrug :) |
| 17:24.03 | cyborg_ar | paste |
| 17:24.06 | cyborg_ar | aw |
| 17:24.23 | c4757p | eyes archivist suspiciously |
| 17:24.37 | archivist | hooomeeee, wernt me guv |
| 17:24.40 | Yatekii | can anyone tell me how i get kicad to export my 3dmodels like the standard ones?? |
| 17:24.50 | c4757p | nope |
| 17:24.51 | c4757p | doesn't know |
| 17:25.00 | Yatekii | and how i can make my paths relative? |
| 17:25.18 | nickoe | Yatekii: standard ones? |
| 17:25.51 | Yatekii | nickoe: the ones feom the github libs are exported properly and visible in the wrl |
| 17:26.16 | Yatekii | the ones i added (absolute path) are not |
| 17:26.54 | nickoe | just use ${KIPRJMOD} as your root |
| 17:27.10 | nickoe | if that is where the model is |
| 17:27.30 | *** join/#kicad Shuggsy_ (~Shuggsy@2600:1005:b003:63e1:a72c:8b08:eef4:2fd2) |
| 17:34.10 | Yatekii | nickoe: is that the location of the module(footprint)? |
| 17:34.21 | nickoe | no, the wrl |
| 17:34.36 | nickoe | Yatekii: I don't understand your problem nor what you want |
| 17:34.45 | nickoe | are you trying to export a VRML assembly? |
| 17:36.54 | *** join/#kicad kraiskil (~kraiskil@a91-152-241-34.elisa-laajakaista.fi) |
| 17:37.53 | *** join/#kicad pazul (~pazul@ptr-2hj4tbi3h8zz81s2j7xokbjug.ip6.access.telenet.be) |
| 17:39.06 | *** join/#kicad ferdna (~ferdna@cpe-69-23-129-73.elp.res.rr.com) |
| 17:43.13 | Yatekii | nickoe: yes |
| 17:43.21 | Yatekii | nickoe: ok, let me put it this way: |
| 17:43.30 | Yatekii | I have a kicad lib I have all modules gathered in |
| 17:43.36 | Yatekii | I import them into every project |
| 17:43.45 | Yatekii | everything is imported relative so it works on every mates pc |
| 17:43.57 | Yatekii | (because ofc the repo is in different places on their pc.) |
| 17:44.01 | Yatekii | now: |
| 17:44.11 | Yatekii | I add a 3d model to one of the modules(footprints) |
| 17:44.38 | Yatekii | and that path is ALWAYS absolute OR relative to the 3dmodels directory from github. |
| 17:44.55 | nickoe | file -> export -> vrml |
| 17:45.14 | Yatekii | now I want to have that model added relative to the .kicad_mod |
| 17:45.44 | Yatekii | nickoe: no the problem is: when I use that footprint in a project (pcb) and I go to the 3dviewer that model is there |
| 17:46.32 | Yatekii | if I now export that pcb to vrml (just they way you mentioned) every 3d model is there (pcb, standard footprints(from github libs); only my custom 3dmodel from my library is missing |
| 17:46.39 | Yatekii | the question is why and how I fix that |
| 17:46.45 | Yatekii | and I assume it has to do with absolute paths |
| 17:47.28 | nickoe | Yatekii: there are multiple export options, use them. |
| 17:47.40 | Yatekii | nickoe: tried em all. its missing with all of the 6 combos |
| 17:47.46 | nickoe | you can make it use relative paths and you can have it copy them to another directory |
| 17:48.41 | Yatekii | yes I know, I tried em both |
| 17:48.49 | Yatekii | and it wont copy my custom footprint |
| 17:49.15 | Yatekii | also whats more: the export problem is one but not being able to add 3d models to my libraries is way worse |
| 17:49.25 | Yatekii | like that I have to add the 3d model in every project again |
| 17:49.28 | Yatekii | which sucks |
| 17:50.21 | nickoe | Yatekii: http://storage4.static.itmages.com/i/16/0914/h_1473875464_5958057_64e0bb4124.png |
| 17:50.53 | nickoe | what do you mean by custom footprint? |
| 17:51.14 | nickoe | does it render in the 3d viewer? (your custom footprint and the 3d model) |
| 17:52.26 | Yatekii | custom footprint == I did the footprint myself and added my own 3d model |
| 17:52.41 | Yatekii | and I didn not use one from the standard library |
| 17:52.48 | Yatekii | and yes the part is visible in the 3d viever |
| 17:52.51 | Yatekii | *viewer |
| 17:53.04 | Yatekii | but it does not get exported to the shapes3D dir |
| 17:53.10 | Yatekii | which fucking drives me crazy |
| 17:53.39 | nickoe | what does your path to the 3d model look like? |
| 17:53.50 | nickoe | is there accented chars maybe? |
| 17:54.10 | Yatekii | no |
| 17:54.14 | nickoe | can you take a screenshot of the 3d properties for the footprint? |
| 17:54.18 | Yatekii | as I said it is displayed in the 3D viewer |
| 17:54.20 | nickoe | or simply copy the path here? |
| 17:54.41 | Yatekii | D:/Data/git/swarm2/mechanical/PCIex1.x3d |
| 17:54.43 | Yatekii | is the path |
| 17:54.50 | Yatekii | ()and yes iwindows annoys me even more) |
| 17:55.05 | nickoe | hmm, maybe the vrml exporter ignores x3d files? |
| 17:55.25 | nickoe | could you try to also add file with .wrl? |
| 17:55.31 | nickoe | and check if that gets exported? |
| 17:55.47 | nickoe | it could be a bug in the exporter |
| 17:56.48 | Yatekii | http://grab.by/SKle |
| 17:56.53 | Yatekii | here is the 3d view |
| 17:57.05 | nickoe | yeah, that looks good. |
| 17:57.18 | Yatekii | http://grab.by/SKlg and as you can see it wont be exported to shapes3D and everything else will be |
| 17:57.20 | nickoe | but could you try to add any wrl file to the footprint and reexport? |
| 17:57.29 | Yatekii | I odn't have the wrl file |
| 17:57.37 | Yatekii | and blender can only import wrl but not export it |
| 17:57.41 | Yatekii | so no clue how to do that |
| 17:57.46 | nickoe | Yatekii: you clearly have all those in the last screenshot |
| 17:57.56 | Yatekii | yes |
| 17:57.58 | Yatekii | as I said |
| 17:58.04 | Yatekii | but the pci-e is missing |
| 17:58.27 | nickoe | try to copy one of those, put it next to your x3d file and add that to the footprint. Also rename the wrl file to foobar.wrl |
| 17:58.48 | nickoe | then theck the shapes3D dir again |
| 18:02.08 | Yatekii | ok its with the wrl |
| 18:03.21 | *** join/#kicad ferdna (~ferdna@cpe-69-23-129-73.elp.res.rr.com) |
| 18:04.45 | decimad | I can't stand eevblog for more than 10 minutes in a row... that guy is transporting so few information in such a large amount of time keeping repeating even the most obvious things... |
| 18:05.29 | c4757p | yeeep |
| 18:05.35 | nickoe | Yatekii: what do you mean? You mean that it works if you add a wrl model? |
| 18:05.44 | Yatekii | yes |
| 18:06.04 | nickoe | Yatekii: ok, I am not sure if it is a bug or something. But please report it. |
| 18:06.37 | nickoe | I would like to ask Cirilo about it. |
| 18:17.19 | nickoe | xzcvczx: just to refresh. To get nightlies working again we need brew toolchaing that is built with 10.9 as the minver? |
| 18:18.02 | *** join/#kicad OneSploit (~OneSploit@c-569672d5.022-26-73746f3.cust.bredbandsbolaget.se) |
| 18:18.38 | OneSploit | Any help ? |
| 18:18.54 | OneSploit | how can I make a component with multiple units, but each unit has different square size? |
| 18:19.07 | OneSploit | it seems when I make square in one unit it's the same over all squares |
| 18:19.15 | OneSploit | besides I set that untis are not interchangable |
| 18:19.32 | OneSploit | pins can be added separetly, but not squares around them |
| 18:21.00 | *** join/#kicad Shuggsy (~Shuggsy@2600:1005:b003:63e1:a72c:8b08:eef4:2fd2) |
| 18:21.48 | *** join/#kicad kyranf (~kyranf@s209-121-157-74.bc.hsia.telus.net) |
| 18:21.53 | *** join/#kicad Wetmelon (~wetmelon@host-73-169.ncasgro.asheville.nc.us.clients.pavlovmedia.com) |
| 18:30.20 | *** join/#kicad ferdna (~ferdna@cpe-69-23-129-73.elp.res.rr.com) |
| 18:32.55 | nickoe | xzcvczx: But if I can build HEAD without problems, why can't Adam then? |
| 18:33.36 | nickoe | OneSploit: there is also a property for the rectangle object itself to be present in all or not |
| 18:33.53 | OneSploit | nickoe: thanks will check |
| 18:34.25 | *** join/#kicad ricardocrudo (~ricardocr@ip5f5bf593.dynamic.kabel-deutschland.de) |
| 18:34.48 | nickoe | OneSploit: it took me a while to figure out the first time too.. |
| 18:35.40 | OneSploit | found it, works as expected now, thanks a lot |
| 18:35.50 | nickoe | np |
| 18:36.11 | OneSploit | should be default behaviour :) |
| 18:40.00 | *** join/#kicad alphaferret (~user@54034F08.catv.pool.telekom.hu) |
| 18:40.30 | alphaferret | am i retarded or did they fuck up the drawings (upper left) of the orientation where the slider is...http://www.digikey.com/product-detail/en/c-k-components/AYZ0203AGRLC/401-2015-1-ND/1640124 |
| 18:42.14 | cyborg_ar | drawing? |
| 18:42.20 | cyborg_ar | those things are just illustrations |
| 18:42.27 | cyborg_ar | the datasheet has the data |
| 18:43.17 | alphaferret | same problem on datasheet |
| 18:43.29 | alphaferret | or same retarded person making the same error |
| 18:43.48 | nickoe | alphaferret: what is the problem? |
| 18:44.07 | alphaferret | ok i think i got it, i'll go by the datasheet |
| 18:44.16 | alphaferret | but the illustration on that link is misleading |
| 18:44.53 | nickoe | it looks there there is two type, one where the pin is up and where it is flat. |
| 18:45.21 | nickoe | alphaferret: you should always confirm the exact model number with the one listed in the datasheet to make sure you get the proper modle |
| 18:46.11 | nickoe | AYZ0102AGRLC is flat while AYZ0202AGRLC is not flat |
| 18:46.18 | decimad | I wonder why someone is considered retarded when he made a mistake! (not relating the current case) |
| 18:46.40 | alphaferret | the link i posted is for AYZ0203AGRLC |
| 18:46.49 | alphaferret | retarded is a colloquiallism |
| 18:47.16 | nickoe | yes, and? I just tried to illustrate my point, not to say anything on the one you selected. |
| 18:47.40 | *** join/#kicad gzamboni_ (~gzamboni@gzamboni.zael.fr) |
| 18:47.43 | nickoe | decimad: because retarded mistakes are retarded |
| 18:48.34 | decimad | Okay, I'll simply take this as given and won't ask further questions ;) |
| 18:48.41 | cyborg_ar | alphaferret, digikey usually has the wrong picture |
| 18:48.51 | cyborg_ar | i wouldnt consider it a mistake |
| 18:49.02 | cyborg_ar | it's there to give you a rough idea of what it looks like |
| 18:50.23 | alphaferret | technically retarded is the right term if you take it literally as meaning slow |
| 18:50.31 | alphaferret | if you call someone a retard it's considered an insult |
| 18:50.34 | *** join/#kicad Novus (~Novus@2605:e000:628b:1400:df2:8679:73d0:e704) |
| 18:50.36 | nickoe | we can simply conclude that alphaferret is retarted then |
| 18:50.37 | nickoe | <PROTECTED> |
| 18:51.00 | alphaferret | no offense taken |
| 18:51.53 | alphaferret | i always considered the currency of getting free advice on irc was taking shit from strangers, with this channel being one of the few exceptions |
| 18:52.09 | decimad | I wonder if there's been arguing when the flight computer told the pilot "retard! retard!" |
| 18:53.36 | nickoe | hehe |
| 18:53.52 | alphaferret | in some circles it could even be a compliment, like if you were a student of tai chi where you have to practice everything really slow, someone could say oh that guy was super retarded today |
| 18:54.34 | alphaferret | or in mime school |
| 18:54.51 | nickoe | do you have experince with tai chi? |
| 18:54.58 | alphaferret | a littlel chi qong |
| 18:55.05 | alphaferret | little |
| 18:59.09 | cyborg_ar | airplanes yell retard if you fly them wrong |
| 19:00.18 | cyborg_ar | https://youtu.be/9svSePWwisQ?t=24s |
| 19:01.29 | alphaferret | classical music passages have retard |
| 19:02.14 | alphaferret | i think if you put dough in refrigerator when raising it it's called retarding |
| 19:02.48 | cyborg_ar | also when tuning a car you either advance or retard the timing |
| 19:03.17 | alphaferret | context rules once again |
| 19:04.36 | *** join/#kicad oofus (~chris@oofus1-adsl.demon.co.uk) |
| 19:07.01 | c4757p | this channel has the strangest conversations sometimes |
| 19:08.26 | cyborg_ar | haha |
| 19:12.15 | decimad | :) |
| 19:16.34 | *** join/#kicad berjan (~chatzilla@2001:67c:2564:a156:2432:e961:735a:5724) |
| 19:17.16 | *** join/#kicad m_w (~Michael@ip68-14-170-91.ok.ok.cox.net) |
| 19:28.06 | nickoe | cyborg_ar: this is also a retarded landing https://www.youtube.com/watch?v=tGCoyRxXcOk |
| 19:44.05 | *** join/#kicad OneSploit_ (~OneSploit@c-569672d5.022-26-73746f3.cust.bredbandsbolaget.se) |
| 19:51.52 | *** join/#kicad rdmitry (~Dmitry@95-37-249-206.dynamic.mts-nn.ru) |
| 19:57.23 | decimad | would anyone be totally opposed to adding the header files to the cmake projects as well btw? |
| 20:00.20 | *** join/#kicad ridesa7__ (~quassel@rrcs-98-101-140-98.midsouth.biz.rr.com) |
| 20:00.24 | nickoe | decimad: to the cmake projects? |
| 20:01.01 | decimad | add_executable( myexe some.c some.h ) |
| 20:01.32 | decimad | this does not bring anything to the table build-wise, but the project generator then knows the headers and which headers belong to which project semantically |
| 20:01.35 | *** join/#kicad kyranf (~kyranf@s209-121-157-74.bc.hsia.telus.net) |
| 20:02.45 | decimad | which in turn lets an ide know in which contexts and with which build switches a header is parsed |
| 20:02.48 | nickoe | hmm, now the footprint editor does not init with the origin in the center... That must be from orsonmmz|away's pathces |
| 20:03.49 | nickoe | decimad: What problem will this solve? |
| 20:04.28 | decimad | I better not tell you :P It's totally important, but I cannot disclose ;) |
| 20:09.32 | kakimir | is there such things in kicad I could filter a group of components with regex? |
| 20:09.48 | kakimir | then manipulate them all with something |
| 20:10.22 | kakimir | and is there any sense in what I'm talking about |
| 20:11.43 | decimad | no, no and yes in that order |
| 20:12.05 | kakimir | is it a good idea? |
| 20:12.15 | decimad | but if you have the regexp handy, you can do it in the file by hand |
| 20:12.23 | kakimir | oh |
| 20:12.44 | decimad | it's not a convenient option, but currently the only |
| 20:12.51 | kakimir | it would be okay |
| 20:13.03 | kakimir | if there was some like.. tools for it |
| 20:13.18 | decimad | There are plenty editors that support regexp match and replace |
| 20:14.11 | kakimir | it's not like first thing you demand from your editor but I think I take a look> |
| 20:14.34 | kakimir | I wonder if any of editors I use have it |
| 20:15.13 | decimad | I think I don't have to tell you it's wisest to make a backup copy before starting to work ;) |
| 20:15.47 | kakimir | why isn't that thing taken care of by kicad |
| 20:16.01 | kakimir | like at least making 2 copies of stuff |
| 20:16.22 | decimad | I mean before you start to hand-edit the file |
| 20:16.27 | kakimir | yes |
| 20:16.30 | decimad | kicad has autosave internally |
| 20:16.41 | decimad | at least pcbnew |
| 20:16.51 | kakimir | but does it have multiple save files/locations then? |
| 20:17.02 | decimad | the autosave is a different file |
| 20:17.06 | kakimir | nice |
| 20:17.20 | kakimir | good to know stuff |
| 20:19.36 | GyrosGeier | lol |
| 20:19.49 | GyrosGeier | boost regulator in a 6 pin BGA package |
| 20:20.01 | GyrosGeier | next to it an inductor that is ten times larger |
| 20:20.02 | *** join/#kicad maiden (~maiden@c83-255-72-208.bredband.comhem.se) |
| 20:20.27 | decimad | can't you like integrate the ic into the inductor? |
| 20:20.39 | *** join/#kicad k\o\w (~fffffff@CPE3085a93a4319-CM00fc8db094f0.cpe.net.cable.rogers.com) |
| 20:21.08 | decimad | iiic ... inductor integrated integrated circuit |
| 20:21.49 | decimad | in case of emergency it can exit like a railgun |
| 20:22.53 | c4757p | yes, you can |
| 20:22.55 | c4757p | http://www.digikey.com/product-detail/en/murata-electronics-north-america/LXDC2HL12A-050/490-5789-1-ND/2783640 |
| 20:23.04 | c4757p | whole thing's a ferrite block :) |
| 20:23.30 | decimad | how much is such a thing? |
| 20:23.36 | c4757p | isn't the price on the page |
| 20:23.38 | c4757p | ? |
| 20:23.48 | c4757p | $1 one-off, about $0.80 in quantity? |
| 20:23.54 | decimad | oh right |
| 20:23.58 | decimad | sorry |
| 20:24.06 | bvernoux | does anyone know a way to export Pick&Place information from kicad ? |
| 20:24.11 | bvernoux | XYRS file |
| 20:24.25 | bvernoux | like the one used by pcb.ng |
| 20:24.27 | c4757p | File -> Fabrication Outputs |
| 20:24.38 | bvernoux | ha it is supported ? |
| 20:25.46 | bvernoux | and ? |
| 20:25.58 | bvernoux | Generate Component Position files ? |
| 20:26.12 | c4757p | could be |
| 20:26.30 | Plugh | Has anyone here ever converted OrCAD schematics and PCB layouts to KiCad without having to start from scratch? |
| 20:27.19 | bvernoux | hmm it is not XYRS format |
| 20:27.37 | c4757p | oops, sorry. I thought it could do XYRS - I remember the macrofab dude talking about that |
| 20:28.38 | bvernoux | and XYRS required pos in mil :( |
| 20:29.18 | bvernoux | format is defined here |
| 20:29.19 | bvernoux | https://pcbng.freshdesk.com/support/solutions/articles/9000054725-optional-files-pick-and-place |
| 20:29.26 | bvernoux | it will be great to support this file format |
| 20:34.51 | decimad | they say they like metric, but mil is supported as well... so I guess it also supports metric units but they don't show how... |
| 20:38.40 | nickoe | are the pos files no good? |
| 20:39.49 | bvernoux | no the pos files does not respect pcb.ng file format |
| 20:39.54 | bvernoux | main problem is unit |
| 20:40.14 | bvernoux | and also the fact pcb.ng requires csv and special names ... |
| 20:44.36 | *** join/#kicad alphaferret (~user@54034F08.catv.pool.telekom.hu) |
| 20:45.12 | alphaferret | mothah f##@ all i gots to do is re-measure all the header connections and it's time to send gerbers |
| 20:45.33 | alphaferret | of course i said that a week ago |
| 20:46.44 | decimad | c4757p how do you tell the digikey filter that the input range should include 12V ? |
| 20:47.00 | c4757p | swearing a bunch, then selecting every input voltage range that includes 12V |
| 20:47.06 | decimad | ... |
| 20:47.23 | decimad | nicely put :) |
| 20:48.23 | *** join/#kicad cirilo (~cirilo@CPE-120-144-53-157.lnse5.win.bigpond.net.au) |
| 20:48.31 | alphaferret | does it matter where you put a phase invert switch on balanced signal... pre input jack, or anywhere before it hits the in-amp |
| 20:52.59 | nickoe | cirilo: hi |
| 20:53.15 | cirilo | nickoe, hi :) |
| 20:53.20 | nickoe | do you know if x3d should be exported when using the VRML export option? It seems that it is not. |
| 20:58.13 | cirilo | nickoe, at the moment no - I need to discuss the PCB VRML model with Mario. If the 3DViewer can create a scenegraph of the PCB then any model (x3d, IDF, STEP, IGES) will be in the exported board |
| 20:58.45 | cirilo | + the board can be exported with all the features shown by the viewer |
| 20:59.52 | cirilo | and we can remove the "Export VRML" item from the PCB menu and put it in th 3DViewer where it makes more sense |
| 21:01.23 | Zhenn | hey again cirilo, c4757p. multiplexing led displays will make them appear dimmer, yeah? what about nixie tubes? |
| 21:02.46 | cirilo | same with gas tubes - they will appear dimmer because they are being switched on and off so the average number of photons coming out is lower |
| 21:03.49 | Zhenn | makes total sense. thanks again. =) |
| 21:05.03 | cirilo | but unless you like the orange glow, why use nixie - I'd use VFD because the technology is still mass-manufactured |
| 21:07.01 | *** join/#kicad pointhi (~pointhi@heim-032-115.raab-heim.uni-linz.ac.at) |
| 21:07.20 | Zhenn | because I didn't know they existed until now! |
| 21:07.32 | Zhenn | (vfd) |
| 21:07.36 | cirilo | + VFDs are much brighter than nixies |
| 21:07.55 | cirilo | heh - nixie is old tech - they'd been around before the second world war |
| 21:08.23 | cirilo | ah, sorry - wiki sez 1955 |
| 21:08.35 | nickoe | cirilo: I think it makes sense to have the export option in pcbnew as well if you decide to add it to the 3d viewer. |
| 21:10.24 | nickoe | cirilo: But then the export is still vrml? |
| 21:10.38 | cirilo | nickoe, it might not be possible to export the same board from pcbnew - but we'll see |
| 21:11.02 | cirilo | yes, the export is VRML but all models are self-contained |
| 21:11.04 | nickoe | Yatekii: so, you need to convert your x3d file to wrl for now. I have used blender for it before. |
| 21:11.29 | nickoe | cirilo: what do you mean with it not being possible? |
| 21:11.53 | Plugh | I just ran across a site SnapEDA that says they have a large collection of components and and footprints that can be used with several different capture/pcb programs. Might be helpful in avoiding having to redo a lot of parts when starting on new projects. |
| 21:12.21 | cirilo | at the moment (and historically) Export VRML and 3DViewer show very different things because Export doesn't have access to the data structures of 3DViewer |
| 21:13.08 | cirilo | at the moment, because of the 3D plugin, Export at least can access all the component model data |
| 21:13.28 | cirilo | (but it doesn't since the Export code has not been touched) |
| 21:14.41 | cirilo | so if I rewrite the Export code I can use the old code which creates its own model of the board + markings and is different from what 3DViewer shows - or if we reimplement Export in 3DViewer we have the possibility of having exactly the same data in the export |
| 21:16.15 | cirilo | the scenegraph library can convert x3d to VRML; there is just no current method for accessing the function in kicad |
| 21:16.29 | cirilo | except to build with special flags that enable all that spam. :) |
| 21:16.29 | nickoe | cirilo: ahh, ok, I see. I forgot that it was "and excercise" for you before the scenegraph stuff :) |
| 21:16.52 | nickoe | *an excercise |
| 21:17.22 | cirilo | I was bored and so many people were complaining about the VRML export being broken - but when was that - 2012? |
| 21:17.28 | cirilo | or maybe even 2011? |
| 21:18.01 | nickoe | :) |
| 21:18.58 | xzcvczx | nickoe: because you are building only for 10.11 whereas adam has to pander to the public and support 10.9 |
| 21:19.27 | xzcvczx | and he can build fine its just run-time which is fragged |
| 21:20.47 | xzcvczx | c4757p: woooooo i won the ppa version numbering fight :) |
| 21:20.56 | nickoe | xzcvczx: ? |
| 21:21.01 | nickoe | where is the fight? |
| 21:21.06 | nickoe | or was |
| 21:21.12 | xzcvczx | https://bugs.launchpad.net/kicad/+bug/1623364 |
| 21:21.50 | cirilo | you won the internetz? |
| 21:21.58 | xzcvczx | indeed |
| 21:22.10 | xzcvczx | it was a hard fought battle |
| 21:22.18 | xzcvczx | with manu casualties |
| 21:22.59 | c4757p | high-fives xzcvczx |
| 21:23.11 | xzcvczx | c4757p: what do you think of new layer visibility switching speed on your mobo :P |
| 21:23.29 | c4757p | almost instantaneous :) |
| 21:23.54 | xzcvczx | you aren' |
| 21:24.04 | xzcvczx | t annoyed that it was so slow while you were doing board :( |
| 21:24.39 | xzcvczx | i want fireworks dammit |
| 21:24.40 | nickoe | so does he name it like https://lists.launchpad.net/kicad-developers/msg23019.html ? |
| 21:25.24 | xzcvczx | no he will probably name it 20160915-git# or 20160915-git |
| 21:25.26 | nickoe | ok |
| 21:25.27 | xzcvczx | or something |
| 21:26.51 | xzcvczx | for some reason i missed that discussion when it was made back in jan, but the point is academic anyway as a nightly is theoretically never older than a stable, and therefore trying to say you want nightly until next stable just don't work |
| 21:29.05 | nickoe | yeah, I do agree. |
| 21:29.16 | nickoe | anything is better thatn 4.1.0 |
| 21:29.40 | cirilo | this seems to be news in the english world: "political donations" buy corporate influence |
| 21:30.04 | cirilo | surely not - corporations only donate because they love democracy so much |
| 21:30.26 | nickoe | What are you reffering to? |
| 21:30.35 | Zhenn | anyone know of a good online vendor of pcb blanks suitable for home etching? the offerings on amazon look pretty weak. |
| 21:32.04 | cirilo | Wisconsin "John Doe Investigation" |
| 21:33.59 | trygvis | Zhenn: I usually look at ebay and/or aliexpress for that kind of stuff |
| 21:34.33 | trygvis | but I found out that PCBs in my local shop was only 10% more expensive (usually it is more than 3 times the pris) |
| 21:35.01 | Zhenn | kk, thanks trygvis. |
| 21:36.37 | decimad | do you experience cursor-dancing too if you have track snapping and grid snapping enabled at the same time hovering over a 45° track? |
| 21:38.48 | trygvis | \o\ /o/ \o\ |
| 21:39.23 | decimad | Is there something to celebrate? |
| 21:40.06 | trygvis | yes, my cmake stuff for generating gerber files is working. yay |
| 21:40.16 | decimad | congrats! |
| 21:40.33 | trygvis | `make` and I have production ready zip files with gerbers |
| 21:40.46 | trygvis | having this upstream would be nice! :) |
| 21:40.50 | *** join/#kicad berjan_ (~chatzilla@2001:67c:2564:a156:39b8:b3a0:9528:dc4b) |
| 21:44.46 | Yatekii | nickoe: blender can only convert in the other direction sadly |
| 21:44.57 | nickoe | Yatekii: no, you can install some plugins |
| 21:45.21 | *** join/#kicad bombledmonk (~bombledmo@64.77.213.245) |
| 21:45.23 | Yatekii | nickoe: oh that might work :) |
| 21:45.46 | Yatekii | nickoe: still, as long as relative paths dont work, its useless :) |
| 21:46.00 | nickoe | they work |
| 21:55.22 | cirilo | nickoe, Yatekii, I wrote to Mario to see what he thinks. If he is too busy or it is too much hassle to export from the 3DViewer I will rewrite the Export VRML code. |
| 21:57.42 | nickoe | cirilo: hmm, ok, cool. :) |
| 21:58.56 | *** join/#kicad Shuggsy (~Shuggsy@c-73-7-71-211.hsd1.ga.comcast.net) |
| 22:08.47 | Yatekii | cirilo: the main problem imo isn't to export it (well that would be cool too) but rather that a model cannot be saved relative to the kicad_mod but only to the kicad project or absolute |
| 22:08.56 | Yatekii | which sucks hell if you want to build up your own lib |
| 22:09.21 | Yatekii | xzcvczx: the buildscript is actually broken for me. when I build it telsss me nothing to be done |
| 22:09.32 | Yatekii | when I do rebuild it tells me glm and swig is missing |
| 22:09.38 | Yatekii | I install those two and it builds fine |
| 22:09.40 | Yatekii | very weird |
| 22:09.50 | xzcvczx | Yatekii: you haven't pulled from my git have you? you still have an old ass script? |
| 22:10.22 | Yatekii | xzcvczx: you sent it to me like < 2M ago |
| 22:10.36 | cirilo | Yatekii, the problem with the current system is that it has links in the VRML file. Any rewrite will remove all that and you will have a single exported file |
| 22:11.02 | xzcvczx | Yatekii: https://github.com/xzcvczx/kicad-app |
| 22:11.16 | Yatekii | cirilo: yes that is nice but then you still have the relative problem. |
| 22:11.23 | nickoe | Yatekii: so you are talking aout teh export kicad_mod option not taking the 3d models with it? |
| 22:11.33 | Yatekii | cirilo: its a problem within kicad not with interaction with any other programs (export) |
| 22:11.49 | cirilo | links were used historically because the model processing code was too awful to provide useful information, but the new scenegraph code can build models in a sensible way |
| 22:11.57 | Yatekii | nickoe: right. the 3d models are ALWAYS within the project. they cannot be within the lib |
| 22:12.06 | cirilo | Yatekii, what do you mean the relative problem? |
| 22:12.07 | nickoe | Yatekii: of course they can |
| 22:12.14 | *** join/#kicad eszett (~Eszett@dslb-084-059-074-203.084.059.pools.vodafone-ip.de) |
| 22:12.34 | nickoe | what is your file tree, and how would you like it? |
| 22:12.40 | Yatekii | nickoe: no they cant. because its only possible to set an absolute path or the standard 3d models folder of the kicad-github libs |
| 22:12.48 | Yatekii | so it works on ONE SINGLE AMCHINE |
| 22:12.55 | eszett | Hi! I have a question. are the powerflags that important? I replaced them by global labels, which is pretty much the same isnt it? |
| 22:13.02 | Yatekii | but as soon as you have it on ANOTHE RMACHINE it will break |
| 22:13.17 | Yatekii | because the path is broken (asolute is very seldom the same on two machines |
| 22:13.20 | cirilo | Yatekii, that's not true with the new plugin system - or even the older system. You can have any number of aliased paths |
| 22:13.21 | Yatekii | ) |
| 22:13.23 | nickoe | Yatekii: what kicad version are you using? |
| 22:13.35 | Yatekii | nickoe: 4.0.4 in the lab |
| 22:13.40 | cirilo | it's only the ancient 2012 'stable' which has that problem |
| 22:13.41 | nickoe | or you can use the KIPRJMOD variable |
| 22:13.49 | Yatekii | nickoe: no I cannot |
| 22:13.53 | nickoe | why? |
| 22:13.55 | Yatekii | that points to the kicad PROJECT |
| 22:14.03 | Yatekii | not the FOOTPRINT/LIBRARY |
| 22:14.13 | Yatekii | which is useless if you want a lib taht works with ANY poject |
| 22:14.44 | cirilo | you can define any path alias you want- ${PATH_A}, ${PATH_B} etc - then your models can be somewhere like ${PATH_A}/subdir_a/subdir_b/mymodel.x3d |
| 22:15.23 | cirilo | So in pcbnew (or footprint editor), select "Configure Paths" and define those paths |
| 22:15.24 | nickoe | you have not told me where you keep the fp lib |
| 22:15.24 | nickoe | in relation to the project I assume it is inside the project dir |
| 22:16.16 | Yatekii | cirilo: yeah and that way I have to define that path for the project eveytime etc ... it would be need if a model can be properly stored relative to the footpirnt/module (no idea why you call it module since it leaves SO MUCH SPACE for confusion) |
| 22:16.18 | cirilo | fp lib only tells you about kicad_mod files though; logically it has nothing to do with the 3D models |
| 22:16.46 | Yatekii | why would you store it with weird env vars/relative to project if it belongs to that footprint and but to that footprint? |
| 22:17.18 | Yatekii | nickoe: no its outside the project dir |
| 22:17.30 | cirilo | It makes no sense to mix 3D models and the footprints, especially for me since I don't care about any types of model except STEP, IGES, and IDF |
| 22:17.43 | Yatekii | nickoe: oh well with going back that path it could work. I still think its ugly |
| 22:18.56 | cirilo | Yatekii, if you want, just have another model tree which has a similar structure to the footprint tree |
| 22:18.59 | Yatekii | cirilo: so you keep fps and the project separate but the models belonging ONLY TO THE FOOTPRINTS within the project/relative to the project? |
| 22:19.06 | Yatekii | WTF that makes absolute zero sense, sorry |
| 22:19.19 | xzcvczx | cirilo: unfortunately i am forced to *shudder* agree with the australian...... down with wrl |
| 22:19.20 | Yatekii | cirilo: that would be okay to me |
| 22:19.28 | cirilo | well, to me it really makes no sense to mix this unrelated data |
| 22:19.37 | Yatekii | who's australian? |
| 22:20.12 | Yatekii | cirilo: now you mix models wit hthe project that don't belong to the project but not with the thing they belong to (the footprints) so why do you say it makes no sense and then still do it? :D |
| 22:20.53 | cirilo | it doesn't matter if you created 3D models for a specific footprint - in general you do not have a 1:1 mapping and there can be many different types of model data - so it is better that model data be kept separate |
| 22:20.55 | xzcvczx | Yatekii: don't ask.... it makes it too easy to mercilessly mock him |
| 22:20.56 | decimad | Also is there any reason why grid snapping is disabled during drag routing? |
| 22:20.56 | xzcvczx | :P |
| 22:21.12 | xzcvczx | decimad: because it would make pns a bitch? |
| 22:21.31 | cirilo | in fact personally I would separate different types of models into their own tree - VRML1/2 + X3D can go in 1 tree since they are all related, and STEP + IGES +IDF can go in another tree |
| 22:21.41 | decimad | xzcvczx why? |
| 22:22.00 | Yatekii | cirilo: yeah but there is no possibility to properly keep it separate. only within the project ;) |
| 22:22.13 | Yatekii | or within ONE SPECIFIC folder (the standard kicad one) |
| 22:22.15 | xzcvczx | decimad: because if your mins were near edge of grid it would never be tightly packed? |
| 22:22.25 | cirilo | Yatekii, I really don't know what you mean - it's trivial to keep it separate |
| 22:23.10 | Yatekii | cirilo: nope. you can only add either: |
| 22:23.13 | Yatekii | an absolute path |
| 22:23.19 | cirilo | if you share files then on the other computer you need to install those files and specify the value for the path aliases |
| 22:23.20 | Yatekii | a path relative to the project |
| 22:23.32 | Yatekii | a path relative to the kicad standard lib 3d models |
| 22:23.40 | cirilo | I wrote the code and I know that's not true - you can put models anywhere you want |
| 22:23.52 | decimad | mins? edge of grid = grid line? |
| 22:23.56 | Yatekii | yeah and how do you put a path relative to tthat? |
| 22:24.00 | Yatekii | I tried it for 2h today. |
| 22:24.01 | Yatekii | doesnt work |
| 22:24.06 | cirilo | even the earlier code which I didn't write allows you to do that - it's just not documented well |
| 22:24.07 | Yatekii | it HAS to be absolute |
| 22:24.23 | cirilo | like I said it's only really ancient code (~2012) which can't do that |
| 22:24.40 | Yatekii | <cirilo> if you share files then on the other computer you need to install those files and specify the value for the path aliases |
| 22:24.48 | cirilo | no, I already told you that you can give names like ${PATH_A}/subdir_1/mymodel.wrl |
| 22:24.49 | Yatekii | thats what I am saying all the time and its bad IMO |
| 22:25.15 | Yatekii | its just an opinion |
| 22:25.28 | decimad | xzcvczx: as usual one could still use shift to override snapping for using drag as a packer-replacemnt |
| 22:25.38 | cirilo | well the only option to that is to put *everything* in the project directory when you share |
| 22:25.40 | Yatekii | it needs 3 metric shittons of proper configuration for absolutely zero use in my opinion :) |
| 22:25.52 | Yatekii | yeah why would I do that? |
| 22:26.03 | xzcvczx | decimad: that would drive me nuts |
| 22:26.09 | decimad | xzcvczx: Also I'm thinking only the segment that is dragged needs snapping |
| 22:26.11 | xzcvczx | decimad: shift to enable grid snapping maybe |
| 22:26.29 | cirilo | you seem to assume that people will use the kicad library project's kicad_mod files - I never do and in fact many professionals never would |
| 22:26.44 | xzcvczx | ^^ |
| 22:27.04 | Yatekii | i dont understand that sentence. |
| 22:27.16 | xzcvczx | i really wish only device lib was included by default |
| 22:27.18 | Yatekii | but it makes ZERO sense to redo every footprint for every project xD |
| 22:27.23 | cirilo | so for me and many other devs, putting 3D models into the kicad_mod directories or even specifying paths relative to that is senseless |
| 22:27.40 | Yatekii | dunno why you think so. |
| 22:27.57 | Yatekii | ALL of my mates at the lab absolutely hat eit |
| 22:28.45 | cirilo | well, if you can propose a more sensible system I'm sure people would accept it |
| 22:29.28 | Yatekii | sensible? |
| 22:29.33 | decimad | xzcvczx: do you work with a grid at all btw? |
| 22:29.42 | xzcvczx | yes |
| 22:29.59 | cirilo | but if people don't even want to configure 1 or 2 path aliases to get their system set up I don't think any programmers can help there |
| 22:30.07 | Yatekii | cirilo: you could just give the option and all would be fine ;) I would even code it myself (not sure if my code is good enough in your oppinion but yeah`^) |
| 22:30.08 | decimad | what do you use it for? |
| 22:30.19 | xzcvczx | component and for some layout areas |
| 22:30.32 | xzcvczx | but when i am routing a bus or something often i want the crap closer together |
| 22:30.38 | xzcvczx | even more so with differential pair |
| 22:31.25 | Yatekii | cirilo: lol "cant even" its just unnecessary and unneeded. and path variables is a thing I dont like. exspecially when its just plain text input and then no real verbosity wher ethe path actually points to to find the error when you misstyped ... |
| 22:31.30 | cirilo | well, you can try - but I have no idea how you determine a path relative to a kicad_mod file |
| 22:32.30 | cirilo | Yatekii, what your asking for is very specific to the way you want to work - if other people believe it's a good way to work then you can do it |
| 22:32.31 | decimad | xzcvczx: I understand that use case... although apart from any grid stuff I think using dragging to do that is somewhat strange (and I know there is no other way currently) |
| 22:32.40 | Yatekii | well the route rhas to load the footprint somewhere. so you know its location. from there you relatively march to the model ... easy as that |
| 22:33.03 | cirilo | otherwise you might have to do like some devs and have a private set of patches to keep kicad working the way you want |
| 22:33.23 | Yatekii | cirilo: I see it the other way around. I think what you like is the minority ;) see, and suddenly I know maybe mine is not the thing everybody wants ;) |
| 22:33.28 | cirilo | well, go write the code and propose a patch |
| 22:33.41 | Yatekii | I just want to be able to do a 100% standalone lib which atm is NOT possible.y |
| 22:33.55 | xzcvczx | cirilo: lol btw there was some talk of 2d kernels in here yesterday :P |
| 22:33.58 | Yatekii | you always need thing set up specificially set up in your project |
| 22:34.23 | *** join/#kicad Peetz0r (~peter@haas-en-berg.nl) |
| 22:34.42 | cirilo | no - and the reason that's not possible now is that kicad has never forced a rule for directory structure and forced all conceptually different data to be tied together |
| 22:34.47 | xzcvczx | sorta wishes that archive project would embed the models in the .kicad_mod files |
| 22:34.51 | decimad | xzcvczx: maybe it would be cool to have some tool where you select one segment and then a second one and everything (that is intersecting the selection line) is packed against the first segment |
| 22:35.02 | cirilo | xzcvczx, oh, did I miss all the fun. :( |
| 22:35.17 | xzcvczx | cirilo: you did indeed, that will teach you for not having a bouncer or simialr :P |
| 22:35.31 | xzcvczx | we all stared at you |
| 22:35.38 | xzcvczx | but you didn't even spare us a glance |
| 22:35.44 | cirilo | meh - little enough time as is, and you know I'm more of a Luddite than Ludd himself |
| 22:36.01 | xzcvczx | decimad: gah i am not a fan of having to select shit in a certain order, makes it a pain to remember which order that is |
| 22:36.53 | decimad | that you get used to... the tablet scheme would be to hold a segment tight with one finger and then swap the other one pushing... |
| 22:37.17 | xzcvczx | lol |
| 22:38.09 | decimad | the way it currently is you need to lock the edge you push against as to not move it |
| 22:38.35 | cirilo | 'k - switching to the other computer :O |
| 22:38.38 | Yatekii | cirilo: yeah I feel the oposite ;) like that I am forced to do it your way ;) instead of choosing :) |
| 22:38.55 | cirilo | Yatekii, it's not my way, it's the historical KiCad way |
| 22:38.57 | xzcvczx | cirilo: yeah ya see humans would want an intermediatary so they don't need to disconnect to do something so simple |
| 22:39.34 | cirilo | and although everyone complains about it, incuding all devs, no one has yet proposed a scheme which people can agree on |
| 22:39.37 | Yatekii | cirilo: doesn't matter. its an enforced way I don't like. just saying because you are talking about how free of choice I am right now ;) |
| 22:40.12 | cirilo | xzcvczx, who said I'm human? :P |
| 22:41.21 | Yatekii | xzcvczx: what about the git repo of your build script you talked about? |
| 22:41.49 | *** join/#kicad bugzc (~1@unaffiliated/bugzc) |
| 22:43.03 | Yatekii | btw xzcvczx it built just fine and VERY fast :D (my laptop had like 45 minutes and my pc just had like < 10mins :D |
| 22:44.08 | decimad | I'm happy that I see so often how fast it builds |
| 22:45.45 | Yatekii | how do you mean that?^^ |
| 22:46.00 | Yatekii | well now I can finally code the things I would like to have yay! |
| 22:46.07 | Yatekii | finally a shortcut to change the grid |
| 22:46.16 | Yatekii | and a nice footprint preview :D |
| 22:46.20 | Yatekii | awesome! |
| 22:46.49 | decimad | I always get complete rebuilds when I prepare for pushing my changes since I need to apply a patchset locally to build with msvc |
| 22:48.40 | Yatekii | ahh kk fun micro$oft world |
| 22:49.01 | decimad | no it's actually code I need to touch that is incorrect but doesn't trigger diagnostics with gcc |
| 22:50.38 | Yatekii | so why dont you fix it once and for all? |
| 22:50.41 | decimad | I send at least two patches to the list, but I think I would have to explain in detail why the current code is incorrect, since it runs well with gcc, and then I'd have to mention msvc and then hell would break loose |
| 22:51.42 | *** join/#kicad Plugh (~kcozens@CPE68b6fc4e50b3-CM68b6fc4e50b0.cpe.net.cable.rogers.com) |
| 22:51.53 | Yatekii | wdll thatsthe sad part about thelinux community :( |
| 22:52.22 | Yatekii | but msvc is weird :P |
| 22:52.53 | Yatekii | also i dont know how code can be broken and still work. |
| 22:53.32 | Yatekii | i mean sure broken stuff can compile and not error/warn but sooner or later the binary will fck up during runtime, no? |
| 22:53.41 | Yatekii | please teach me senpai |
| 22:53.46 | decimad | Object slicing that is not detected and access outside of a bitset array |
| 22:54.36 | decimad | the latter working, because the bitset actually using a multiple of byte |
| 22:55.00 | decimad | still if you have 13 bits in a bitset and then you access the 14th, it's a mistake |
| 22:55.54 | *** join/#kicad superlou (~superlou@ool-457533dc.dyn.optonline.net) |
| 22:56.56 | decimad | and then there's one spot where the code relies on gcc stl loading a header implicitly where msvc doesn't |
| 22:57.34 | Yatekii | uhm yeah i see your point (i actually didnt know what object slicing is. well i did but didnt know the name) |
| 22:57.49 | decimad | oh, and the timing code uses the posix api |
| 22:57.54 | Yatekii | but i dont get why you would have to defend yourself :s |
| 22:58.07 | Yatekii | or even the need of mentioning msvc |
| 22:58.26 | decimad | because people would say: it works well with gcc, why bother? :) |
| 22:58.28 | Yatekii | oh ok that one is not so trivial i guess :P |
| 22:58.37 | Yatekii | ... thats the dumb people |
| 22:59.41 | Yatekii | actually i know that behavior from a lot of my ms friends ... "but pressing the green arrow runs it just fine inside vs with debug mode" "yeah but release in standalone doesnt" "er uhm ..." :P |
| 22:59.42 | decimad | I replaced the timing with std::chrono, but mingw's std::chrono implementation is bad |
| 22:59.53 | Yatekii | hmm ok |
| 23:00.15 | decimad | so here I am rebuilding when I pushed something ;) |
| 23:00.21 | Yatekii | :P |
| 23:00.33 | Yatekii | lotsfun |
| 23:01.09 | Yatekii | but i guess that 10mins of background cpu time once in a while dont matter right? its just annoying :) |
| 23:01.32 | decimad | Well it's 10 minutes I need to watch pr0n instead of working ;) |
| 23:02.14 | DocScrutinizer05 | >>...use shift to override snapping... <xzcvczx> decimad: shift to enable grid snapping maybe<< decent *full* "key"mapping (actually input mapping) to arbitrary functions - FTW |
| 23:02.23 | Yatekii | damn :P |
| 23:03.04 | decimad | DocScrutinizer05: you want to press Ctrl+F5 to override snapping? |
| 23:03.14 | DocScrutinizer05 | sure, why not? |
| 23:03.17 | Yatekii | decimad: or code a new feature ;) |
| 23:03.27 | DocScrutinizer05 | I want to be free to define that if I want |
| 23:03.44 | Yatekii | lol the f keys are the most useless shyte in the world |
| 23:03.54 | Yatekii | i still dont get it when apps use it |
| 23:04.53 | decimad | DocScrutinizer05: While I do see your point I didn't really want to refactor the command system just to get snapping for segment dragging ;) |
| 23:05.41 | DocScrutinizer05 | sure. It only makes sense as a generic kicad feature, not for a "silly" specific single usecase |
| 23:06.10 | decimad | I think that's stuff that could be done when eeschema and pcbnew are on a same level so they could be combines into a common environment so we'd have tabbed windows ;) |
| 23:08.00 | decimad | Yatekii: I have to admit, when all cores are compiling, the msvc code helpers (go to definition, show signature, show type etc.) are not really responsive anymore |
| 23:09.03 | decimad | If I had an 8 core I'd probably compile on 6 and have still 2 left to work with... but currently that would mean 2 instead of 4 |
| 23:09.12 | Yatekii | well yeah vs ... |
| 23:09.27 | Yatekii | ye |
| 23:09.48 | Yatekii | one towork withis enough tbh |
| 23:10.10 | Yatekii | well ima go to bed :) gn8 & till tomorrow |
| 23:12.31 | *** join/#kicad rohitksingh (~rohit@119.82.107.167) |
| 23:13.22 | decimad | gn8 |
| 23:13.26 | DocScrutinizer05 | decimad: ((that's stuff that could be done when...)) ACK. That was the purpose of me chiming in: suggest considering it when due. A further suggestion: don't bother about multiple intricate input mappers with overlapping redundant functionality, just make **everything** a macro, no matter if you press a plain "e" or F11 or ctrl-Del |
| 23:14.13 | DocScrutinizer05 | ..or right mouse button |
| 23:15.26 | *** join/#kicad ayjay_t (~ajp_@24-151-28-56.dhcp.nwtn.ct.charter.com) |
| 23:17.33 | decimad | DocScrutinizer05: that's a technicality reallly... doesn't bring anything to the table but: unifies on one side (command dispatch), but makes special cases for the ui necessary (nobody wants to see fn script() MoveCursor end as an action in the config ui, right?) |
| 23:18.06 | DocScrutinizer05 | yeah, right, scripts need an alias |
| 23:18.32 | decimad | So I'd rather say: have fixed-function commands and scripts... but they all run through the same thing |
| 23:18.48 | DocScrutinizer05 | wouldn't mind it, yep |
| 23:19.13 | decimad | And actually that was the idea behing pcbnew's toolmanager, wasn't it? |
| 23:19.43 | decimad | I only looked at it in the context of specific interactive tools yet |
| 23:19.49 | DocScrutinizer05 | actually I prefer calling scripts being a genuine function like "fn script(MoveCursor, "right")" |
| 23:20.09 | DocScrutinizer05 | yes, toolmanager works the way I'd prefer |
| 23:20.36 | DocScrutinizer05 | at least from a user's perspective |
| 23:21.31 | DocScrutinizer05 | assigning arbitrary functions to a key, like "cursor-right" or "runscript(...)" |
| 23:21.37 | decimad | Omg, Tom's started a discussion there... :( |
| 23:23.01 | DocScrutinizer05 | actually that only `moves first (and only) line of code of a virtual script into the key map manager´ |
| 23:24.20 | DocScrutinizer05 | this would cover keymapping, macros, and scripts |
| 23:26.55 | DocScrutinizer05 | there should probably be a "push" and "pop" for the complete keymapping, so a key sequence's first key could do a push and then redefine keymapping temporarily for next keypress |
| 23:33.12 | decimad | That should be built into the keystroke-tree so you could even pop up continuations if the user holds the keys longer |
| 23:33.43 | decimad | I see we have a dev there that would like to implement it ;) |
| 23:35.04 | DocScrutinizer05 | I'd actually like to implement that, if only I'd understand a word of the kicad code |
| 23:38.57 | *** join/#kicad mroszko (~mroszko@ool-4350d3a1.dyn.optonline.net) |
| 23:42.15 | *** join/#kicad ridesa7___ (~quassel@pool-74-110-107-3.nrflva.fios.verizon.net) |
| 23:44.45 | *** join/#kicad C47 (~Xwx47@189.242.29.24) |
| 23:46.25 | *** join/#kicad cirilo (~cirilo@CPE-120-144-53-157.lnse5.win.bigpond.net.au) |
| 23:50.24 | *** join/#kicad mroszko_ (~mroszko@413335d3.cst.lightpath.net) |
| 23:54.45 | decimad | that doesn't count, since you could just create a bare wx app and implement it there... so it could be ported ;) |
| 23:55.47 | decimad | that's how I'd do it anyways, since you don't have to dig around bunches of code that you could break during development |
