**** BEGIN LOGGING AT Thu Feb 02 02:59:57 2012 Feb 02 03:50:38 Brybry, is it alive again? Feb 02 03:51:04 is what alive Feb 02 03:51:49 oh, my touchpad? yeah, I just had to reset it with power+home, not a big deal Feb 02 03:51:59 ya that helps Feb 02 03:52:11 now I have a pdk app running that doesn't let it sleep while I run kismet to see if there are other stability issues Feb 02 03:52:25 53 minutes so far Feb 02 03:52:31 ~8000 packets Feb 02 04:30:11 oil, ping Feb 02 04:30:16 PuffTheMagic, what should I look into to make http://img401.imageshack.us/img401/8734/wterm20120102194957.png look like http://img214.imageshack.us/img214/738/puttykismet.png in terms of the lines? Feb 02 04:30:47 Brybry, your term setting Feb 02 04:30:49 TERM Feb 02 04:31:07 and in the prefs code you should be able to test if there is localstorage (or at worst do a version check) and then use localstorage/cookie setting depending Feb 02 04:31:19 that's TERM=xterm for both of those Feb 02 04:31:23 it looks way worse with vt100 :( Feb 02 04:32:00 try linux,rxvt,xterm-color,vt102 Feb 02 04:32:10 it's curses so I didn't expect it to work right out of the box Feb 02 04:32:47 Brybry, also run vttest on the box where it "works well" Feb 02 04:32:52 and do the charater set test Feb 02 04:33:03 and see how many of them support line drawing symbols Feb 02 04:34:24 PuffTheMagic: hey Feb 02 04:36:03 oil, so I got a enyo hybrid kind that works on all webos devices :D.... Feb 02 04:36:12 oil, but now I need a preferences kind that works on all webos devices Feb 02 04:36:31 1.4.5 and probably <2.2 dont have localStorage Feb 02 04:36:40 maybe its just 1.4.5 Feb 02 04:37:36 less then 1.x doesnt Feb 02 04:37:37 2.x does Feb 02 04:37:48 hmm Feb 02 04:42:07 Cookies works on everything, no? Feb 02 04:43:02 probably, but since using enyo "we" oil/I have been using localstorage for some reason Feb 02 04:43:12 rwhitby: yeah Feb 02 04:43:31 cause localstorage is easier/better? Feb 02 04:46:45 and at the time the decision was made we didn't know they were going to put enyo on phones Feb 02 04:46:54 :( i have to reremember how to use cookies Feb 02 04:47:03 its got a weird getter/setter interface Feb 02 04:47:14 instead of just localStorage['key'] = value Feb 02 04:47:16 oil, im not blaming you for this Feb 02 04:47:26 i didnt think you were Feb 02 05:00:48 do they have enyo on 145? it gets the bing maps update? Feb 02 05:04:32 ya Feb 02 05:23:37 sans prefs issues, wTerm now runs on my pixi, veer, pre3 and TP Feb 02 05:24:36 PuffTheMagic: hard keyboard stuff too? Feb 02 05:26:37 no that stuff is buggy Feb 02 05:26:47 but the term renders Feb 02 05:26:51 which is the big milestone Feb 02 05:26:57 in an enyo hybrid Feb 02 05:50:52 hmmm, draws fine when I do echo -e "\033(0mlqqqqqqqqqqqqkj\033(B" Feb 02 05:56:59 ok pixi is pissing me off Feb 02 05:57:02 wterm is sent: ^[[0m^N^[[34m^[[40mq putty is sent: ^[(0^[[0m^[[34m^[[40mq Feb 02 05:57:18 2012-02-02T05:50:26.750371Z [1652] palm-webos-device user.err us.ryanhope.wterm.plugin[6658]: Cannot create SDL framebuffer: Couldn't set console screen info Feb 02 05:57:18 2012-02-02T05:50:26.763291Z [1652] palm-webos-device user.crit us.ryanhope.wterm.plugin[6658]: WTerm not started. Feb 02 05:58:34 so m_surface = SDL_SetVideoMode(0, 0, 0, SDL_OPENGL); fails? Feb 02 05:58:47 for pixi in plugin mode Feb 02 05:58:53 its fine in standalone mode Feb 02 06:03:49 try SDL_SWSURFACE instead of SDL_OPENGL just to see if it gets past that (it should definitely break a little later though) and if it still breaks then maybe it needs width/height set based off of screen metrics Feb 02 06:04:17 I would guess it's more likely the height/width with the plugin bit than opengl though Feb 02 06:04:43 i've tried the screen metrics Feb 02 06:05:06 seems like a perms issue Feb 02 06:05:14 that error is related to an ioctl that fails Feb 02 06:25:07 toshiba thrive (10" uses tegra 1ghz SoC like original galaxy tab) 1GB RAM $280 http://www.woot.com/ Feb 02 06:25:21 not a touchpad deal but what is? Feb 02 06:25:57 eh Feb 02 06:26:13 thats expensive for a refurb Feb 02 06:26:16 I picked up a xoom for 230, and even that was overpriced. Feb 02 06:26:42 its just... not a touchpad Feb 02 06:27:02 #dealofthecentury Feb 02 06:27:57 halfhalo, why does this happen Feb 02 06:27:58 Cannot call method 'appendChild' of undefined Feb 02 06:28:34 my bet is either scope issues or your parent variable is not defined. Feb 02 06:28:45 http://pastebin.com/ty9sxBjy Feb 02 06:28:52 this works on webos 2.2 Feb 02 06:28:54 and 3.x Feb 02 06:28:59 but its failing on 2.1 Feb 02 06:29:26 line 21 Feb 02 06:29:55 document or document.head may not exist. you may be able to go through window to get what you need though Feb 02 06:31:22 switching head with body worked Feb 02 06:33:07 sounds like enyo is being loaded in different locations in that case Feb 02 06:36:28 body works on 3.x too Feb 02 06:37:19 cool Feb 02 06:45:27 Brybry, ping Feb 02 06:46:05 back to finishing my little package grabber script Feb 02 06:50:10 * PuffTheMagic puts his pre- and pre2 on charges for first time in over a year Feb 02 06:50:37 going for wTerm support on 6 webos devices Feb 02 06:50:39 :D Feb 02 06:52:54 nice Feb 02 06:55:17 but looks like i didnt connect my wifi antenna last time i assembled my pre- Feb 02 06:55:22 or a messed up some tokens Feb 02 06:55:30 so... no testing on that Feb 02 06:56:43 palm-install ftw Feb 02 06:57:02 i need the enyo update Feb 02 06:57:26 stick enyo in the ipk to test Feb 02 06:57:26 or can i just copy enyo from git Feb 02 07:06:59 booo Feb 02 07:07:01 2010-12-13T09:09:10.940917Z [1012] palm-webos-device user.err us.ryanhope.wterm.plugin[9057]: Cannot create SDL framebuffer: Could not initialize EGL Feb 02 07:07:09 ^^ that is for pre- Feb 02 07:07:24 prob a similar issue with the pixi Feb 02 07:08:39 nope, this wont even run in standalone right now Feb 02 07:08:43 at least the pixi did Feb 02 07:12:26 pres on pre2 Feb 02 07:12:27 also Feb 02 14:26:03 PuffTheMagic, still need me? Feb 02 16:47:30 * Brybry writes quick and dirty paste Feb 02 16:47:44 I am tired of typing out escape sequences by hand Feb 02 17:03:17 stick the echo ... in a .sh and . it? Feb 02 17:12:00 what up Feb 02 17:16:29 Brybry: i got wterm "working" on all devices Feb 02 17:16:40 and when i say working i mean, i got the enyo hybrid stuff working Feb 02 17:16:44 it starts the plugins Feb 02 17:17:00 however on my Pixi it only runs in standlone Feb 02 17:17:07 and on my pre- i cant get it to run at all Feb 02 17:17:09 the Term Feb 02 17:17:18 what, paste took all of like 2 minutes to write Feb 02 17:18:11 Brybry: cool, just know it might need to change in the future when we tackle full mouse support Feb 02 17:18:23 no mouse stuff involved Feb 02 17:18:26 I didn't do copy Feb 02 17:18:43 how are u triggering paste? nothing to do with touching the screen? Feb 02 17:18:52 in the browser, sure Feb 02 17:18:57 'paste' menu option with enyo.dom.getClipboard(enyo.bind(this, this.handleClipboard)); handleClipboard(clipData) { this.$.terminal.inject(clipData)} Feb 02 17:19:04 i know how paste works in webos Feb 02 17:19:05 ive done it Feb 02 17:19:22 i just want to know how from a UI pov you are triggering the paste Feb 02 17:19:30 our dropdown menu Feb 02 17:19:40 oh Feb 02 17:21:46 i used wterm to scp something this morning Feb 02 17:21:50 it made me feel good Feb 02 17:22:14 trying to figure out why kismet hates me for line drawing, I know we can draw the lines as echo -e "\033(0mlqqqqqqqqqqqqkj\033(B" works great Feb 02 17:22:42 Brybry: did u do the vttest i told you about Feb 02 17:22:52 yes Feb 02 17:23:06 but like I said, the above works great. I don't think it's a charset issue Feb 02 17:23:07 I could be wrong Feb 02 17:23:54 right the above works because we load the line drawing chars in 1 charset Feb 02 17:24:05 im just curious if your term has them in more than 1 charset Feb 02 17:27:18 wterm vttest: http://img824.imageshack.us/img824/2376/wterm20120102223537.png putty vttest: http://img404.imageshack.us/img404/1599/puttyvttest.png Feb 02 17:27:33 Brybry: i just tried vttest on my mac, and it has line drawing chars on 2 charsets Feb 02 17:27:42 so that could be the problem Feb 02 17:28:42 Brybry: on my mac i have line drawing chars on charset 2 Feb 02 17:28:55 wterm right now only implements on charset 0 Feb 02 17:33:32 uhm Feb 02 17:33:39 yay Feb 02 17:33:44 i think we have support for slots G1-G4 Feb 02 17:33:44 stbuehler: we have a problem Feb 02 17:34:01 stbuehler: "support" yes, but we only fill G0 with the line drawing chars Feb 02 17:34:24 you can load charsets into G* slots, and then activate them Feb 02 17:34:31 only one can be active obviously Feb 02 17:34:57 it doesn't help that kismet(curses?) sends some different escape codes to putty than wterm (both TERM=xterm) though both sets of sequences I get should draw lines, I believe Feb 02 17:35:00 I'm still parsing through it Feb 02 17:36:10 stbuehler: i know that, what im saying is that if kismet sets G2 and tries to draw lines, in wterm its gonna be ascii chars Feb 02 17:36:27 with how we have the charsets filled in right now Feb 02 17:37:04 well, as i said i think i added support for G* Feb 02 17:37:10 imho it should work :) Feb 02 17:37:17 stbuehler: what do u mean support Feb 02 17:37:24 the ability to swtich to, yes Feb 02 17:37:28 will it draw lines Feb 02 17:37:30 no Feb 02 17:37:34 why wouldn't it? Feb 02 17:37:40 because we didnt fill them in Feb 02 17:37:49 well, i think i did Feb 02 17:37:54 but perhaps i'm wrong Feb 02 17:38:03 i thought only G0 has line drawing Feb 02 17:38:04 i basically copied what konsole is doing Feb 02 17:38:22 i think they are only slots that can be selected Feb 02 17:41:27 stbuehler: in applyCharset we only modify charset0 for line drawing chars Feb 02 17:42:14 Brybry: in terminalstate.cpp below line 812 add case '2': Feb 02 17:42:17 see if that helps Feb 02 17:43:35 PuffTheMagic: wrong Feb 02 17:43:42 .charset is the currently active charset Feb 02 17:44:05 .charsets is the array of available charsets in G0-G3 Feb 02 17:44:23 and .charset_ndx is the index of the active slot Feb 02 17:44:43 i dont get how what I suggested is wrong Feb 02 17:45:06 if i wanted to add linedrawing chats to charset 2 how would you do it? Feb 02 17:45:31 hm Feb 02 17:45:38 i guess we talked about different things Feb 02 17:45:41 s/chats/chars/ Feb 02 17:46:06 we support 'B' (default afaik), 'A' (uk?) and '0' (graphics) afaik Feb 02 17:46:18 right Feb 02 17:46:34 and on my term on my mac I have line chars on '2' Feb 02 17:46:34 that is why i was confused when you said we only support charset 0 Feb 02 17:46:52 so i was just suggeting that might help Brybry's kismet issue Feb 02 17:46:53 but idk Feb 02 17:46:54 and i thought you were talking about the slot G0 Feb 02 17:47:00 no :D Feb 02 17:47:07 but yes, mappings for '1' and '2' would be nice Feb 02 17:47:42 but we might also be missing a command that sets the proper SHIFT state Feb 02 17:47:49 which might also be causing his issues Feb 02 17:47:50 idk Feb 02 17:48:32 one thing is I get http://pastebin.com/RutqFY2N vs http://pastebin.com/r22vf3t8 (note: these are raw and don't reset your terminal after they run) Feb 02 17:48:43 no, we have shift states for G0-G3 Feb 02 17:49:14 they should both draw lines but I don't really understand yet how kismet/curses 'knows' to send different values unless it's some stty modes or something that it picks up on Feb 02 17:49:24 i'll be back later, and have a look at it if you didn't fix it already :) Feb 02 17:49:39 Brybry: kismet is ncurses and ncurses uses termcap/terminfo Feb 02 17:49:48 which knows what to send based on your TERM value Feb 02 17:49:52 both terms are the same Feb 02 17:50:26 and obviously some of those things are mouse commands Feb 02 17:50:41 or so they looked t ome Feb 02 18:10:37 ok Feb 02 18:11:23 case CS_GRAPHICS_MODE_SET: //ESC[;...;m -> if (values[i] == 0) ->m_currentGraphicsState.nGraphicsMode = 0; instead of m_currentGraphicsState = TSGraphicsState(); fixes it Feb 02 18:11:27 in vtterminalstate.cpp Feb 02 18:11:38 but that probably breaks other things Feb 02 18:18:29 Brybry: nice, just do some tests Feb 02 18:19:03 htop,emacs,mc,vim,vttest are my tests apps Feb 02 18:19:07 and cmatrix Feb 02 18:20:55 oil: 1.4.5 does have localStorage Feb 02 18:21:01 but the lsvar is different Feb 02 18:21:12 its undefined_prefs Feb 02 18:23:37 maybe thats just cause enyo.fetchAppInfo().id is undefined? Feb 02 18:29:14 ok looks like localStorage really doesnt exist Feb 02 18:29:36 the check in the prefs object we use needs a tweaking Feb 02 18:55:35 typeof localStorage != "undefined" probably Feb 02 18:57:13 javascript object existence checking in general really sucks, so many different nonstandard ways of doing it Feb 02 19:07:31 Brybry: you find any breakage from your fix? Feb 02 19:08:22 Brybry: also, i cant use typeof localStorage on 1.4.5 Feb 02 19:08:26 it comes back as an object Feb 02 19:08:37 thing is it never saves/writes it to a file Feb 02 19:11:53 no but I'm looking at docs thinking if it's correct or not Feb 02 19:12:09 there are different ways I could do it Feb 02 19:14:28 ah, see. I didn't reset foreground/background color Feb 02 19:14:55 I feel like y'all were talking about this just yesterday Feb 02 19:17:48 there could very well be erroneous charset resetting in our code Feb 02 19:17:50 i believe that Feb 02 19:18:17 yeah, that's what was happening Feb 02 19:18:23 but I think it was on purpose Feb 02 19:18:35 ehhh Feb 02 19:18:42 resetting of graphics states yes Feb 02 19:18:43 stbuehler might know Feb 02 19:18:52 resetting of charsets, uninformed Feb 02 19:18:57 use git blame Feb 02 19:19:01 and see it tracks back to me Feb 02 19:19:08 I could instead make a constructor that accepts arguments and pass it the previous charset, I dunno Feb 02 19:19:11 if it was, I prob didnt resarch it much Feb 02 19:33:44 Brybry: just pushed a fix to prefs that works on 1.4.5 Feb 02 19:34:38 halfhalo: pinger Feb 02 19:34:48 hmm? Feb 02 19:35:25 enyo is exploding on me some more on 1.4.5, i need your input/help Feb 02 19:35:31 hmmm Feb 02 19:36:43 https://github.com/PuffTheMagic/wTerm/blob/master/src/enyo/vkb.js Feb 02 19:36:53 the function that starts on line 57 Feb 02 19:37:05 really pisses off 1.4.5 Feb 02 19:37:22 2012-02-02T19:32:09.505025Z [34617] palm-webos-device user.crit LunaSysMgr: {LunaSysMgrJS}: us.ryanhope.wterm: Uncaught TypeError: Object function (layout) { ^I^I^Iif (name != this._layoutName) return; ^I^I^Ivar components = [], i, j, comps, c, e; ^I^I^Ifor (i = 0; i < layout.length; i++) { ^I^I^I^Irow = layout[i]; ^I^I^I^Icomps = []; ^I^I^I^Ifor (j = 0; j < row.length; j++) { ^I^I^I^I^Ie = row[j]; ^I^I^I^I^Iif (e.content || e.symbols) Feb 02 19:38:29 hmmmm Feb 02 19:39:55 gimme a few, about to leave work and get food/go home Feb 02 19:40:20 actually i think i fixed it Feb 02 19:40:41 i moved the annon function to a real function Feb 02 19:40:48 man, you guys are crazy with this wTerm development Feb 02 19:40:50 and then used enyo.bind() instead of .bind(this) Feb 02 19:41:03 PuffTheMagic, I don't think I have seen you working this hard on something since the early days of wIRC Feb 02 19:41:23 now if only I could get someone to write a ventrilo app ;) Feb 02 19:41:35 cryptk: i've always wanted to work on a term Feb 02 19:41:46 and i got other people interested too Feb 02 19:41:53 so that makes it more fun and easier to keep working Feb 02 19:41:58 I do have libventrillo3 compiling and working, I even had my original Pre connecting to a ventrilo server and playing MP3's in it, lol Feb 02 19:42:03 although i need to do some work for school at some point Feb 02 19:42:08 haha Feb 02 19:42:18 school is for people who want a job doing something other than Linux Feb 02 19:47:44 now to find out why wTerm doesnt play nice with SDL in cards Feb 02 19:54:56 With the webOS 1.4.5 release, there is limited PDK application support for Pixi devices. Full-screen applications built using OpenGL for display should work. Applications using SDL for display initially draw, but the screen does not subsequently update. Combined JS/Plug-in applications will not work - the JS executes, but the Plug-in component is not created. Feb 02 19:55:01 damn Feb 02 19:59:18 only one solutions: provide the keyboard with sdl :D Feb 02 19:59:36 when and where to reset which char attribute i don't know Feb 02 20:00:18 but i guess we need something like m_currentGraphicsState.noVisualAttributes() that returns a state with bold/.. reset but keeps the colors Feb 02 20:02:39 cryptk: ventrilo? like use your phone as a push to talk voip thing? Feb 02 20:02:51 destinal, yep Feb 02 20:02:57 it would be able to connect to a ventrilo server Feb 02 20:03:03 I have already done it Feb 02 20:03:13 had my phone connect, as well as stream MP3's into the server Feb 02 20:03:23 but I didn't have an app for it, it was all done through novaterm Feb 02 20:03:53 cryptk: thought they were binary only Feb 02 20:04:00 they are Feb 02 20:04:03 but... Feb 02 20:04:12 the guys who made mangler, which is a linux ventrilo client Feb 02 20:04:15 ahh Feb 02 20:04:20 they made libventrilo3, reversed the protocol Feb 02 20:04:47 nice. someone should do that with teamspeak3 too Feb 02 20:04:55 cryptk: they have all kinds of reasons to use mumble instead though :) Feb 02 20:05:04 yep Feb 02 20:05:07 how does that handle gsm codecs? Feb 02 20:05:11 isn't gsm non-free? Feb 02 20:05:11 mumble server is crap imho Feb 02 20:05:17 speex is crap imho Feb 02 20:05:17 in fact even though they made mangler, they still recommend that people use mumble instead Feb 02 20:05:25 change the ip -> need to edit sqlite db manually Feb 02 20:06:10 php guys fixed dos, but added a remote code execution. nice :D Feb 02 20:18:17 im so bummed the pixi cant handled the plugin :( Feb 02 20:18:26 stbuehler: have you tried HEAD on your Pre- yet? Feb 02 20:21:35 not yet Feb 02 20:23:50 PuffTheMagic: pixi can't but pre can? Feb 02 20:24:10 on 1.4.5? Feb 02 20:24:11 destinal: well on my pre- last night I was getting an EGL error Feb 02 20:24:17 idk what thats about Feb 02 20:26:29 wterm.plugin[4119]: Cannot execute child shell. Feb 02 20:32:07 next try made it reboot. Feb 02 20:32:48 wterm.plugin[4453]: Cannot execute child shell: No such file or directory Feb 02 20:36:33 wterm.plugin[4604]: Cannot execute child shell '12': No such file or directory Feb 02 20:36:34 aha Feb 02 20:38:46 might be my jailer stuff Feb 02 20:41:24 and i see wrong characters Feb 02 20:41:40 could be a problem with the shader / float precision Feb 02 20:43:00 execute shell '12' wtf Feb 02 20:43:04 that is the font size Feb 02 20:43:11 what isnt passing the exec preoperly Feb 02 20:44:23 was my wrapper thing Feb 02 20:44:34 ah ya, you shouldnt need any of that now Feb 02 20:44:47 remove your wrapper, uninstall wterm and do a fresh install Feb 02 20:44:50 it should all just work Feb 02 20:44:58 it does Feb 02 20:45:01 that part Feb 02 20:45:46 numbers don't work Feb 02 20:47:34 it misses the "H" and has "P" twice Feb 02 20:47:42 so verything between those is moved one left Feb 02 20:48:09 misses X again Feb 02 20:49:35 hmm i didnt push a commit Feb 02 20:49:51 hm? Feb 02 20:50:04 it isn't a bug on the tp Feb 02 20:50:23 i transmit character values in a color channel of a texture Feb 02 20:50:41 the opengl spec says that the input is 8-bit, but the imlementation can do whatever it wants with it Feb 02 20:50:47 i dont get how it runs for you at all Feb 02 20:50:55 i get javascript erros Feb 02 20:50:56 it could for example drop to 4 bit precision Feb 02 20:50:56 with HEAD Feb 02 20:51:02 im on 2.1 Feb 02 20:51:06 oooooh Feb 02 20:51:22 1.4.5 had too many bugs Feb 02 20:51:36 and i needed nodejs :) Feb 02 20:52:00 so u have 2.1 on a pre- Feb 02 20:52:04 yes Feb 02 20:54:21 stbuehler: vtterminalstate.cpp:229 do we purposefully reset charset there or can I make it not? Feb 02 20:57:01 i think thats an error Feb 02 20:57:06 to reset charset there Feb 02 20:57:16 charset is not part of the graphicsstate Feb 02 21:01:30 the VT420 doesn't know charsets '1' and '2' Feb 02 21:01:34 hm Feb 02 21:01:54 i think a reset should just reset everything :) Feb 02 21:02:59 if we want to follow xterms behavior i suggest reading xterm source :) Feb 02 21:03:14 we arent talking about ESCc here Feb 02 21:03:35 ESC[m is just the graphics stufs Feb 02 21:03:47 u can change charset with ESC[*m Feb 02 21:03:53 so i dont think it should be resetting them Feb 02 21:03:57 s/can/cant/ Feb 02 21:16:22 xterm source clears INVERSE | BOLD | BLINK | UNDERLINE | INVISIBLE flags and resets FG/BG color Feb 02 21:16:35 yay Feb 02 21:20:54 hm right Feb 02 21:24:43 can I just added a void reset() { nGraphicsMode = 0; foregroundColor = TS_COLOR_FOREGROUND; backgroundColor = TS_COLOR_BACKGROUND; } to TSCellGraphicsState or does that increase the struct size for all of the TSCells when we'll only call it from TSGraphicsState Feb 02 21:25:07 trying to keep the default settings for any type in one place =x Feb 02 21:26:32 non virtual functions don't add to struct size Feb 02 21:26:51 although a function can be virtual by inheritance too Feb 02 21:27:09 i guess the charset stuff should be moved out of graphicstate Feb 02 21:27:27 the question is whether charset stuff can be saved/restored with any operation Feb 02 21:28:57 stbuehler, prob just the xterm alternate screens tuff Feb 02 21:28:59 stuff Feb 02 21:29:46 hm, konsole has it in save/restore Cursor Feb 02 21:30:30 although it doesn't look consistent what they are restoring Feb 02 21:31:33 they just save whether the current charset had graphics ("SPEC") (and "pound"); but they don't backup the actual charsets or the index of the current one Feb 02 21:34:32 xterm does as well I think (in save cursor) Feb 02 21:35:56 and xterm does not map '2' as graphics Feb 02 21:36:19 neither does konsole Feb 02 21:36:29 stbuehler, i know it doesnt, that was mac's console that does Feb 02 21:36:39 and i dont really see anything wrong with that Feb 02 21:37:25 hm k Feb 02 21:37:31 is there any spec on those? Feb 02 21:37:38 couldn't find 1 and 2 in vt420 rm Feb 02 21:38:06 i've only seen vt102 spec for the charsets Feb 02 21:38:24 is head supposed to be broken Feb 02 21:38:32 :D Feb 02 21:39:16 it shouldnt be Feb 02 21:39:22 Brybry, is it for you Feb 02 21:39:31 http://vt100.net/docs/vt102-ug/table5-11.html Feb 02 21:39:34 http://vt100.net/docs/vt102-ug/table5-12.html Feb 02 21:39:36 http://vt100.net/docs/vt102-ug/table5-13.html Feb 02 21:39:40 http://vt100.net/docs/vt102-ug/table5-15.html Feb 02 21:39:42 charsets Feb 02 21:42:24 vkb doesn't show up for me on head but I'm still looking into it, maybe my reset --hard didn't work out right Feb 02 21:44:16 well, i'm going to ignore the vt52 charset :P Feb 02 21:44:28 still didn't found '1' and '2' specs Feb 02 21:44:40 Brybry: had the problem too once, restarting wterm fixed it Feb 02 21:45:46 Brybry, i was trying to fix something on 1.4.5 with that latest commit Feb 02 21:45:59 i didnt test it on 2.x+ yet Feb 02 21:46:36 no JS errors though, it's all very weird Feb 02 21:46:44 i removed the _t suffixes now Feb 02 21:46:57 well,i will have after rebase :D Feb 02 21:47:11 ah, [20120202-15:46:43.793904] error: Uncaught TypeError: Cannot read property 'PalmSystem' of null, /usr/palm/frameworks/enyo/1.0/framework/build/enyo-build.js I can get that now if I spam show keyboard Feb 02 21:47:16 pretty useless error ._. Feb 02 21:49:43 make vkb loading happy on 1.4.5 Feb 02 21:49:51 its that commit that prob broke vkb for somet reason Feb 02 21:51:41 i suppose that can just get reverted Feb 02 21:51:49 since it doesnt look like we can support 1.4.5 Feb 02 21:53:21 * Brybry shrugs Feb 02 21:53:32 I'm going to go eat and then I'll hunt it down when I come back if you haven't reverted it or something Feb 02 21:54:09 i reverted it Feb 02 21:54:11 vkb shows again Feb 02 22:11:19 ok, i splitted charset handling from graphic state Feb 02 22:11:43 also graphic state has now a "color()" method to only select the colors from the current state Feb 02 22:11:50 you might want to use this in some of the erase functions Feb 02 22:25:10 stbuehler, this was probably related to the random mc weirdness Feb 02 22:42:25 ok Feb 02 22:42:41 konsole only saves the value of the currently selected charset, and only restores that one Feb 02 22:42:50 that fixes kismet Feb 02 22:43:13 at least the graphic chars Feb 02 22:54:25 i thought the "stuck key" thing was fixed Feb 02 22:54:36 but if i touch and move away, the key still is down Feb 02 23:04:52 ya its not fixed Feb 02 23:27:45 yay, tp crash again -.- Feb 02 23:40:32 stbuehler, what are u doing to crash it Feb 02 23:41:59 and jailer broken again Feb 02 23:42:08 just make ... test Feb 03 00:14:45 stbuehler, i still dont get how you have jail issues Feb 03 00:14:58 stbuehler, so.... xterm-256 color support? Feb 03 00:56:36 PuffTheMagic: uhm. do we really need that? it would cost another texture Feb 03 01:00:42 need, no, cool, yet Feb 03 01:00:43 yes Feb 03 01:00:44 :D Feb 03 01:07:24 perhaps some other time :) Feb 03 01:07:38 and gn8 Feb 03 01:08:49 night Feb 03 01:23:43 so what is the process of promoting something from the alpha feed to beta or public? Feb 03 01:23:56 git mv .. Feb 03 01:24:35 could someone do that for Xserver 9.7 in the alpha? Feb 03 01:25:00 since dtzwill has not been around lately Feb 03 01:25:01 has it been tested? Feb 03 01:25:16 yeah by a few people with out any issues reported Feb 03 01:25:29 looks like it's in beta already, not alpha Feb 03 01:25:55 must have missed that promotion Feb 03 01:26:54 does beta to Public require more testers and time? Feb 03 01:31:45 ya Feb 03 01:31:59 but I use it daily on my TP and have seen no issues Feb 03 01:33:13 I do as well but since I worked on it. I figured I needed more then my own testing. Feb 03 01:33:13 I'd leave it to dtzWill to make the decision on whether it's ready for public Feb 03 01:35:37 ok that works, if anyone by chance sees him around, could someone remind him that it is still in the beta feed Feb 03 01:46:07 I also need to get xserver to recognise an Opal as needing a virtual keyboard Feb 03 01:50:47 an 0pal? Feb 03 01:51:01 TouchPad Go 7" Feb 03 01:51:31 oh dang, totally forgot that was the name of the mythical device Feb 03 01:53:40 egaudet, now that enyo is avail for all devices and now that root plugins work on all devices I think I am gonna convert freeTether to a plugin and loose all the annoying service nonsense Feb 03 01:55:31 PuffTheMagic: you should be able to put a dependency on org.webosinternals.enyo now, since it won't overwrite an existing enyo installation. Feb 03 01:58:14 rwhitby, is there anyway we can add some sort of flag (extra metadata) in the feed like ENYO=1, that preware can detect and do a quick check to see if it needs to install enyo Feb 03 01:58:20 instead of making it a hard coded dep Feb 03 01:58:57 nope Feb 03 01:59:58 well, not without someone doing a whole heap of work on the service, app, and testing on the full range of devices and webOS versions on which Preware runs ... Feb 03 02:00:39 and hopefully making it somewhat generic, so we don't get 100 FEATURE= code additions for every little last difference between device capabilities Feb 03 02:21:25 could you cheat and make on device enyo a fake ipk that satisfies the dependency? Feb 03 02:32:13 PuffTheMagic, you could do that in the postinst though Feb 03 02:32:37 or better yet, have the app give a graceful error message if they don't have enyo Feb 03 02:33:00 something like "Enyo not found, if you have an older device you may need to install it from preware" Feb 03 02:33:11 i dont see why a flag like there is for REQUIRES LUNA RESTART couldnt be added Feb 03 02:33:25 a simple hasEnyo() service call would be easy to add Feb 03 02:33:36 and i dont see how it would have 100 difference device varriants Feb 03 02:33:40 PuffTheMagic, because REQUIRES LUNA RESTART is the same for every device Feb 03 02:33:56 it isn't device specific Feb 03 02:34:04 and it doesn't need to check if luna is installed Feb 03 02:34:06 etc etc etc Feb 03 02:34:21 an enyo check would need to be device specific Feb 03 02:34:30 its installed in the same place on all devices Feb 03 02:34:42 MIN_ENYO_VERSION=1.0 Feb 03 02:34:48 pretty easy Feb 03 02:35:07 now that flag would make sense actually Feb 03 02:35:16 since enyo 2.0 apps may not run on enyo 1.0 Feb 03 02:35:28 [ -d /usr/palm/frameworks/enyo/${MIN_ENYO_VERSION} ] **** ENDING LOGGING AT Fri Feb 03 02:59:57 2012