| 00:00:19 | SpaceCoaster | GoRK: it's a serial bus that can be shared, we use two GPIOS, one to enable/disable the DSP and one for the Serial Flash. They are both normally disabled. |
| 00:00:48 | BZFlag | would be nice to locate at the end, and then a large kernel/small ramdisk or a large ramdisk/small kernel config would both work on small memory devices. |
| 00:01:00 | GoRK | SpaceCoaster: cool well i will ahve to study it, the circut looked pretty simple |
| 00:01:10 | SpaceCoaster | BZFlag: yeah I am assumming that the memory is all fragmented into 1M or 2M chunks. |
| 00:01:24 | GoRK | SpaceCoaster: the circut for the lart with the burr-brown dac |
| 00:01:53 | SpaceCoaster | GoRK: the atmel manual is really clear about how it wants to be drive, good drawings of clocks etc. |
| 00:01:56 | BZFlag | so assume the kernel will put them together and tell it the offset into the spliced reqion. |
| 00:02:07 | BZFlag | then load it framented at the end. |
| 00:02:42 | BZFlag | so if it's 3M (like it can be on the tux) it would go to a 5M offset on an 8M tux. |
| 00:03:34 | SpaceCoaster | But if we fix it at 4M nothing can hit it, can it? |
| 00:03:35 | GPSFan | GoRK: SpaceCoaster: the DSP & serial flash are connected to gpio's |
| 00:04:10 | SpaceCoaster | GPSFan: yes, their chip enables are. |
| 00:04:16 | GPSFan | GoRK: SpaceCoaster: serial port 4 (SSP) is connected to the UCB1200 |
| 00:04:23 | BZFlag | a really large kernel would by the time it decompresses no? hmm.. not at present,as the kernel only get's 4 |
| 00:04:43 | BZFlag | but picture the same feature on a 4M machine and it becomes more relevant. |
| 00:05:30 | GoRK | hrm. the lart has both a ucb1x00 and the dac so they are probably connected in a similar manner |
| 00:06:19 | BZFlag | also consider that once the jffs2 patches are in we will likely switch to an uncompressed kernel. |
| 00:07:00 | GoRK | see http://www.lart.tudelft.nl/lartware/ksb/ksb_rev2.pdf page 6 for the lart's audio stuff |
| 00:07:38 | SpaceCoaster | BZFlag: when the kernel is running the jffs2 filesystem is used in situ, from the flash, is that true? |
| 00:08:04 | BZFlag | blob could then load an uncompressed kernel directly from jffs2 into memory. or it could be configured to load and decompress a gzipped uncompressed kernel. |
| 00:08:27 | BZFlag | yes. jffs2 is always in flash (well, mtd device) not memory. |
| 00:09:07 | SpaceCoaster | so the issue that we are dealing with is just programming the jffs2 into flash from a fragmented memory, so it's just a blob issue. |
| 00:09:22 | BZFlag | So I think a gzipped vmlinux and gzipped ramdisk would also be likely for some devices. |
| 00:09:41 | GPSFan | GoRK: In Lart's KSB board, the UCB1200 is connected the same way as on the Tux pins 39-42 on the UCB1200 |
| 00:10:06 | BZFlag | jffs2 is in flash.. it's loading a kernel and possibly a ramdisk into fragmented memory. |
| 00:10:18 | BZFlag | the kernel fill find jffs2 in flash by itself. |
| 00:10:23 | GoRK | GPSFan: yeah but there's also some seperate circut for pcm audio using a dac... |
| 00:10:24 | BZFlag | s/fill/will/ |
| 00:10:43 | BZFlag | oh, you mean for flashing.... yeah. that too. |
| 00:10:50 | GPSFan | GoRK: I'm off to dinner, I'll look at it in a bit. :>) |
| 00:10:58 | GoRK | GPSFan: enjoy |
| 00:11:54 | SpaceCoaster | BZFlag: I think you have a wider view of all this, fitting it in a bigger plan. |
| 00:12:01 | BZFlag | I wonder if erikm would consider enabling the mmu in blob and then disabling it before jumping to the kernel. then we'd have just one block in blob. ;-) |
| 00:12:08 | BZFlag | SpaceCoaster: ;-) |
| 00:12:36 | BZFlag | perhaps. I'm waiting on a few sa devices that I hope to do blob/kernel work on soon. |
| 00:13:52 | BZFlag | I figure if we can determine a clean plan that will work for most any device, then we get to make it the default in blob and the kernel easier and have less maintainace and more testing. |
| 00:14:40 | SpaceCoaster | BZFlag: do you want to put together some scenarios on the wiki? |
| 00:16:07 | BZFlag | SpaceCoaster: perhaps. I will definitely be more active on this when even one of these contract actually pays Lineo any money. |
| 00:16:17 | BZFlag | I may also be redoing the zaurus work. |
| 00:16:39 | BZFlag | they have paid us, but I think there is a lot of room for improvement. |
| 00:18:09 | BZFlag | zaurus uses arm loader + cf updater + diagnostics + cramfs + ext2 + 2 rd minix partitions. |
| 00:18:12 | BZFlag | yuck. |
| 00:18:23 | SpaceCoaster | BZFlag: woa |
| 00:18:29 | BZFlag | I'd change it over to blob + one jffs2 |
| 00:18:50 | SpaceCoaster | BZFlag: I would be interested in CF booting and even tftp for development. |
| 00:20:01 | SpaceCoaster | I wondered why blob was allocated 128k, guess I now know. |
| 00:20:09 | BZFlag | yes. If I can get some time here I will add cf+tftp to blob. |
| 00:20:21 | BZFlag | probably even [xz]modem and nfsboot |
| 00:21:54 | BZFlag | heh. blob gets the first flash block. and mtd likes all blocks to be the same. so that means that blob might as well use 128k cause that how big the first region is. it's actually in 4 blocks on the tux, but as I said, jffs2/mtd does not really want the small blocks. |
| 00:26:07 | GoRK | blob will probably need it by the time you get all that stuff you want in there :) |
| 00:26:34 | SpaceCoaster | GoRK: that's what I was thinking, expanding to meet the available space. |
| 00:29:42 | SpaceCoaster | BZFlag: at the moment nothing boots, I guess I have half of one idea and half of another implemented, hurrah for CVS. |
| 00:33:09 | SpaceCoaster | BZFlag: do you know anyone who works with pocketlinux or kaffe on ARM? |
| 00:34:35 | BZFlag | SpaceCoaster: not yet. |
| 00:35:07 | BZFlag | I have a zaurus here with the tao jvm on it ;-) |
| 00:36:29 | SpaceCoaster | My work has been directed to developing an app in Java but when I got enough memory the blackdown VM was way too slow. The tao jvm, is it any good? |
| 00:37:53 | GoRK | SpaceCoaster: what is the current status of the non dsp part of the wheaties driver? eg the serial flash? |
| 00:39:06 | SpaceCoaster | I last looked at it about 10 days ago, it has corruption of the last byte in each 256 page, so not great! |
| 00:39:24 | prpplague | SpaceCoaster: you done anything with the NET+ARM processor? |
| 00:39:48 | SpaceCoaster | prpplague: no, fraid not. |
| 00:40:23 | SpaceCoaster | GoRK: did you have plans for the serial flash? |
| 00:47:09 | GoRK | SpaceCoaster: not yet, just wondering |
| 00:47:12 | prpplague | SpaceCoaster: hmm found a printserver in our office that uses that core and a brand of router we carry uses it |
| 00:48:14 | SpaceCoaster | GoRK: it can be looked at, must be a stupid bug, the pages are 264 bytes so this is in the middle, I do the copying. |
| 00:48:26 | prpplague | SpaceCoaster: http://www.netsilicon.com/EmbWeb/products/netuClinux.asp |
| 00:49:48 | prpplague | needs sleep, cisco crap to deal with in the morning |
| 00:52:01 | SpaceCoaster | GoRK: I was debatting if it should stay in the 2.4.1X release but I think it would result in more hassle and less progress, I would like to make it use DMA so it doesn't mess with the audio. How is that these days? |
| 00:52:31 | SpaceCoaster | s/debatting/debating/ |
| 00:57:58 | GoRK | SpaceCoaster: the new audio driver doesnt seem to be very bugged by wheaties |
| 00:59:14 | SpaceCoaster | good, the audio has a 8 word fifo, it should be ok for 1ms at 8khz |
| 01:05:06 | SpaceCoaster | GoRK: any chance of placing an advance order for your audio board/ |
| 01:05:29 | BZFlag | SpaceCoaster: have not done enough in the tao jvm to get an opinion. it's close source so getting one that works for you might be tough. |
| 01:06:28 | BZFlag | Check the Zaurus development site to see if they have it for download. I know it's available for pocketpc, but I have not seen a public Linux arm release. might be one someplace. |
| 01:06:40 | SpaceCoaster | BZFlag: pocketlinux's kaffe is open source and getting one that works is very tough :-) |
| 02:20:09 | GPSFan | GoRK: still about? |
| 02:23:15 | Tangent | GPSFan : re-assembled your Tux yet? |
| 02:23:45 | GPSFan | Tangent: not yet, still ringing stuff out. :-( maybe this weekend |
| 02:24:43 | Tangent | GPSFan : Cool... BTW.. can you tell me what happenes to the TX line from uart2 (The IrDA one I mean if it's not 2) |
| 02:25:12 | Tangent | GPSFan : Is it used for anything, or can we add IR TX capability in much the same way as we did for serial? |
| 02:25:47 | GPSFan | Tangent: sure, check out the new rev of the UCB1200's the TXD for the IR port is unconnected. |
| 02:26:16 | GPSFan | Tangent: pin 31 od J601 |
| 02:26:22 | GPSFan | s/od/of |
| 02:26:47 | Tangent | GPSFan : Cool.. thanks... not that I actually want ir tx capability at the moment, but someone else was asking for it |
| 02:27:10 | Tangent | GPSFan : And I've already found a schematic for a transmitter that's pretty much suitable |
| 02:27:31 | GPSFan | Tangent: ;>) can't hurt to ponder new possibilities. |
| 02:28:10 | Tangent | GPSFan : Yep... and you gotta admit, the Tux would make a cute touchscreen remote control in webpad form. |
| 02:29:17 | GPSFan | Tangent: you could watch & control your tv/dvd/cd/vcr... and surf the web while controlling your house. ;>) |
| 02:29:43 | Tangent | GPSFan : Yeah.. that sort of thing would be pretty cute... |
| 02:30:02 | Tangent | GPSFan : And my Pronto _has_ started playing up lately |
| 02:30:30 | Tangent | GPSFan : It's just a shame that the hardware hack takes so much less time than writing software to use it |
| 02:31:28 | GPSFan | Tangent: yep, I've not done any real software development in a looooooong time. |
| 02:37:32 | Tangent | GPSFan : I've not done any real software development ever... largely because I don't find coding particularly interesting |
| 02:38:03 | Tangent | GPSFan : I much prefer breaking things that other people have made :) |
| 02:39:51 | GPSFan | Tangent: at one time I had to write some 8080 ASM code to support some hardware I designed, and not that I remember it, I wrote some 1802 ASM code as well. |
| 02:40:08 | GPSFan | s/not/now |
| 02:40:28 | Tangent | GPSFan : device drivers and firmware are a different matter... I'm talking about stuff with databases, and file handles, and GUIs |
| 02:40:41 | Tangent | GPSFan : That stuff is interesting... |
| 02:41:37 | Tangent | GPSFan : Stuff where you create it all yourself in ASM, or low level stuff to interface with real physical hardware is cool... |
| 02:41:44 | GPSFan | Tangent: never, avoid it like the plague. ;>(( there are many much more talented SW types just itching to write all that code, who am I to deny them their fun? |
| 02:42:08 | Tangent | GPSFan : I just like to live my life below layer 3 of the OSI model |
| 02:43:12 | GPSFan | Tangent: ;>) |
| 03:09:23 | Tangent | goes to bed... night all |
| 03:09:33 | GPSFan | night.. |
| 03:23:18 | StarKruzr | So how's the project doing, *? |
| 03:26:35 | GPSFan | StarKruzr: lots of talented people havinb a ball. ;>) |
| 03:26:49 | GPSFan | s/havinb/having |
| 03:30:55 | StarKruzr | Cool |
| 03:31:06 | StarKruzr | Is there a list somewhere of what work needs to be done? |
| 03:32:40 | GPSFan | StarKruzr: read the wiki, and get a feel for the project, look at the schematics if you are hardware orientated, by one,2,3,4,... think of a good app and go to town. |
| 06:53:12 | GoRK | helooo |
| 06:54:05 | GoRK | Tangent: what are you up to today? ir? |
| 06:54:48 | GoRK | err never mind |
| 06:54:50 | GoRK | lalala |
| 06:54:52 | GoRK | anyone up |
| 08:14:57 | Russ | hey GoRK, andersee |
| 08:28:13 | andersee | howdy Russ |
| 08:28:23 | GoRK | hey |
| 08:28:25 | GoRK | what's the happs? |
| 08:28:31 | andersee | Trying to fix some bind issues for BZFlag |
| 08:29:05 | GoRK | i realized this audio transformer was much larger than my pcb allowed for. actually i realized how freaking small i made myi pcb.. so anyway im fixing that now |
| 08:30:28 | andersee | heh |
| 08:30:43 | andersee | Pretty monsterous huge transformer? |
| 08:31:07 | GoRK | hot particularly... .8"x.6" |
| 08:35:41 | andersee | thats not too bad |
| 08:42:57 | rfbrown | GoRK: Hello! I had a look around for information about suppressing ground noise. Didn't find any alternatives to the transformer, unfortunately. |
| 08:44:24 | GoRK | rfbrown: well... i also thought that there are a few people who want to put this in their car as a mp3 player or voice prompting navigation system... they will very likely need an isolation transformer to get rid of the noise inside of that environment |
| 08:44:52 | GoRK | so... even if there is an alternative, this probably will be the best course of action. the tux's grounding is fairly dirty it seems |
| 08:48:43 | GoRK | or it might just be the new kernel :) |
| 08:51:21 | GoRK | hmm thsese boards are way cheaper for qty > 100 |
| 08:51:47 | andersee | how much? |
| 08:52:05 | GoRK | they will probably end up costing folks about $5 |
| 08:52:23 | GoRK | just for the board |
| 08:53:23 | GoRK | that's w/ solder mask and silkscreen... without those they are considerably cheaper |
| 08:56:24 | andersee | whats solder mask do? |
| 08:56:49 | GoRK | it's the gree coating. only the places solder is supposed to go get exposed |
| 08:57:07 | GoRK | without it the copper traces are bare |
| 08:59:31 | andersee | ahh. yeah we want that |
| 09:17:55 | GoRK | ok about to make my second mouser order for test components |
| 11:26:13 | MicroChip32 | is anyone alive that can help me with basic shell scripting (aka accessing passed parms) |
| 11:30:51 | MicroChip32 | ok better question: is anyone alive? |
| 11:33:20 | cory | *checks pulse* |
| 11:35:14 | MicroChip32 | cory: just as i thought, no life here ;) |
| 11:35:20 | cory | ;p |
| 11:35:49 | MicroChip32 | well i could have said no [intelligent] life here **runs for cover** |
| 11:36:05 | cory | yes, I know, but I thought of that before you said it |
| 11:36:21 | cory | is a telepath... I'm Troi's brother |
| 11:36:33 | MicroChip32 | =P |
| 11:37:01 | cory | at least I'm not telepathetic. hahah. |
| 11:37:21 | MicroChip32 | lol |
| 11:39:40 | MicroChip32 | btw, i thought of it before i said it too (the [intelligent] comment ) |
| 11:40:50 | cory | I can imagine.. I didn't think you had anyone else help you think of it |
| 11:43:50 | cory | ok, this got old fast :) |
| 14:04:34 | MicroChip32 | releases a tux-virus to see if anyone is awake |
| 14:19:26 | Tangent | releases a virus killer |
| 14:19:44 | Tangent | MicroChip32 : I'm awake.. but I don't know why I bothered |
| 14:20:09 | prpplague | Tangent: look out windows users, it'll probably kill the entire os |
| 14:20:39 | Tangent | prpplague : Maybe we should release a virus for windows users... |
| 14:20:58 | MicroChip32 | Tangent: hehehe |
| 14:23:55 | prpplague | argh, its now been 6 days since i've been able to do any ARM development |
| 14:24:27 | Tangent | prpplague : Why? |
| 14:25:02 | prpplague | lines, training, and routers, ohh my |
| 14:25:03 | prpplague | lines, training, and routers, ohh my |
| 14:25:25 | prpplague | has been playing network guru |
| 14:25:44 | Tangent | prpplague : Network guru is fun... that's my favorite job |
| 14:29:05 | prpplague | hmm so that is what that cable is for...... |
| 14:33:50 | MicroChip32 | hey Tangent |
| 14:34:11 | Tangent | MicroChip32 : Ah.. so you are awake :) |
| 14:34:26 | MicroChip32 | not really hehehe you feel like poking at some c ? |
| 14:34:55 | prpplague | MicroChip32: poking at or poking fun of? |
| 14:35:03 | MicroChip32 | hehehe well whatever works ;) |
| 14:35:31 | MicroChip32 | http://home.mctech.org/pub/mytest.tgz |
| 14:35:46 | Tangent | MicroChip32 : Not at this precise moment... I've just walked in, and my ears are still so cold that they hurt |
| 14:36:01 | MicroChip32 | hehe Tangent needs a hot tub ;) |
| 14:36:06 | Tangent | MicroChip32 : Not that that has anything to do with C coding... but I just don't feel like it right now |
| 14:37:01 | MicroChip32 | no worries |
| 14:38:04 | MicroChip32 | hmmm DigExt ? |
| 14:42:51 | Tangent | MicroChip32 : DigExt? |
| 14:44:02 | MicroChip32 | someone grabbed the file with a browser labeled digext |
| 14:44:44 | Tangent | MicroChip32 : It wasn't me |
| 14:48:41 | prpplague | Tangent: i grabed the file with netscape |
| 14:49:03 | prpplague | sorry that was MicroChip32 |
| 14:49:32 | MicroChip32 | prpplague: from linux right? |
| 14:49:37 | prpplague | MicroChip32: ya |
| 14:49:53 | MicroChip32 | nope this wasnt you, this person was ms 98 ie 5 |
| 14:50:31 | prpplague | ahh |
| 14:50:39 | MicroChip32 | probably someone with and auto link grabber |
| 14:51:22 | MicroChip32 | no matter, was just curious about the extra tag on the browser idenitfyier |
| 14:51:43 | prpplague | you know you work too much when you realize you keep a pillow and blanket in your desk drawer |
| 14:51:54 | MicroChip32 | lol |
| 14:52:12 | MicroChip32 | prpplague what does it mean when you have a whole bed next to your desk ? |
| 14:52:45 | prpplague | MicroChip32: that your boss is a little more accepting of you sleeping at the office |
| 14:52:53 | MicroChip32 | lol hmmmm |
| 14:53:16 | MicroChip32 | or, as expressed in perl ... my $office = $desk; |
| 14:53:22 | MicroChip32 | ack no no |
| 14:53:27 | MicroChip32 | or, as expressed in perl ... my $office = $bedroom |
| 14:53:59 | prpplague | anyone seen that cdw commercial where the the chairman of a board says "anyone else have more un-resonable demands for the it department?" |
| 14:55:30 | prpplague | anyone here ever deen to barbados? |
| 14:55:37 | prpplague | s/deen/been |
| 14:56:04 | MicroChip32 | i think everyone but you me and tangent are sleeping |
| 14:56:13 | MicroChip32 | or else deep inside their tux's |
| 14:56:18 | prpplague | lol |
| 14:56:35 | Tangent | MicroChip32 : Maybe sleeping inside their Tuxes... there's a big shortage of IT work right now :) |
| 14:56:52 | MicroChip32 | hehe |
| 14:57:19 | prpplague | MicroChip32: i'm walking a guy over the phone installing rhat with kickstart disk, everything in the caribean takes forever, something that should take 10 minutes takes 10 hours |
| 14:57:51 | MicroChip32 | hmmmm sounds like a good place to take a date ... <G> |
| 15:00:03 | prpplague | lol |
| 15:03:52 | prpplague | http://www.netsilicon.com/EmbWeb/products/netarm.asp |
| 15:04:05 | prpplague | seen that MicroChip32 |
| 15:05:40 | MicroChip32 | no but at first glance it looks cool |
| 15:06:21 | prpplague | i found a print server in the office that uses it |
| 15:08:29 | prpplague | they even have a sdk for uclinux |
| 15:08:32 | MicroChip32 | looks like something Tangent needs for his home control stuff ;) |
| 15:10:46 | prpplague | i didn't find any reference to it having an mmu though, and since they use uclinux, i would guess that it doesn't |
| 15:20:14 | Tangent | Hrm.. those pins look mightily close together.. I'd have to buy a better soldering iron |
| 15:21:29 | prpplague | lol |
| 15:23:06 | prpplague | Tangent: it's my understanding they have a version of the cpu on a pga socket |
| 15:24:27 | prpplague | this looks cool - http://www.pdabuzz.com/#newsitem1005329972,8153, |
| 15:36:39 | prpplague | MicroChip32: did you get this url - http://www.pdabuzz.com/#newsitem1005329972,8153, |
| 15:38:10 | Tangent | is just reading the simcity review |
| 15:40:19 | MicroChip32 | prpplague i see it, but on my bandwidth i dont browse much *grumble* |
| 15:40:25 | Squart | Does anyone know of some sourcecode for a serial PIC programmer? preferably for DJGPP |
| 15:40:59 | Tangent | Squart : I'm sure I've seen it around somewhere... |
| 15:41:34 | Squart | Tangent: ;-) let's have a look... |
| 15:42:27 | Tangent | searches through his bookmarks |
| 15:45:09 | Tangent | Squart : Will this do you? http://www.jdm.homepage.dk/easypic.htm |
| 15:46:38 | MicroChip32 | prpplague did you look at that code i posted earlier ? |
| 15:48:08 | prpplague | MicroChip32: only briefly, i'm still working on some routers in barbados right now |
| 15:49:52 | MicroChip32 | oh yea ;) |
| 15:50:19 | MicroChip32 | wonders if those "routers" are wearing bikinis <G> |
| 15:51:49 | Squart | Tangent: Yaiks, it is pascal source! Ah well, done that ages ago too.... |
| 15:52:44 | Tangent | Squart : I'm finding plenty stuff just searching in google... but I've got sidetracked now by JAL |
| 15:56:50 | Squart | Tangent: Yeah, sorry I found some usefull stuff as well. But I especially wanted to know if someone knew some djgpp source |
| 15:57:18 | Tangent | Squart : Not me.. sorry.. I'm using MPLAB for everything, as I actually paid money for it |
| 15:57:49 | Squart | Tangent: You did? why, it is free? |
| 15:58:39 | Tangent | Squart : I got it from Maplin, with the Programmer.. Has nice 40-pin socket, and programs all the different PICS |
| 15:59:26 | Tangent | Squart : If I knew then what I know now, I probably wouldn't have bought it... but that's just the way life is |
| 16:00:12 | Squart | Ah, I see. I borrowed a programmer from a colleage, but the programmer software is for Win95 and the comp I am using for it has win3.1 ;-( |
| 16:03:02 | Tangent | Squart : heh.. I just pulled apart my PIC programmer for a look.. unsurprisingly, it's PIC powered :) |
| 16:05:20 | Squart | Tangent: Odd, a programmer needs almost nothing.. |
| 16:05:32 | Tangent | Squart : It's got 20 ICs in it |
| 16:07:48 | Squart | Tangent: ??!! what does it do? control a plane? |
| 16:08:24 | Squart | is checking out how black his pizza is becoming |
| 16:08:59 | Tangent | Squart - for starters, it's got stuff for controlling the voltage to the pic... it's got a nice little switchmode converter hooked up to parts of the pic, so i guess they're selecting it based upon the chip that's inserted |
| 16:09:16 | Tangent | brb.. reboot needed.. shift key has started functioning as backspace again |
| 16:24:57 | ctr | any solaris experts here? |
| 16:26:49 | Tangent | ctr : Depends how much expertise you need |
| 16:27:06 | ctr | ok |
| 16:27:16 | ctr | i have intalled a library in a non standard path |
| 16:27:20 | ctr | (/tmp/ferite/sun) |
| 16:27:49 | ctr | and i want to now build another package that depends on it |
| 16:28:17 | MicroChip32 | solaris? isnt that the big bright ball in the sky. seems you'd want #astronomy for that <G> |
| 16:28:38 | ctr | but it's being a pita |
| 16:28:41 | ctr | MicroChip32: heh :) |
| 16:30:41 | Tangent | ctr : Ah.. i don't really understand things like that.. I know how to set my library path, and run ldconfig, but I guess that's not sufficient? |
| 16:32:25 | ctr | hrmm |
| 16:32:26 | ctr | nm |
| 16:32:34 | ctr | got a little bit futher down the road now :) |
| 16:33:19 | ctr | now it's a libtool problem |
| 16:33:22 | ctr | blum |
| 16:47:23 | Tangent | BZFlag : Morning |
| 16:48:41 | Squart | OK this is off-topic, but if anyone likes to answer... whenever I ftp get a file on my win3.1 box from my debian server, the files get corrupted somehow. exe files faile to execute when downloaded. Does anyone know why? |
| 16:48:54 | Squart | BZFlag: evening... |
| 16:49:16 | MicroChip32 | Squart make sure youre downloading as binary (for exe's) or ascii (for text/etc) |
| 16:49:31 | MicroChip32 | g'mornin/ev'nin BZFlag |
| 16:49:36 | Squart | uChip: Doh... |
| 16:49:56 | MicroChip32 | Squart: *chuckle* well, thats the usual cause |
| 16:50:15 | Squart | how stupid of me, never even crossed my mind |
| 16:52:17 | Squart | I was beginning to think there was something wrong with my win3.1 box, until I did a good 'ol floppy transfer |
| 16:57:41 | Squart | BZFlag: as soon as my PalPay checks out I'll be placing an order for a Tux ;) |
| 17:24:16 | ctr | how do i create a tarball under solaris? |
| 17:27:05 | Tangent | ctr : Same way as normal as long as tar is installed |
| 17:31:10 | ctr | yeah |
| 17:31:13 | ctr | i found gnu tar |
| 17:31:23 | ctr | i'm not in the mood to wrestle normal tar |
| 17:44:31 | pattieja | prpplague: afternoon |
| 17:44:37 | prpplague | pattieja: howdy |
| 17:45:09 | pattieja | does UPS deliver on Saturday? |
| 17:45:29 | MicroChip32 | pattieja around here they do |
| 17:46:47 | pattieja | wonder about St. Louis? |
| 18:01:33 | AlHaz | for an extra $6 or $7 they do |
| 18:01:50 | AlHaz | it is a union shop after all |
| 18:13:17 | SpaceCoaster | GoRK: want a change to tuxphone which fixes the wheaties problems? |
| 18:29:58 | BZFlag | bbiab |
| 18:53:09 | prpplague | anyone know of any docs about implementing a parallel port on ARM ? |
| 19:16:59 | prpplague | howdy u-potato-chip |
| 19:21:29 | prpplague | MicroChip32: what do you think - http://www.planex.net/product/print/max100.htm |
| 19:21:36 | prpplague | MicroChip32: nice hack? |
| 20:16:46 | prpplague | anyone got experience with uclinux? |
| 20:17:14 | ctr | nup |
| 20:19:24 | pattieja | prpplague: what'ya need to know? |
| 20:19:34 | pattieja | (only compiled staticly, not shared libraries, though) |
| 20:19:50 | pattieja | prpplague: sorry. thinking of uClibc not uClinux. |
| 20:19:54 | pattieja | nevermind |
| 20:19:56 | prpplague | pattieja: np |
| 20:20:20 | prpplague | i need a small print server that can run linux and costs less that $150 |
| 20:21:04 | GoRK | SpaceCoaster: yeah what's up? did you add some delays to tuxphone or was there some other issues? |
| 20:21:14 | prpplague | i've found several units that are are based, but they all have no mmu and also are very resources 1meg ram 1meg flash |
| 20:21:34 | prpplague | s/are/arm |
| 20:22:24 | GoRK | prpplague: i dont see you finding one with more goodies that costs less than $150-- |
| 20:23:11 | prpplague | GoRK: ya i know, from what i can tell the units all can run uclinux but don't know for sure |
| 20:23:45 | prpplague | i'd settle for a print server i could at least modify the code |
| 20:24:43 | GoRK | hmmm.. build one out of a tux? :) LART has schematics for adding a parallel port to sa1100 |
| 20:25:11 | prpplague | GoRK: i'd be hard to get 5,000 tuxs per year |
| 20:25:20 | GoRK | prpplague: yes |
| 20:25:58 | prpplague | GoRK: ya i've got a couple of LART board in production now for testing |
| 20:26:50 | prpplague | its just a shame that most print server OEM's won't even bother fixing their code |
| 20:27:21 | GoRK | i dont know if the parallel port is good for printers or not... they used it for moving high datarate video in between lart's |
| 20:33:13 | GoRK | spppppppppacecoaster!!!!!!!!! |
| 20:39:52 | SpaceCoaster | GoRK: hi |
| 20:40:49 | GoRK | SpaceCoaster: hello |
| 20:40:56 | GoRK | SpaceCoaster: what does the tuxphone patch do? |
| 20:41:43 | SpaceCoaster | Ah, that was just to get you excited, I wasn't seeing the problems that others had and it was because I had modified tuxphone to use stderr for debug. |
| 20:42:17 | SpaceCoaster | stderr is normally unbuffered so I guess it introduced delays. |
| 20:42:46 | GoRK | SpaceCoaster: ahhh . heh i guess we should be using stderr for debug... but anyway did you take it out and then see the problems? |
| 20:43:34 | SpaceCoaster | I have put a delay between reads into wheaties.c and this works on all forms of tuxphone. Can I create a drivers directory in tuxscreen.net CVS and do my development there? |
| 20:44:10 | GoRK | i dont see why not. i'd call it kerneldrivers though |
| 20:44:29 | GoRK | ask tim to be sure though :) |
| 20:44:54 | GoRK | wheaties.c in the kernel driver right? can you give me the line to try on 2.4.13? |
| 20:45:46 | SpaceCoaster | sure, add a udelay(100) just before the return in read_message(). |
| 20:47:40 | GoRK | cool |
| 20:47:46 | SpaceCoaster | does it work? |
| 20:47:54 | ctr | so is there tone dialing for the tuxphone now? |
| 20:48:12 | SpaceCoaster | ctr: where have you been boris? |
| 20:48:16 | ctr | uhm |
| 20:48:19 | ctr | busy with real life |
| 20:48:29 | ctr | :) |
| 20:48:49 | ctr | spenind time with my gf |
| 20:48:55 | ctr | university work =/ |
| 20:49:00 | GoRK | SpaceCoaster: dunno still compiling |
| 20:49:28 | SpaceCoaster | man, tone dialing, music playing, it's all happened, don't give me that, GoRK tell this guy whats been happening! |
| 20:49:32 | as2 | Evnin all. |
| 20:49:47 | BZFlag | SpaceCoaster: kernel would be fine. |
| 20:49:48 | GoRK | heh tone dialing |
| 20:50:44 | ctr | can i use the speaker phone? |
| 20:50:55 | SpaceCoaster | BZFlag: yeh, I want others to test/change/improve these things. |
| 20:51:09 | GoRK | ctr: you can do everything except put people on hold |
| 20:51:22 | ctr | rock |
| 20:51:31 | BZFlag | no problem. I mean create a kernel module in ccs and check things in there. |
| 20:51:35 | BZFlag | er cvs |
| 20:51:41 | BZFlag | brb |
| 20:51:45 | ctr | is there a jffs2 image and kernel availible with this all work for us mere mortals? |
| 20:52:32 | SpaceCoaster | BZFlag: I have three drivers, spi, serial_flash, wheaties, they all build great as modules, so I'll put in a Makefile to allow that. |
| 20:52:35 | GoRK | ctr: not yet. the new picogui image will have fbtuxcal and pgtuxphone to do all the fun stuff hopefully this weekend |
| 20:52:49 | ctr | GoRK: excellent |
| 20:52:52 | BZFlag | hmm... /me wishes his dns would update. |
| 20:53:05 | BZFlag | SpaceCoaster: great! |
| 20:53:42 | GoRK | ctr: i have been messing with newer kernels as of late though... i think we lose the ability to calibrate the same way we are currently using in the new kernel |
| 20:54:15 | ctr | GoRK: bugger |
| 20:54:22 | GoRK | so fbtuxcal may be a bit up a creek :-/ though the values it saves could be used by the input drivers of the gui's, i guess |
| 20:54:26 | ctr | GoRK: know why? or still hunting it down? |
| 20:55:19 | GoRK | ctr: the ucb1200-ts driver changed compeletely. it doesnt work the same way anymore.. and for the kernel purists the calibration stuff didnt belong in the kernel anyway, i guess |
| 20:56:03 | ctr | i see |
| 20:56:20 | GoRK | ctr: but accessing the touchscreen also completely hard locks the kernel still so ... :) |
| 20:56:27 | SpaceCoaster | GoRK: hmm |
| 20:56:56 | ctr | really? |
| 20:57:01 | ctr | uhm |
| 20:57:07 | ctr | thats not good =/ |
| 20:57:42 | GoRK | SpaceCoaster: you want to tackle that? it may be related to the original problem where some ts's didn't work... the kernel seperates MCP code from ucb1x00 code now... the resets are a little weird i think.. |
| 20:58:10 | SpaceCoaster | sure, what kernel patch are we using these days? |
| 20:58:13 | GoRK | since we were doing things that affected ucb1200 right in the middle of things that dealt with the mcp last time |
| 20:58:42 | GoRK | SpaceCoaster: russh.dhs.org/files/tuxpatches .... there is a patch to 2.4.13-ac5-rmk2 |
| 20:58:44 | GoRK | lemme look |
| 20:58:59 | GoRK | http://russ.dhs.org/files/tuxpatches/patch-2.4.13-ac5-rmk2-tux2 |
| 20:59:24 | GoRK | of course youll need some other files to patch up to that :) |
| 20:59:36 | SpaceCoaster | doh |
| 20:59:57 | GoRK | 2.4.13 original, -ac5 patch ac5-rmk2 patch |
| 21:00:30 | GoRK | and there are a couple lines you have to change after you get it patched up |
| 21:01:10 | SpaceCoaster | Come on I can build a kernel, the extra patches are interesting tho' |
| 21:01:44 | GoRK | :) hrm the udelay(100) didnt seem to completely solve things.. speakerphone button still didnt work |
| 21:01:55 | ctr | i love having headphones |
| 21:02:40 | GoRK | picking up the handset doesnt either |
| 21:02:44 | SpaceCoaster | Ho, hum, let me check this stuff in and then we can sync our sources. |
| 21:09:10 | GoRK | SpaceCoaster: if i print debug on stderr it mostly works. but a lot of the writes are still getting eaten.. in particular the 100+ byte write the speakerphone button makes |
| 21:11:04 | GoRK | SpaceCoaster: also if you are running stuff on jffs2 root you have to comment out the /dev/mdtblock2 line from fstab otherwise it wont get past mount -a |
| 21:12:32 | SpaceCoaster | GoRK: checkout the wheaties.c from CVS/kernel on tuxscreen.net, we will work from that, I originally (way back) had huge delays, maybe we still need them. |
| 21:13:03 | SpaceCoaster | GoRK: I haven't really been changing much at all. |
| 21:14:22 | pattieja | is away: see you Monday |
| 21:15:25 | GoRK | SpaceCoaster: you accidentally synced your .flags files in |
| 21:20:34 | SpaceCoaster | oops, CVS neophyte I'm afraid. |
| 21:21:02 | SpaceCoaster | you can keep the .flags files for free :-) |
| 21:21:13 | GoRK | im so happy |
| 21:21:57 | SpaceCoaster | do I just cvs delete them ? |
| 21:22:12 | GoRK | yeah they'll go into the Attic but they'll be as good as gone |
| 21:22:21 | GoRK | you can put them in .cvsignore |
| 21:28:59 | ctr | isn;t it cvs remove? |
| 21:29:56 | SpaceCoaster | ctr: it is indeed, thanks |
| 21:30:30 | ctr | heh |
| 21:30:31 | ctr | np :) |
| 21:30:51 | ctr | ferite now builds almost cleanly on solaris |
| 21:35:05 | GoRK | can i somehow give make modules_install a target directory? |
| 21:35:46 | SpaceCoaster | make INSTALL_MOD_PATH=/foobar modules_install |
| 21:35:54 | GoRK | sweeet |
| 21:36:23 | SpaceCoaster | foobar is the root, it has a subdirectory lib/modules/... |
| 21:47:51 | prpplague | howdy mallum |
| 21:54:43 | GoRK | SpaceCoaster: im getting: drivers/mtd/mtdlink.o(.text.init+0xf8): undefined reference to `serial_flash_setup' |
| 21:55:19 | GoRK | oh whoops |
| 21:56:15 | GoRK | guess i need NAND flash support for serial_flash now huh? |
| 21:56:47 | SpaceCoaster | comment _setup and _cleanup out of mtdlink.o, I've asked russ to remove it, if we use modules for serial flash then we don't need to worry about initialisation. |
| 21:57:51 | mallum | prpplague: hi there |
| 21:58:22 | GoRK | out of mtdlink.o? |
| 21:58:46 | SpaceCoaster | nah, it's one of the other files in that directory |
| 21:59:43 | GoRK | nftlcore.c? |
| 22:00:53 | GoRK | i cant find it :( |
| 22:01:04 | SpaceCoaster | one second, still patching |
| 22:01:43 | SpaceCoaster | drivers/mtd/maps/sa1100-flash.c |
| 22:02:07 | prpplague | whooo, 4 routers down 2 to go |
| 22:02:18 | SpaceCoaster | I had to put it in there otherwise the serial flash went in as device 0, and life ended. |
| 22:02:35 | GoRK | heh |
| 22:03:40 | prpplague | i was hope'n that somehting bad would happen to these routers and i'd have to go fix them, no such luck |
| 22:04:52 | SpaceCoaster | prpplague: what broke 'em? |
| 22:05:15 | prpplague | SpaceCoaster: new install, 3 56k p-t-p lines |
| 22:06:11 | SpaceCoaster | prpplague: do you do this stuff from home? |
| 22:06:49 | prpplague | SpaceCoaster: i could, however my 4 year old some can make that difficult, so i drive into the office |
| 22:06:57 | prpplague | s/some/son |
| 22:07:43 | GoRK | SpaceCoaster: new warnings in this about using malloc.h ... deprecated in favor of slab.h |
| 22:09:05 | SpaceCoaster | GoRK: do you reckon it works the same? |
| 22:09:19 | GoRK | SpaceCoaster: dunno. just letting you know about it :) |
| 22:09:31 | GoRK | SpaceCoaster: this all finally compiled up as modules and all that good st uff |
| 22:10:28 | ctr | anyone here listen to digitally imported? |
| 22:10:54 | GoRK | ctr: yeah i have even played it on the tux :) |
| 22:11:01 | prpplague | ctr: not that i know of |
| 22:11:22 | GoRK | ctr: if you are talking about the shoutcast station |
| 22:12:44 | SpaceCoaster | GoRK: got my resistor/capacitor pack from radioshack, they always ship in two parts on by USPS (2-3 days) and the other by UPS, 8-10 days, sigh, gonna hack together a shanip for internal audio, no phones etc. |
| 22:20:55 | GoRK | SpaceCoaster: still problems with the speakerphone on this current set of drivers and kernel 2.4.13 |
| 22:21:34 | GoRK | SpaceCoaster: (this is all with debugging off |
| 22:21:51 | GoRK | spacecoater: in tuxphone) |
| 22:22:23 | GoRK | SpaceCoaster: anyway when you push the button you should get dialtone on speakerphone and the led should turn on |
| 22:22:42 | GoRK | SpaceCoaster: with debug printing to stderr the dialtone comes up but the led doesnt show |
| 22:22:51 | SpaceCoaster | ok, 2.4.13-ac5-rmk2-tux2, installed |
| 22:23:45 | prpplague | BZFlag: you done anything with uclinux? |
| 22:23:50 | SpaceCoaster | speakerphone worked first time but then DSP hung until reset. |
| 22:24:49 | SpaceCoaster | GoRK: what was that problem with mount, mine worked fine |
| 22:25:13 | GoRK | hmm are you nfsroot or jffs2 root? |
| 22:25:38 | GoRK | busybox mount -a doesnt seem to behave properly |
| 22:25:47 | SpaceCoaster | I use pivot_root to switch to nfs |
| 22:26:01 | GoRK | yeah then you dont have a problem |
| 22:26:12 | SpaceCoaster | cool |
| 22:26:28 | GoRK | if you want to use jffs2 root you have to comment out the line in your /etc/fstab |
| 22:27:20 | SpaceCoaster | ok so you are happy with the 1.16 version of wheaties.c but nothing since, is that correct? |
| 22:27:33 | GoRK | well hang on |
| 22:28:04 | GoRK | actually i was using 1.15 |
| 22:28:29 | GoRK | is this on cvs somewhere that i can explore diffs? |
| 22:29:31 | SpaceCoaster | only the latest is on cvs, I can give you my wheaties.c,v from my CVS, it has my development history. |
| 22:29:59 | GoRK | well im fairly sure some of the delays need to be reinstated but not sure which ones |
| 22:30:04 | GoRK | is udelay blocking? |
| 22:30:09 | SpaceCoaster | yes |
| 22:30:18 | GoRK | is it possible to delay without blocking? |
| 22:31:16 | SpaceCoaster | yeah you can delay in multiples of 10m (jiffies) but not in the interrupt handler. |
| 22:31:23 | SpaceCoaster | Is this a new problem? |
| 22:31:28 | SpaceCoaster | With the new kernel? |
| 22:31:48 | GoRK | is what a new problem? |
| 22:31:54 | SpaceCoaster | the hang |
| 22:32:10 | SpaceCoaster | the DSP losing sync and hanging tuxphone |
| 22:32:44 | GoRK | i am not having it |
| 22:35:49 | SpaceCoaster | ok let's describe it as "push the button you should get dialtone on speakerphone and the led should turn on", that problem has it always happened for you or is it just happening with the new kernel? |
| 22:38:46 | GoRK | well i always used 1.15 on the old kernels |
| 22:40:08 | SpaceCoaster | and did 1.15 always work? |
| 22:41:14 | GoRK | yes |
| 22:41:48 | GoRK | there should be a happy medium somewhere though |
| 22:42:00 | SpaceCoaster | ok could you put your 1.15 into CVS on TuxScreen and we will work from that |
| 22:42:32 | prpplague | well, look like i'm 4 out 6 today, telco's got the last 56k circuit screwed up |
| 22:42:42 | prpplague | heads home for some sleep |
| 22:42:50 | SpaceCoaster | prpplague: see ya |
| 22:43:06 | prpplague | later |
| 22:43:20 | GoRK | SpaceCoaster: will it even build as a module??!? |
| 22:43:37 | GoRK | oh yeah i guess it will |
| 22:43:40 | SpaceCoaster | GoRK: should do |
| 22:45:11 | GoRK | wheaties_init: $Revision: 1.15 $ installed on new kernel |
| 22:45:26 | GoRK | ooh yeah rocks |
| 22:45:37 | GoRK | ok ill put it on cvs |
| 22:46:36 | SpaceCoaster | GoRK: can you rev it as 1.15? |
| 22:47:42 | GoRK | i dont know how to do that |
| 22:48:06 | SpaceCoaster | me neither, commit -r1.15 ? |
| 22:48:49 | SpaceCoaster | the manual says that should "Commit to revision." |
| 22:51:50 | GoRK | the new audio driver doesnt skip when tuxphone is going either |
| 22:52:13 | GoRK | that is of paramount importance |
| 22:53:36 | SpaceCoaster | did it skip with the crap wheaties.c? |
| 22:54:08 | GoRK | the crap one being the new low delay one? |
| 22:54:12 | GoRK | no it didnt |
| 22:54:21 | GoRK | but i was afraid the longer delays might cause it to skip |
| 22:55:34 | SpaceCoaster | Gonna build my a temporary shanip2 tonight so audio tomorrow! |
| 22:55:52 | GoRK | cool :) |
| 22:57:29 | SpaceCoaster | 1.15 has udelay(10 * 1000); |
| 22:57:29 | SpaceCoaster | in write_messgaa |
| 22:57:34 | SpaceCoaster | write_message |
| 22:58:12 | GoRK | wow |
| 22:58:21 | GoRK | is that the only difference? |
| 22:58:26 | SpaceCoaster | I guess the audio is buffering up lots and they are being sent out on interrupts. |
| 22:59:30 | SpaceCoaster | no, not the only difference, I didn't realise that the version you were using was so far behind mine, I thought I had sent you 1.16, maybe you didn't try it, you were pretty excited at all the stuff that started to work. |
| 23:00:01 | SpaceCoaster | GoRK: the interrupts will interrupt a udelay() |
| 23:01:05 | GoRK | oh ok. that is good then |
| 23:02:50 | SpaceCoaster | the interrupt handler in wheaties reads bits with udelay(10) for each bit. Ok, now we have a working baseline, eh? |
| 23:04:15 | GoRK | right |
| 23:05:15 | SpaceCoaster | GoRK: now we have a simple test (the speakerphone) I should be able to make progress on making reads and writes interrupt drive and avoid all major delays that way. Let's move onto the touchscreen, ok? |
| 23:05:37 | GoRK | SpaceCoaster: yes but speakerphone button without debug |
| 23:06:00 | GoRK | debug off in tuxphone that is |
| 23:06:03 | GoRK | ok touchscreen |
| 23:06:34 | GoRK | first to get ucb1x00 to do anything at all on our platform you have to add a machine_is_shannon() to the if clause at the bottom of drivers/misc/ucb1x00-core.c |
| 23:06:43 | GoRK | err wait mcp-sa1100.c |
| 23:07:46 | SpaceCoaster | ok, done. |
| 23:15:21 | GoRK | ok i think i got it committed properly |
| 23:15:56 | GoRK | it axed your copyright too... sorry :) |
| 23:16:30 | SpaceCoaster | looks good, no worries, just as is is good, it has a bug when used as a module, rmmod traps, try it. |
| 23:23:15 | GoRK | ill take your word for it: :) im taking off for the evening though... got to go to some parties :) i think someone is having a toga party too.. ill have to check that out lol |
| 23:23:22 | GoRK | wonder if otis day and the knights are playing |
| 23:24:35 | GoRK | ttyl |
| 23:24:46 | SpaceCoaster | have fun |
| 23:25:39 | GoRK | the sourceforge cvsweb diff viewer is telling |
| 23:25:57 | GoRK | probably everything except the delay line was bugfixes |
| 23:26:01 | GoRK | :) |
| 23:26:03 | GoRK | later |
| 23:35:24 | MacAtack | brb getting a better irc client |
| 23:37:15 | MacAtack | me |
| 23:41:01 | MacAtack | s/printer/print company/ |
| 23:41:08 | MacAtack | oops |