**** BEGIN LOGGING AT Wed Jan 11 02:59:57 2012 Jan 11 03:20:12 nopers not sure what's wrong. Jan 11 03:28:18 hmmmmm Jan 11 03:28:23 well partial fix Jan 11 03:28:28 wait no Jan 11 03:28:29 hahaha Jan 11 03:28:30 sorry >.< Jan 11 03:28:34 bbl Jan 11 03:28:41 best i can think it has to do with kerning Jan 11 03:38:32 dtzWill, also why do u have to load the font 4 times Jan 11 03:56:54 everyone's a critic Jan 11 03:56:56 haha Jan 11 03:56:59 nah that's badness for sure Jan 11 03:57:06 b/c each has their own attributes Jan 11 03:57:20 but shouldn't happen, and some refactoring to fix that would be a good idea for sure Jan 11 03:59:02 (that's definitely bad, just something that didn't make the cut first time through) Jan 11 04:04:58 PuffTheMagic: what fonts have the properties you want on terminals you trust? Jan 11 04:08:41 okay so it seems like it for sure is kerning-related Jan 11 04:09:36 i can make the bars line up, for example, by telling SDL_TTF to render {'A', c, 'B', 0} and using the width/height to pull out 'c' from the resulting surface Jan 11 04:09:48 the problem is then some characters aren't adjusted right Jan 11 04:09:58 maybe the answer is to just render each glyph and use the glyph metrics to position things ourselves? Jan 11 04:11:29 ran into this earlier, it's why i could just tell SDL_TTF to render all the characters at once and use that for the whole row Jan 11 04:11:54 dtzWill: forget it Jan 11 04:12:11 I guess we will just support 1 font Jan 11 04:12:33 I don't see anyone needing to change it Jan 11 04:12:45 the one I hs Jan 11 04:12:53 the one I h Jan 11 04:12:57 grr Jan 11 04:13:16 the kne I have pushed now looks  perfect **** BEGIN LOGGING AT Wed Jan 11 05:01:37 2012 Jan 11 06:09:03 dtzWill, doubt u are around, but if u know what is causing the font loading issue that is all that is blocking a new release Jan 11 06:09:31 PuffTheMagic: i've never once seen it Jan 11 06:09:43 (so haven't spent any time debugging it, unfortunateyl) Jan 11 06:09:50 did you see what brybry said about open fd on the font? Jan 11 06:10:01 pull latest, run make test Jan 11 06:10:02 i don't use palm-install for testing, still just scp the wterm binary over Jan 11 06:10:07 then while the app is running Jan 11 06:10:08 do make test Jan 11 06:10:11 and it fails Jan 11 06:10:15 this never happened before Jan 11 06:11:02 and ya I see that Brybry says the filesize is 0 Jan 11 06:11:08 and this only happens when wterm is running Jan 11 06:11:24 but im not sure why the file is still open Jan 11 06:15:01 dtzWill, idk how to go about debugging this either :/ Jan 11 06:24:13 dtzWill, if u are rendering the fonts to a texture why do they have to remain open Jan 11 06:24:45 haha, they absolutely don't Jan 11 06:24:48 but they did before Jan 11 06:24:54 so i didn't go out of my way to close them Jan 11 06:25:04 "let's make this work, first" being a guiding principle as usual :) Jan 11 06:25:13 well, im not sure why it worked before Jan 11 06:25:20 but its def having a cow now Jan 11 06:25:52 well we're also opening the same font multiple times Jan 11 06:25:55 where before perhaps we did not Jan 11 06:26:03 yes that is also true Jan 11 06:26:17 TTF_GetError might show things Jan 11 06:26:18 I dunno Jan 11 06:26:23 i was also curious why we need to open it 4 times Jan 11 06:26:36 why cant it be opened once, and attributes set 4 times Jan 11 06:26:50 each time before it renders Jan 11 06:27:13 PuffTheMagic: lol Jan 11 06:27:30 "historical reasons" Jan 11 06:28:41 PuffTheMagic: because writing the sdlfontgl stuff (and the many iterations it took to get something that had the desired performance and fit our usage) filled my buffer and I didn't fix "all the things" related to it simultaneously Jan 11 06:28:57 but yes, we could definitely open fonts, set attrs, render into texture, and close Jan 11 06:29:06 hell we could not open fonts at all until they're needed and then do the same, etc Jan 11 06:29:15 that's all a minor refactoring away :) Jan 11 06:29:30 dtzWill, im not trying to harass, just trying to understand and debug Jan 11 06:29:32 nothing personal Jan 11 06:29:46 sorry, and thank you for the clarification. there's no good reason, is the answer :D Jan 11 06:30:17 i'm in the middle of actually defending myself in a different context ATM so my apologies for being defensive :) Jan 11 06:30:58 i'll poke at it in a bit and see if i can't shake anything loose :) Jan 11 06:31:16 its ok Jan 11 06:31:21 i will hack at this for a few more min Jan 11 06:34:25 also i'm grumpy i can't make the fonts work Jan 11 06:34:42 i can fix the alignment issue (the "right" way, but then there are gaps) Jan 11 06:34:52 who knew fixed-width fonts could be so obnoxious Jan 11 06:36:40 i bet its the memcpy of the fonts that is breaking shit Jan 11 06:36:53 well we're memcpy'ing pointers :( Jan 11 06:36:56 i mean Jan 11 06:37:12 :/ Jan 11 06:48:20 in clearGL() Jan 11 06:48:44 Brybry: ? Jan 11 06:48:47 what about it Jan 11 06:48:59 we free our copy of the fonts theree Jan 11 06:49:04 probably needs to be the TTF_Close function Jan 11 06:49:25 (on each pointer in the list) Jan 11 06:49:27 we free one copy of it there, don't we free them in sdlcore::closeFonts() too? Jan 11 06:49:47 the sdlglfont copy was never meant to own their lifetime Jan 11 06:49:50 I think close fonts just frees theoriginals Jan 11 06:49:56 * PuffTheMagic doesnt understand 1/2 of the new font code ;) Jan 11 06:50:08 memcpy on a pointer gives you the same pointer Jan 11 06:50:19 not lke we're asking TTF for a new copy or allocating anything Jan 11 06:50:34 PuffTheMagic: me eitehr ;) Jan 11 06:50:34 :P Jan 11 06:51:08 Brybry: sorry, i mean if we're leaking fonts something is wrong, and maybe i'm missing something. Jan 11 06:51:15 give it a go? Jan 11 06:51:22 I don't think we're leaking fonts Jan 11 06:51:51 I think SDL might be tracking a file a handle or something and we're losing that/it doesn't clean it up? I have no idea Jan 11 06:51:59 I've have to dig into the SDL_ttf code Jan 11 06:52:04 * PuffTheMagic doesnt understand 1/2 of the new font code ;) Jan 11 06:52:07 oops Jan 11 06:52:07 yaeh Jan 11 06:52:15 I see malloc and my eyes glaze over Jan 11 06:52:28 Brybry, thats the part that huts me too Jan 11 06:52:32 I'm like, "this isn't new and delete!" Jan 11 06:52:50 haha well the existing code uses malloc/free Jan 11 06:52:55 so i figured i'd follow suit :( Jan 11 06:53:29 dtzWill, havent u learned we should follow anything the past dev did ;) Jan 11 06:53:38 shouldnt Jan 11 06:53:45 haha Jan 11 06:53:46 well Jan 11 06:53:57 malloc/free aren't so bad Jan 11 06:53:58 lol Jan 11 06:54:19 Brybry, there are def files open somewhere Jan 11 07:00:52 * PuffTheMagic thinks createFonts() should be part of sdlfontgl Jan 11 07:01:07 and avoid all the memalloc shit Jan 11 07:26:06 so the problem is that when using palm-install to install the package while wterm is running, the font somehow doesn't get extract properly? Jan 11 07:26:22 strange that it can replace the binary, but not the font :( Jan 11 07:26:49 maybe it's b/c the font used to start with the letter 'A' so it'd get extracted first Jan 11 07:26:52 * dtzWill jokes Jan 11 07:44:59 you can do echo "blah" > font (or try to cp something over it and watch cp belch) to test it Jan 11 07:45:26 keep a copy of it around though ;) Jan 11 07:47:12 I couldn't get TTF_CloseFont to work, kept crashing and giving some null mutex error which means nothing to me Jan 11 16:24:09 dtzWill, just added blink support Jan 11 16:24:27 dtzWill, if u think there is a more efficient way to do it let me know Jan 11 16:24:45 and/or fix it ;) Jan 11 16:28:31 serious question: Jan 11 16:28:32 why? Jan 11 16:28:32 haha Jan 11 16:28:33 XD Jan 11 16:29:54 no, that looks fine to me Jan 11 16:30:02 \o/ Jan 11 16:31:37 off to a meeting, bbiab. haha glad you implemented it it's fun if nothing else :D Jan 11 16:32:42 dtzWill, completeness sake? Jan 11 16:33:22 dtzWill, i think it might be possible to set a higher level global flag which turns to 1 if there are are cells with blink enabled Jan 11 16:33:32 only then will the blink thread run Jan 11 16:33:35 else it waits Jan 11 16:34:51 idk that might be hard Jan 11 18:00:32 almost have autowrap tests fixed in vttest Jan 11 18:00:41 it just gets pissed off in 1 spot Jan 11 18:00:48 \o/ Jan 11 18:01:14 dtzWill, so u really werent going to add blink? Jan 11 18:01:53 PuffTheMagic: i hadn't planned /not/ to Jan 11 18:02:01 but wasn't spending time thinking about how to add that, no Jan 11 18:02:08 lol Jan 11 18:02:19 your comment seemed like u had already decided not to Jan 11 18:02:43 there has to be a way to improve it, so that the thread waits until its needed Jan 11 18:02:54 instead of always injecting an event ever .5s Jan 11 18:03:09 hahaha Jan 11 18:03:37 probably, yes. Jan 11 18:06:54 TerminalState::insertChar() is a mess Jan 11 18:29:05 dtzWill, how did your meeting go Jan 11 18:30:17 good, although i have no idea how i'll get this all done..... and classes haven't even started yet Jan 11 18:30:18 xD Jan 11 18:30:40 what program are you in that has qualifier exams? Jan 11 18:31:08 CS at UIUC Jan 11 18:31:20 graduate or undergrad? Jan 11 18:31:26 oh, phd program if that was what you were asking. Jan 11 18:31:32 ya ok Jan 11 18:31:37 im so glad id ont have qualifiers Jan 11 18:31:41 for my program Jan 11 18:32:18 im stressed enough conference deadlines and writing journal papers Jan 11 18:33:59 god damn -> __(cup(region + 1, width), printf("%c\010 %c", on_left[i], on_right[i])); Jan 11 18:34:05 backspace at right margin Jan 11 18:34:12 wterm cant do this Jan 11 18:34:14 lol Jan 11 18:38:51 :( Jan 11 18:39:25 for the uninitiated, you're saying if you use CUP to put the cursor on the right edge of the screen, and try to backspace..it won't? Jan 11 18:39:28 lol >.< Jan 11 18:40:07 if i remove that space it looks perfect Jan 11 18:41:07 well this is an autowrap test Jan 11 18:41:18 so in this case autowrap is prob supposed to be off Jan 11 18:41:28 ah Jan 11 18:41:39 i was thinking maybe we weren't good enough to un-wrap wrapped text or something. Jan 11 18:41:49 so inserting that first %c should keep the cursor at the margin Jan 11 18:41:52 shrug, i know nothing of the related code, just running my mouth haha Jan 11 18:41:57 then backspace would move it back to width-1 Jan 11 18:42:04 space moves it back to margin Jan 11 18:42:09 and then last char shows up Jan 11 18:42:18 now that I explain this I have an idea where the bug is Jan 11 18:42:29 thinking out loud always helps Jan 11 18:42:36 err if you write past margin, doesn't it write the character on the marign? Jan 11 18:42:45 (remember seeing some code to that effect) Jan 11 18:42:50 oh yay ideas, gl then ^.^ Jan 11 18:42:56 and yes it's remarkably helpful sometimes hahaha Jan 11 18:49:19 hmm Jan 11 18:49:22 autowrap is on though Jan 11 18:49:31 so now i am confused Jan 11 18:50:30 dtzWill, think you might get a chance to work out the font loading issue today/tonight? Jan 11 18:51:18 maaayybbee :) Jan 11 18:51:34 does the "echo 0 > font.ttf" test work as well as the ipk reinstall test? Jan 11 18:52:06 im not sure i understand how the echo 0 is a test of anything Jan 11 18:52:10 oh Jan 11 18:52:14 try writing to the font file Jan 11 18:52:17 the same way the ipkg reinstall might Jan 11 18:52:29 or maybe just try unlinking it, idk Jan 11 18:52:52 the only 'real' problem i know of is that when you reinstall via ipk the font ends up being zero bytes, right? Jan 11 18:52:58 and as such doesn't load :( Jan 11 18:53:01 right Jan 11 18:53:13 solution: add dep to a font package that puts it in /usr/share/fonts Jan 11 18:53:14 \o/ Jan 11 18:53:15 :P Jan 11 18:53:23 not really a solution Jan 11 18:53:26 kk i'll see if i can poke at it Jan 11 18:53:28 oh not at all Jan 11 18:53:29 haha Jan 11 18:53:35 did the echo Jan 11 18:53:37 file is now 0 Jan 11 18:53:43 haha >.< Jan 11 18:53:43 did that while running wterm Jan 11 18:53:47 idk what that tested Jan 11 18:54:03 j/w, when you reinstall the ipk does anything interesting show up in /var/log/messages? Jan 11 18:54:16 not that I saw Jan 11 18:54:19 kk Jan 11 19:07:21 hmm well i fixed the test i was on Jan 11 19:07:32 idk if i broke anything else though in the process Jan 11 19:15:00 i dont get why the htop mem meter is always full Jan 11 19:20:54 I think the bug might have been there for longer than we think Jan 11 19:21:14 (the font handle thing) Jan 11 19:22:20 so I sorta want to implement a scroll buffer now Jan 11 19:22:40 but maybe that should wait Jan 11 19:25:05 Brybry: yeah i'm interested in what caused it too Jan 11 19:25:43 honest opinion is i don't see how the sdlfontgl stuff did it, although maybe i poked the wrong demons with some change :) Jan 11 19:25:50 has anyone tried /not/ opening the same font 4x? Jan 11 19:25:51 lol Jan 11 19:26:19 i don't mean 'and make everything else work' i just mean as a debugging measure. ignore the need for bold/etc or something Jan 11 19:26:28 it's in my build of sdlterminal from like dec 30th so I don't think it's sdlfontgl Jan 11 19:27:02 the bug is, you mean? Jan 11 19:27:04 yeah Jan 11 19:27:10 oh and that's before i touched anything. hooray :P Jan 11 19:27:11 that's pre-anything you added Jan 11 19:27:33 dtzWill, i tried and i broke all the code doing that Jan 11 19:27:37 so i reset --hard Jan 11 19:31:41 heh kk Jan 11 19:31:50 well nvm if it's not after my changes Jan 11 19:31:57 since we used to open different fonts Jan 11 19:32:21 but a refactoring to do something like load the fonts on demand, and immediately close once in the cache Jan 11 19:32:24 might solve that all Jan 11 19:32:33 Brybry, i've never hit this bug before Jan 11 19:32:57 how come u never mentioned it before Jan 11 19:33:01 I hadn't either but normally I just copied binaries Jan 11 19:33:32 pass sdlfontgl "font specifiers" instead (w/e information one needs to construct and equiv TTF_Font... font path and fontstyle flags?) Jan 11 19:49:34 as a temporary ugly measure we could do the font installing in the postInstall, like make wterm look for ./fonts/wtermFont.ttf and then on post install copy (or move) ./fonts/blah to ./fonts/wtermFont.ttf Jan 11 19:50:31 cause I imagine it'll have killed the wterm process by then (or even further we could check if wterm is running, kill it, wait for it to be killed, move font, whatever) Jan 11 19:51:51 or we could just not leave the file open all the time Jan 11 19:52:22 agreed but =( Jan 11 19:52:48 it looked to me like CloseFont was supposed to take care of that =/ Jan 11 19:54:34 from palm's patched sdl_ttf: https://gist.github.com/076372b69b7592d5c7f4 Jan 11 20:00:34 hmm Jan 11 20:01:41 I don't think it happened in wterm 2.0 Jan 11 20:01:55 0.2.0? Jan 11 20:02:18 yeah Jan 11 20:02:20 it doesn't Jan 11 20:05:20 Brybry, you should 'git bisect' it then Jan 11 20:05:38 well, actually, let me make sure of that Jan 11 20:05:48 maybe I tried to overwrite an unused font Jan 11 20:06:01 i dont get why u guys dont use palm-install Jan 11 20:06:02 though back to back installs with it running definitely didn't crash Jan 11 20:06:13 I use palm-install most of the time now Jan 11 20:06:40 palm-install is slower Jan 11 20:06:52 ohh, a second of your life Jan 11 20:06:58 they add up! Jan 11 20:07:21 but when installed in the real world it uses palm-install mechanisms Jan 11 20:07:26 so.... Jan 11 20:07:29 fail Jan 11 20:10:01 hmm, some of these fonts don't overwrite Jan 11 20:10:10 (with echo) Jan 11 20:10:22 lol i hope it's permissions on the files Jan 11 20:10:23 lol Jan 11 20:10:30 ...somehow. :) Jan 11 20:10:42 they look the same permissions wise Jan 11 20:12:55 just implemented reversed screen state Jan 11 20:13:20 i now believe everything in menu 2 works Jan 11 20:13:24 perfectly Jan 11 20:16:27 :D:D Jan 11 20:19:13 and there is 1 evil test in menu 1 that it doesnt pass still Jan 11 20:19:29 hmm, I'm starting to think the echo test might be bad Jan 11 20:19:29 but i dont think it can pass with our current parser Jan 11 20:19:46 because its basically a test of nested escape codes Jan 11 20:20:03 and i think htop actually uses those a few times Jan 11 20:20:05 for cursor movements Jan 11 20:24:20 see, 0.2.6 doesn't crash either Jan 11 20:24:29 so palm-install is really the only way to test Jan 11 20:24:56 I wonder if the file handles are being left open after wterm exit Jan 11 20:30:15 Brybry, i thought that was what we already agreed on Jan 11 20:32:07 Brybry, do u know how to use git bisect? Jan 11 20:32:49 nope, I'm sure the documentation for it is pretty good though Jan 11 20:36:27 http://book.git-scm.com/5_finding_issues_-_git_bisect.html Jan 11 21:20:17 https://gist.github.com/3d5f4d1223a731cd3d8c was what git bisect led me to Jan 11 21:24:44 thats about when it broke for me if i recall Jan 11 21:25:51 dtzWill, ^^ Jan 11 21:30:47 Brybry, was that the last good or first bad commit? Jan 11 21:30:58 that's the bad one Jan 11 21:31:08 088f586e65a3f86dee7c3bc7ce968e1c373f6769 should be last good Jan 11 21:31:14 or at least of what git bisect told me Jan 11 21:31:54 it usually does a good job Jan 11 21:32:33 it was pretty impressed Jan 11 21:32:49 i used to use that for kernel hacking all the time Jan 11 21:42:32 ok, it looks like it has something to do with opening the same font twice without first closing it Jan 11 21:43:11 :D Jan 11 21:43:21 aka if I change the file names for each open font to be unique files it doesn't do bad things Jan 11 21:43:50 dtzWill, i would say i told you so but im better than that Jan 11 21:43:58 ;) Jan 11 21:44:15 I guess I should see if that works with the latest stuff Jan 11 21:45:35 GAH Jan 11 21:45:36 lol Jan 11 21:46:03 PuffTheMagic: my #1 most suspected source was TTF not playing nice with us opening the same font multiple times Jan 11 21:46:05 never disagreed! Jan 11 21:46:14 you're the ones beign scared of malloc and memcpy :P Jan 11 21:46:46 git bisect ftw Jan 11 21:46:48 and suggested until i had a chance to figure it out someone actually verify it's not the "opening same font multiple times" idea Jan 11 21:46:52 >.< Jan 11 21:47:00 so don't gimme that ;) :D Jan 11 21:47:06 but glad it's resolved \o/ Jan 11 21:47:15 means release can come, and sounds like it'll be quite the release :D Jan 11 21:47:35 but yes git bisect ftw haha Jan 11 21:48:36 yeah, the memcpy stuff works fine as long as no duplicate opens Jan 11 21:49:35 I have no idea how to fix the code to work with only one open though =p Jan 11 21:49:50 memcpy has nothing to do with it? right? :( Jan 11 21:49:51 lol Jan 11 21:49:59 right, memcpy shouldn't matter Jan 11 21:50:05 okay Jan 11 21:50:06 great Jan 11 21:50:07 haha Jan 11 22:01:33 almost have insert/replace mode Jan 11 22:01:37 working Jan 11 22:05:02 PuffTheMagic, if I wanted to make a 'clean' branch with only one change/fix could I do something like: git branch foo; git reset --hard PTM/master; and then cherry pick the fix commit from elsewhere? Jan 11 22:05:24 ya i did that sort of thing all the time Jan 11 22:09:22 running vim on my gentoo box though ssh in wterm -> http://ompldr.org/vYzdkaw/wterm_2012-11-01_170738.png Jan 11 22:09:24 ka6sox-farfarawa, ^^ Jan 11 22:26:26 Brybry, got a fix brewing? Jan 11 22:28:14 I have that "orientation not obeying rotation lock" bug fixed (I actually fixed it like 3 days ago) but I'm still trying to work out cherry-pick Jan 11 22:30:06 are u trying to cherry pick the fix into a new branch Jan 11 22:30:16 or do u have the fix in a branch and want to cherry pick it into master? Jan 11 22:34:05 Brybry, ? ^ Jan 11 22:35:15 I was trying to cherry-pick from one branch into another new branch Jan 11 22:35:24 I can get it to work with master Jan 11 22:37:19 so switch to the branch with your fix Jan 11 22:37:28 use git log to figure out the commit id Jan 11 22:37:32 then switch back to master Jan 11 22:37:34 and cherry pick it Jan 11 22:40:51 what worked->checkout master, cherry pick, resolve, add it (but not commit), make new branch, commit Jan 11 22:47:08 guess I needed a fetch in there first but it worked out ok either way Jan 11 22:50:21 pull request sent! Jan 11 22:54:41 merged Jan 11 22:55:10 is anyone working on the font crash Jan 11 22:59:08 if not will take another stab at it Jan 11 22:59:14 but i dont want to duplicate effort Jan 12 00:08:26 hmm meant to say that here. oh well. Jan 12 00:23:50 \o/ i'm back Jan 12 00:24:10 i know you all were very concerned >_> Jan 12 00:24:10 :D Jan 12 00:27:21 (should stbuehler be in here/) Jan 12 00:27:29 *here? Jan 12 00:43:58 dtzWill, i invited him Jan 12 00:44:13 coolies Jan 12 00:44:22 dtzWill, now fix them fonts ;) Jan 12 00:44:47 in due time :) Jan 12 00:48:29 PuffTheMagic: that reminds me of two things I wanted to bring up regarding wterm: 1)credit/branding and 2)donations Jan 12 00:49:57 both are potentially sticky subjects :) Jan 12 00:50:06 but hopefully we don't have to make them that way Jan 12 00:50:16 ya I was going to bring up the donations too, was going to do the 10% kickback to wosi then split donations evenly periodically Jan 12 00:50:41 but im not worried much about the donations, i dont think its gonna bring in much Jan 12 00:50:49 that was my inclination Jan 12 00:50:50 i've got like 22.02 so far Jan 12 00:51:00 and as long as it's low as far as i'm concerned it's all yours Jan 12 00:51:15 (idk how other contributors feel) Jan 12 00:51:35 but personally despite the time i've put in, this project wouldn't *exist* without you so there's a certain bootstrap of sorts hehe :P :) Jan 12 00:51:44 its not like freeTether where people are canceling their broadband to tether Jan 12 00:51:59 and then feel like they need to give me a months payment lol Jan 12 00:52:08 yeah, i'm more interested in not being rocked rather than actually making $$ on time spent on wterm Jan 12 00:52:47 i'm happy being paid nothing, but if (somehow) it was pulling in $$ then i'd feel kinda used, whether that's what happened or not :P :) Jan 12 00:53:35 seems silly not to have the donation link there but if the donation rate continues like it is, i dont see it breaking $100 for a few months Jan 12 00:53:57 depends if we fix enough things and/or get a p|c break or so Jan 12 00:54:18 esp with the rate we're fixing things, hopefully it hits 'feature complete' soonish Jan 12 00:54:25 ya a p|c break might help, but I sorta feel like wterm is gonna be in the same boat as Spaz Jan 12 00:54:37 no one donates for spaz Jan 12 00:54:54 maybe we can rope in some wosi PR, usually works pretty well. Jan 12 00:54:58 i have no idea what wirc got cause it all went to wosi Jan 12 00:55:11 freeTether was sorta an odd exception Jan 12 00:55:25 and preware/phd are sorta outliers too Jan 12 00:55:29 i thought i bought spaz in the app cat, lol. Jan 12 00:55:31 apparently not Jan 12 00:55:52 maybe u did, but funkatron swears no one ever donates Jan 12 00:56:14 well it's marked free in the app cat atm, so maybe i'm wrong Jan 12 00:56:28 oh there was a 'special edition' Jan 12 00:56:46 anyway anyway the point is virtually no one contributes financially and i'm not disagreeing with that conclusion Jan 12 00:57:11 so I will leave the donations in paypal for a while and we can revisit later if u want, the money wont be going anywhere Jan 12 00:57:31 as for the branding, if u want the appid to change I suppose thats not a big deal either Jan 12 00:57:34 like i said i'm not trying to make bank on wterm, if any of us are in it for the $$ then we need to reevaluate the our decisions regarding the time/$$ ratio Jan 12 00:57:46 lol Jan 12 00:58:18 dtzWill, i was pretty much just expecting to offset the cost of a case of beer now and then Jan 12 00:58:22 nah, appid is annoying to change, although if we want to put it in webosinternals that's cool. i don't really care about the appid. neat little cookie for you for making it Jan 12 00:58:24 PuffTheMagic: yeppers Jan 12 00:58:46 the best thing you can do for an OSS developer is pay for the beer he drinks while he works on the project Jan 12 00:58:46 :D Jan 12 00:59:06 maybe not best, but that's all i hope for usually Jan 12 00:59:27 but alright, i just wanted to bring that up since it hadn't been discussed at all Jan 12 01:00:01 i have no concerns with how things are, but also wanted to know if tomorrow i was gonna read an article about the amazing "PuffTerm" ;) Jan 12 01:00:09 lol Jan 12 01:01:07 well the readme.md has listed the contributors Jan 12 01:01:10 might push to add main contributors to a place that people not looking at the source code can see O:) Jan 12 01:01:15 yeah that was nice, ty for that :) Jan 12 01:01:27 unrelated to this topic, those screenies are a bit dated right? Jan 12 01:01:43 dtzWill, ya screenies can get updated as people want Jan 12 01:02:05 dtzWill, i was going to put more info in the about down the road Jan 12 01:02:13 well sooner rather than latter Jan 12 01:02:18 (does htop still have those rendering artifacts?!) Jan 12 01:02:21 but i've been busy on vttest stuff Jan 12 01:02:49 PuffTheMagic: okay. i didn't mean to be in any way attacking, and there's no particular hurry :) Jan 12 01:02:53 dtzWill, there are a few things about htop that are still not 100% perfect depending on the TERM u set Jan 12 01:03:21 dtzWill, im not taking it as an attack, im just saying, its on my todo list, which is like a mile long Jan 12 01:03:29 hahaha kk Jan 12 01:03:37 good to hear, and ty Jan 12 01:03:53 i was gonna add a first use with change log like thing too Jan 12 01:03:56 PuffTheMagic: hrmm, kk. also, \o/ on fixing all those vttest stuff! Jan 12 01:04:21 dtzWill, wTerm cant compete with xterm if it cant pass basic tests Jan 12 01:04:26 well htop works good to me, hehe. TERM=xterm, fwiw Jan 12 01:04:31 those are mandatory requirements Jan 12 01:04:39 PuffTheMagic: indeed Jan 12 01:04:44 dtzWill, yes it works good, but there are some things that are not perfect Jan 12 01:04:49 having xterm dev on wterm's side helps a bit :) Jan 12 01:04:55 i'm hopeful wterm can entirely replace xterm Jan 12 01:04:56 like that green bar witht he headers Jan 12 01:05:09 that should extend across the whole term Jan 12 01:05:15 xterm was always a hack. i like the x server an awful lot for various reasons, but... xterm by itself is just obnoxious haha Jan 12 01:05:21 hmm Jan 12 01:05:26 PuffTheMagic: i see, interesting. Jan 12 01:05:39 (you're right, it's not on my latest build either) Jan 12 01:06:00 im pretty sure there is an erase sequence that should be using the previous cells graphics state Jan 12 01:06:32 PuffTheMagic: while talking random things, how do you feel about a native keyboard? or allowing user to switch to default built-in vkb? Jan 12 01:07:10 the reason i ask is that the enyo one is wonderful and complete, but compared to the builtin one tends to lag a bit and i find myself using xterm for that reason Jan 12 01:07:12 i didnt really ever imagine anyone would want to use the native vkb with it being so limited Jan 12 01:07:21 * Brybry is working on bluetooth keyboard stuff Jan 12 01:07:40 Brybry: oh yes, bt support is a must. it "mostly" works, right? modulo that issue on github? Jan 12 01:07:55 (esc/arrow keys?) Jan 12 01:08:02 dtzWill, if it is truly lagging its probably related to my code (which can probably be improved) Jan 12 01:08:16 most of the keyboard stuff has been first cut code Jan 12 01:08:24 oh, okay. Jan 12 01:08:24 working enough for me to debug the term/vt stuff Jan 12 01:08:40 i had no idea, i know...very little enyo/js and so had no idea what its status was Jan 12 01:08:44 kk i see Jan 12 01:08:56 if we can make it work well then anything else is a hack and confusing Jan 12 01:09:03 so sounds good to me Jan 12 01:09:35 I'm probably doing things that don't need to be done/may end up being just for me Jan 12 01:09:51 Brybry: nothing wrong with that :D Jan 12 01:10:15 but so far you'll basically be able to pick from a list of bt devices to set your default bt keyboard and then it'll automatically put away the virtual keyboard/resize on bt keyboard detect (or bring it back on disconnect) Jan 12 01:10:21 dtzWill, and as far as an article about PuffTerm ;) that would only happen because of my tweeting, i've tried to say "we" instead of "I" since you and Brybry started hacking on it Jan 12 01:10:24 like the native one does Jan 12 01:11:08 PuffTheMagic: wonderful, and very much appreciated. Jan 12 01:11:17 (PuffTerm has a nice ring to it....:P) Jan 12 01:11:17 and also just knowing the bt keyboard is connected allows me to do cleaner handling of things like the arrow keys on the HP one Jan 12 01:11:22 Brybry: oh sweet Jan 12 01:11:28 the arrow keys are really annoying though Jan 12 01:11:39 dtzWill, well wterm is sorta overused according to google ;) Jan 12 01:11:44 they have different key values depending on if they're a plugin or standalone Jan 12 01:11:46 Brybry: yeppers, I almost always have my tp+bt keyboard on me Jan 12 01:11:47 I don't understand that at all =( Jan 12 01:12:09 Brybry: lol :(. can we just set it to recognize both? or do they conflict with other things...? Jan 12 01:12:26 but that does make debugging obnoxious, althuogh if it's just arrow keys syslog wouldn't be so bad Jan 12 01:12:28 I don't think they conflict with anything but I was thinking they *might*? Jan 12 01:12:55 hmm. Jan 12 01:13:05 have any of you run vttest recently since I fixed so many of the tests Jan 12 01:13:24 nope Jan 12 01:13:33 PuffTheMagic: nope, still at work. haven't updated my tree at all today Jan 12 01:13:36 nearly all of menu 1,2 and 8 work Jan 12 01:13:46 which is basically all of vt100/102 Jan 12 01:14:49 dtzWill, http://ompldr.org/vYzdocQ Jan 12 01:15:39 PuffTheMagic: you'll hit $100 in no time ... just need more of wrea's 1c Jan 12 01:16:02 well nothing is gonna happen while CES is going on Jan 12 01:16:07 or the week or so after Jan 12 01:17:02 oh, yeah Jan 12 01:17:06 and I did key repeat stuff Jan 12 01:17:07 forgot about that Jan 12 01:17:14 like on the C side Jan 12 01:17:56 well send a merge request Jan 12 01:18:00 I basically looked into the SDL source and used their code to make a function to simulate fake key presses the same as they do Jan 12 01:18:05 it still needs tidying up Jan 12 01:18:20 PuffTheMagic: quick git q... I pull'd from wTerm, and had a conflicted Makefile.inc ... how do I restore that Jan 12 01:18:35 git checkout and git reset didn't seem to restore it Jan 12 01:18:40 git reset --hard Jan 12 01:18:47 if u dont have changes u care about Jan 12 01:19:03 dwc-, so arent going to fork officially? Jan 12 01:19:07 I don't. I had just changed the path to the compiler Jan 12 01:19:36 thanks, that fixed it Jan 12 01:20:00 PuffTheMagic: I haven't even looked at the code yet... no need for me to go sign up for github yet Jan 12 01:20:17 well, ok, I looked at that bit in the plugin last week ago Jan 12 01:20:22 so u are just too cool for ipks? Jan 12 01:21:02 ? Jan 12 01:22:49 dwc-: pulling source with possibility of forking, but deferring doing so (and github account creation) until you have a reason to do so? Jan 12 01:23:30 more or less Jan 12 01:24:16 dtzWill, he just wants to have my code available so he can troll and make crack marks about how i suck Jan 12 01:24:17 pulling src also means better searching than webUI... and building my own when puff says "oh, you need the version not in preware yet" Jan 12 01:25:22 haha Jan 12 01:30:00 this is the best escape code ever echo -e -n "\033#8" Jan 12 01:33:20 what's it do? Jan 12 01:33:31 try it Jan 12 01:33:50 haha Jan 12 01:34:17 what's the fascination with a screen full of E? Jan 12 01:34:35 #7 and #9 are different functions... so why E Jan 12 01:34:58 its the screen alignment patters Jan 12 01:35:05 prob didnt matter what the char was Jan 12 01:35:12 but its DEC escape code Jan 12 01:35:17 so they must like Es Jan 12 01:45:25 dtzWill, when do u experience these vkb lags Jan 12 01:45:38 except for sometings when it first starts up it seems pretty damn responsive to me Jan 12 01:46:42 the keys themselves Jan 12 01:46:48 perhaps autorepeat will alleviate that Jan 12 01:47:13 i dont understand what you mean by "the keys themselves" Jan 12 01:47:40 the letters seems to show as soon as i press a key Jan 12 01:48:51 https://github.com/Brybry/wTerm/commit/fff71d4159915720d347143094cd8330c8d86621 is working key repeat but I'm not really happy with my changes to the event loop Jan 12 01:53:22 PuffTheMagic: I see... Jan 12 01:53:42 maybe i'm just perceiving it all wrong because of the enyo buttons lagging display-wise Jan 12 01:53:48 but if i watch the term they work regardless of what i see Jan 12 01:55:48 the visual state changes of the keys you talking about? Jan 12 01:56:17 mmhmm Jan 12 01:56:23 hmmm Jan 12 01:56:29 ncmpc does not render properly Jan 12 01:56:46 i sorta wish wterm had tabs so i could do more at once Jan 12 01:57:59 screen! Jan 12 01:58:24 but, yeah, when I don't use screen I'm like, "Why can't I just bring up a second wterm?" Jan 12 02:01:36 heh, integrated screen support would be sweeeeeet Jan 12 02:01:37 :P Jan 12 02:38:32 dtzWill, ping **** ENDING LOGGING AT Thu Jan 12 02:59:57 2012