| 00:04.19 | *** join/#riftuidev Baanano (~chatzilla@104.166.221.87.dynamic.jazztel.es) |
| 00:53.18 | Ivnedar | my head hurts trying to work out crypto code in c# |
| 00:54.26 | Ivnedar | mostly because everything is a CStringT<wchar_t,ATL::StrTraitATL<wchar_t,ATL::ChTraitsCRT<wchar_t> > > |
| 01:17.49 | Imhothar | that's the MFC string class |
| 01:21.54 | Ivnedar | yeah |
| 01:22.03 | Ivnedar | just makes it a pain to read :) |
| 01:31.36 | Imhothar | just ignore all the stuff between <> |
| 01:31.44 | Imhothar | it isn't really important |
| 01:31.50 | Imhothar | not for you probably |
| 01:31.51 | Ivnedar | yup, understand that |
| 01:32.02 | Ivnedar | I could probably just strip it out to make the rest more readable |
| 02:06.43 | *** join/#riftuidev Imhothar (bcae88f2@gateway/web/freenode/ip.188.174.136.242) |
| 05:26.07 | Thorarin | ROFL, my girlfriend got a survey "because she hasn't played Rift for a while" |
| 05:26.35 | Thorarin | We were playing up to the point where we went to bed last night... |
| 05:26.59 | Ivnedar | first time in a while? |
| 05:27.04 | Ivnedar | or pretty regularly playing? |
| 05:27.20 | Thorarin | Every day.. |
| 05:27.30 | Ivnedar | was it a real survey? not a phishing one? |
| 05:27.38 | Thorarin | Yeah it's real |
| 05:28.36 | Thorarin | Reason for not playing Rift: Other - I need sleep ;) |
| 05:28.56 | Ivnedar | lol |
| 05:29.18 | Ivnedar | "I'm sorry that I did not continue to play beyond the 14 hours straight I played last night, I really needed to sleep" |
| 05:33.22 | Thorarin | Well, it was only 5 hours I think, but still :) |
| 05:55.52 | *** join/#riftuidev Alystin (~alystin@ip-5-146-170-181.unitymediagroup.de) |
| 06:07.03 | *** join/#riftuidev Alystin (~alystin@ip-5-146-170-181.unitymediagroup.de) |
| 06:07.21 | ZorbaTHut | I wonder if something's wrong with our survey code :V |
| 06:07.51 | Ivnedar | Nah, they were slacking off and stopping after 5 hours |
| 06:08.11 | Ivnedar | or it went to an old address that was registered to play long ago? |
| 06:08.27 | Ivnedar | wonder if I've gotten one, since I haven't played in weeks :( |
| 06:08.33 | Ivnedar | not by choice, so don't hate on me! |
| 06:09.59 | ZorbaTHut | if I had to guess, I'd have assumed the old-address thing |
| 06:11.10 | Ivnedar | I'm currently having a very bizarre experience |
| 06:11.24 | Ivnedar | I need to fix up some ldap linkages between our voicemail system and phone system |
| 06:11.40 | Ivnedar | I knew I'd done it before, so hunted around for any clues I left from last time |
| 06:11.54 | Ivnedar | found a document in my home directory describing it, thought I should put it into our wiki for next time |
| 06:12.01 | Ivnedar | found it was already there, in the place where I was about to put it |
| 06:12.28 | Ivnedar | so following that through, every time something odd would happen, like an error popped up in the process, the next line in the doco mentioned the error and explained why it happened |
| 06:12.41 | Ivnedar | having a convo with my past self is odd |
| 06:12.55 | ZorbaTHut | (insert aliens meme here) |
| 06:13.39 | Ivnedar | I'm not exactly doing things that are supported by cisco |
| 06:13.54 | Ivnedar | but since they can't fix the root cause, and I have a fix, I use it |
| 06:14.06 | Ivnedar | so some errors come back because the tools used didn't expect to be used in the way they are :) |
| 06:14.49 | ZorbaTHut | that reminds me of scaleform |
| 06:14.50 | ZorbaTHut | like |
| 06:14.50 | ZorbaTHut | all of it |
| 06:15.00 | Ivnedar | things like talking to the underlying database with raw sql as the tools provided build up select statements, but expose where it sends the select statement, so I can send update statements, and they work |
| 06:15.09 | Ivnedar | but the tool errors because no rows are returned |
| 06:15.53 | Ivnedar | I'm odd in that I enjoy working these kinds of things out :) |
| 06:16.03 | Ivnedar | I have issues with taking "It's not possible" as the final answer |
| 06:23.23 | ZorbaTHut | apparently the rendering engineers before me had a tendency to respond to difficult things with "that's not possible" |
| 06:23.55 | ZorbaTHut | I've surprised our artists and designers by saying "*everything* is possible, given the engineer budget, cpu budget, gpu budget, and memory budget - if you can justify it to the project lead, we'll make it happen" |
| 06:24.06 | Ivnedar | pretty much my response too :) |
| 06:24.24 | ZorbaTHut | as a result I now have a few wacky "this will make our lives so much easier" projects to tackle, but so it goes :V |
| 06:24.41 | Ivnedar | sounds very very familiar |
| 06:25.12 | Ivnedar | currently working on a 3rd party library to talk to microsoft's office365 password sync interface |
| 06:25.50 | Ivnedar | so we can send password hashes to them instead of plaintext passwords |
| 06:25.54 | Ivnedar | it's fun :) |
| 06:59.11 | Alystin | morning all :) |
| 07:00.31 | Alystin | Cairenn: is there some way to get the addons on riftui programmatically or do i have to parse the public html pages? |
| 07:00.48 | Alystin | ie xml lists other that lastest.xml |
| 07:01.09 | Cairenn | dolby ^^ ? |
| 07:02.27 | Alystin | ? |
| 07:02.54 | Cairenn | I'm not sure if he has anything else set up on RiftUI to do that or not, so asking. ;) |
| 07:03.22 | Alystin | ah i see^^ |
| 07:03.26 | Cairenn | :) |
| 07:07.36 | Alystin | dunno if it was even possible to parse all the stuff... i have no idea how minion would do it |
| 08:50.07 | tubsawubs | Hello, anyone around? |
| 09:00.42 | Ivnedar | yup |
| 09:07.15 | Alystin | yeah, kinda |
| 09:07.36 | Alystin | how would i reference something, opposed to copy the whole thing? |
| 09:11.22 | Ivnedar | what object type? |
| 09:11.30 | Alystin | table |
| 09:11.39 | Ivnedar | that's what it does by default |
| 09:12.02 | Alystin | oh, cool |
| 09:12.10 | Ivnedar | a = {}; b = a |
| 09:12.21 | Ivnedar | a and b both point at the same table |
| 09:12.45 | Alystin | i'm confusing myself so much >.< |
| 09:15.04 | Alystin | I have a table with quest ids which i want to display in a list with two buttons next to it... |
| 10:01.37 | tubsawubs | Hi, given people are alive now; there's no uv/texcoord support as far as I can see - is there any way to use "mask" objects with arbitrary shapes (non-rectangles) to get some of the way there? |
| 10:03.14 | Ivnedar | not non-rectangle |
| 10:05.20 | tubsawubs | Otherwise is the "accepted" way to do spinny-cooldown-indicators to have progressively more "progressed" textures as an alpha overlay? |
| 10:05.23 | tubsawubs | My god those words were bad. |
| 10:05.52 | Ivnedar | pretty much |
| 10:06.10 | Ivnedar | pretty horrible, the new canvas frame type that's coming down the track may help there |
| 10:06.17 | Ivnedar | but at the moment, that's your only option |
| 10:06.28 | tubsawubs | yuck, so for a 15sec cooldown/duration I need, say, 60 odd textures? :< |
| 10:06.41 | Ivnedar | depends on how smooth you want it |
| 10:06.48 | tubsawubs | I can blend between them |
| 10:06.49 | Ivnedar | http://forums.riftgame.com/technical-discussions/addon-api-development/342768-spiral-texture.html |
| 10:06.58 | tubsawubs | but .25 sec granularity is pretty rough already |
| 10:06.58 | Ivnedar | older convo possibly exactly what you're looking for? |
| 10:07.35 | Ivnedar | with how gadgets used to do it |
| 10:07.41 | Ivnedar | mask frames are pretty expensive though |
| 10:07.45 | tubsawubs | Ah yeah, was gonna ask about texmaps |
| 10:07.50 | tubsawubs | but tbh individual textures are fine |
| 10:08.43 | tubsawubs | Does :SetTexture("addon", foo) force a load or is there an internal cache?/ |
| 10:10.16 | Ivnedar | from memory it's loaded on first request then cached |
| 10:10.39 | tubsawubs | Ah excellent, at least that's something I don't need to do from scratch. :-) |
| 10:11.37 | tubsawubs | PS: Does the king of "clickboxhealer" ever appear in this channel? |
| 10:12.19 | Ivnedar | yeah |
| 10:12.39 | tubsawubs | I want to add proper support for hots and debuffs like weakened soul to it (so you can see the cooldown/duration, not just a binary "yes/no"), but dunno how open people are to contributions. |
| 10:13.20 | Ivnedar | cbh is Solsis isn't it? |
| 10:13.34 | tubsawubs | That's what the curse page says, yes. |
| 10:14.54 | Ivnedar | last on 4-5 days ago |
| 10:15.33 | tubsawubs | Okey dokey. |
| 10:16.45 | tubsawubs | (Tell me to shut up if I'm being annoying), is there way to reload addon[s] without shafting my chat? |
| 10:17.25 | Ivnedar | turn on /log and read back chat through that before using /reloadui is the only thing that comes to mind |
| 10:17.51 | Ivnedar | or use the mobile client to watch guild chat while reloading ui and losing in game history |
| 10:18.10 | tubsawubs | Mobile client hates winphone, alas. |
| 10:18.14 | tubsawubs | Logging seems fine. |
| 10:18.21 | Ivnedar | It hates most phones :D |
| 10:18.29 | Ivnedar | been pretty unusable for the last few weeks |
| 10:18.52 | tubsawubs | Presumably I can write a small addon to load the last X lines of chat from a log and spam them to the console? |
| 10:18.59 | tubsawubs | Or is there no-no file access? |
| 10:19.21 | Ivnedar | no file access |
| 10:19.24 | tubsawubs | bah |
| 10:19.47 | tubsawubs | So you can't dump logs anywhere? |
| 10:19.56 | tubsawubs | Other than the persistant storage provided by RIFT? |
| 10:20.24 | Ivnedar | only other thing is spam the logs to the chat console and log that |
| 10:23.43 | tubsawubs | Ahh this explains why none of the dps meters do accumulated data. Though that should still work in-memory right? |
| 10:24.07 | Ivnedar | you can push history into saved variables and reload that, but usually the data isn't that useful |
| 10:24.25 | Ivnedar | or it builds up way too quickly to be of any use |
| 10:25.03 | tubsawubs | I thought there was no on-demand load/save for variables, unless I'm reading an out of date forum post. |
| 10:25.29 | Ivnedar | saved variables are loaded in addon startup and saved in addon shutdown |
| 10:25.48 | Ivnedar | the ones declared as saved variables in your .toc file |
| 10:25.53 | tubsawubs | *nods* |
| 10:26.34 | tubsawubs | Mm, you don't need history of every fight, just a rolling total (so I can see overall damage/healing/etc), which only doubles the data for N fights, not N times the data. |
| 10:27.24 | tubsawubs | As an aside, why does everyone hate cooldown bars (in favour of Karuul)? |
| 10:27.41 | Ivnedar | yup, could do that, not sure of knowing how much damage total you've done since you started playing rift would help, but may do for someone :) |
| 10:28.09 | Ivnedar | they usually takes up too much screen realestate, they're added movement on the screen that you don't need |
| 10:28.46 | Ivnedar | I want to know a buff/ability needs attention just before it needs attention, not for the duration of it being up |
| 10:32.09 | Ivnedar | addons tend to start out as a "I want x, I wonder if I can get x, I can get x, I wonder if anyone else wants x" |
| 10:34.05 | tubsawubs | True dat |
| 10:34.27 | tubsawubs | http://hoggerandfriends.com/a2.jpg I like cooldown bars because it shows me "priority" |
| 10:34.57 | tubsawubs | Like, "Do I burn a GCD on refreshing a dot or hitting my big hitter?" |
| 10:35.20 | tubsawubs | So I need them in absolute relation, rather than relative relation. |
| 10:35.28 | Ivnedar | fair enough :) |
| 10:35.42 | tubsawubs | All the bars cap at Nseconds (8 is my preference). I only care about the order of things in the next 8 seconds. |
| 10:36.01 | tubsawubs | More useful for some souls/classes than others I guess. |
| 10:36.11 | Ivnedar | yeah, and play styles |
| 10:36.23 | Ivnedar | some people just want to mash buttons, others want a rotation that they know by heart |
| 10:36.32 | tubsawubs | PS: Sorry about all the noob questions - I write graphics drivers for a living, I'm not really used to *consuming* APIs. |
| 10:36.51 | Ivnedar | not a problem, we've all been there, or are still there :) |
| 10:37.08 | tubsawubs | Especially not used to crazy lua. :D |
| 10:37.59 | tubsawubs | I fell in love with cooldown bars playing feral dps when it was too much stuff to track in my head, kinda just use them for everything now. (I say cd bars, cd/buff/debuff bars). |
| 10:38.18 | tubsawubs | Oh, not damage total since you started playing, but for example, since you started a dungeon |
| 10:38.32 | tubsawubs | Gives me a feel for how balanced my boss/trash damage is. |
| 10:38.38 | Ivnedar | *nod* |
| 10:41.59 | Alystin | raises hand at the "still be there"-part |
| 10:42.38 | Ivnedar | me too, most of the time :) |
| 10:42.44 | tubsawubs | Still-be-there? |
| 10:42.52 | Alystin | it's so frustrating to have something in mind and just not getting it right |
| 10:42.54 | Ivnedar | noob stage |
| 10:43.31 | tubsawubs | Ahhh. |
| 10:43.36 | Alystin | i even gave up on the optimizing your code part for now |
| 10:44.36 | tubsawubs | No need to optimize until you need to. |
| 10:44.43 | Alystin | i guess |
| 10:44.57 | Alystin | lua is pretty self-optimizing anyways^^ |
| 10:45.08 | tubsawubs | Until WATCHDOG or your favourite product manager starts crying, I code slow and simple. |
| 10:45.48 | Alystin | anyone has some experience with LibSimpleWidgets? |
| 10:46.02 | tubsawubs | I suffer from NIH so, sorry, no. :-( |
| 10:46.10 | Alystin | especially the scrollView-part - can't get it right |
| 10:46.13 | Alystin | NIH? |
| 10:46.18 | tubsawubs | Not Invented Here. |
| 10:46.22 | Alystin | xD |
| 10:58.23 | tubsawubs | I hate github with a passion. :-( |
| 11:26.35 | Alystin | hm. Stuff with the same layer should be visible in the order it was created, right? |
| 11:26.56 | Alystin | so if i place a frame and then a text, i should be able to see the text? |
| 11:29.35 | Alystin | ah nvm |
| 11:39.40 | Alystin | are all the Inspect.* commands bad to use? |
| 11:40.17 | Alystin | ie. would it be bad to use it in a loop to request quest titles on the fly rather than saving them in a table? |
| 11:41.14 | Ivnedar | I still don't know the answer to that |
| 11:41.54 | Ivnedar | depends on how often you're doing it I guess |
| 11:42.09 | Ivnedar | in your case, it'd be fine to loop over them when you first see them for that session |
| 11:42.19 | Ivnedar | and just keep track of them from that point, cached yourself |
| 11:47.01 | Ivnedar | just saying how I'd do it :D |
| 11:58.08 | Alystin | that's cool :D |
| 11:58.48 | Alystin | Command.Event.Attach(Event.Quest.Abandon, recheck, "QuestAb") would call the function recheck(element), right? |
| 11:59.26 | Ivnedar | recheck(handle, elements) |
| 11:59.38 | Alystin | ugh friggin' handle |
| 12:00.00 | Ivnedar | elements being a {qXXXXXXXX... = {quest table}} table |
| 12:01.01 | Alystin | my only fault was that i missed the handle and tried to read that^^ |
| 12:25.19 | Alystin | ugh... How would i embed a function in a button without writing a whole function body |
| 12:27.09 | Alystin | oh nvm, i had a mistype >.< |
| 12:35.50 | Alystin | i still haven't found a way to remove frame lements without redrawing the whole thing though :/ |
| 13:13.52 | Ivnedar | oh far out, finally tracked down what was chewing up all my 3g data |
| 13:14.24 | Ivnedar | skype was installed, apparently it's a pretty constant 4kbps when not in use |
| 13:14.42 | Ivnedar | no wonder my quota was being flattened :( |
| 13:14.45 | tubsawubs | :( |
| 13:14.48 | tubsawubs | Stupid mobile data. |
| 13:17.23 | Ivnedar | I've been wracking my brain trying to work out what I'd missed when I started using mobile data |
| 13:18.02 | Ivnedar | ended up using ~100mb/day when it's connected, 2gb quota doesn't last long if I'm trying to do other things on top of that |
| 13:26.03 | tubsawubs | I'm on "all you can use" at the moment, but I probably don't use much more than a few mb a month :-( |
| 13:26.15 | tubsawubs | Waste of money, only a few months of contract left though. |
| 13:27.02 | Ivnedar | I only have a few more days to survive until I get real internet again |
| 13:46.31 | *** join/#riftuidev Thinkofdeath (~Thinkofde@198.199.127.128) |
| 13:51.23 | tubsawubs | Have you recently moved house? |
| 13:54.00 | Ivnedar | yeah, a few weeks ago |
| 13:54.23 | tubsawubs | Ah, yeah, painful when that happens. |
| 13:54.32 | tubsawubs | But then you get fresh new internet and everything is happy happy happy. |
| 13:54.41 | Ivnedar | looking forward to that bit |
| 13:55.11 | Ivnedar | I do a lot of work from home, so trying to manage a relatively small quota is 'fun' |
| 14:28.19 | tubsawubs | yuck, even at 150ms granularity the textures look a bit |
| 14:28.22 | tubsawubs | jumpy |
| 14:40.32 | tubsawubs | Do you reckon I can load 100 16x16 textures in a single event before WATCHDOG shouts at me? |
| 15:30.54 | Alystin | the only way to "destroy" a frame is =nil, right= |
| 15:31.33 | Ivnedar | that only dereferences it, it still "exists" until next reloadui |
| 15:32.07 | Alystin | hrm |
| 15:32.23 | Alystin | so how would i remove a frame from the screen without just hiding it? |
| 15:32.59 | Ivnedar | what's wrong with hiding it? |
| 15:33.38 | Alystin | it's lingering^^ |
| 15:33.53 | Alystin | and i guess it would take up space where something else should be |
| 15:34.34 | Ivnedar | set its position to be off the screen somewhere until you reuse it? |
| 15:34.44 | Ivnedar | not sure how it'd take up space |
| 15:35.02 | Alystin | hm, that would definitely work |
| 15:35.20 | Ivnedar | or is it making a mouse click dead spot where it was? |
| 15:35.29 | Alystin | dunno, gotta try |
| 15:35.29 | Ivnedar | if so, change its mouse interaction state before/after hiding it |
| 15:36.00 | Alystin | my code is a bloody mess -.- |
| 15:37.29 | Ivnedar | I think I need to uninstall solitaire some time |
| 15:37.45 | Ivnedar | 52004 games played since last stat reset |
| 15:38.40 | Ivnedar | it's after 1am, so I'm early to a work web call thing that starts at 2am, so churning through games while I wait |
| 15:40.01 | tubsawubs | 1am? |
| 15:40.11 | Ivnedar | 1:40am now |
| 15:40.19 | tubsawubs | From what colony do you hail? :S |
| 15:41.14 | tubsawubs | Australia? |
| 15:41.18 | Ivnedar | yeah |
| 15:41.37 | tubsawubs | Well you might have won this test |
| 15:41.40 | Alystin | hm. i'll have to move up stuff to fill the gap... |
| 15:41.40 | tubsawubs | BUT THE ASHES REMAIN OURS |
| 15:41.47 | Ivnedar | lol |
| 15:41.48 | tubsawubs | bwahahaha |
| 15:43.08 | tubsawubs | Alystin: don't worry about cleaning up frames. If you really need to, maintain a list of "deleted" frames, then have a button to reloadui and flush them. |
| 15:43.09 | Ivnedar | Alystin: what are you moving? |
| 15:43.18 | Ivnedar | eh, reuse them instead :P |
| 15:43.21 | Ivnedar | recycling ftw! |
| 15:43.26 | tubsawubs | Or that. |
| 15:43.41 | Alystin | well... kind hard to describe |
| 15:44.21 | Alystin | i have something like that: Window->ScrollView->Frame->List of frames with thext and buttons inside |
| 15:44.52 | Alystin | i need to remove the last element |
| 15:45.01 | Alystin | meaning a frame with some stuff in it |
| 15:45.45 | Ivnedar | each set of text/button is a quest? |
| 15:46.10 | Alystin | yep |
| 15:47.28 | Ivnedar | I'd not tightly couple quest to frame gset |
| 15:47.33 | Ivnedar | s/gset/set/ |
| 15:47.34 | Alystin | http://imgur.com/woFyg6d <- |
| 15:48.05 | Alystin | if i'd abandon the first quest, i'd need to move up the rest |
| 15:48.19 | Ivnedar | so when you update the list, set the text/questsharemacro on all the frames down the list |
| 15:48.24 | Ivnedar | since it's done that infrequently |
| 15:48.46 | Ivnedar | and then use the frames as display only, not state |
| 15:48.54 | Ivnedar | if that makes sense |
| 15:48.59 | Alystin | not really^^ |
| 15:49.21 | Ivnedar | I'd keep the real quest state structure in a seperate table |
| 15:49.23 | tubsawubs | "Re-assign quests to frames every time you add/remove a quest"? |
| 15:49.40 | Ivnedar | then have a function that pushes that table into the display list |
| 15:49.51 | Ivnedar | adding new , adding new frames to the end if it grows, or hiding frames if it doesn't |
| 15:49.56 | Alystin | ah, so essentially rebuild the frame |
| 15:50.29 | Ivnedar | so see a frame as "quest 1" instead of "qB36267272323" |
| 15:50.51 | Ivnedar | and whatever the current "quest 1" is in your table is pushed into the text/macro for "quest 1" |
| 15:51.40 | Alystin | http://pastebin.com/UKavLyW7 |
| 15:53.12 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 15:53.13 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 15:53.59 | Alystin | so actually a frame would be items[1] i.e. |
| 15:54.09 | Alystin | with btns and stuff |
| 15:56.22 | Ivnedar | so I'd split it into two sections. Init code to create the scrollview cradle, and update code that takes a list of quests and manages the rows |
| 15:57.01 | tubsawubs | seebs = the api king? |
| 15:58.22 | Ivnedar | so whenever there's a change to quests, you can push the new list into that function, and it either creates new rows if more quests are added, or hides old rows if quests are removed |
| 15:58.26 | seebs | Not sure what an API king is, but I'm not the developer or anything, I just made an addon that makes an HTML copy of the docs. |
| 15:59.39 | tubsawubs | The king of the API. Keeper of the functions. Master of enums. |
| 15:59.43 | Ivnedar | Alystin: I've never worked with LibSimpleWidgets, so that may not be the best way to do it |
| 16:04.34 | Alystin | the only thing i need it for is the non-deformed window and the scrollview, since i've got no idea how to realize it with the api stuff |
| 16:06.55 | seebs | Nope, I think that's Zorba. |
| 16:07.46 | tubsawubs | Hmm, but seeeeebs.net is where I go to ponder themysteries of the API? |
| 16:24.11 | seebs | That's just the output of my DocExport addon. I maintain(ed) a copy as a convenience thing. |
| 16:26.40 | ZorbaTHut | tubsawubs, ironically, I've been taking a break from writing the addon API in order to do a lot of rendering work on Rift. I feel like we've swapped places :P |
| 16:27.39 | Ivnedar | The real API king is off fighting a holy war |
| 16:32.00 | tubsawubs | ZorbaTHut: rendering as in addon rendering or game rendering? |
| 16:32.08 | ZorbaTHut | tub, game rendering |
| 16:32.21 | ZorbaTHut | my main task right now is "make rift run faster", and a good chunk of the issues have been rendering issues |
| 16:32.43 | tubsawubs | ZorbaTHut: I know all about making games run faster. Unfortunately. I implement OpenGL drivers. :-( |
| 16:32.56 | ZorbaTHut | in fairness there's a lot of similarity at this point |
| 16:33.31 | tubsawubs | Aye, indeed. Though possibly not for us (I do mobile graphics). |
| 16:33.40 | ZorbaTHut | ah, that starts reducing similarity quickly :) |
| 16:33.51 | ZorbaTHut | mobile devices don't exactly love deferred rendering |
| 16:33.55 | ZorbaTHut | (yet, at least - maybe someday) |
| 16:33.57 | tubsawubs | Well |
| 16:33.58 | tubsawubs | Hilariously |
| 16:34.07 | tubsawubs | You guys stole that phrase from us. :-( |
| 16:34.38 | tubsawubs | PowerVR, "tile based deferred rendering" since forever. But that's a totally different type of "deferred". |
| 16:34.44 | ZorbaTHut | hah, really? I suspect we didn't even know about it - was it largely a graphics-driver-developer-internal phrase? |
| 16:35.49 | tubsawubs | Yep. It just leads to hilarious conversations in the office when we get the wrong meaning nowadays. :-) |
| 16:36.00 | ZorbaTHut | welp. sorry about ganking your phrase :D |
| 16:37.19 | tubsawubs | You should be more sorry I don't have an excuse to RIFT at work, being evil DX that you are. |
| 16:38.23 | ZorbaTHut | I'd actually love to add opengl support so we could in theory harness Google's NaCl. or OpenGL/DX10 support so we could take advantage of some newer graphics card features |
| 16:38.41 | ZorbaTHut | one thing I *love* about OpenGL is that you don't need to rewrite your entire rendering layer to use new flashy options |
| 16:38.53 | tubsawubs | You make it sound like Trion only has a few people working on the rendering side? |
| 16:39.46 | ZorbaTHut | well, uh . . . that's sort of inaccurate . . . "a few" would be an overestimate. It's me. I'm the rendering team. :P |
| 16:40.09 | tubsawubs | Madness. How big is Trion? :s |
| 16:40.25 | ZorbaTHut | not anywhere near as big as you'd think, actually - it's a compact company that hires *very* good people |
| 16:40.30 | ZorbaTHut | (I don't know exact numbers) |
| 16:40.44 | ZorbaTHut | there's one other guy who does some shader development, but he also spends a lot of time on vfx |
| 16:41.12 | seebs | Rift team's not huge, especially not after the layoffs, but they do seem to get a lot of things right in terms of development/content-design splits, and efficient development processes. |
| 16:42.35 | ZorbaTHut | one thing that helps a lot is that we tend to hire people who can work under their own guidance, and then hire management who don't micromanage. I think I get maybe one actual instruction from management per week, roughly, and it's usually just bumping around priorities. |
| 16:42.57 | ZorbaTHut | half the time it's me going to them and saying "hey, order these priorities for me plz" |
| 16:43.29 | seebs | My boss (not at Trion) is awesome. |
| 16:45.35 | Baanano | didn't imagine you were the only one working on rendering :S what about that guy that got bashed for announcing the pixel granularity feature? |
| 16:45.45 | seebs | I asked him what to focus on, once, and he said "focus your priorities across the board". |
| 16:46.04 | ZorbaTHut | he was actually just a temp on rift rendering, he's moved to other trion projects (that's what he was really hired for) |
| 16:46.20 | seebs | He then waited a minute or two and gave a real answer. |
| 16:46.52 | seebs | At one point, the VP of our department was another guy in the same small office I was in, and he'd take us out to lunch occasionally. |
| 16:46.55 | *** join/#riftuidev Kaelten (~Kaelten@WoWUIDev/WoWAce/WoWIFA/CurseStaff/kaelten) |
| 16:46.55 | *** mode/#riftuidev [+o Kaelten] by ChanServ |
| 16:47.02 | ZorbaTHut | oh hey, actually, looks like it was pithos who announced it? that guy was never rendering, he's just our engineering lead |
| 16:48.18 | seebs | He at one point told me that my boss has the unique attribute that the people reporting to him have the best productivity in the department. |
| 16:48.23 | seebs | Doesn't matter who's reporting to him. |
| 16:48.35 | seebs | Shuffle people around, and within a month, the people reporting to him have the best productivity. |
| 16:49.38 | tubsawubs | Hrrrm |
| 16:50.25 | seebs | He's very good at extracting information from employees about what they need to get their jobs done, and keeping everyone else from bugging us so we can do it. |
| 16:50.51 | tubsawubs | Is performance that bad in RIFT? I never noticed framerates dropping too far, and I don't have the fanciest rig. |
| 16:51.35 | ZorbaTHut | they're lower than they should be - there are a lot of inefficiencies |
| 16:51.44 | tubsawubs | seebs: Indeed, a good boss is probably the most important aspect of a job, imo. |
| 16:52.09 | seebs | Varies a lot, but basically, it gets noticably worse performance than you'd expect on a lot of rigs, and when a lot is happening, GPU usage can be <10%. |
| 16:52.18 | tubsawubs | Ouch. |
| 16:52.22 | seebs | Also, in my limited experience, it's moderately pathological on ATI hardware. |
| 16:52.24 | tubsawubs | GPU usage down? Blame the driver! |
| 16:52.37 | tubsawubs | I have an AMD gpu, something decent but not too modern. |
| 16:52.37 | ZorbaTHut | sadly, the "gpu usage <10%" isn't going away in 2.4 - but I do know what's happening and it'll be a high priority for 2.5 |
| 16:52.50 | seebs | I used to have a Radeon-based desktop and an nVidia-based laptop. They got basically comparable performance in Rift. |
| 16:53.11 | seebs | In every other game that could knock them under 60fps, the desktop was about 2x-3x faster than the laptop. |
| 16:53.15 | ZorbaTHut | it actually wouldn't surprise me if 2.4 fixes *that* issue. No promises. |
| 16:53.27 | seebs | Cool! |
| 16:53.34 | tubsawubs | ZorbaTHut: if you're Mr Renderer, does that mean work on the addon API is somewhat frozen? |
| 16:53.43 | seebs | I mean, the direct impact is pretty minimal on me at this point, but I appreciate it in principle. :) |
| 16:53.57 | ZorbaTHut | tub, right now, yes :/ I'm trying to get more time for it but right now the rendering stuff is more important |
| 16:54.05 | ZorbaTHut | it's definitely not abandoned, just frozen in metaphorical carbonite. |
| 16:55.01 | tubsawubs | Well, it gets the job done, the only issues I'm running into are the lack of rendering calls, which are work-aroundable. |
| 16:55.21 | ZorbaTHut | that one actually is improved in 2.4 - there's a new frame type showing up called Canvas that allows somewhat-arbitrary geometry |
| 16:55.22 | Alystin | Zorba you make it sound like the Rift team is around 5.10 ppl^^ |
| 16:55.27 | Alystin | *5-10 |
| 16:55.36 | tubsawubs | Aye, Ivnedar mentioned Canvas. |
| 16:55.54 | ZorbaTHut | it's not an opengl context or anything like that, but it's still a good deal of extra power |
| 16:55.58 | Baanano | sometimes i feel it's our fault (for not having produced enough high quality addons) that the addon api seems to be one of the lowest priorities :( |
| 16:56.10 | tubsawubs | Does that have something like an "Event.Canvas.Draw" we can subscribe to with some basic primitives like "drawLine"? |
| 16:56.15 | Alystin | wel, at least it gets the job doen pretty good |
| 16:56.29 | ZorbaTHut | baanano, eh, it's hard to say. I suspect if there were a lot of really amazing addons it'd be a bit higher priority, but of course it's really tough to *make* really amazing addons with holes remaining in the addon system |
| 16:56.44 | tubsawubs | ZorbaTHut: I am gonna try my hardest with that, at least. :> |
| 16:56.52 | ZorbaTHut | tub, not event-based, you set geometry on it and it draws that. changing the shown geometry is unfortunately kind of expensive |
| 16:57.15 | tubsawubs | Ah so a render-to-texture then display? |
| 16:57.15 | ZorbaTHut | I'm thinking of adding a geometry-compile step that does the parsing and stores it in a faster structure that can be used on demand, but we'll have to see if it's worthwhile |
| 16:57.35 | ZorbaTHut | more like "set geometry to render, underlying software does something black-box with it" :V |
| 16:57.44 | tubsawubs | Sounds like WPF :x |
| 16:57.51 | ZorbaTHut | Scaleform, really |
| 16:57.52 | tubsawubs | Or at least, a particular rendering mode in WPF |
| 16:58.34 | tubsawubs | From what I've poked around with so far, uv coords (still with rect targets) would go a long way round most issues. |
| 16:58.41 | seebs | Trion's pretty decent about dev priorities, although I think there have been times when "catch up on fixing bugs" was not as high a priority as I'd have liked. |
| 16:58.58 | ZorbaTHut | annoyingly I actually can't provide exact UV coords - instead, Canvas allows a generic 3x2 matrix transformation to be applied to textures |
| 17:00.00 | tubsawubs | Will that allow for texmaps or just arbitrary scale/rotate? |
| 17:00.27 | Baanano | hope for rift 2 they let you design and implement the ui layer and get rid of scaleform :P |
| 17:00.37 | tubsawubs | RIFT 3, surely? :p |
| 17:00.40 | ZorbaTHut | it's basically arbitrary scale, rotate, skew, translate, and crop-to-arbitrary-geometry |
| 17:01.14 | tubsawubs | Better than nothing *shrug* |
| 17:02.45 | tubsawubs | Also are there memory/frame constraints per-addon? |
| 17:03.27 | ZorbaTHut | if you crash the client, you used too much memory :V beyond that, nope |
| 17:03.42 | tubsawubs | Aha, good. |
| 17:03.55 | ZorbaTHut | I actually can't track per-addon memory usage, it would require massive changes to lua |
| 17:04.27 | tubsawubs | Wanting to do a dps analyser/visualiser, but given we can't do custom drawing, I have to use tons of frames for my bars. |
| 17:04.37 | ZorbaTHut | Canvas will absolutely help with that |
| 17:04.53 | tubsawubs | I thought updating the geometry would be painful? |
| 17:05.05 | ZorbaTHut | yes, but less painful than updating a ton of frames to display the same thing :V |
| 17:05.24 | ZorbaTHut | I'm saying "it's slow" in the sense of "if you can get away with using one or two frames or textures, you should probably just use the frames or textures" |
| 17:05.26 | tubsawubs | Unless I can create an absolutely giant canvas and scale/translate it to only show a subsection? |
| 17:05.49 | ZorbaTHut | but if you're using five or more, canvas is probably a better choice |
| 17:05.58 | Baanano | i'm excited about drawing charts using canvas |
| 17:06.17 | ZorbaTHut | I think what it comes down to is that people will have to play around with performance once it shows up :) |
| 17:07.31 | tubsawubs | Well, my last bug just got marked as passed, so it's home time (and by home time, I mean addon time). |
| 17:08.17 | tubsawubs | Gonna get my spinny cooldown indicators working. Though I have to use 100 tiny textures *shakes fist at ZorbaTHut* |
| 17:10.29 | *** join/#riftuidev Imhothar (bcae6f0e@gateway/web/freenode/ip.188.174.111.14) |
| 17:29.25 | *** join/#riftuidev tubsahubs (~t7@cpc20-heme11-2-0-cust99.9-1.cable.virginmedia.com) |
| 17:29.31 | tubsahubs | Bloody learner drivers. |
| 17:33.50 | tubsahubs | Actual question: "Event.Buff.Change" fires on cooldown/duration changing or just stacks etc? |
| 17:41.15 | Imhothar | when any value changes that's part of Inspect.Buff.Detail |
| 17:41.49 | tubsahubs | Cool, including remaining. Cheers. |
| 17:42.10 | Imhothar | stack changing works differently, there you get Event.Buff.Remove followed by Event.Buff.Add because a 2-stack buff is a different thing than th 3-stack version |
| 17:42.55 | tubsahubs | Ahh oki. I have it working but want to simplify by understanding exactly which events fire when. |
| 17:43.38 | Thorarin | Meh, why does RIFT show ping so inaccurately :P |
| 17:43.57 | Imhothar | because it can? |
| 17:44.06 | Thorarin | It times things with a resolution of 16ms, normal PC clock resolution :( |
| 17:44.19 | Thorarin | So my ping is always 16ms or 31ms :P |
| 17:44.33 | Imhothar | the client has to take 2 timestamps |
| 17:44.56 | Thorarin | But it's using the wrong API to do it :P |
| 17:45.44 | tubsahubs | tbh a 16ms resolution is fine for ping - why do you care if you have another 3ms or so? :x |
| 17:45.50 | Imhothar | the high performance counter would probably be a bit overkill |
| 17:46.28 | Imhothar | guess the guys didn't assume people's ping would be lower than the system clock resolution ;) |
| 17:48.32 | Thorarin | Not much lower on my cable.. |
| 17:48.51 | Thorarin | Hovers somewhere around 15 I guess |
| 17:50.52 | Imhothar | why even care |
| 17:51.48 | Imhothar | whether it's 15 or 30 doesn't change much |
| 17:57.13 | seebs | Oh, wow. |
| 17:57.27 | seebs | Someone in another channel just exposed me to Lua's "weak tables", which I did not know existed. |
| 17:58.40 | Imhothar | I think I use weak tables for my item detail cache, keyed by itemid |
| 17:58.55 | tubsahubs | I prefer monf tables. |
| 17:58.59 | Imhothar | with __mode="k" |
| 17:59.11 | Baanano | would those ever be collected imhothar? |
| 17:59.27 | Imhothar | they would when the itemid disappeared |
| 17:59.30 | seebs | Someone's doing a thing in one of the other Lua dev environments (moai) using a weak table as a resource cache, and things are getting gc'd from it at times when they are still in use by something else. My suspicion is that the "still in use" is in the C++ side and doesn't create an actual Lua "reference". |
| 17:59.56 | seebs | ... Oh, interesting. Apparently the key/value pairs still exist post-gc. That is odd. |
| 18:00.06 | Imhothar | it depends on the mode |
| 18:00.09 | Baanano | but if itemid is a string... i thought strings were never collected |
| 18:00.26 | Imhothar | everything is collected if its unused |
| 18:00.48 | Imhothar | but if either the key or the value is collected the entire enty is removed from the table |
| 18:01.57 | Baanano | so i understood it wrong, nice to learn something new :) |
| 18:02.11 | seebs | I do not understand the k/v distinction, if either is gc'd the entry is removed, right? |
| 18:02.13 | Baanano | i was using tables/userdata only |
| 18:02.30 | Imhothar | the mode decides whether the key or value keeps a reference to the object |
| 18:03.22 | Imhothar | setting mode to "k" makes the key a weak reference, so the table entry can only disappear if the key is collected, becasuse the value is still a strong ref |
| 18:03.33 | seebs | Ahh. |
| 18:03.50 | seebs | And v means that the entry can only disappear if the value is collected, and kv means it can disappear if either is collected. |
| 18:04.12 | Imhothar | aye |
| 18:04.19 | seebs | Well, strictly speaking, it will always disappear if either is collected, but "k" would mean that the value can't be collected ever. |
| 18:04.33 | Imhothar | correct |
| 18:06.07 | seebs | The guy is doing a thing with a weak table (__mode = "v") which is being used as a resource cache. |
| 18:06.21 | Imhothar | that's a bit weird |
| 18:06.27 | Imhothar | I'd rather make the key weak |
| 18:06.34 | seebs | And apparently, what is happening is basically: t.foo = loadResource("foo") |
| 18:06.55 | seebs | The intent then is that as long as t.foo is non-nil, the resource is still loaded, so you don't have to reload it, and can use it and pass it around as a reference. |
| 18:07.21 | seebs | And if the resource isn't being used by anything, it can be gc'd... And if it is, t.foo becomes nil, so the check for whether to load it or not will determine that it needs loading. |
| 18:07.23 | Imhothar | hmm yah that should work |
| 18:07.56 | seebs | The behavior observed is that, stepping through the code, he can see that the value has been gc'd, but the reference in the table is still present. |
| 18:08.06 | Imhothar | unless he screws up the C++ side |
| 18:10.03 | Imhothar | well the GC runs in cycles |
| 18:10.25 | seebs | ... Heh, he just observed that if he calls the collector a few times the problem disappears. |
| 18:10.31 | Imhothar | the default Lua GC is mark-and-sweep |
| 18:10.41 | seebs | So it may be there's an intervening state where the object is invalidated but the reference hasn't been pruned yet. |
| 18:11.18 | Imhothar | does he change the mode of the table multiple times? that can screw things up |
| 18:12.08 | Imhothar | vchanging __mode after it's been set is technically not supported |
| 18:12.22 | Imhothar | undefined behavior land |
| 18:13.25 | seebs | Thanks. |
| 19:28.07 | seebs | http://forum.square-enix.com/ffxiv/threads/63381-Forum-Design-and-Features-Update-%28Aug.-22%29 <-- this is possibly one of the most fascinating forum policies I have ever seen. |
| 19:32.13 | seebs | The 1K limit is obviously stupid, but the throttles based on level are... intriguing. |
| 19:49.43 | tubsahubs | hmm |
| 22:01.13 | Alystin | Good night, guys :) |
| 22:02.42 | *** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net) |
