**** BEGIN LOGGING AT Wed Mar 25 02:59:57 2009 Mar 25 04:08:42 * mwester wanders over here, and says hi to raster Mar 25 04:13:35 mwester: is chooky an idiot? Mar 25 04:13:47 did i say something wrong to tick him off? Mar 25 04:13:49 seriously Mar 25 04:13:52 i cant spot it Mar 25 04:19:08 he's rather a jerk who seems to think he owns a channel and can "police" the activity/conversation on same. Mar 25 04:19:21 yeah Mar 25 04:19:31 other than that..i was tyring to find what i said that ticked him off Mar 25 04:19:41 Are just back for a while, or are you done with Korea? Mar 25 04:20:26 i'm gonna ignore it Mar 25 04:20:52 the png icon file for openmoko-contacts doesn't load for the desktop Mar 25 04:21:09 oh i was just in korea for 1 week beginning of feb Mar 25 04:21:12 then in brazil for 1 week Mar 25 04:21:17 will be in kr beginning of may Mar 25 04:21:41 Blu3: can you tell me the icon from the .desktop? Mar 25 04:21:44 Brazil? Cool! Mar 25 04:21:51 yeah, lemme hook up my fone Mar 25 04:22:47 Blu3: hopefully you will actually be pleasant, unlike chocky :) Mar 25 04:22:54 so at least i can get an idea of what/why is going on Mar 25 04:23:10 i've been pleasant ever since you've known me many years back :] Mar 25 04:24:03 Blu3: i know :) i just am sometimes amazed at how people can behave on irc Mar 25 04:24:48 http://stuph.org/openmoko-contacts.png Mar 25 04:25:14 i haven't inspected it so i have no bearing on why it won't load Mar 25 04:28:41 ooh noo Mar 25 04:28:47 i wanted the like from the .desktop file Mar 25 04:28:56 ie Icon=openmoko-contacts Mar 25 04:29:00 or whatever the line is Mar 25 04:29:00 the content of my desktop file? Mar 25 04:29:03 one sec Mar 25 04:29:07 and where in the fs that file actually is Mar 25 04:29:09 yeah Mar 25 04:29:10 bingo Mar 25 04:29:38 Icon=openmoko-dialer.png Mar 25 04:29:42 that's what works Mar 25 04:29:48 (whihc i was tryng to get chocky to give but he decided to accuse me of patronising him and launch into a tirade about me not listening) Mar 25 04:29:49 Icon=openmoko-contacts was the original Mar 25 04:29:49 hmm Mar 25 04:29:51 ok that works Mar 25 04:30:01 but what was it before (when it wasnt working) Mar 25 04:30:02 ? Mar 25 04:30:02 Icon=openmoko-contacts.png didn't work either Mar 25 04:30:05 ok Mar 25 04:30:15 where is dialer.png and where is contacts.png ? Mar 25 04:30:28 the icons (both of the above) are in /usr/share/icons/shr/86x86/apps/ Mar 25 04:31:02 aaah hmm Mar 25 04:31:04 both there Mar 25 04:31:07 crap Mar 25 04:31:10 therte goes that theory Mar 25 04:31:24 locate tells me there are also: Mar 25 04:31:26 /usr/share/icons/openmoko-standard/128x128/apps/openmoko-dialer.png Mar 25 04:31:26 /usr/share/icons/openmoko-standard/22x22/apps/openmoko-dialer.png Mar 25 04:31:26 /usr/share/icons/openmoko-standard/32x32/apps/openmoko-dialer.png Mar 25 04:31:26 /usr/share/icons/openmoko-standard/48x48/apps/openmoko-dialer.png Mar 25 04:31:45 crap Mar 25 04:31:53 but contacts.png is only in the above ../shr/86x86/ Mar 25 04:31:54 ok thi sisnt going to be an easy solve Mar 25 04:31:58 *nod* Mar 25 04:32:16 ok Mar 25 04:32:23 can i ask a favor - so9 i can reproduce this sanely Mar 25 04:32:29 sure Mar 25 04:32:40 1. can u tar up openmoko-contacts.desktop (is it that?) Mar 25 04:32:47 ok Mar 25 04:32:48 yes Mar 25 04:33:03 and also tar up /usr/share/icons/openmoko-standard Mar 25 04:33:08 ok Mar 25 04:33:18 so i can install here and hunt down at leats on current svn if it hunts this down right Mar 25 04:46:37 got some tarballs? Mar 25 04:46:50 http://stuph.org/broken-icon-issue.tar.bz2 Mar 25 04:46:53 i am wondering if this is a problem with the icon cache and its hash Mar 25 04:46:54 all in one Mar 25 04:46:58 cool Mar 25 04:47:32 i specifically included (and only included) contacts and dialer files Mar 25 04:48:43 poking into it Mar 25 04:50:15 the odd thing is, if i specify the contacts icon in dialer, then the dialer comes up correctly with the contacts icon Mar 25 04:50:24 i just can't get contacts to come up with the contacts icon Mar 25 04:51:43 hmm, odd, i "fixed" it Mar 25 04:51:48 hmm Mar 25 04:51:52 wlel no icon here Mar 25 04:51:57 but.. these icons are inside themes Mar 25 04:52:09 -Categories=Application;Office;PIM; Mar 25 04:52:10 +Categories=Application;PIM;Office; Mar 25 04:52:20 i've noticed this in other files Mar 25 04:52:27 eh? Mar 25 04:52:32 that made the icon work? Mar 25 04:52:33 the icon didn't show up unless Office was the -last- category listed Mar 25 04:52:35 yep Mar 25 04:52:38 oh Mar 25 04:52:42 u mean icon image Mar 25 04:52:44 or icon at all Mar 25 04:52:48 ie label with blank space above Mar 25 04:52:55 vs label + the right icon image? Mar 25 04:53:07 unless Office is the -last- entry for categories, then the icon doesn't show up Mar 25 04:53:58 this also prevents some desktop entries from showing up period. i.e. tangogps Mar 25 04:54:23 i had to tack on a Categories=Office; at the end of the .desktop file to get it to show up on the desktop Mar 25 04:54:49 ok Mar 25 04:54:52 i'd look at the section of code managing the desktop that deals with the categories Mar 25 04:54:56 thats more a matter for the system applications.menu Mar 25 04:54:58 oe doesnt provide one Mar 25 04:54:59 it seems to be inadvertently filtering things Mar 25 04:55:05 so the e-wm pacages happen to install a rough one Mar 25 04:55:13 thats one level of problem Mar 25 04:55:19 the other is finding the right image Mar 25 04:55:43 and i think the above may be a bug that was fixed recently Mar 25 04:55:46 the last e-wm/e-wm-utils deleted my /etc/xdg/menus/applications.menu Mar 25 04:55:54 i had to copy it off someone Mar 25 04:56:04 (or opkg did...) Mar 25 04:56:38 ok Mar 25 04:56:50 dialer appears here with no changes to category Mar 25 04:56:57 as does contacts Mar 25 04:57:04 it gest a default file icon (white paper) Mar 25 04:57:06 not the real one Mar 25 04:57:22 yup Mar 25 04:57:42 but as soon as you put Office as the last category, then the icon shows Mar 25 04:58:08 yuop Mar 25 04:59:25 i thnik that was a bug where things didnt get added to lists and hashes properly Mar 25 04:59:28 i think its fixed Mar 25 04:59:34 ok Mar 25 04:59:34 it seems to be here Mar 25 04:59:39 if u use the oe applications.menu Mar 25 04:59:47 and the .desktops for dialler/contatcs u provides Mar 25 04:59:52 and i unbstall the icosn fromy our tarball Mar 25 04:59:58 contacts and dialler are in my list Mar 25 05:00:21 i'm very unfamiliar with how the menu system works on here, i discover these things mostly by accident Mar 25 05:00:43 trial and error has gotten me a long ways in a couple days :) Mar 25 05:01:35 :) Mar 25 05:01:43 well i think this is fixed upstream in svn Mar 25 05:01:49 the icon doesnt show Mar 25 05:01:55 but thats because its inside theme dirs Mar 25 05:02:01 and those theme dirs wont show up as theme options Mar 25 05:07:26 aaah Mar 25 05:07:27 thats why Mar 25 05:07:32 the oe themes dont install .desktops Mar 25 05:07:52 index.theme diles actually Mar 25 05:07:59 they are .desktop (ini) formatted Mar 25 05:08:02 and describe the theme Mar 25 05:08:08 if u dont have one - u cant chose it Mar 25 05:09:51 ok Mar 25 05:10:07 so to have hte icon the right image.. u need to select the openmoko theme Mar 25 05:10:28 i'm using illume Mar 25 05:10:46 and the shr/86x86 icons are the only ones that are being used at present Mar 25 05:11:02 and now i added an index.theme to the openmoko-standard dir aand editied it appropriately ... Mar 25 05:11:04 i can select it Mar 25 05:11:10 and.. dialer and contacts have their proper icons Mar 25 05:11:18 *nod* Mar 25 05:11:29 shr is also an icon theme and would need the index.theme to make it correct Mar 25 05:11:34 it would work the same way Mar 25 05:12:07 so with an oe applications.menu Mar 25 05:12:37 both contacts and dialer icons appear Mar 25 05:12:38 with labels Mar 25 05:13:05 and i changed contatcs ot be Mar 25 05:13:09 Icon=openmoko-contacts Mar 25 05:13:14 category is unchanged Mar 25 05:13:19 Categories=Application;Office;PIM; Mar 25 05:13:28 what process is controlling the cache, how do i flush it? Mar 25 05:13:34 and it displays with its own correct icon (when i chose openmoko-standard as my icon theme) Mar 25 05:13:44 there is no process - its part of efreet Mar 25 05:13:47 its in ram Mar 25 05:13:50 ok Mar 25 05:14:03 efreet is a library implementing the xdg freedesktop etc. standards Mar 25 05:14:12 it handles hunting directories for files Mar 25 05:14:17 and buildign hashes/lists in ram Mar 25 05:14:21 e just asks efreet for this info Mar 25 05:14:28 efreet can rais events when thngs change Mar 25 05:14:35 (eg .desktop get modified, deleted or added) Mar 25 05:14:35 and this caching survives reboot? Mar 25 05:14:39 and e just responds Mar 25 05:14:44 no it doesnt Mar 25 05:14:51 its in ram and loaded from disk every time e starts Mar 25 05:14:54 (or restarts) Mar 25 05:14:54 ok, it can't be a cache bug then Mar 25 05:15:03 thus sending e a HUP signal is enough Mar 25 05:15:06 because i've restarted ~6 times in the last 48 hours Mar 25 05:15:15 full cold boot Mar 25 05:15:17 hmm Mar 25 05:15:23 let me send u some files Mar 25 05:23:40 http://www.rasterman.com/works-with-svn-icons.tar.gz Mar 25 05:23:42 try that Mar 25 05:23:45 untar from / Mar 25 05:23:53 you will ALSO want to (to get the icon images working) Mar 25 05:24:12 1. go to e's modules config dialog and find the configuration - Icon theme one and load it Mar 25 05:24:24 and 2. go to e's config under look and see icon theme Mar 25 05:24:41 use that - select shr or openmoko as icon theme (now i added index.theme files you can and it works) Mar 25 05:24:52 for me - with latest svn. it works like a charm Mar 25 05:25:27 k, gimme a few. other things on my plate at the moment Mar 25 05:27:05 raster: reclaimed domain? Mar 25 05:27:36 what domain was it? Mar 25 05:28:26 DocScrutinizer: reclaimed? Mar 25 05:29:16 well dunno. DNS been down? Mar 25 05:30:31 which dns for what domain? Mar 25 05:31:00 didn't you mention trouble with some DNS not resolving lately Mar 25 05:31:25 oh that was for al lcd supplier Mar 25 05:31:27 an Mar 25 05:31:36 they had a vaguely ok list of "small lcd's" Mar 25 05:31:39 like 2-5" Mar 25 05:31:39 ahh, tjought it's been Urs Mar 25 05:31:44 no no :) Mar 25 05:32:03 http://www.trulylcds.com/tft/index.html Mar 25 05:32:35 unknown server ;D Mar 25 05:33:08 yup Mar 25 05:33:20 i was just wondering if people know more nice listings of lcd's Mar 25 05:34:33 raster: btw still interested in the piezo thing? Mar 25 05:34:39 yes Mar 25 05:34:43 k Mar 25 05:34:50 i wanted to just get some more lcd sizes/specs Mar 25 05:35:00 see if i can find a large one with really slim "frame" Mar 25 05:35:06 to build around Mar 25 05:35:24 then need to inject some reality into it like battery sizes and board sizes, component thickness Mar 25 05:35:29 and allow for some padding etc. Mar 25 05:35:42 i see Mar 25 05:36:33 checked OLED? Mar 25 05:37:46 so far no - as i havent fouind any that are good res, the right size with enough specs Mar 25 05:38:20 still no off-the-shelf technology Mar 25 05:41:58 let's cut some sony oled tv-sets to size ;D Mar 25 05:42:41 hahahaha Mar 25 05:42:47 well there is Mar 25 05:42:54 i have a pdf for 1 oled creen here Mar 25 05:42:56 3.1" Mar 25 05:43:21 3.4" 480x727 Mar 25 05:43:29 sounds good Mar 25 05:43:35 not great dpi Mar 25 05:43:37 MTBF? Mar 25 05:43:38 but "ok" Mar 25 05:46:05 ah, and "bezel"? Mar 25 05:46:39 its big Mar 25 05:46:41 bad bad bad beze Mar 25 05:46:45 p18 Mar 25 05:47:05 :( Mar 25 05:54:28 is initramfs compiled in freerunner kernel now? Mar 25 05:54:56 I found CONFIG_INITRAMFS_SOURCE="" in .config Mar 25 05:55:16 does it mean initramfs is not embedded in the kernel? Mar 25 05:58:57 There is no initramfs in the standard Om kernels Mar 25 07:01:59 xrn, mwester: one can't disable initramfs at all. If the source is not specified, dummy initramfs filesystem is supplied. Mar 25 07:13:09 * PaulFertser wonders where borked /etc/asound.conf in FSO/SHR comes from. Mar 25 07:25:33 PaulFertser: what should it look like? (asound.conf) Mar 25 07:26:03 mrmoku: hm, it shouldn't be supplied at all i guess? Mar 25 07:26:12 mrmoku: recent alsa has sane defaults already Mar 25 07:26:17 mrmoku: unlike SHR :p Mar 25 07:26:56 mrmoku: the config OE supplies is incorrect. Uses dmix as the default output device. This is what makes recording from default device impossible. Mar 25 07:27:55 PaulFertser: fine then... because it is part of the alsa-states package... which we do not want anyway Mar 25 07:28:53 PaulFertser: hmm... isn't dmix needed to have parallel access to sound hw? Mar 25 07:28:54 mrmoku: in fact i'm thinking a lot about more sane way to solve alsa states issues. Too bad i have to do a day-time job today, but probably i'll be sane enough tomorrow to hack on it. Mar 25 07:29:30 PaulFertser: ok... me too is almost on his way to the office :( thanks for pointing out Mar 25 07:30:08 mrmoku: i guess default alsa conf has dmix already properly pre-configured. They have some kind of "templates" to do this on system-level. And if user needs some tweaking, he can use custom asound.conf, but for default device he should enable dsnoop too, to make recording work. Mar 25 07:31:06 PaulFertser: ok... will try to get rid of alsa-state pack :-) Mar 25 07:31:42 mrmoku: that's not a definitive advice, as i'm not an alsa-magician, but still. It makes problems, something should be done with it. Mar 25 07:32:56 PaulFertser: ok. If it makes problems we have to convince DocScrutinizer anyway to do a sane asound.conf :P Mar 25 07:33:10 so we can remove the insane one and see what happens Mar 25 07:37:25 morning Mar 25 07:37:33 mrmoku: stop talking behind my back when I'm asleep ;-) Mar 25 07:38:10 DocScrutinizer: continue to sleep... have not finished yet ;) Mar 25 07:38:15 spaetz: morning Mar 25 07:40:45 * spaetz checks if libeflvala repaired itself over night Mar 25 07:42:29 spaetz: finally managed to compile your vala-settings Mar 25 07:42:50 need to setup the cross compile env now Mar 25 07:46:56 mrmoku: you in fact need dmix on FR. usually this is done somewhere in /usr/share/alsa/*.... Mar 25 07:47:19 for mic you need to use dsnoop, or at very least Mar 25 07:47:37 split upstream and downstream for default Mar 25 07:47:57 and insert dmix for downstream (playback) only Mar 25 07:48:21 while piping upstream directly fro hw:0,0 Mar 25 07:48:48 dmix is not suited for recording Mar 25 07:49:10 dsnoop does for record what dmix does for playback Mar 25 07:49:29 DocScrutinizer: i thought alsalib supplies a configuration like that by default already. Mar 25 07:49:50 asound.conf isnät taking care about recording at all Mar 25 07:50:23 that's not alsalib, that's config in /usr/share/alsa Mar 25 07:51:00 well you might call that part of alsalib Mar 25 07:51:57 mrmoku: I have problems with libeflvala Mar 25 07:52:10 configure fails with that: http://dpaste.com/18853/html/ Mar 25 07:52:26 night again Mar 25 07:52:43 it installs fine on my desktop, but in the OE environment, it fails with Mar 25 07:52:46 checking for C compiler default output file name... a.out Mar 25 07:52:49 checking whether the C compiler works... configure: error: cannot run C compiled programs. Mar 25 07:52:52 If you meant to cross compile, use `--host'. Mar 25 07:53:08 it compiles just fine if I add a "--host=arm-linux" to configure Mar 25 07:53:21 good Mar 25 07:53:22 so I am at a loss why OE doesn't get it right Mar 25 07:53:23 morning Mar 25 07:53:36 morning Mar 25 07:54:02 spaetz: but there is a recipe for it... Mar 25 07:54:06 and that does not work? Mar 25 07:54:12 I know, I have written it :-) Mar 25 07:54:33 spaetz: I'm talking about the one in upstream OE (which might have been written by mickey|zzZZzz instead) :P Mar 25 07:54:50 ahh, haven't seen that. Mar 25 07:54:52 * spaetz checks Mar 25 07:55:23 it's *not* in the ms5.5 branch... main branch instead Mar 25 07:55:32 but compare that one to yoursw Mar 25 07:55:33 -w Mar 25 07:55:44 ptitjes: morning Mar 25 07:55:45 nahh, I don't see any eflvala Mar 25 07:55:59 * mrmoku thinks ptitjes is talking tooo much with DocScrutinizer Mar 25 07:56:18 could you not talk so loud, men! Mar 25 07:56:21 it is early! Mar 25 07:56:35 yup !!! Mar 25 07:56:45 spaetz: openembedded/recipes/freesmartphone/libeflvala_git.bb Mar 25 07:56:54 OE did move packages --> recipes Mar 25 07:57:06 why do I talk too much with you, DocScrutinizer ? Mar 25 07:57:16 ptitjes: because you stop to sleep too :P Mar 25 07:57:24 arf Mar 25 07:57:24 in the dev branch? Mar 25 07:57:28 spaetz: yep Mar 25 07:57:40 * spaetz just checked MS5.5 branch so far. will look Mar 25 07:57:44 mrmoku: did not have the choice! (if that were me...) Mar 25 07:58:33 * DocScrutinizer missed his 15min window after his goodnight coffee. Considers staying awake now :-S Mar 25 07:58:58 mrmoku: bah :-) Why did I spent all day yesterday fiddling with automake/conf? ... Mar 25 07:59:05 * spaetz checks that recipe Mar 25 07:59:11 DocScrutinizer: wow! arn't you mad ? Mar 25 07:59:17 spaetz: to learn something? :P Mar 25 07:59:25 bah :-) probably Mar 25 07:59:29 call me maddoc Mar 25 07:59:49 arfarfarf Mar 25 07:59:51 :D Mar 25 08:01:22 senile bettflucht Mar 25 08:01:48 hehe Mar 25 08:02:05 * mrmoku wonders how old DocScrutinizer is... to already feel 'senil' Mar 25 08:02:48 sometimes I just feel too old for all this sh.... Mar 25 08:02:50 ;-) Mar 25 08:03:40 well... I feel old when around midnight I have to go to bed... because I'm f.....g tired... and you guys still keep on talking for some hour :P Mar 25 08:04:45 anyway... Mar 25 08:04:49 * mrmoku --> office Mar 25 08:04:51 dammit, mickey .bb files works and it looks nearly as mine Mar 25 08:04:53 cu Mar 25 08:05:09 spaetz: hehe... compare it closely ;) Mar 25 08:05:11 cu Mar 25 08:05:20 cu Mar 25 08:09:04 nytowl: can you merge classes/vala.bbclass (from dev branch) to the FSO-MS5.5 branch? Mar 25 08:10:16 mrmoku: cu Mar 25 08:11:57 spaetz: are you working on packaging vala ? that is cool! Mar 25 08:13:03 ptitjes: I worked on libeflvala.bb and nearly became a nerve wreck and now I see that mickey has a working .bb file in the dev.OE branch Mar 25 08:13:19 arf Mar 25 08:13:24 so, I just copy that over until it hits the shr-unstable branch Mar 25 08:13:40 but yes, vala is in OE Mar 25 08:14:11 spaetz: does it use git.fso/vala.git#test ? :D Mar 25 08:16:36 nah, vala proper is still the one in FSO-MS5.5 which is vala-0.5.6 Mar 25 08:16:49 there is one vala-0.5.7.bb in the dev branch Mar 25 08:17:22 no git versions afaik Mar 25 08:17:33 so we would have to package that ourselves Mar 25 08:19:44 mrmoku: working at OM is speeding up aging by a factor 20 ;D Mar 25 08:20:08 hehe Mar 25 08:20:57 DocScrutinizer: Hehe Mar 25 08:21:08 DocScrutinizer: BTW where should I send the Cs and Rs? Mar 25 08:21:45 They also sent me wire. Do you want some of that as well? Mar 25 08:21:45 I'll toss you my addr later on Mar 25 08:21:52 Great, thanks Mar 25 08:22:15 ehlo stefan_schmidt Mar 25 08:22:22 wire :D no thanks Mar 25 08:22:28 moin Mar 25 08:22:32 Okay :-) Mar 25 08:23:00 DocScrutinizer: For replacement devices you'll need to talk to Steve/Chelsea yourself, though Mar 25 08:23:11 k Mar 25 08:23:26 mrmoku: have you managed to get vala-settings running on the FR? Mar 25 08:23:38 I'm still a little doubtful that they'll send me the devices... Mar 25 08:23:49 due to lack of proper DBus methods it's pretty much doing nothing on the desktop Mar 25 08:24:03 alphaone: r u going to accept mail delivery? Mar 25 08:24:14 DocScrutinizer: Yeah Mar 25 08:24:23 There were quite some requests for that Mar 25 08:24:35 seen Mar 25 08:24:56 I just need to figure out how to best handle the fees. Mar 25 08:25:29 that's the point Mar 25 08:25:42 well one of them Mar 25 08:25:55 Put stamps in the package :) Mar 25 08:26:00 Good old style :) Mar 25 08:26:12 Unless I can use OM dhl account... Mar 25 08:26:16 But I doubt it Mar 25 08:26:43 * stefan_schmidt too Mar 25 08:26:59 if they mail a FR they can put in a 10EUR ntoe too Mar 25 08:30:21 DocScrutinizer: Also I have been asked whether I could perform the Bass fix as well. Mar 25 08:30:37 At the moment I would tend to not do that Mar 25 08:30:39 alphaone: can you? ;) Mar 25 08:30:57 PaulFertser: Should be possible Mar 25 08:31:20 alphaone: too troublesome and time-consuming for a mass rework imho. Mar 25 08:31:40 yeah Mar 25 08:31:47 stefan_schmidt, alphaone: hello Mar 25 08:31:49 And too unsupported from OM Mar 25 08:31:51 ptitjes: Mar 25 08:31:54 hi :-) Mar 25 08:32:05 it's even much more work, chances are you get new buzz on headphones, and OM seems not to support that Mar 25 08:32:10 hi ptitjes Mar 25 08:32:19 what do you think of the planning like it is now ? Mar 25 08:32:22 guys Mar 25 08:32:24 PaulFertser: exactly Mar 25 08:32:24 ptitjes: Sending a mail to linuxhotel right now Mar 25 08:32:33 stefan_schmidt: cool Mar 25 08:32:42 * stefan_schmidt crosses fingers they have süpace for us Mar 25 08:32:44 space Mar 25 08:33:01 wanted to know if they have a little sound system in the conference room... Mar 25 08:33:16 ptitjes: For what? Mar 25 08:33:18 because I can take some stuff from my work Mar 25 08:33:45 if we are 30, 40, I will take some HF microphones Mar 25 08:33:58 ptitjes: Well, it a location where they normally have $BIGDOLLAR seminars. So I think they will have Mar 25 08:34:11 could you ask that anyway ? Mar 25 08:34:28 ptitjes: Do you really think we will get that much people? Mar 25 08:34:33 Can do Mar 25 08:36:03 stefan_schmidt: we are already 13, I know of one people that is not on the list. And in fact, a french that is on the list wrote me a mail directly to ask whether it was private or no. So I think when the date/place will be set, some other might tinker Mar 25 08:36:59 stefan_schmidt: so I would say we will be almost 20 Mar 25 08:37:29 ptitjes: yeah, 20 was what I was thinking about, too Mar 25 08:37:30 stefan_schmidt: and 20 me already make a lot of noise Mar 25 08:37:31 :D Mar 25 08:38:55 stefan_schmidt: do you think we should announce on the om-devel ml ? Mar 25 08:39:18 ptitjes: good question Mar 25 08:39:37 I like some focus there to get some stuff done Mar 25 08:39:48 Not sure if more people help with that... Mar 25 08:40:24 Let me first get an idea how much people we could have in the linuxhotel. Maybe this is a limiting factor Mar 25 08:40:26 stefan_schmidt: also juergbi (Vala's lead) might be interested to come too... I talked to him and seeing|talking with guys moving completely to Vala might interest him. So I have to keep him in touch. And obviously it is interesting for use too (FSO and SHR) Mar 25 08:41:14 ptitjes: That should be no problem. Appreciated even :) Mar 25 08:41:40 stefan_schmidt: I do understand your point. (stuff get done) But I would also like that to be an event to gather the community (you know what I mean) Mar 25 08:41:52 ptitjes: i know Mar 25 08:41:57 stefan_schmidt: if we can attract new devs that would be very cool Mar 25 08:42:01 ptitjes: Balance again Mar 25 08:42:26 ptitjes: Let's wait for an answer from the linuxhotel before we go further on this Mar 25 08:42:32 stefan_schmidt: yeah Mar 25 08:42:53 you are right this is the way to do things (first see if it is ok for the) Mar 25 08:43:10 stefan_schmidt: BTW I added lots of place for workshops have you seen ? Mar 25 08:43:11 :D Mar 25 08:43:39 do you want more ? Mar 25 08:44:20 ptitjes: nah, looks good :) Mar 25 08:44:34 so men: call for paper ;) Mar 25 08:44:47 gtg walk dog and go working Mar 25 08:44:54 cu /me ptitjes|work Mar 25 08:44:59 arf Mar 25 08:45:03 * ptitjes ptitjes|work Mar 25 08:46:10 stefan_schmidt: BTW I may take my tent Mar 25 08:46:31 weather will be beautiful at that time (in fact I hope) Mar 25 08:46:44 so... Mar 25 08:46:58 and I think some others will do the same... Mar 25 08:47:00 bye Mar 25 08:47:06 hmm Mar 25 08:47:27 sometimes I just feel too old for all this sh.... Mar 25 08:47:43 spaetz: no... not on the fr... just my desktop... and there only the main menu... the rest needs some strange dbus stuff I have not available on my desk :P Mar 25 08:48:44 right. I fixed it yesteraday so at least the power module screen comes up even without dbus Mar 25 08:49:06 yeah... that one yes... without much content :-) Mar 25 08:49:16 by adding some dbus error checking. But it disables all widgets that don't have dbus available, so it looks pretty empty :) Mar 25 08:49:55 can you trigger a feed rebuild by now? or is fakeroot stilla cting up? Mar 25 08:50:23 connectivity page can enable/disable GSM/BT/Wifi by now Mar 25 08:50:44 still pondering what should be next on the priority list :) Mar 25 08:51:39 spaetz: feed is building... just the image not Mar 25 08:51:46 can update and trigger if you want Mar 25 08:52:05 ahh, k. would be cool. to see if vala-settings builds fine on the build host now Mar 25 08:52:11 probably just need to update Mar 25 08:52:24 as I modified libeflvala half an hour ago Mar 25 08:53:32 build is running Mar 25 08:53:37 thanks Mar 25 08:54:43 * spaetz wondrs what can be done about the fakeroot problem Mar 25 08:56:28 spaetz: welll. there was khorben here yesterday... looks like he had the same problem and solved it Mar 25 08:56:34 but me was away :( Mar 25 08:56:59 mmh Mar 25 08:57:08 ~seen khorben Mar 25 08:57:08 khorben was last seen on IRC in channel #openmoko-cdevel, 16h 48m 47s ago, saying: 'hi'. Mar 25 08:58:41 hehe... and me moves back to home office :-) Mar 25 08:58:50 hehe Mar 25 08:58:58 * mrmoku likes home office :D Mar 25 08:59:00 short office stay Mar 25 08:59:21 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=229050 Mar 25 08:59:24 something like this? Mar 25 08:59:29 yeah... my colleageue got an appointment I did not know off... so Mar 25 08:59:32 nothing to do here Mar 25 08:59:44 multithreaded apps cause this fakeroot issue? Mar 25 09:00:59 spaetz: yeah... looks like our issue Mar 25 09:01:10 what I don't understand... it worked until some day ago Mar 25 09:01:29 and what I don't know... how to switch OE to do tcp instead Mar 25 09:01:38 spaetz: ponder about that while I'm moving home :P Mar 25 09:01:44 bbiab Mar 25 09:07:02 morning Mar 25 09:08:44 mrmoku: We believe that the bug you reported is fixed in the latest version of Mar 25 09:08:51 fakeroot Source-Version: 1.5.13 Mar 25 09:09:21 also, I had a left over parenthesis in the configure.in in vala-settings which made the compile fail Mar 25 09:09:33 I pushed a fix. can you retriger a feed build please? Mar 25 09:09:57 what fakeroot version is on the build host? Mar 25 09:17:24 spaetz: :-) Mar 25 09:17:43 spaetz: the nice thing about my office is... it's just 5min away by bike Mar 25 09:18:19 hi Mar 25 09:18:47 spaetz: OE is using fakeroot-native me thinks... Mar 25 09:18:52 that's really cool. Mar 25 09:19:13 * spaetz checks the version of fakeroot-native in OE Mar 25 09:19:15 and that would be 1.9.6 Mar 25 09:19:18 5 min by bike! Mar 25 09:19:27 /home/serdar/openmoko/shr/shr-unstable/tmp/staging/armv4t-angstrom-linux-gnueabi/usr/include/usb.h:330: error: previous declaration of 'usb_get_busses' was here Mar 25 09:19:35 what does this mean? Mar 25 09:19:44 takes me longer to get to our coffee machine downstairs :) Mar 25 09:19:46 spaetz: yep... that's one kind of luxus you can't pay for :-) Mar 25 09:19:49 or better, how do I get rid of this? Mar 25 09:20:33 serdar1: when do you get this... pastebin whole error log? Mar 25 09:21:09 fakeroot 0:1.9.6-r1 Mar 25 09:21:11 ok... build is running again Mar 25 09:21:12 fakeroot-native 0:1.9.6-r1 Mar 25 09:21:37 ok, so if this has been fixed in 1.5.13 we'd need to bump the fakeroot version quite a bit Mar 25 09:22:12 err Mar 25 09:22:14 no Mar 25 09:22:15 ??? me thinks 1.9 > 1.5 Mar 25 09:22:20 * spaetz needs more coffee Mar 25 09:22:31 spaetz: well have I nice trip downstairs then :P Mar 25 09:22:38 s/I/a/ Mar 25 09:22:38 mrmoku meant: spaetz: well have a nice trip downstairs then :P Mar 25 09:22:45 :) Mar 25 09:23:15 anyway. dev. branch has fakeroot_1.12.1.bb Mar 25 09:23:22 perhaps just bumping the version helps? Mar 25 09:23:44 spaetz: hmm... could give it a try, yeah Mar 25 09:26:59 spaetz: nop... not yet Mar 25 09:27:14 fakeroot (1.11) unstable; urgency=medium Mar 25 09:27:14 * communicate.c: patch from dann frazier to fix race condition when Mar 25 09:27:14 SysV IPC is used with a threaded application. closes: #499142. Mar 25 09:27:24 error: elm not found in specified Vala API directories Mar 25 09:27:24 error: evas not found in specified Vala API directories Mar 25 09:27:24 error: eina not found in specified Vala API directories Mar 25 09:27:47 ok, then libeflvala has not been installed correctly Mar 25 09:27:56 -c clean libeflvala Mar 25 09:28:13 heyho Mar 25 09:28:16 the .bb file that existed until this morning did not complain but it also did not install anything Mar 25 09:28:54 it *should* work with the current libeflvala. (I did not bump the PR version, just copied mickey's .bb file) Mar 25 09:29:11 sorry to bother you so much with my test apps :) Mar 25 09:29:20 morning morphis Mar 25 09:29:41 spaetz, I see you are very active with libeflvala :) Mar 25 09:30:02 morphis: ... a bit Mar 25 09:31:09 http://rafb.net/p/30qRY641.html can someone help me with this? Mar 25 09:31:48 hm FSOSHR'09 will be on 22/23/24 may ... :( Mar 25 09:32:13 where? Mar 25 09:32:35 checking LinuxHotel in Essen Mar 25 09:32:36 the place is open Mar 25 09:32:53 Essen sounds nice Mar 25 09:33:55 and a price is known? Mar 25 09:34:07 spaetz: you will be there? Mar 25 09:34:15 doubtful Mar 25 09:34:37 why? Mar 25 09:38:32 I travel already 3 weekends this months Mar 25 09:38:46 and I'd like to see my family at least once a months :) Mar 25 09:38:57 * mrmoku thinks spaetz is setting wrong priorities :P Mar 25 09:38:59 ok :) Mar 25 09:39:15 (not for the family... for the other travels) Mar 25 09:39:25 (family always has priority :) Mar 25 09:39:38 but I'm still thinking of going there Mar 25 09:39:51 other travels are day-job related (or visiting friends) Mar 25 09:39:52 hm maybe I can attend only on the 23rd Mar 25 09:40:48 * spaetz sees that mrmoku rebuild libeflvala on the build host , yay Mar 25 09:40:58 serdar1: sorry... don't know how to help with your build problem Mar 25 09:41:25 spaetz: yeah... step one is done Mar 25 09:41:35 now step2... vala-settings Mar 25 09:41:49 crossing thumbs. exciting Mar 25 09:42:04 My first usage of autotools Mar 25 09:42:20 suceeded Mar 25 09:42:32 \o/ Mar 25 09:42:35 the cleaning ;) Mar 25 09:42:50 ahh Mar 25 09:42:54 now cross you fingers :-) Mar 25 09:43:47 shit Mar 25 09:43:47 nop... was not it Mar 25 09:44:20 * spaetz wonders why I specify explitic --vapidirs in my command Mar 25 09:44:36 and why does it compile on my box flawlessly? Mar 25 09:44:55 * spaetz removes those Mar 25 09:46:27 ok pushed, rebuilding locally Mar 25 09:49:25 ok, removing my local libeflvala installation, I get the same build failure now Mar 25 09:49:35 somehow it had picked up my desktop installation Mar 25 09:52:52 I can see /shr-unstable/tmp/staging/armv4t-angstrom-linux-gnueabi/usr/share/vala/vapi/eina.vapi Mar 25 09:53:24 somehow it isn't found by the valac compiler Mar 25 09:57:19 spaetz: try setting --vapidir=${STAGING_DATADIR}/vala/vapi Mar 25 09:57:48 hmm... or the equivalent autoconf thingie Mar 25 09:57:54 mmh, good hint Mar 25 09:59:14 mmh STAGING_DATADIR is not set, but OPIEDIR points to OPIEDIR=/home/spaetz/src/shrbuildenv/shr-unstable/tmp/staging/armv4t-angstrom-li Mar 25 09:59:17 nux-gnueabi Mar 25 09:59:25 which is nearly right. Mar 25 09:59:56 * spaetz checks how to get STAGING_DATADIR set Mar 25 10:01:06 spaetz: do the autoconf way... which should be $(datadir)/vala/vapi Mar 25 10:02:54 right Mar 25 10:07:37 using $(datadir)/vala/vapi in Makefile.am results in an incredible --vapidir /usr/share/vala/vapi Mar 25 10:07:45 which is not correct either. Mar 25 10:07:50 I hate bitbake Mar 25 10:08:56 spaetz: why is that not correct? Mar 25 10:09:13 it needs to use >/shr-unstable/tmp/staging/armv4t-angstrom-linux-gnueabi/usr/share/ Mar 25 10:09:47 hmm... Mar 25 10:10:05 which is what $(datadir) should point to in OE build... yeah Mar 25 10:10:12 exactly Mar 25 10:10:43 I just did a -c clean -c build so it was using bitbake and still set $datadir to my local host datadir Mar 25 10:12:28 mickey|zzZZzz: awake? Mar 25 10:14:08 spaetz: think I know... Mar 25 10:14:17 ? Mar 25 10:15:19 $srcdir/configure "$@" Mar 25 10:15:54 see the "$@" Mar 25 10:16:02 you just do ./configure Mar 25 10:16:14 so bitbake can't pass arguments to it and it uses defaults Mar 25 10:16:24 (that's my theory at least) Mar 25 10:16:31 I've seen that line in some of mickey's autogen.sh files Mar 25 10:16:44 that's where I looked yeah :P Mar 25 10:16:44 I wonder if I should try to add it there... Mar 25 10:17:10 I am not even sure if autogen.sh is being called by OE or if it just calls autoreconf. But I'll try it out Mar 25 10:17:14 just try to add "@" as parm for configure in autogen.sh Mar 25 10:17:21 hmm what to do with this bluez problem? Mar 25 10:18:06 serdar1: is that a build from scratch? or some older buildtree Mar 25 10:18:19 actually, i think OE will call configure separately. Still worth a try Mar 25 10:18:57 spaetz: well you can check what it does... in run_configure Mar 25 10:28:00 mrmoku: it's from yesterday or from the day before Mar 25 10:29:51 serdar1: updated or from scratch Mar 25 10:29:54 ? Mar 25 10:29:59 scratch Mar 25 10:30:20 hmm... Mar 25 10:31:17 serdar1: you should ask mickey|zzZZzz when he passes Mar 25 10:31:38 mickey|zzZZzz: wake up! Mar 25 10:32:06 :) Mar 25 10:50:09 Serdar: you got me wrong. the usb.h should be in that bluez-package you're compilingfffjjffjfjf Mar 25 10:55:26 Hey there. Mar 25 10:56:44 Ainulindale: hey Mar 25 10:57:50 Ainulindale: did khorben tell you something? I was away when he queried me... and he was gone when I was back :( Mar 25 11:02:46 nah Mar 25 11:03:21 spaetz found out that there is a newer fakeroot in org.openembedded.dev... Mar 25 11:03:28 will try that one... Mar 25 11:03:38 Ainulindale: ping! should I schedule another IRQ for my patch? :) Mar 25 11:03:53 Raaaah forgot about that again Mar 25 11:04:34 I'll keep pestering you :) Mar 25 11:05:07 mrmoku: cool Mar 25 11:05:33 mmh, I pass STAGING_DATADIR to ./configure now Mar 25 11:45:34 ptitjes, Ainulindale: LinuxHotel reservation is positive. :) Mar 25 11:46:24 YAY Mar 25 11:46:27 ptitjes: For people who like to sleep in the hotel rooms they could offer 20 place right now. That get less over the time of course. So we need to send them a list asap for these rooms Mar 25 11:46:42 I'll end a mail with the details after lunch. Mar 25 11:56:40 stefan_schmidt: well for me it's ok Mar 25 12:02:47 stefan_schmidt: great :-) Mar 25 12:04:32 back! Mar 25 12:04:45 stefan_schmidt: good news Mar 25 12:05:13 ptitjes: "good news everyone". Mar 25 12:07:20 Ainulindale: ? Mar 25 12:07:55 Futurama reference. Mar 25 12:12:51 Ainulindale: arf Mar 25 12:12:54 too tired Mar 25 12:13:49 freesmartphone.org: 03mickey 07libeflvala * r681bad1ab327 10/vapi/ecore.vapi: vapi: EcoreFdHandler is new class, might make constructor a static method (inline with GDir and friends?) Mar 25 12:13:50 freesmartphone.org: 03mickey 07libeflvala * r7ed3c2a7b278 10/examples/library/mainloops.vala: Mar 25 12:13:50 freesmartphone.org: example/library: read from E mainloop Mar 25 12:13:50 freesmartphone.org: bidirectional communication indication seems to work now Mar 25 12:13:52 freesmartphone.org: 03mickey 07libeflvala * r77446381b648 10/ (7 files in 3 dirs): eflvala: move first bits into library Mar 25 12:13:55 freesmartphone.org: 03mickey 07libeflvala * rfa97bd857224 10/ (Makefile.am configure.ac examples/Makefile.am): make library optional; improve wording Mar 25 12:14:13 hmm... new fakeroot needs new libtool... :( Mar 25 12:14:26 Ainulindale: what do you think of this ? Mar 25 12:14:28 (09:41:03) ptitjes: stefan_schmidt: I do understand your point. (stuff get done) But I would also like that to be an event to gather the community (you know what I mean) Mar 25 12:14:28 (09:41:15) stefan_schmidt: ptitjes: i know Mar 25 12:14:28 (09:41:20) ptitjes: stefan_schmidt: if we can attract new devs that would be very cool Mar 25 12:14:28 (09:41:24) stefan_schmidt: ptitjes: Balance again Mar 25 12:14:56 * mrmoku thinks more like stefan_schmidt Mar 25 12:15:18 mrmoku: so no announce on om-devel list ? Mar 25 12:15:51 mrmoku: I'm open but I think this is a point to speak about Mar 25 12:17:00 ptitjes: hmm... is shr-devel not enough (and some fso list)? Mar 25 12:17:09 people interested in SHR read the list me thinks... Mar 25 12:17:59 and we don't want to do an 'advertise SHR' event (at least that's what I think), but 'get stuff done' (to use stefan_schmidt's words) Mar 25 12:18:40 of course having new devs is important too.... but I guess you won't get them by posting to om-devel... maybe Mar 25 12:18:59 mrmoku: hum yes that is a good point Mar 25 12:19:32 ptitjes: lets start with some 'core dev convention' and start expanding next year :P Mar 25 12:20:02 aggreed Mar 25 12:20:28 We can do a "get stuff done" event and mail the om-devel list for that Mar 25 12:20:39 It's not intended to be a demo convention that's for sure Mar 25 12:20:49 but that doesn't mean we can't attract new developers advertising the event Mar 25 12:21:04 So I disagree with your agreement Mar 25 12:21:09 :-) Mar 25 12:21:46 mrmoku: don't be affraid! he was like that all the evening yesterday Mar 25 12:21:49 lol :D Mar 25 12:21:50 at least it should be *clear* what it is... an event for devs that want to get things done Mar 25 12:21:59 ptitjes: yeah... I'm getting used to it :P Mar 25 12:22:20 we won't do anything at this convention Mar 25 12:22:25 we will share, we will talk Mar 25 12:22:32 but we won't be able to truly decide Mar 25 12:22:40 Because in the end we don't have to decide Mar 25 12:22:42 The community has Mar 25 12:22:47 which is the most important part of development Mar 25 12:22:50 We can plan ahead, and start to write down things Mar 25 12:22:55 And then explain it to the list Mar 25 12:22:59 For everyone to know Mar 25 12:23:04 But we won't get anything done per se Mar 25 12:24:17 mrmoku: please do not forget the rope for the convention Mar 25 12:24:37 ptitjes: for Ainulindale ? Mar 25 12:24:55 I heard there were enough trees in the park :D Mar 25 12:25:24 ptitjes: you french think always about revolution :P Mar 25 12:25:32 Heh, guys, I'm doing my job, if you're not happy with it, fork SHR ! Else, bear my opinion. (Though if you don't bear it it doesn't matter to me, as I will still stand by it) Mar 25 12:25:46 Ainulindale: !!!! Mar 25 12:25:51 What? Mar 25 12:26:06 humour! :) Mar 25 12:26:13 Duh, you can't even spot a troll when you see one Mar 25 12:26:25 No wonder you're not able to understand me Mar 25 12:27:33 Ainulindale: I can tell you're really prettier and smarter than a troll Mar 25 12:27:45 Plus my ass is as sweet as a towel Mar 25 12:27:50 To sum up, anyway Mar 25 12:28:04 The whole point of this convention isn't to decide as is Mar 25 12:28:10 To decide between us what we would like to do, sure Mar 25 12:28:14 Ainulindale: arf! I did not know this but I'm not interested in knowing more :) Mar 25 12:28:18 But this has to go through the normal "community" process Mar 25 12:28:29 So the main point is to share, exchange, plan ahead Mar 25 12:28:45 And be able to share technical point of views, talk about what we have, what we lack Mar 25 12:29:14 Because as is I can already picture you talking mickey|aalborg and alphaone aside and convincing them to redevelop frameworkd using brainfuck or caml Mar 25 12:29:23 (Because you're like that) Mar 25 12:30:05 Being enthusiast is great, just ensure that you're not overdoing it, that's what I keep telling you =) Mar 25 12:30:32 Else we will keep zigzaging between ideas and concepts and will never manage to deliver something coherent Mar 25 12:30:50 (But again) Mar 25 12:30:54 (This is my opinion) Mar 25 12:31:07 (And I'm not god) Mar 25 12:31:14 I do aggree Mar 25 12:31:16 (As I'm better than him/it anyway) Mar 25 12:31:26 hi mickey|aalborg Mar 25 12:33:01 yo serdar1 Mar 25 12:33:50 ptitjes: You agree with what? Mar 25 12:33:54 Just to be sure :-) Mar 25 12:34:10 (Ah, and ptitjes, this is far more comfortable as you can't interrupt me thinking that you know what I will tell :-p) Mar 25 12:34:28 mickey|aalborg: do we have a problem with bluez? Mar 25 12:34:40 (But please be sure that I'm not condescending or anything, just trying to make sure we understand each other clearly) Mar 25 12:34:59 serdar1: no idea Mar 25 12:35:02 do we? Mar 25 12:36:27 Ainulindale: you are not go Mar 25 12:36:29 god Mar 25 12:36:30 :) Mar 25 12:36:36 Ainulindale: no seriously Mar 25 12:36:44 Ainulindale: I completely agree about that community thingie... though I thought our community is shr-devel? Mar 25 12:36:58 I know I'm an enthusiast Mar 25 12:37:11 However there are missing things here and there Mar 25 12:37:15 do you have some kind of access to ml management of shr-devel? to see the list of recipients? Mar 25 12:37:57 so either we are pissing code that have no way to evolve in one year because this has not been thought in the long term Mar 25 12:38:10 either we are doing quality|long term stuff Mar 25 12:38:20 and here is why I propose the things that I propose Mar 25 12:38:34 and as I understoud it you like what I propose Mar 25 12:39:19 but anyway I make great efforts to not go in every directions :) Mar 25 12:41:16 ptitjes: and I agree with you that we have to plan ahead Mar 25 12:41:27 I just disagree on the consequences of that planification Mar 25 12:41:34 To me, we can't decide just because we meet each other Mar 25 12:41:40 We can plan, that we can do Mar 25 12:41:46 Then we can inform others we intend to do that Mar 25 12:41:50 If nobody reacts, we do that Mar 25 12:41:58 Else, we justify ourselves or try to compromise Mar 25 12:42:11 In any case, the fact that we keep others informed is good for everyone Mar 25 12:42:20 Because if yesterday we decided to do a backend for openvibe in FSO Mar 25 12:42:28 There's no way mickey|aalborg or anyone would be informed about that Mar 25 12:42:32 But for a mail Mar 25 12:42:39 Or any formal transcript of our talk Mar 25 12:42:47 I'm boring and I know that with mails Mar 25 12:43:03 But this isn't optional if we want to make sure everybody knows what is going on Mar 25 12:43:23 (Proof of what I say, I had no idea about some stuff we talked about yesterday) Mar 25 12:43:51 Anyway, I'm going to make myself a coffee injection, as I need it badly Mar 25 12:44:23 * ptitjes takes a coffee too Mar 25 12:44:36 * mrmoku already had his coffee Mar 25 12:45:41 * mickey|aalborg looks around to see whether this airport serves coffee... Mar 25 12:48:48 coffee at an airport! Mar 25 12:48:51 unheard of :-D Mar 25 12:49:44 mrmoku: "his"? Mar 25 12:49:50 You take only one coffee? Mar 25 12:50:03 Man that's why you keep being overwhelmed by ptitjes :-p Mar 25 12:50:12 :D Mar 25 12:50:18 Ainulindale: well... the one after lunch... had lots of coffee before Mar 25 12:50:27 This is my almost my fifth coffe :) Mar 25 12:50:31 remember I'm married with an italian women Mar 25 12:50:48 haaa Mar 25 12:50:56 Ainulindale: BTW we are talking of mails Mar 25 12:51:27 could you please do the (little) work needed to open a shr user ml ??? Mar 25 12:51:44 ptitjes: why that? Mar 25 12:52:14 ptitjes: the problem is that we have to think about that before as we would need to migrate users from current hosting to new hosting first Mar 25 12:52:16 keep the devel list for devel Mar 25 12:52:25 mrmoku: there are too much noise Mar 25 12:52:46 Nah you're exaggerating again Mar 25 12:53:00 Anyway, this implies a migration Mar 25 12:53:12 I'll do that but first bearstech has to tell me if they can host it Mar 25 12:53:14 And if they can Mar 25 12:53:18 They have to install it Mar 25 12:54:50 by the way ptitjes Mar 25 12:54:52 http://www.scriblink.com/index.jsp?act=phome&roomid=3334&KEY=BFD4E70E0EEA270D3EABC7356A72BBB8 Mar 25 12:54:57 this might be of interest to you too mrmoku Mar 25 12:55:11 yeah!! Mar 25 12:55:45 took me about two seconds to find it using google Mar 25 12:55:57 Java 1.5 or greater required (www.java.com). If you are using a corporate computer, a firewall may be blocking the site. Mar 25 12:56:02 :-) Mar 25 12:56:07 Ainulindale: because the actual ml is hosted on om ? Mar 25 12:56:14 ptitjes: ahem Mar 25 12:56:36 oh yes Mar 25 12:56:37 :) Mar 25 12:56:59 never noticed that :) Mar 25 12:57:31 Ainulindale: so yes that would be a good opportunity to get two mls Mar 25 12:57:52 Ainulindale: could you ask people from bearstech these days ? Mar 25 12:58:30 I already did that Mar 25 13:02:44 We're drawing obscene things Mar 25 13:03:17 *you* are drawing obscene things :P Mar 25 13:04:37 :-) Mar 25 13:04:47 plus I do maths Mar 25 13:04:59 and you're not very good at it ;) Mar 25 13:05:03 Looks like latex the math editor by the way Mar 25 13:06:24 Ainulindale: it is writen in the help tab of the math editor. it is a letex math editor Mar 25 13:07:04 s/letex/latex Mar 25 13:08:20 who connected to the whiteboard? :-) Mar 25 13:09:30 By talking about obscenity you might them curious ;) Mar 25 13:09:37 s/might/made/ Mar 25 13:09:37 mrmoku meant: By talking about obscenity you made them curious ;) Mar 25 13:10:22 Ainulindale, mrmoku, ptitjes: My feeling tells me that we have the major part of active devs already on the list, right? I have no problem if we send a mail to om-devel. We may be limited due to space reasons which then would mean that people that are already on the list come first. Mar 25 13:10:32 stefan_schmidt: yep Mar 25 13:10:39 * stefan_schmidt will write a mail about the linux hotel status now Mar 25 13:10:49 stefan_schmidt: I aggree Mar 25 13:12:04 stefan_schmidt: did they give a max number of people Mar 25 13:12:05 ? Mar 25 13:12:39 ptitjes: Not yet. I'll ask thme right away. Mar 25 13:18:16 *yawn* Mar 25 13:22:28 hmm, reading about **coffee** :) Mar 25 13:30:26 ok, the room has enough space for 20 people Mar 25 13:30:43 If needed we should be able to get up to 22 or 23 Mar 25 13:31:05 On the wiki we have 14 so fare. Means still 6 to go. Looks ok. Mar 25 13:31:08 Mails is out Mar 25 13:33:35 mmh coffeee Mar 25 13:33:51 mickey|aalborg: will you be online later today? Mar 25 13:34:12 need some bitbake advice on how to use STAGING_DATADIR in a Makefile Mar 25 13:35:15 probably not, my plane departing @ 16 and when I'm home I need some rest Mar 25 13:35:18 mickey|aalborg: http://rafb.net/p/30qRY641.html Mar 25 13:35:18 mrmoku: I see that fakeroot-1.12.1 failed too :-( Mar 25 13:35:29 yep Mar 25 13:35:35 mickey|aalborg: ok, do you have a quick hint then? Mar 25 13:35:43 --vapidir ??? Mar 25 13:35:45 ptitjes: mailing are being created Mar 25 13:35:56 I need the STAGINF_DATADIR/vala/vapi there Mar 25 13:35:58 +lists Mar 25 13:36:12 check that libusb1 has been built Mar 25 13:36:20 mickey|aalborg: is Mar 25 13:36:28 mickey|aalborg: I did it myself yesterday Mar 25 13:36:39 then for some reason it picks up the wrong one Mar 25 13:36:45 the new one has the fixed declaration Mar 25 13:36:52 mickey|aalborg: with bitbake -c clean libusb; bitbake -c build libusb I hope this was right Mar 25 13:36:54 Ainulindale: cool! Mar 25 13:36:58 hmm Mar 25 13:37:06 is libusb = libusb1? Mar 25 13:37:23 oh there's a 1 Mar 25 13:39:05 mickey|aalborg: just out of curiosity (interesting topic). I've seen you were quite interested in using Vala tighly enough to avoid the use of GLib but rather rely on pure posix. Am I wrong ? If this is the way fso|you take, ow do you achieve this cleany ? Mar 25 13:39:35 s/ow/how/ Mar 25 13:39:36 ptitjes meant: mickey|aalborg: just out of curiosity (interesting topic). I've seen you were quite interested in using Vala tighly enough to avoid the use of GLib but rather rely on pure posix. Am I wrong ? If this is the way fso|you take, how do you achieve this cleany... Mar 25 13:40:15 ERROR: Please set the 'PERSISTENT_DIR' or 'CACHE' variable. Mar 25 13:40:18 wtf? Mar 25 13:40:29 ptitjes: I don't mind GLib, we're going to use it in lots of places. I do think though that I don't need te additional overhead of IOChannels if pure posix I/O is sufficient. Mar 25 13:40:44 i.e. if all i want is read from a file Mar 25 13:40:50 i don't need encoding Mar 25 13:40:51 nor filestreams Mar 25 13:41:06 (i.e. sysfs file) Mar 25 13:41:09 ok Mar 25 13:42:23 I need to set this variables? Mar 25 13:42:28 with what? Mar 25 13:48:14 vampi Mar 25 13:48:56 DNS zones for shr-project have been modified ptitjes Mar 25 13:49:12 admin interface will be up in no time at lists.shr-project.org Mar 25 13:49:23 When it'll be up I'll do the migration thing Mar 25 13:49:52 wow :-) Mar 25 13:51:07 What? Mar 25 13:52:19 freesmartphone.org: 03mickey 07libeflvala * r0417b49d649b 10/ (3 files in 2 dirs): examples: remove direct writing into the queue from mainloop example Mar 25 13:56:17 freesmartphone.org: 03mickey 07libeflvala * rd2869fbeb707 10/README: README: Elementary.h no longer needs patching Mar 25 13:58:42 Ainulindale: great! Mar 25 13:59:18 Ainulindale, mrmoku: was wondering if we should not create directly four lists as does fso. Mar 25 13:59:24 user Mar 25 13:59:24 dev Mar 25 13:59:28 kernel Mar 25 13:59:34 announce Mar 25 13:59:44 et commits!!! Mar 25 14:00:10 I'd welcome if you'd share the kernel list with us Mar 25 14:00:19 smartphones-kernel is for generic userland/kernel interfacing Mar 25 14:00:28 an attempt to bring different smartphone kernel hackers together Mar 25 14:00:30 ha! nice! Mar 25 14:00:36 mickey|aalborg: I rebitbaked now libusb1 too but I get the same problem Mar 25 14:00:37 working on getting openezx and htc folks on board as well Mar 25 14:00:54 héhé :) Mar 25 14:01:16 serdar1: that's sad. try the patch to workaround. Mar 25 14:01:16 I would dream to have an htc :) with FSO/SHR on it :) Mar 25 14:01:25 ptitjes: that's 5 ;-D Mar 25 14:01:26 it must not be a dream Mar 25 14:01:27 ptitjes: I was wondering about announce and for now we don't need it per se Mar 25 14:01:28 we're close Mar 25 14:01:30 which patch? Mar 25 14:01:33 the HTC Touch Pro rocks da house Mar 25 14:01:38 it just needs more standard interfaces Mar 25 14:01:40 As for kernel we need to talk about that Mar 25 14:01:49 As I agree with mickey|aalborg Mar 25 14:01:54 yeah Ainulindale announce was the fifth intruder in my list :) Mar 25 14:02:01 so user, dev, commit Mar 25 14:02:03 serdar1: i patched the declarations some weeks ago when I stumbled over this problem. should be in the git log somewhere Mar 25 14:02:08 commit is useless to me Mar 25 14:02:11 Ainulindale: wow... finally ml's will be under our control :-) Mar 25 14:02:17 Ainulindale: not to me! Mar 25 14:02:26 Too bad for you then Mar 25 14:02:27 Ainulindale: and other people that do code Mar 25 14:02:29 :D Mar 25 14:02:45 register with CIAbot Mar 25 14:02:47 I meant those who really work, you know Mar 25 14:02:49 For now we have no infra to do that anyway, so no for now Mar 25 14:02:50 :) Mar 25 14:02:50 i'd love to see your commits here Mar 25 14:03:03 mickey|aalborg: what do I need to do? Mar 25 14:03:20 two things, first register a CIAbot at cia.cx Mar 25 14:03:25 mickey|aalborg: isn't this a little-bit too much noise for openmoko-cdevel ? Mar 25 14:03:33 second, use a canned git script that triggers ciabot Mar 25 14:03:37 mickey|aalborg: there are already fso commits Mar 25 14:03:40 cia.cx doesn't exist Mar 25 14:03:47 well, whatever it's called these days :) Mar 25 14:03:52 ptitjes: those that really work use git to know whats going on :P Mar 25 14:04:03 ok cia.vc it seems Mar 25 14:04:09 we should think about renaming the channel anyways Mar 25 14:04:13 shr is so much more than openmoko Mar 25 14:04:14 imo Mar 25 14:04:15 heh Mar 25 14:04:21 héhéhé Mar 25 14:04:22 as is FSO Mar 25 14:04:22 and have accesses too Mar 25 14:04:29 why not move to #fso Mar 25 14:04:32 as well as vhosts Mar 25 14:04:32 ? Mar 25 14:04:42 opie?? ;D Mar 25 14:05:01 :p Mar 25 14:05:02 ptitjes: ptitjes@shr/developer/dumbass for example Mar 25 14:05:38 * spaetz doesn't like -cdeve as channel suffix anyway Mar 25 14:06:16 what's wrong with cdevel? Mar 25 14:06:40 no, it's -cdeve, right? at least that's what i joined Mar 25 14:06:51 mickey|aalborg: any pointer for this: Mar 25 14:06:51 (15:02:53) mickey|aalborg: second, use a canned git script that triggers ciabot Mar 25 14:07:18 duh Mar 25 14:07:25 char limit? Mar 25 14:07:31 probably :) Mar 25 14:07:32 shit Mar 25 14:07:59 ptitjes: want me to find out? Mar 25 14:08:09 ptitjes: e.g. http://cia.vc/clients/git/ciabot.bash Mar 25 14:08:16 mrmoku: do you know that ? Mar 25 14:08:24 i was about to suggest #fsoshrudevel-ometal Mar 25 14:08:54 ;) Mar 25 14:08:55 ptitjes: I was researching the cia stuff about two months ago... Mar 25 14:09:06 Now, how do I install this crap Mar 25 14:09:17 ptitjes: I'm already doing that you know Mar 25 14:09:26 So please don't do it, as I already registered the project and such Mar 25 14:09:40 cool Mar 25 14:10:22 Ainulindale: \o/ Mar 25 14:11:19 Ok seems it's configured. Mar 25 14:11:27 mickey|aalborg: do I have to replace that in all the subrepos? Mar 25 14:11:51 Ainulindale: the hook... yeah, me thinks Mar 25 14:11:53 s/replace/replicate Mar 25 14:11:57 +/ Mar 25 14:12:00 i have a git script from somewhere that does the CIAbot notification and commit mailing at once Mar 25 14:12:12 well the perl script does that Mar 25 14:12:13 you need to call it from every hook if you want different repos to appear in the message Mar 25 14:12:14 :o Mar 25 14:12:15 for now I put my mail in there Mar 25 14:12:30 there are ciabot scripts that do not transport the repo entitty Mar 25 14:12:38 so check that you have the "full" one Mar 25 14:13:44 Well let us see. Mar 25 14:15:15 Well doesn't seem to work. Mar 25 14:15:53 Ah ok I know why I think. Mar 25 14:16:40 mickey|aalborg: do you know of similar online services for bug bots ? Mar 25 14:16:43 Hmmm doesn't seem to work properly. Mar 25 14:16:57 ptitjes: not offhand Mar 25 14:17:09 we used to have one in #openmoko Mar 25 14:17:15 but i don't remember whether this was a private thing or not Mar 25 14:17:16 But there was too many bugs? Mar 25 14:17:16 Ainulindale: you know why you think? ne me though ;D Mar 25 14:17:31 Well I installed the perl script Mar 25 14:17:36 It's executable and all Mar 25 14:17:38 l8er Mar 25 14:17:41 And I commited to shr-overlay Mar 25 14:17:41 boarding soon Mar 25 14:18:20 Ah damn. I know why. Mar 25 14:18:22 why are sms not moved from SIM to nand? see #openmoko Mar 25 14:18:49 can anybody help out? Mar 25 14:19:52 DocScrutinizer: ??? that will come with opimd... someday :P Mar 25 14:20:07 oh o.O Mar 25 14:20:23 thought that's in FSO right now Mar 25 14:20:47 that is but we do not have any gui for that yet Mar 25 14:22:01 ptitjes: so what is user supposed to do when SIM is filled to limit? Mar 25 14:22:05 not even any non-gui stuff :P Mar 25 14:22:15 DocScrutinizer: DELETE Mar 25 14:22:24 what???? Mar 25 14:22:34 DocScrutinizer: remove SMS... Mar 25 14:22:48 ah you can remove on SIM Mar 25 14:22:51 ok Mar 25 14:22:58 yep Mar 25 14:23:02 too bad, but still... Mar 25 14:24:36 DocScrutinizer: yeah that's what all phone do :( Mar 25 14:25:09 nonsense. usually SMS are stored in phone mem Mar 25 14:25:18 not on SIM Mar 25 14:25:44 DocScrutinizer: yeah but when the phone mem is full, you have to delete. then you don't have any backup! Mar 25 14:25:48 I'm not sure calypso has any mem for that Mar 25 14:25:59 I want transparent backup|sync on my laptop Mar 25 14:26:05 s/on/with/ Mar 25 14:26:05 ptitjes meant: I want transparent backup|sync with my laptop Mar 25 14:26:58 (installing RPC::XML...) Mar 25 14:27:09 I want SIM being moved from SIM to NAND as soon as reception notification is processed Mar 25 14:28:26 s/SIM/SMS/ Mar 25 14:28:26 DocScrutinizer meant: I want SMS being moved from SIM to NAND as soon as reception notification is processed Mar 25 14:28:50 someone bitbaking lately? Mar 25 14:29:01 SHR: 03ainulindale * r64acf3df5737 10/test.commit: After RPC::XML install, this test should work. Mar 25 14:29:03 do you have some problems with python-ecore? Mar 25 14:29:13 Ok so we lack the git path. Mar 25 14:30:46 ptitjes: I don't get your point with phonemem full Mar 25 14:31:38 1. SIM isn't any backup with the max 8 SMS you can store Mar 25 14:32:11 2. every time any mem is full deleting is all you can do Mar 25 14:32:41 3. phonemem is rather large compared to standard size of SMS Mar 25 14:34:46 and 4. I want SMS in system mem (nand, usd), not in phonemem nor at very least on SIM Mar 25 14:35:44 DocScrutinizer: well that might be a thing of user preference... it is nice to have them in SIM when you switch the SIM between phones Mar 25 14:35:59 * mrmoku wants full control what to store where :-) Mar 25 14:36:07 what for??? 8 SMS? Mar 25 14:36:24 well I have at least 20 SMS on my SIM Mar 25 14:36:32 * mrmoku counts Mar 25 14:36:34 it is sim dependant Mar 25 14:36:40 I think I can store 10 on mine Mar 25 14:36:41 maybe short ones ;) Mar 25 14:37:01 mrmoku, the max sms value is showed on shr-settings Ithink Mar 25 14:37:34 Sharwin_F: hehe, good tip Mar 25 14:37:38 counting gave me 19 Mar 25 14:37:43 so I exagerated a log :P Mar 25 14:37:46 s/log/lot Mar 25 14:38:24 and shr-settings tells me I can store 30 Mar 25 14:38:33 Ok back online. Mar 25 14:38:40 I can tore 20 on mine Mar 25 14:38:42 DocScrutinizer: and S in SMS stands for short, no ? :P Mar 25 14:38:43 *store Mar 25 14:39:13 the concept of storing SMS on SIM is GSM-plaeozen and was abundoned when ther's been a few more k of flash in phone Mar 25 14:40:01 DocScrutinizer: anyway we have support for it Mar 25 14:40:13 you can't receive any more SMS when storage is filled, so storing them on SIM is a really bad idea Mar 25 14:40:20 DocScrutinizer: in fact it ever arrives by that way as this is a read buffer for clients Mar 25 14:40:25 I think the best would be to have all sms by default on phonemem and then copy the ones you want into the SIM, which may be handful Mar 25 14:40:32 DocScrutinizer: the idea is to manage it intelligently Mar 25 14:40:35 alphaone? Mar 25 14:40:53 if one need to keep on particular message in addition to buffering he should have the possibility Mar 25 14:41:21 ptitjes: I'm not talking about buffer Mar 25 14:41:27 hum 'intelligently' is a stupid word sorry Mar 25 14:41:42 I'm talking about freeing buffer maybe Mar 25 14:42:18 DocScrutinizer: then I'd like to have different selectable|combinable policies Mar 25 14:42:33 Nokia fixed that bug around 8y ago Mar 25 14:42:53 Ainulindale: yes? Mar 25 14:43:06 alphaone: do you know what you did to output the repo path for ciabot? Mar 25 14:43:19 (could you give me optionnaly your pl file and update hook?) Mar 25 14:43:22 I didn't do anything Mar 25 14:43:28 when 711p had the bug of not receiving more than 8 SMS despite empty phone storage Mar 25 14:43:28 Was a different script afaik Mar 25 14:43:52 Do you have access to it? Mar 25 14:43:55 7110 even (iirc) Mar 25 14:44:33 Ainulindale: Could be Mar 25 14:44:50 WIll have a look in an hour or two. Mar 25 14:44:56 Gotta attend a lecture now Mar 25 14:44:58 bye Mar 25 14:45:10 ok :-) Mar 25 14:46:57 DocScrutinizer: I was wondering if it would be possible to make a companion modem pc board that would fit in a larger freerunner (like openmokast). to have a second gsm+sim, communicating via, say, usbnet to the fr ? Mar 25 14:47:29 you're kidding on me? Mar 25 14:48:53 DocScrutinizer: I don't doubt it is possible. I would like to know from your expertise how easy or not it would be :) Mar 25 14:49:18 it's for sure possible (for me ;) to *design* one, but I don't think it's possible to *make* Mar 25 14:49:55 DocScrutinizer: say another arm+gsm chip Mar 25 14:50:30 ptitjes: sorry I don't get the picture Mar 25 14:50:50 fr - usb - additionnal arm + gsm Mar 25 14:50:58 what do you have in mind? Mar 25 14:51:18 sure, just hook up two FR by usb Mar 25 14:51:20 np Mar 25 14:51:37 DocScrutinizer: this is one use we were talking about yesterday with Ainulindale Mar 25 14:51:55 DocScrutinizer: we could even use adhoc Wifi Mar 25 14:52:04 sure Mar 25 14:52:17 DocScrutinizer: I like this use case a lot :) Mar 25 14:52:18 still I don't see the design goal Mar 25 14:52:29 for the integrated one ? Mar 25 14:52:55 integrated? Mar 25 14:53:12 (when I talk about complex use cases with ptitjes, it's as if I was throwing a bone at h ime) Mar 25 14:53:16 s/h ime/him/ Mar 25 14:53:16 DocScrutinizer: be able to have enough power to do software voicemail Mar 25 14:53:17 Ainulindale meant: (when I talk about complex use cases with ptitjes, it's as if I was throwing a bone at him) Mar 25 14:53:25 please be more specific. I can't answer this sort of fuzzy questions Mar 25 14:53:40 Ainulindale: ppp Mar 25 14:53:45 :p:p:p Mar 25 14:53:58 if-down Mar 25 14:54:56 DocScrutinizer: the idea would be the same use case than 2 connected phone. but both in the same case (package) Mar 25 14:55:20 glue them together Mar 25 14:55:41 DocScrutinizer: you would have two sim (say work & personnal). one master sim in the real phone, one in the daugther board Mar 25 14:55:45 :) Mar 25 14:55:56 but wait a minute, maybe two gsm ant that near to each other aren't a good idea Mar 25 14:56:14 Ok I redevelopped this thing. Mar 25 14:56:23 Ainulindale: could you launch a process on bearstech ? Mar 25 14:56:42 Ainulindale: a bug irc bot ? Mar 25 14:56:48 ptitjes: buy a dualsim Mar 25 14:56:58 or twinsim Mar 25 14:57:09 DocScrutinizer: a dual sim phone or adpater ? Mar 25 14:57:22 * DocScrutinizer alway confusing those Mar 25 14:57:34 DocScrutinizer: because with the adapter you can't use both simultaneously, am I wrong ? Mar 25 14:57:45 there are dualsims, which is basically two sims in one Mar 25 14:58:04 nope you cant. nor with dualsim Mar 25 14:58:56 SHR: 03ainulindale * r3a37622d155c 10/test.commit: New test with customized CIAbot for repositories. Mar 25 14:59:07 Damn. Didn't get my branch. Mar 25 14:59:11 freesmartphone.org: 03jluebbe 07framework * r87b827ce9909 10/ (3 files in 2 dirs): ophoned: use prefrences to configure the headset Mar 25 14:59:12 freesmartphone.org: 03jluebbe 07framework * r66e764b8a0fe 10/framework/ (4 files in 2 dirs): opreferencesd: enforce types in SetValue Mar 25 14:59:17 but afaik you can select line 1/2 for a dualsim Mar 25 15:00:03 for outbound calls Mar 25 15:00:17 I would love a phone that could handle two sims simultaneously, I had to carry two phones for years at my last job it was a pita, one personal line, one provided by the business. I suppose another user case would be for least cost routing between competing mobile networks. Mar 25 15:00:27 for inbound you forward business number to private one Mar 25 15:00:57 ask your provider about dual/twin-sim plans Mar 25 15:02:09 (one is two sims with same number, the other one *one* sim with *two* numbers) Mar 25 15:02:44 SHR: 03ainulindale 07shr-overlay * r646dd01c8553 10/test.commit: Test was inconclusive. Hope this one won't be. Mar 25 15:02:53 \o/ Mar 25 15:02:55 where's the problem with voicemail? Mar 25 15:03:12 DocScrutinizer: interesting! I had no idea this would be possible Mar 25 15:03:12 Ainulindale: :))) Mar 25 15:03:32 Ok, this is done with. Mar 25 15:03:38 SHR: 03ainulindale 07shr-overlay * r3f104aef422b 10/test.commit: Clean things up. Mar 25 15:03:50 two sims with same number have different implementations Mar 25 15:03:58 houra! Mar 25 15:04:03 Currently 4 projects are tracked Mar 25 15:04:09 shr-settings, ophonekitd, shr-overlay & shr-makefile Mar 25 15:04:12 either all phones ringing, or just the last one to register Mar 25 15:04:28 hm Mar 25 15:04:31 My solution doesn't handle branchs properly but who cares. Mar 25 15:04:39 For now it'll work; Mar 25 15:04:41 I cannot compile shr-unstable :'( Mar 25 15:04:54 DocScrutinizer: one problem with that is is that my personal and work numbers are on different networks and also anything that confuses the billing of the two would have to involve the company accounts department and all the pain that causes Mar 25 15:05:05 Ainulindale: forget branches. noone need this crap XD Mar 25 15:05:32 DocScrutinizer: ptitjes will complain. You know that. Mar 25 15:05:35 He always complains. Mar 25 15:05:39 tig|: so you're out of luck Mar 25 15:05:41 heh Mar 25 15:05:48 hey guys.. anyone who can explain me this message: edje_cc: Error. Cannot create buffer engine canvas for image load. Mar 25 15:06:05 tig|: or get a phone with native dual sim support Mar 25 15:06:27 Ainulindale: in fact we could share a samba disk. why do we need vcs ?? Mar 25 15:06:33 there are a few out there Mar 25 15:06:51 ptitjes: well your solution is troublesome that would imply infrastructure Mar 25 15:07:04 I'd rather use your dog to go back and forth your flat and mine with an USB key Mar 25 15:07:09 with the sources on it of course Mar 25 15:07:22 Or maybe you could print them and send them to me over the mail? Mar 25 15:07:31 DocScrutinizer: so I repeat my question. why do you think we cannot 'make' on from the fr ? Mar 25 15:07:38 hehe we called this "turnshuhnetz" Mar 25 15:07:51 *schuh Mar 25 15:08:09 "sports shoe network" Mar 25 15:08:22 Ainulindale: I love that! there was no better vpn before dog-transport-protocol Mar 25 15:08:23 very popular in the 80s ;) Mar 25 15:08:35 ptitjes: well when I was studying at the university Mar 25 15:08:57 We did some basic maths to prove that a dog with an hard drive would be faster than fiber over a certain distance Mar 25 15:09:10 héhéhé Mar 25 15:09:13 maybe *you* can, ptitjes. I don't see how this should look like and work Mar 25 15:09:35 given the hw err.. given Mar 25 15:09:58 hmm Mar 25 15:10:05 DocScrutinizer: well at first I was kindof simply thinking put a second fr card in an fr Mar 25 15:10:27 I cannot find the patch from mickey Mar 25 15:10:32 won't work due to RF situation Mar 25 15:10:32 DocScrutinizer: but that would be a pitty to have a second screen... Mar 25 15:10:46 DocScrutinizer: but that may be two dual same screen phone Mar 25 15:10:49 :) Mar 25 15:11:09 DocScrutinizer: oh! really ? that is bad Mar 25 15:11:14 sorry have to leave now Mar 25 15:11:23 bye Mar 25 15:14:57 ptitjes: could you please send me your dog? I have to commit something :P Mar 25 15:17:29 héhé Mar 25 15:17:34 Voyou go! Mar 25 15:17:52 Ainulindale: and on your server ? could you run a bot ? Mar 25 15:19:28 What for? Mar 25 15:19:31 Logging? Mar 25 15:20:02 bug tracker Mar 25 15:20:25 Ah, that. Well you're talking about details (again). Don't you have something more useful to do? =) Mar 25 15:20:50 Ainulindale: I like that because it ease community bug wrangling Mar 25 15:21:37 Yeah, it could also be great if we had naked ladies free of charge to give us some massages Mar 25 15:21:44 And if I won to the lottery Mar 25 15:21:50 One thing after another will you? Mar 25 15:21:52 good ideas! :) Mar 25 15:22:00 héhéhé Mar 25 15:24:05 SHR: 03Sebastian 07shr-overlay * r53ab7c574f88 10/openembedded/packages/ (2 files in 2 dirs): disable-library in libeflvala. set explicit --vapidir in vala-settings Mar 25 15:25:07 hey, that is a CIA tracker for the SHR-overlay? Mar 25 15:25:12 I think I missed something here Mar 25 15:25:38 No Mar 25 15:25:43 I'm manually pasting the messages Mar 25 15:25:50 Thought it would be useful Mar 25 15:26:06 muahaha Mar 25 15:26:17 hehe Mar 25 15:26:17 (d'hib) Mar 25 15:27:43 spaetz: you missed nothing. this was the first meaningful commit Mar 25 15:28:07 Sheesh Mar 25 15:28:09 that's CIA'd Mar 25 15:30:28 Shhh, here he is Mar 25 15:31:43 hu? Mar 25 15:32:10 hmm as there won't be livestreams @fsoshrudcon would you mind doing an amateur audio recording of the opening talks and put them on the net? Mar 25 15:32:31 bumbl: we don't know if there won't be Mar 25 15:32:39 I think it's possible given linuxhotel facilities Mar 25 15:32:55 stefan_schmidt: could you confirm? Mar 25 15:33:01 it's written in fritz Mar 25 15:33:03 I can't understand Mar 25 15:33:17 Ainulindale: Confirm what? Mar 25 15:33:46 What are their facilities Mar 25 15:34:19 http://www.linuxhotel.de/tagungsraeume.html <= Mar 25 15:34:21 the chair on the right Mar 25 15:34:23 it'll be mine Mar 25 15:34:31 I'll bring the white cat to go with it Mar 25 15:34:42 (I always go around with my evil laugh so I won't forget that) Mar 25 15:34:44 Ainulindale: read last line of the FSOSHR' wiki page Mar 25 15:34:56 Ainulindale: I can't confirm this. It's already booked for me :) Mar 25 15:35:02 FSOSHRUDCON you mean ptitjes? Mar 25 15:35:15 stefan_schmidt: confirm what they have to stream the convention I meant Mar 25 15:35:30 Ainulindale: ah, now it makes more sense. :) Mar 25 15:35:48 Streaming, hmm. I don't think they offer a camera or such. Mar 25 15:35:51 Seems they have SDSL Mar 25 15:35:52 Ainulindale: well it does not make sense if you have to prepare much - waste of time != good Mar 25 15:35:56 Ainulindale: I shorcut it everytime, soon it will be F|S' Mar 25 15:35:56 So if we bring a camera Mar 25 15:35:58 It'll be doable Mar 25 15:36:08 If somebody will care about it we should be able to bring camera and mirco to do it. Mar 25 15:36:13 bumbl: well it's not that time consuming Mar 25 15:36:20 Ainulindale: I may get a good quality camera Mar 25 15:36:23 I'm sure we have enough brain cells to do that Mar 25 15:36:27 Just don't underestimate this. Streaming always turns out harder in the end. :) Mar 25 15:36:42 stefan_schmidt: well audio streaming is easy Mar 25 15:36:52 I could even ask bearstech to set up icecast Mar 25 15:37:03 streaming presentations is as easy Mar 25 15:37:04 I'm all for it as long as I don't have to care. Mar 25 15:37:11 Yeah so am I Mar 25 15:37:13 haha Mar 25 15:37:14 We have an SDSL line there so it should be possible Mar 25 15:37:23 That's why I'll name ptitjes the official referent for that Mar 25 15:37:28 That way it'll keep him busy Mar 25 15:37:31 ANd he won't bother us Mar 25 15:37:32 arggghhhh Mar 25 15:37:34 :-p Mar 25 15:37:35 * ptitjes runs Mar 25 15:37:37 We still would need an proxy outside for the load Mar 25 15:37:47 Yeah, bearstech could do that Mar 25 15:37:54 Who will care about the program anyway? Mar 25 15:38:15 Cool, if somebody takes care of it it would be a nice service Mar 25 15:38:32 Just asked them Mar 25 15:38:33 I think having recordings accessible online would be a good thing Mar 25 15:38:54 And pictures of the party Mar 25 15:39:01 I'd like to see pictures of ptitjes drunk somewhere Mar 25 15:39:03 yeah sure! Mar 25 15:39:06 Wearing nothing but his hat Mar 25 15:39:16 dreamer! Mar 25 15:39:23 me drunk ? Mar 25 15:39:25 So you think you're dream material? Mar 25 15:39:44 I'd rather say "freaking annoying nightmarish pain in the ass" Mar 25 15:39:55 But you already know that :-p Mar 25 15:40:06 I'll take that as a compliment Mar 25 15:43:06 youhou Mar 25 15:43:10 thank you Ainulindale Mar 25 15:43:36 #363 Mar 25 15:43:43 bug 363 Mar 25 15:43:55 SHR-WugBranler: bug 363 Mar 25 15:43:55 ptitjes: Error: "bug" is not a valid command. Mar 25 15:44:00 @!@ ticket 363 Mar 25 15:44:00 Ainulindale: Error: "!@" is not a valid command. Mar 25 15:44:05 @ ticket 363 Mar 25 15:44:06 Ainulindale: http://trac.shr-project.org/trac/ticket/363 Mar 25 15:44:29 @ 363 Mar 25 15:44:29 ptitjes: Error: "363" is not a valid command. Mar 25 15:44:33 arf Mar 25 15:44:47 @ ticket help Mar 25 15:44:47 ptitjes: Error: 'help' is not a valid integer. Mar 25 15:44:58 mrmoku: ok, both libeflvala and vala-settings should be fixed up so they compile Mar 25 15:45:10 spaetz: ok Mar 25 15:45:22 mickey had some ew stuff introduced that is not in any vala release version, so I had to disable some more stuff there too Mar 25 15:45:24 ptitjes: query Mar 25 15:45:27 (to the bot) Mar 25 15:45:28 (please) Mar 25 15:45:34 need to leave office now. BBL (probably) Mar 25 15:45:55 and let me end the conf Mar 25 15:48:53 tracfeed: Ticket #367 (/etc/resolv.conf gets spammed by nameservers) created  || Ticket #131 (QWO quick writing alternative to qwerty) closed  || Ticket #131 (QWO quick writing alternative to qwerty) updated Mar 25 15:50:07 tracfeed: Ticket #368 (Tell ptitjes to focus.) created Mar 25 15:50:39 alphaone|gone: I can set that one for you too and give you the conf for it Mar 25 15:50:51 mickey|zzZZzz, stefan_schmidt: ping Mar 25 15:50:54 That way you'll know =) Mar 25 15:51:01 Ainulindale: pong Mar 25 15:51:09 Ainulindale: Would be great Mar 25 15:51:10 tracfeed: Ticket #368 (Tell ptitjes to focus.) closed Mar 25 15:51:21 (FYI it's automatically checking it) Mar 25 15:51:32 Ainulindale: alphaone|gone Is soledring on some of his pet projects right now Mar 25 15:51:40 Yeah I got that :-) Mar 25 15:51:45 that's sweet Mar 25 15:53:02 spaetz, mrmoku, ptitjes: anything else while I'm at it? Mar 25 15:53:07 Maybe I should bring you a coffee? Mar 25 15:53:25 * mrmoku is still waiting for ptitjes's dog.... Mar 25 15:53:29 :-p Mar 25 15:53:39 mrmoku: there is some lag I think Mar 25 15:54:11 fidonet ;D Mar 25 15:54:22 hehe... Paris - Munich is not the distance where the dog wins the fiber me thinks ;) Mar 25 15:54:35 well it depends on the amount of data Mar 25 15:54:44 Is your changeset in petabytes? Mar 25 15:54:55 not yet... working on that :P Mar 25 15:54:59 "here fido, here fidoooh" Mar 25 15:55:30 "get me the slippers" Mar 25 15:55:38 SICK Mar 25 15:56:32 :D Mar 25 15:56:35 ptitjes: do you like the name wugbranler by the way? Mar 25 15:57:07 love it! Mar 25 15:57:10 happy document freedom day Mar 25 15:57:23 w00t? Mar 25 15:58:29 Ainulindale: you did not set the WONTFIX status before closing my ticket 368 Mar 25 15:58:41 ptitjes: I thought about naming it "Harpo" but I was wondering if you would like it Mar 25 15:58:44 :-> Mar 25 15:58:54 ptitjes: That's because I fixed it. Mar 25 15:59:04 I FIXED YOU HARD! Mar 25 15:59:11 huh! Mar 25 15:59:15 lol Mar 25 16:00:02 (by the way, standard supybot plugins, all of that) Mar 25 16:00:23 * mrmoku tries Mar 25 16:00:33 There's also a channel logger feature if we have somewhere to set that up Mar 25 16:00:50 Ainulindale: yeah! Mar 25 16:00:51 Because right now it's at home Mar 25 16:00:55 And I'm moving on saturday Mar 25 16:00:55 that would be great Mar 25 16:00:55 SHR: 03mok 07shr-overlay * raa12dca75b6c 10/openembedded/conf/distro/include/shr-autorev-unstable.inc: autorev: disable preferred version for libtool Mar 25 16:01:07 SHR: 03mok 07shr-overlay * rf4196ef97683 10/openembedded/packages/ (2 files in 2 dirs): Merge branch 'master' of git+ssh://git@shr.bearstech.com/shr-overlay Mar 25 16:01:27 mrmoku: pull before commit/push Mar 25 16:01:36 Baaad, baaad moku Mar 25 16:01:56 Ainulindale: bah... this is not the way it works :P Mar 25 16:02:14 Stop it, else I'll spank you Mar 25 16:02:32 spank me, spank me... I LOVE THAT ;) Mar 25 16:02:44 Meh. Mar 25 16:02:54 ptitjes has a bad influence on you Mar 25 16:03:12 Ainulindale: seriously speaking... yeah. for that small commit you're right... could have pulled before committing Mar 25 16:03:22 generally it is not always possible though Mar 25 16:03:22 and we only in non-RL yet!! Mar 25 16:03:27 Well I'm mostly right all of the time Mar 25 16:03:30 s/in/know in/ Mar 25 16:03:31 ptitjes meant: and we only know in non-RL yet!! Mar 25 16:04:03 Ainulindale: soo.... tell me what to do about that stupid fakeroot problem then ;) Mar 25 16:04:14 I have no idea. Mar 25 16:07:32 morphis: hi Mar 25 16:09:48 bumbl: hey Mar 25 16:10:42 Hm, frameworkd is hung and we still don't have an integrated pdb... Mar 25 16:13:08 All threads are at pthread_cond_wait and one thread in sem_wait... Doesn't tell much. Mar 25 16:13:14 Any ideas how to debug this? Mar 25 16:13:32 ptitjes: http://www.youtube.com/watch?v=pJqoRaphiEk Mar 25 16:13:38 morphis: two questions: why the hell python does not want to import pyelm if pythonpath is not set but it imports evas, edje,... which are in the same dir perfectly Mar 25 16:13:58 and any eta for List/genlist in pyelm? Mar 25 16:14:59 dont know .. hm delete the python-elementary egg and install it again, maybe its fix after that (I had also something like that sometime ago) Mar 25 16:15:12 List is implemented already and should work Mar 25 16:15:26 I testet it in my local version of the test app Mar 25 16:15:35 s/testet/tested/ Mar 25 16:15:39 morphis meant: I tested it in my local version of the test app Mar 25 16:16:41 about genlist ... hm maybe I can finish it until next weekend Mar 25 16:17:31 Ainulindale: wow! Mar 25 16:18:23 perfect Mar 25 16:18:29 yeah I knew you'd enjoy that =) Mar 25 16:18:35 that's when i should have time again Mar 25 16:19:33 Who was the brain behind shr-settings? Mar 25 16:19:41 dos1 Mar 25 16:19:47 ok Mar 25 16:20:01 (and the FSM of course) Mar 25 16:20:04 * stefan_schmidt is beating connman in shape. Would be great to have a setting for wifi Mar 25 16:20:30 hi is the shr developmen done in the testing or unstable branch ? Mar 25 16:20:44 unstable Mar 25 16:21:13 bumbl: time for what? :) Mar 25 16:21:43 coding and playing with the moko Mar 25 16:21:56 bumbl: yeah and with python-elementary :) Mar 25 16:22:10 at the moment i should actually learn for a test tomrrow and the day after tomorrow Mar 25 16:23:00 stefan_schmidt, can you tell harald about the spam problem in the gnufiish trac system? Mar 25 16:23:17 bumbl: good luck! Mar 25 16:23:21 hm Mar 25 16:23:23 i am lazy Mar 25 16:23:39 i will probably start studying tomorrow sharp before the test ;) Mar 25 16:23:40 morphis: I did already. :/ Will do again when he is back from his holidays. Mar 25 16:24:24 me too but until now I passed all semester without a second try for any test :) Mar 25 16:24:37 stefan_schmidt, ok Mar 25 16:26:54 stefan_schmidt, is he active somewhere in irc or have a jabber account? Mar 25 16:27:48 morphis: IRC not really, jabber from time to time. Mar 25 16:28:28 ah ok Mar 25 16:38:38 stefan_schmidt: Actually /me was attending a talk about Contiki Mar 25 16:38:51 pet project soldering comes now :-) Mar 25 16:39:29 alphaone: ah, right Mar 25 16:40:01 Any configure guru here? Mar 25 16:40:05 Was nice Mar 25 16:40:23 Seems like ignoring patches is normal there... Mar 25 16:40:26 Is ac_cv_path_DHCLIENT=/sbin/dhclient the same we --with-dhclient=/sbin/dhclient ? Mar 25 16:40:37 * stefan_schmidt has never seen the first notation Mar 25 16:42:10 you scared him Mar 25 16:42:57 autofoo scares people, not me :) Still the best we have. Mar 25 16:43:07 hehe Mar 25 16:43:27 --with-foo will map so some strange variable, but I don't know which one... Mar 25 16:43:46 SHR: 03ptitjes 07ophonekitd * r81b121f2ec79 10/ (ChangeLog Makefile.am autogen.sh gitlog-to-changelog): Generate Changelog from git log (script from Jim Meyering) Mar 25 16:43:54 Best look at other autofoo stuff for reference. That's how everyone else gets by as well :-) Mar 25 16:46:43 SHR: 03ptitjes 07ophonekitd * rf69eda5dffd6 10/src/Makefile.am: Fix org.shr.Usage service generation and GIR file generation Mar 25 16:47:28 Say, what does shr.Usage do what fso.Usage doesn't? Mar 25 16:47:59 SHR: 03ptitjes 07ophonekitd * rea3fb14ec47b 10/src/ (8 files): Use new libmodulo LifecycleError and ContainerError Mar 25 16:50:02 alphaone: shr.Usage has not been ported to Vala yet, so I'm not sure to answer correctly( Ainulindale and mrmoku please ??) Mar 25 16:50:04 :) Mar 25 16:50:27 alphaone: the goal is to place ophonekitd as the central place for SHR resource handling Mar 25 16:50:39 Hmm, okay Mar 25 16:50:39 And as ophonekitd requests the GSM resource Mar 25 16:50:44 and as if we truly want to release it Mar 25 16:50:49 we have to do it on behalf of ophonekitd Mar 25 16:50:53 ousaged is necessary Mar 25 16:50:53 So what is fso Usage interface missing? Mar 25 16:51:07 A way to request/release a resource on behalf of another process, maybe? Mar 25 16:51:54 hmm Mar 25 16:52:08 Would need to think about that Mar 25 16:52:14 That is problematic Mar 25 16:52:20 And that's why I put that in ophonekitd Mar 25 16:52:36 The only other way would have been to kill ophonekitd Mar 25 16:52:46 Or send it a signal somehow (which sums up to dbus anyway) Mar 25 16:54:33 There's no true way (that I know of) to release a resource on a "master" level Mar 25 16:54:45 E.g. two processes access the GSM resource at the same time Mar 25 16:54:49 And I'm boarding a plane Mar 25 16:54:54 I want to release the GSM resource Mar 25 16:54:56 I have to release both Mar 25 16:55:00 Hence my issue Mar 25 16:56:02 Actually for plane boarding I would just set the resource policy to disabled Mar 25 16:56:11 pull the battery. Mar 25 16:56:26 Since with auto you can't be sure that noone will request it again Mar 25 16:56:30 * mwester mutters something about lack of consistent kernel support for rfkill and GSM Mar 25 16:56:45 alphaone: and then if I want to release the resource in ophonekitd and not in the other software? Mar 25 16:58:14 so I am on the way home Mar 25 16:58:21 cya Mar 25 16:58:23 Ainulindale: well, depends. If resource management is auto it will stay on as long as one process needs the resource. Mar 25 16:58:40 I think I still don't quite understand your dilemma Mar 25 16:59:49 alphaone: if I need to disable the GSM handling in ophonekitd from an external software (e.g. shr-settings) Mar 25 16:59:53 I still need that kind of thing Mar 25 16:59:57 As i can't do it from FSO Mar 25 17:00:17 Ainulindale: it seems we could set the resource to disabled Mar 25 17:00:23 as alphaone just said Mar 25 17:00:30 that would fullfil our use case! Mar 25 17:00:42 Nope Mar 25 17:00:42 That would be globally disabled Mar 25 17:00:47 One of our use cases Mar 25 17:00:49 Not all of them Mar 25 17:01:08 Ainulindale: and in fact as alphaone says we can't prevent non-SHR apps to also request the resource on its behalf Mar 25 17:01:15 If you just want ophonekitd to release the resource and keep other processes using it you'll need something like that, yeah Mar 25 17:01:15 And that is wrong Mar 25 17:01:36 what is wrong? Mar 25 17:01:42 alphaone: what ptitjes said Mar 25 17:01:52 err Mar 25 17:01:53 how can you prevent it ? Mar 25 17:02:01 I misread what you said Mar 25 17:02:06 I saw "can" Mar 25 17:02:08 not "can't" Mar 25 17:02:11 So yes, we can't prevent Mar 25 17:02:15 And that is normal Mar 25 17:02:33 One thing we can't do is on an external point of view sever the link between a requester and its resource Mar 25 17:02:44 Ainulindale: so for your plane use case our actual strategy doesn't fit Mar 25 17:02:55 So is there a good use case in which you want ophonekitd to not use GSM, but other processes might? Mar 25 17:03:15 ptitjes: There's no solution for my two processes use case Mar 25 17:03:21 That's why to me ophonekitd has to be central Mar 25 17:03:28 Ainulindale: There is Mar 25 17:03:35 alphaone: Manual GPRS script? Mar 25 17:03:38 (for example) Mar 25 17:03:46 That's what resource policy disabled is there for Mar 25 17:03:58 No power to the modem -> no GPRS script Mar 25 17:04:01 alphaone: I wasn't talking about this one Mar 25 17:04:10 I was talking about the "no GSM call handling but GPRS please sir" Mar 25 17:04:12 which one then? Mar 25 17:04:20 ah, okay Mar 25 17:04:33 This use case is a bit crooked, yes, but still Mar 25 17:04:43 I'm sure I can come up with more realistic use cases Mar 25 17:04:48 I just don't have the time now to think about it Mar 25 17:04:52 But even then you would still need to handle PIN and registration Mar 25 17:05:02 alphaone: yeah but my script could do that anyway Mar 25 17:05:06 My point was Mar 25 17:05:13 If I have Resource <=> Requester Mar 25 17:05:20 I can't cut that link today if I'm not the requester itself Mar 25 17:05:42 alphaone: more realistic use case Mar 25 17:05:48 let us say we have a camera on the phone Mar 25 17:05:51 and I requested the resource Mar 25 17:05:57 then I'm calling someone Mar 25 17:06:08 And I'm rerequesting that resource somehow to do video conf Mar 25 17:06:43 I have to be ophonekitd to rerelease that resource without releasing the other one Mar 25 17:07:10 some kind of take over ? Mar 25 17:07:31 Well it would seem logical to me to say Mar 25 17:07:38 I want to list that resource users Mar 25 17:07:45 Then I want to sever the link with that specific user Mar 25 17:07:53 (force release, in fact) Mar 25 17:08:52 No, not that link alone Mar 25 17:09:05 Well that's my point Mar 25 17:10:02 Okay, that use case sounds interesting. Mar 25 17:10:18 Might be good to have that in ousaged at some point Mar 25 17:10:19 Don't you know I always say interesting stuff? :-p Mar 25 17:10:38 Ainulindale: I should know by now :-) Mar 25 17:11:04 (I hope you know that when I'm arrogant that way I'm purely acting) Mar 25 17:11:16 (Just to make sure) Mar 25 17:13:27 Ainulindale: I'll know for sure when I meet your arrogant ass at FUDCON ;-) Mar 25 17:13:51 But, yeah, I gathered that much. I'm dull, but not that dull. Mar 25 17:13:52 Meh :-) Mar 25 17:13:59 Must be all the solder fumes Mar 25 17:14:07 Well people often think I'm an arrogant piece of shit imbued with himself Mar 25 17:14:10 Really looking forward to the meetup Mar 25 17:14:26 So now and then I tend to make sure that people know that I'm not that type of guy Mar 25 17:14:31 Just an annoying ass at most Mar 25 17:14:38 Kind of like ptitjes but cuter Mar 25 17:14:44 (and with a bigger penis too) Mar 25 17:15:22 TMI Mar 25 17:15:39 TMI ? Mar 25 17:15:41 Ah yes. Mar 25 17:15:44 Too much info. Mar 25 17:15:46 :-) Mar 25 17:16:16 Well I tell you that because we set up a wet pants contest for May 23rd night Mar 25 17:16:22 So you'll be able to verify that information Mar 25 17:16:52 (Sometimes I also say things which aren't that dumb) Mar 25 17:18:01 well anyway alphaone Mar 25 17:18:11 @ rss tracfeed announce Mar 25 17:18:12 Ainulindale: Error: 'announce' is not a valid integer. Mar 25 17:18:20 damn ! Mar 25 17:18:51 @rss announce tracfeed Mar 25 17:18:53 Ainulindale: The operation succeeded. Mar 25 17:18:59 SHR-WugBranler: Thanks, dumbass. Mar 25 17:19:00 Ainulindale: Error: "Thanks," is not a valid command. Mar 25 17:19:26 There's nothing to announce it seems. Mar 25 17:19:29 Well anyway Mar 25 17:19:36 @ ticket 42 Mar 25 17:19:37 Ainulindale: http://trac.shr-project.org/trac/ticket/42 Mar 25 17:19:45 that's what I was willing to show you alphaone Mar 25 17:19:54 As I told to stefan_schmidt Mar 25 17:20:00 I can give you the conf for that and set that up for FSO too Mar 25 17:20:04 If you have a server it can run upon Mar 25 17:20:09 As I'll move by saturday Mar 25 17:22:43 That sounds interesting Mar 25 17:22:50 I've got a server Mar 25 17:22:59 fso git as well Mar 25 17:23:07 Though I don't know what you need Mar 25 17:23:42 Will that use fso profiles or own stuff? Mar 25 17:24:44 what do you mean? Mar 25 17:25:20 what does the profile selection actually do? Mar 25 17:25:44 Profile? Mar 25 17:27:50 alphaone? Mar 25 17:28:05 http://trac.shr-project.org/trac/ticket/42 Mar 25 17:28:07 Ainulindale: Mar 25 17:28:41 errr Mar 25 17:28:46 I was just showing the bot Mar 25 17:28:48 Not the ticket Mar 25 17:29:05 ah! Mar 25 17:29:08 Heh :-) Mar 25 17:29:12 Cool Mar 25 17:29:25 It's using trac RSS to output the tickets here Mar 25 17:31:08 stefan_schmidt: i'm kind of reimplementing shr-settinggs in vala and i'd be interested in connman integration too Mar 25 17:31:45 spaetz: cool, I'll let you know when I have something ready you can work with Mar 25 17:32:02 k Mar 25 17:32:02 Ainulindale: http://totalueberwachung.de/~alphaone/ciabot.pl Mar 25 17:32:11 alphaone: meanwhile I modified it =) Mar 25 17:32:17 but thanks anyway Mar 25 17:32:31 spaetz: did you mail the ML about that? Mar 25 17:32:51 k Mar 25 17:33:19 Ainulindale: no, so far it was an experiment to learn vala Mar 25 17:33:35 spaetz: I know, but anyway, I stand my ground :-) Mar 25 17:33:41 and it did not build until this afternoon Mar 25 17:33:41 Mail is necessary! :-) Mar 25 17:33:50 (I was talking about that with ptitjes yesterday) Mar 25 17:33:52 (that's why) Mar 25 17:37:41 hehe... Mar 25 17:37:47 * mrmoku won the fight with fakeroot ;) Mar 25 17:39:09 * mrmoku --> doing some german with the big one Mar 25 17:42:30 so once it's in the feed i'announce on the list Mar 25 17:42:43 mrmoku: yay Mar 25 17:46:34 oups I was lost in space-time continuum Mar 25 18:08:28 * mrmoku thinks *ptitjes* watches too much tv... *not* spaetz ;) Mar 25 18:09:09 erm. oops? openmoko-dialer3 exits after starting a phone call, keypad is no longer accessible and you can't hang up :-X Mar 25 18:09:12 mrmoku: wow! you are days of backlog late ?? Mar 25 18:09:39 ptitjes: well no... I'm referring to your space-time continuum comment :P Mar 25 18:10:02 (and remembered the 'spaetz watches too much tv' accusation) Mar 25 18:10:46 Blu3: there should be three buttons in the 'active call' screen, no? Mar 25 18:11:04 as soon as you press [call], it instantiates the call and exits Mar 25 18:11:15 doesn't go to the active call screen Mar 25 18:11:27 yep... and when the call is active there should be the active call screen... Mar 25 18:11:32 if not thats wrong Mar 25 18:11:46 correct Mar 25 18:12:13 it used to be, but something seems to have broken from my last update. looking thru git diffs now Mar 25 18:12:22 Blu3: is ophonekitd started and still running? Mar 25 18:12:48 hrm. no, it's not running right now.. Mar 25 18:12:56 i wonder why Mar 25 18:13:34 timestamp on the logfile is just over two hours ago Mar 25 18:14:11 at least that explains why you get no active call screen ;) Mar 25 18:14:24 mrmoku: :D Mar 25 18:15:10 hmmmmm. it coincides with the time i turned gprs on Mar 25 18:15:14 :/ Mar 25 18:15:29 indeed Mar 25 18:15:55 nothing to indicate why it died Mar 25 18:17:50 * mrmoku is in good mood, because he has the power back to build images :D Mar 25 18:17:58 :D Mar 25 18:18:02 spaetz: err, what do you mean by "once it's in the feed"? Mar 25 18:20:19 i asked mrmoku if i can put vala-settings in the repo feed for a while, so people can try it out easily Mar 25 18:20:38 mrmoku: what was the right move to win the battle? Mar 25 18:20:51 no sense in announciing it if normal people can't even test it Mar 25 18:21:07 spaetz: well was there no meaning in telling people we started SHR when we started it? Mar 25 18:22:18 (I know you see where I'm going with that) Mar 25 18:22:31 PaulFertser: setting ASSUME_PROVIDED=fakeroot-native and symlink fakeroot-ng to fakeroot somewhere into the path before /usr/bin ;) Mar 25 18:22:53 :) Mar 25 18:23:07 i se it. i just like to have /something/ before i announce it... Mar 25 18:23:53 i was not even sure whether it's going to be a settings app until today Mar 25 18:25:25 Ainulindale: to defend spaetz... it started as 'vala learning project'... and not as 'lets replace shr-settings with vala project' Mar 25 18:26:52 PaulFertser: btw. ASSUME_PROVIDED tells OE not to build a package, but instead use it from the system Mar 25 18:27:27 mrmoku: :) obviosly Mar 25 18:28:52 PaulFertser: btw... now that we all know that you teach embedded stuff... no excuse anymore to not join us at FSOSHRUDCON'09 ;) Mar 25 18:29:21 mrmoku: lame reason, to be honest ;) Mar 25 18:29:22 PaulFertser: I'm more frightened you will get fired if you don't visit the most important conf in the embedded sector this year :P Mar 25 18:32:14 Paul was going to be at the soldering station there, wasn't he???! :p Mar 25 18:32:32 mrmoku: btw, can you suggest any other fun to do if i manage to get to the Germany? I mean, flying there for just the short 3 days is a bit hm... ineffective ;) Mar 25 18:32:46 mrmoku: I know, you don't have to defend him as I'm not accusing him of anything :-) Mar 25 18:32:48 mwester: alphaone will be there with soldering equipment for sure. Mar 25 18:32:54 It's just that I don't want him to be as troublesome as ptitjes :-) Mar 25 18:33:02 c.f. the ringtone paradigm :-) Mar 25 18:33:18 I just want to burn the mail idea into your brain :-) Mar 25 18:33:31 Ainulindale: defense a priori :P Mar 25 18:33:55 well, a priori, a posteriori, or hugh laurie, it's the same Mar 25 18:33:58 I'm not accusing him! Mar 25 18:34:03 :-) Mar 25 18:34:35 PaulFertser: depends a bit what you consider being fun... I'm not sure about that in your case :-) Mar 25 18:38:14 mrmoku: yeah, tough question. I never have fun, you know. Mar 25 18:40:08 PaulFertser: well... then you could stay for some more days *not* having fun :P Mar 25 18:40:22 spaetz: yeah... startup is quite impressive :D Mar 25 18:49:09 what startup ? Mar 25 18:49:37 mrmoku: vala-settings ? Mar 25 18:51:02 Ainulindale: what was the last episode we watch yesterday ? Mar 25 18:53:16 can't recall Mar 25 18:53:28 Were you to ask me the last beer I'd recall that Mar 25 18:53:33 But not the last episode Mar 25 18:55:24 ptitjes: yep Mar 25 18:58:00 cool, yep quite a difference to loading python first Mar 25 18:58:55 Ainulindale: I can't remember what the beer was neither. a blond Mar 25 18:59:03 mrmoku: how faster ? Mar 25 18:59:14 +much Mar 25 19:00:57 has anybody got experience on how long it takes to build SHR on a T61 Thinkpad? Mar 25 19:02:10 ptitjes: my guess would be twice as fast... Mar 25 19:02:23 just my feeling though... nothing measured Mar 25 19:02:48 nezza-_-: the image? Mar 25 19:03:08 nezza-_-: i am running an r61 here, 3gb ram, 1.66 dual core Mar 25 19:03:23 ptitjes: the difference is even bigger when actually starting one of the settings modules Mar 25 19:03:24 and it took bout 3 days with lots of interruptions cause of failed packages Mar 25 19:03:36 i'd say the pure build time is around 24h Mar 25 19:04:33 I'm building on T60... and IIRC the last build from scratch was under 20h Mar 25 19:04:35 yep, module windows are much quicker too. as they are built into the main app Mar 25 19:08:35 i guess, i could improve the startup time by making some of the initial dbus calls async Mar 25 19:09:24 although there's still a difficulty that i need both a e mainloop and a glib mainloop Mar 25 19:11:19 spaetz: mickey|zzZZzz is working on that me thinks Mar 25 19:13:11 yep, i think so. but that is using valac HEADstuff, so i couldn't try it yet Mar 25 19:15:56 spaetz: why don't you use the valac HEAD too? Mar 25 19:19:48 because i use what OE provides :-) Mar 25 19:20:23 no need to deviate from upstream if not necessary Mar 25 19:22:41 ahh. ok... thought it would be in OE and your problem is your devel host Mar 25 19:22:45 freesmartphone.org: 03mickey 07framework * ra8e7bfd7d95c 10/ (8 files in 5 dirs): Mar 25 19:22:45 freesmartphone.org: ogsmd: [FREESCALE NEPTUNE] fix call handling to report outgoing calls properly Mar 25 19:22:45 freesmartphone.org: misc other fixes Mar 25 19:22:57 org.openembedded.dev has vala-native_git... Mar 25 20:15:33 tracfeed: Ticket #321 (no sound (mknod dsp c 14 3)) updated Mar 25 20:25:40 mmh, still needs vala-git Mar 25 20:26:17 vala-native is just used to compile vala, which is written in vala... Mar 25 20:26:56 yep... just wanted to comment that vala-native is there as git version... Mar 25 20:27:04 in the hope that arm-vala will be next Mar 25 20:27:43 yep, would be good to have more up-to-date stuff Mar 25 20:29:32 mmh, fakeroot-native failing again? Mar 25 20:31:59 spaetz: where? Mar 25 20:33:01 http://tinderbox.openembedded.net/packages/443307/ Mar 25 20:33:41 last build on tinderbox Mar 25 20:34:21 well... its not me Mar 25 20:34:27 Ainulindale: building? Mar 25 20:35:13 ahh... no its the old log me thinks Mar 25 20:35:28 the log for the image somehow is not there Mar 25 20:36:03 mmh, ok Mar 25 20:36:57 gone for today. will send a mail to list tomorrow asking for vala-settings feedback Mar 25 20:53:02 Hey, it seems that Brian Code's asound.conf is much saner than the default used in FSO/SHR: http://www.koolu.org/asound.conf Mar 25 20:53:10 mrmoku: Are you interested in commiting it? Mar 25 20:54:59 what's a WubBranler? Mar 25 21:02:42 huh, brian? Mar 25 21:03:19 DocScrutinizer: yes, i wonder why he didn't told anybody. But you can see that he used exactly what was suggested on the alsa wiki: dmix+dsnoop Mar 25 21:05:45 mwester: Ainulindale's SHR trac bot Mar 25 21:06:00 @ ticket 333 Mar 25 21:06:01 mrmoku: http://trac.shr-project.org/trac/ticket/333 Mar 25 21:06:14 PaulFertser: looking at it Mar 25 21:06:40 hmm looks good. though I wonder whether it's wise to use dsnoop at all. usually you dont want to access mic concurrently from different apps Mar 25 21:07:59 also with dsnoop you can't use your own arbitrary samplerate anymore (at least natively for hw, as well as for opening default) Mar 25 21:09:20 afaik this config should fail when opening default for record with -r 8000 Mar 25 21:10:07 DocScrutinizer: why do we have a special config at all? I thought modern desktop distributions just use default alsa configuration. Mar 25 21:12:29 hmm, a third time: it's seetup in /usr/share/alsa. If config there is incorrect it should be fixed there. the file content would remain the same mostly, as file linked above Mar 25 21:14:13 since alsa0.9 (iirc) there should be a dmix definition for default in /usr/share/alsa Mar 25 21:15:23 DocScrutinizer: yes, yes, yes :) I thought that this /usr/share/alsa files come from example alsa configuration from the upstream. Mar 25 21:15:40 DocScrutinizer: and we shouldn't modify it unless really needed. Mar 25 21:15:51 yup, ack Mar 25 21:18:17 PaulFertser: anyway look at /usr/share/alsa/cards on any standard disti. It's no example code, it's a collection of supported devices, each one with it's own special settings Mar 25 21:21:08 hi Mar 25 21:21:13 on OM/FSO this looks quite strange to me Mar 25 21:21:32 as there's just an aliases file Mar 25 21:22:13 DocScrutinizer: yes, but it's probably needed only if defaults should be tweaked. Like softvol used etc. Dmix is configured in one place for all the cards as i see it. Mar 25 21:23:25 user/share/alsa is lib-data, it's not to tweak anything Mar 25 21:25:33 it's owner root (what regretably totally gets lost on our borked user setup on Neo) Mar 25 21:26:03 DocScrutinizer: i mean, with some cards the defaults are ok, but for some cards additional consideration is needed and alsa supplies cards/cardname.conf for that. Mar 25 21:27:12 nah, it's more like describing card properties there, like "front" "5.1" etc pp Mar 25 21:27:33 bluez4 fails to upgrade with opkg since 2 days ago aprox. Mar 25 21:27:43 DocScrutinizer: ok, do we need that for wolfson? Mar 25 21:28:01 YUP, for sure! Mar 25 21:28:19 But why it works fine without it? Mar 25 21:29:12 because userland is both stupid and busy, and does all the setup by hand or working around it Mar 25 21:29:35 buzzword alsactl scenarios Mar 25 21:31:02 basically there should be a device handset-mic_2_stereoADC that does all the mixer setup on opening it Mar 25 21:31:48 exactly there in usr/sh/alsa/card/8753_gta02 Mar 25 21:32:01 DocScrutinizer: ah, and we should have it there. I see now what you're talking about. :) Mar 25 21:32:45 *phooo* Mar 25 21:32:51 :D Mar 25 21:32:52 hmm Mar 25 21:32:54 DocScrutinizer: i'm currently reading a lot about glib and dbus to try to finally implement something fun to manage the states... Mar 25 21:33:04 why is shr-unstable so much broken :'( Mar 25 21:33:29 PaulFertser: there's basically no need to "manage states" Mar 25 21:33:44 alsa should do by itself Mar 25 21:33:47 DocScrutinizer: do you want to discuss it? ;) Mar 25 21:34:22 no, I'm about to release a RFC, you remember Mar 25 21:35:11 DocScrutinizer: we can use our soundcard in many compatible and incompatible ways. Alsa has no internal way to know which usage is compatible with which. Mar 25 21:35:58 problem is: amount of info that needs to go into this RFC is ever increasing Mar 25 21:36:40 alsa has a rudimentary way to handle "compatibility" issues Mar 25 21:38:20 PaulFertser: see http://www.alsa-project.org/alsa-doc/alsa-lib/pcm_plugins.html#pcm_plugins_hooks ""lock true"" Mar 25 21:38:36 in example code Mar 25 21:43:36 DocScrutinizer: hm, two states can be compatible even if they need the same alsa control in the same state. Mar 25 21:44:06 DocScrutinizer: moreover, can those hooks fail?.. Mar 25 21:44:23 That is, refuse to open a device if it's not compatible. Mar 25 21:44:25 i know. but I'm not going to write a draft of my RFC here on IRC ;D Mar 25 21:45:30 DocScrutinizer: i'm thinking about implementing an external arbiter daemon that can be called from whatever app, including from alsa hooks. Do you think it'll be a waste of time? Mar 25 21:45:48 you remember we discussed possibility to call arbitrary exe from hooks? Mar 25 21:45:59 DocScrutinizer: yes Mar 25 21:46:14 DocScrutinizer: and that can be a dbus call to an arbiter daemon. Mar 25 21:46:38 basically no, we'll need sth like this. we need a very clear spec though before starting Mar 25 21:47:03 (waste) Mar 25 21:48:27 probably we even need st like this (dbus), as the alsa lib is user, and the arbiter needs to be systemwide Mar 25 21:56:54 PaulFertser: you see, alsactl was a convention-hack to showoff, and we are stuck with it still after years. I'm strongly averse to do the same mistake a second time Mar 25 21:59:09 PaulFertser: I'd suggest to follow the standard RFC procedure, interesting parties setting up a decent plan, then start to implement Mar 25 22:01:13 DocScrutinizer: i'll just play around for now. The thing is: i may have a bad mood tomorrow and the week later and want implement anything (or even discuss anything) at all. Or i can have a rush and write some prototype. If i get something that i like, i'll write an RFC description, if not, i won't make any fuzz. That's the plan: if i fail, nobody'll notice ;) Mar 25 22:01:24 for now we have only one broken concept (alsactl scenarios) we need to bother how to migrate from there Mar 25 22:01:35 I have some problem with my SHR-testing, Settings shows only black screen. How I put wlan on in commandline? iwconfig doesn't show eth0 ? Mar 25 22:02:02 aapo: use mdbus to turn on wifi (look for the odeviced bus). Mar 25 22:03:57 PaulFertser: I do not know how to use mdbus Mar 25 22:04:22 PaulFertser: sure go ahead. research prototypes are a good thing for sure. My hope was we can do the RFC as a cooperation work Mar 25 22:04:25 aapo: It's easy: mdbus -s, then mdbus -s busname, then mdbus -s busname interfacename, etc... Mar 25 22:06:07 DocScrutinizer: Sure, if you want to discuss anything before publishing your RFC, i'll be glad to take part. Mar 25 22:06:26 great :-) Mar 25 22:11:16 PaulFertser: this is only wifi related. mdbus -s org.freesmartphone.Device.PowerControl.Wifi eth0 -> Service name not found Mar 25 22:11:54 aapo: You're doing it wrong. You need to supply bus, interface, method names and the method arguments. Mar 25 22:12:51 aapo: http://wiki.openmoko.org/wiki/OpenmokoFramework/mdbus Mar 25 22:13:34 thanks Mar 25 22:16:38 is there somebody who wants bug reports? I'm too lazy to send ticket Mar 25 22:19:01 A) that 'Sim author' which pops on start, if I try to just close it, it messed the screen (It dies automatically when I just ignore it and let it alone) Mar 25 22:21:23 B) Settings -> Internet = wlan on/off. Wlan != Internet. (I could want wlan without internet access. And I can use Internet without wlan e.g. usb/bt) Mar 25 22:37:21 uhm Mar 25 22:37:48 why should people care to support your issues if you don't even care to communicate them properly? Mar 25 22:44:04 aapo, in newer versions that category is called Connectivity Mar 25 22:47:48 Sharwin_F, good to know. I flashed today and I thougth testing image is the most newest Mar 25 22:51:34 aapo, unstable one is newer :) Mar 25 22:52:58 ok. Mar 25 23:01:49 freesmartphone.org: 03mickey 07libeflvala * r0b8312af7c25 10/ (3 files in 2 dirs): library: start with better thread class Mar 25 23:03:05 mickey|zzZZzz, my workaround is: bitbake -c clean bluez4; bitbake -c build bluez4 Mar 25 23:56:30 Hey, quick question. I saw a post in a message board somewhere about using a Trie data structure to store the dictionary for the Illume predictive keyboard. Are you still thinking about doing this? If you are, I believe I could implement a Trie in C. If I do this and then give it to you will you use it? Mar 25 23:57:32 jhass840: ping Ainulindale, mrmoku Mar 25 23:59:02 Ainulindale: ping Mar 25 23:59:06 mrmoku: ping Mar 26 00:00:35 or mail it to the mailinglist of shr, works also :) (but they're pretty active on mirc, so they'll also read it) Mar 26 00:02:39 jhass840: feel free to impliment. but here coems the catch. u need to be able to mmap() the trie direct from disk without loading and allocing in ram Mar 26 00:02:53 also every trie entry needs to be able to hold not just a byte but a short or int Mar 26 00:03:17 right, so it can be internationalized I'm assuming Mar 26 00:03:24 correct Mar 26 00:03:29 and mmaped() to keep memory usage down Mar 26 00:03:45 also keep the layout such that you dont page in too many pages on a search Mar 26 00:04:07 also.. u need code to write the tree Mar 26 00:05:05 Hmmm, I don't want to damage your faith in me but I'm not familiar with mmap. I suppose I'll just have to learn it! Mar 26 00:05:23 :) Mar 26 00:05:26 its not hard Mar 26 00:05:29 u know what it does? Mar 26 00:05:54 I'm assuming something with pointers to locations in files? Mar 26 00:06:08 yes Mar 26 00:06:15 basically the file becomes a block of memory Mar 26 00:06:30 so he file contents are now just @ pointer X Mar 26 00:06:38 and onwards byte after byte until file ends Mar 26 00:07:00 so u use the file as if u just fopen+ fread it into a malloc'd blob of ram Mar 26 00:07:08 okay, and this is better because passing pointers around uses less memory than passing Strings Mar 26 00:07:14 but instead of the malloc+read the kernel handles paging it in and out of ram as needed Mar 26 00:07:20 yup Mar 26 00:07:26 and more to the point that it saves ram Mar 26 00:07:34 it means that thekernel will page in whats needed - when Mar 26 00:07:42 and when its not used anymore it can be re-used for something else Mar 26 00:08:05 So it's sort of similar to an ISAM database, at least in regards to storage Mar 26 00:08:09 now the problem is.. if its a trie ina mmaped file Mar 26 00:08:12 u cant use pointer Mar 26 00:08:15 u need "indexes" Mar 26 00:08:21 whihc are maybe bytes or words from the start of the file Mar 26 00:09:22 so u need to - whenever using some indirect reference, need to add it to the base pointer of the mmaped file Mar 26 00:09:23 etc. Mar 26 00:09:39 anyway - its just different Mar 26 00:09:50 no idea what an isam db is - never seen one Mar 26 00:11:09 raster: that's ancient talk Mar 26 00:11:12 ;-) Mar 26 00:11:26 maybbe even SIEMENS Mar 26 00:11:49 index sequential access mode Mar 26 00:12:43 Okay, so if I understand you correctly (and correct me if I don't) I have one pointer that is mapped to a file. And in that file I need to have static indicies so that I can Index that pointer like an array? Mar 26 00:14:39 DocScrutinizer: I believe Active Directory uses an ISAM database, or at least is used to Mar 26 00:16:05 jhass840: may be. Last I heard of (and used) ISAM was arounf 1986 Mar 26 00:16:38 on SIEMENS BS-M Mar 26 00:17:13 DocScrutinizer: My Data Structures professor made us implement a simple ISAM database for a project Mar 26 00:18:54 raster: Okay, so what I'm thinking about is implementing the Trie as a standalone data structure similar to an a class in C++ or Java. And then you guys can plug what I've done into your existing code, does that sound okay? Mar 26 00:18:58 hmm on second thought it's even extremely unlikely it's sth specific to SIEMENS, as it's an english acronym, and afair the only one used on whole BS-M docs Mar 26 00:19:23 Mar 26 00:20:13 whereas there were things like "kellerspeicherzeiger" in the rest of it Mar 26 00:20:57 DocScrutinizer: what do you think about my implementation idea? Mar 26 00:21:17 I didn't follow, sorry Mar 26 00:22:22 jhass840: DocScrutinizer is the hardware wonder; software is just something that makes his work emit smoke ;) Mar 26 00:22:43 hehehe Mar 26 00:22:48 oh Mar 26 00:23:04 Zorkman: well then, what do you think? Mar 26 00:23:56 Keep in mind that we're still operating under the pretense that I can pull this off. I believe I can, so we shall see Mar 26 00:26:08 jhass840: to be honest i really do lack the nescessary knowledge to judge your plan Mar 26 00:27:06 raster is the e-wonder; and Ainulindale or mrmoku or ptitjes will surely be able to tell you if it's a good plan Mar 26 00:28:10 oh, well in that case I'll just wait for one of them to show up. I plan on implementing this data structure anyway, simply because I find it fascinating. For now I'll just pretend they'll use it Mar 26 00:28:42 okie Mar 26 00:29:19 jhass840: the problem is.. lookups are progressive Mar 26 00:29:48 so the code using it needs complete access in walking the trie Mar 26 00:30:36 and getting all the current possible sub-branches to limit permutations Mar 26 00:31:32 well I can provide whatever methods you need in that regard Mar 26 00:31:36 i really suggest u look at the dict code currently Mar 26 00:31:50 as there is a definite blending of the dict lookup and its input data Mar 26 00:31:55 u cant isolate them completelyt Mar 26 00:32:29 I forgot to mention that I'm not using wikipedia's definition of a Trie. I have a data structures textbook that defines it much better Mar 26 00:32:45 it doesnt matter Mar 26 00:32:48 whatever you do Mar 26 00:32:58 1. it needs to be able to be mmap()ed Mar 26 00:33:04 2. it needs to be relatively compact Mar 26 00:34:02 3. it needs to allow for an "int" per character in a string Mar 26 00:34:45 Okay, so what I'll do is write it with those things in mind to the best of my ability and then if you don't like what I've done I will change it until it is correct Mar 26 00:35:23 4. it will need to be possible to walk the tree easily and quickyl from the api Mar 26 00:37:28 but think about the input first Mar 26 00:37:35 and the outout it wants Mar 26 00:38:33 Oh, I should mention that from what I understand the data structure will only be able to offer one selection for each String not multiples as it currently does. The way searching is defined, you either find what you are looking for or you fall off the Trie Mar 26 00:38:58 that is a problem Mar 26 00:39:09 and that basically neuters the usefulness of it Mar 26 00:39:48 well, I thought so too but then I realized that this means that the search will always pick the right word from the dictionary Mar 26 00:40:02 aaaah no Mar 26 00:40:09 did u read my explanation of how it works? Mar 26 00:40:15 it CANT pick the "right word" Mar 26 00:40:20 as tyhere is not just 1 "right word" Mar 26 00:40:23 as its fuzzy matching Mar 26 00:40:27 there could be multiple Mar 26 00:40:37 it can order them from most to least likely Mar 26 00:40:51 but the word you want is not always the "most likely" Mar 26 00:41:49 right, at which point you ignore it and keep typing until it gives you what you want. Mar 26 00:42:14 no no Mar 26 00:42:22 you have already typed all of the word Mar 26 00:42:25 there is nothnig more u can do Mar 26 00:42:31 look at your keyboard Mar 26 00:42:43 now imagine your finger was a big mallet Mar 26 00:42:54 every time you try and press "g" Mar 26 00:43:06 you actually press r, t, y, f, g, h, v, b, n Mar 26 00:43:33 make that a regex Mar 26 00:43:48 but the "press point' is CLOSER to "f" than "g" Mar 26 00:43:50 for example Mar 26 00:43:56 thus f is more likely your intent than g Mar 26 00:43:57 even with emphasis Mar 26 00:44:26 OH, I thought you were talking about searching the dictionary not determining what key was typed on the keyboard Mar 26 00:44:34 so you have a combination of input keys Mar 26 00:44:37 per char Mar 26 00:44:51 and likelihood per key Mar 26 00:44:56 and u have input strigns int he dict Mar 26 00:45:01 each with likelihood too Mar 26 00:45:14 jhass840: oh yeah. each full word int he trie needs to be able to store a probability count Mar 26 00:45:17 an integer Mar 26 00:45:37 jhass840: no - searching dict and determining key pressed goes hand in hand Mar 26 00:46:46 imagine you had to handle "keybouncing" as well ;) a nightmare Mar 26 00:47:48 keybouncing? Mar 26 00:48:24 leave or doublepress a key Mar 26 00:48:45 llke that Mar 26 00:48:49 hmm Mar 26 00:48:57 double-press is.. a double press in illume world Mar 26 00:49:00 well in the kbd Mar 26 00:49:08 it takes that raw as-is Mar 26 00:49:46 iff thhe nmber of kystrokes doesnt fiit Mar 26 00:50:13 jhass840: Mar 26 00:50:14 http://lists.openmoko.org/nabble.html#nabble-td2115715 Mar 26 00:50:22 the search starts get nasty Mar 26 00:52:00 raster: okay, so using your previous example. I can't just search the Trie for Strings beginning with "g" and be done with it. I also need to search for any possible strings beginning with "v", "t", "b", and "f" and eliminate each of them in turn as there are no longer any possible Strings beginning with those letters Mar 26 00:52:10 sorry that took so long to type Mar 26 00:52:11 http://lists.openmoko.org/nabble.html#nabble-td2115715|a2116855 Mar 26 00:52:14 specificalyl that one Mar 26 00:52:15 explains it Mar 26 00:52:31 and yes Mar 26 00:52:41 and the code need to know ifthere is anythign beginning with "t" Mar 26 00:52:46 and the same for any level of the tree Mar 26 00:53:02 also the inpuit chars come with values Mar 26 00:53:10 ie "distance from press point" Mar 26 00:53:18 0 == closest to where you pressed Mar 26 00:53:25 and then 1, 2, 3, 4 etc. gets further away Mar 26 00:53:31 (so less likely u wanted that key) Mar 26 00:53:40 read that post i have on the kbd logic Mar 26 00:53:43 the logic is all the same Mar 26 00:53:56 its simply the mechanics of how the dict is stored and walked/looked up that will change Mar 26 00:54:14 okay so I would give search preference to the one with the lowest weight Mar 26 00:54:46 well in reality the search code will be walking the trie itself Mar 26 00:54:57 raster: That link you gave me took me to an address not found page Mar 26 00:55:03 as it needs to sum up the distances of all chars fo far in the string Mar 26 00:55:45 http://lists.openmoko.org/nabble.html#nabble-td2115715 Mar 26 00:55:46 ok go there Mar 26 00:55:48 scroll down Mar 26 00:56:03 Jan 06, 2009; 08:56pm Mar 26 00:56:03 Re: Openmoko keyboard mockup Mar 26 00:56:05 that one Mar 26 00:56:34 pipermail ftw Mar 26 00:56:52 fsck nabble Mar 26 00:58:54 need to launch FiFo to display a jspage for mail, as konqueror will bork :((( Mar 26 00:59:07 the last one I see on Jan 6th is at 7:28pm Mar 26 01:01:26 search for Mar 26 01:01:52 not possible without compositing. not to mention the absolute HORROR of now Mar 26 01:02:04 its about 1/3 the way down the page Mar 26 01:02:09 from me Mar 26 01:02:22 ok. let me draw u a keyboard (or part of it): Mar 26 01:02:22 q w e r t y u i o p Mar 26 01:02:22 a s d f g h j k l Mar 26 01:02:22 z x c v b n m Mar 26 01:02:25 that stuff Mar 26 01:20:16 raster: Okay, I read that and it did clarify a lot of what I must do. Is it okay if I run it by you? Mar 26 01:20:37 sure Mar 26 01:20:38 :) Mar 26 01:20:40 I apologize If I'm annoying you Mar 26 01:20:52 it just saved me writing it all again here Mar 26 01:21:06 it just is that lookups are not just a regular "find me a word" Mar 26 01:21:24 but its a lot of fuzz and black magic, generating lists of possible matches Mar 26 01:21:25 right, because you have to account for input error Mar 26 01:21:45 using a combo of how likely the letters in that word are (compared to where you pressed) Mar 26 01:21:53 and how common the word is in english Mar 26 01:22:37 bingo Mar 26 01:22:43 I'm assuming neighborhood is defined as any key or keys immediately bordering the area of a touch Mar 26 01:22:45 if input wasnt prone to error Mar 26 01:22:54 or ambiguity Mar 26 01:23:02 you wouldnt need a dict lookup Mar 26 01:23:08 yeah Mar 26 01:23:14 well within N distance Mar 26 01:23:22 the kbd higher level code figures that all out Mar 26 01:23:29 in the end in produces a list of presses Mar 26 01:23:41 and each press can have 1 or more "letters" assigned to it Mar 26 01:23:52 and each leter will have a different distance from your press point Mar 26 01:24:01 this set of letters + distance per letter Mar 26 01:24:12 is your input set of "possibilities" Mar 26 01:24:39 i think i illustrated that reasonably well? Mar 26 01:25:03 e0, d1, r2, s2, w3, f8 Mar 26 01:25:03 k1, i1, u4, j7, m8, l3, o1 Mar 26 01:25:03 h1, y2, u8, j9, n3, b6, g2 Mar 26 01:25:05 for example Mar 26 01:25:27 (3 letter word - very simple single digit distances) Mar 26 01:25:42 expand as necessary for more keys pressed in the word Mar 26 01:26:15 and if the "fuzz" distance that the vkbd searches for "possible hits" increases or decreates the # of leyyers per keystroke will increase or decrease Mar 26 01:32:10 raster: how is dict-word likelihood related to keyhit-likelihood? Mar 26 01:32:29 DocScrutinizer: how do i combine them? Mar 26 01:32:37 yup Mar 26 01:32:43 dig 4 * 100 / 35 = 11 Mar 26 01:32:51 as the lower the number, the more likely Mar 26 01:32:59 i divide by "frequency" Mar 26 01:33:12 as frequency is "the bigger, the more used it is" Mar 26 01:33:20 of course you can add all sorts of multipliers here Mar 26 01:33:32 weighting key hit probabiliy vs dict frequency Mar 26 01:33:56 you could do things like square the distance Mar 26 01:33:59 and all sorts of stuff Mar 26 01:34:05 yeah Mar 26 01:34:17 gbut the core algorithm doesnt change - u are just changing your weighting formula Mar 26 01:34:26 i'm more than happy for there to be debate on that formula Mar 26 01:34:32 ie is a square of distance better than linear? Mar 26 01:34:43 or how much to weight dict freq # vs key dist Mar 26 01:34:46 even statistical values like deviation of key distance Mar 26 01:34:58 i'd even be happy to make thos e config opts runtime with a config dialog Mar 26 01:35:20 yeah Mar 26 01:35:34 but thats all really where open soruce is cool Mar 26 01:35:41 if u have an idea to make it "more accurate" Mar 26 01:35:44 slide it in Mar 26 01:35:54 itss literally like 4 lines of code where it does that Mar 26 01:36:03 e.g a word with ALL keys off by 1 might bee a better guess than the one with better over-all distance Mar 26 01:36:18 makign it pluggable/configurable and everyone putting in their "idea" on how it works best and offering it as an option.. seems good to me Mar 26 01:36:31 hmm Mar 26 01:36:37 in the end i think u'll find it wont be Mar 26 01:36:40 but its an example Mar 26 01:36:44 miund u distance is just that Mar 26 01:36:45 distance Mar 26 01:36:49 it doesnt cover direction Mar 26 01:37:00 your idea would be valid if it ALSO covered direction Mar 26 01:37:05 I know, was just an example Mar 26 01:37:18 ie same distanc e AND the angle (vector) is the same Mar 26 01:37:21 had no better at hand Mar 26 01:37:29 (ie your finger is just transposed) Mar 26 01:37:52 i think though the core of it is more a matter of linea vs squared/cubed Mar 26 01:37:58 or sqrt of distance Mar 26 01:38:09 and weighting of key distance vs word frequency Mar 26 01:38:13 analyze pressure aka finger-vs-stylus to automaticaly ajust dist threshold? Mar 26 01:38:13 Howdy ya'll Mar 26 01:38:22 hmm Mar 26 01:38:28 we dont get any pressure info Mar 26 01:38:32 thus - canty do much Mar 26 01:38:45 Who else here is hungry? :) Mar 26 01:38:52 I thought ts-driver could do Mar 26 01:38:58 one thing the code doesnt store is timestamps Mar 26 01:39:09 ie time @ which the press happens Mar 26 01:39:13 and how long it is Mar 26 01:39:14 also a nice idea Mar 26 01:39:19 if u are "doubling" a letter Mar 26 01:39:25 yup Mar 26 01:39:25 eg "leTTer" Mar 26 01:39:28 got it Mar 26 01:39:40 chances are u press the screen really quickly as u dont have to reposition your finger at all Mar 26 01:39:50 so if 1 strokes are much closer together than the rest Mar 26 01:39:56 and within a threshold Mar 26 01:40:05 u can possibly hint at the matcher that that was intended as a "double" Mar 26 01:40:13 and no - dont get pressure Mar 26 01:40:16 x doesnt report it Mar 26 01:40:20 just a mouse Mar 26 01:40:27 technically it could offer an xinput device Mar 26 01:40:37 xinput devicves can support tonens fo parameters/info Mar 26 01:40:41 like angle, pressure, size etc. Mar 26 01:40:46 but xglmao doesnt do that Mar 26 01:40:48 xglamo Mar 26 01:40:55 bah Mar 26 01:41:00 it just provides basic raw press/release and movement Mar 26 01:41:28 anyway Mar 26 01:41:32 the stuff could improve Mar 26 01:41:33 as i said Mar 26 01:41:46 letter doubling would make acase for tracking the timestamp Mar 26 01:41:58 (well relative timestamp from the previous key press) Mar 26 01:42:03 which my code doesnt do Mar 26 01:42:23 yeah, just had my curious 5min. You should got used to it Mar 26 01:42:37 tracking press time (how long it was held down before a release) might be useful info later too Mar 26 01:43:16 raster: So when I type a character say "g" for example a single press is generated containing an ordered pair of (letter, weight) where a lower weight has higher precedence [("g",0) ("t", 1), ("v",2), ("b",3), ("h",4)] and that press is concatenated to a list containing every press since I started typing a word. Then that list is iterated over generating Strings of characters from each press, starting with the most frequent cha Mar 26 01:45:25 yup Mar 26 01:45:31 and the dict is asked if this exists Mar 26 01:45:47 as it can eliminate firther matches if a branch of the tree doesnt exist Mar 26 01:45:57 ie if nothgn starts with exv Mar 26 01:46:09 then no point checking exvot etc. etc. Mar 26 01:46:17 right Mar 26 01:46:25 ie further purmutations starting with a dead-end head Mar 26 01:46:35 thus the need for the dict to be much less "opaque" and abstracted Mar 26 01:46:46 it needs to make use of this ability to weed out matches for sheer speed sake Mar 26 01:47:39 yeah, I didn't realize how many searches are actually made. This makes it much clearer Mar 26 01:48:41 yeah Mar 26 01:48:49 for every keystroke u make it needs to search Mar 26 01:49:00 and that could generate 1000's of permutations Mar 26 01:49:06 unless u weed it down early Mar 26 01:51:57 don't present result (thus not blocking) for 1. char? Mar 26 01:52:42 there's few reasonable guessing on one-letter words anyway Mar 26 01:52:53 When the Trie realizes that a word doesn't exist does it assign the word some dummy frequency value like say 200 so that it ends up at the end of the list of possibilites or does it simply delete it from that list? Mar 26 01:56:09 DocScrutinizer: it has a small delay Mar 26 01:56:15 i thnk 0.25 sec Mar 26 01:56:24 if u havent pressed in 0.25 sec it does a lookup Mar 26 01:56:51 ie if u type fast it will avoid excessive lookups Mar 26 01:57:03 So how's the glamo tonight? Mar 26 01:57:15 jhass840: if a word doesnt exist - it should not be added to the list Mar 26 01:57:25 glamorous ;D Mar 26 01:57:33 anything that starts with an "impossible letter combo simply is to be avoided Mar 26 01:57:39 but thats part of the lookup process Mar 26 01:57:45 thus u nee to be able to walk the trie Mar 26 02:00:47 okay, so the Trie takes the list of presses and searches those possibilities in the manner previously stated and if it finds the word it adds that word and it's associated frequency to the list of possibilities and if not the word is not added. Then when all possibilites have been examined the list of possibilites is sorted according to frequency and displayed Mar 26 02:00:58 hmm Mar 26 02:01:02 i'm not sure the trie takes it Mar 26 02:01:17 it simply implements an api to allow walking of the trie Mar 26 02:01:34 and being able to select sub-trie nodes to walk to Mar 26 02:01:58 freesmartphone.org: 03mickey 07libeflvala * r3b799c198219 10/ (12 files in 5 dirs): library: move example code into library and strip example Mar 26 02:15:53 raster: Perhaps: presses[press_0[["g", 0], ["t", 1], ["v", 2], ["b", 3], ["h", 4]], press_1[["o", 0], ["x", 1], ["y", 2]], ..., press_n] -> search_string_1 = presses[0][0][0] + presses[1][0][0] + ... + presses[n][0][0] -> trie.search(search_string_1) -> if found then possibilities.add(search_string_1, probability) else do not add (Pseudo-code) Mar 26 02:19:43 the code outside the trie need to do the walking Mar 26 02:20:00 as it needs to calculate the "popularity" of a string based on all the input Mar 26 02:28:49 Ah, so the Trie just provides an abstraction to the data inside the dictionary file and the code outside the Trie access it directly (no methods, no information hiding) Mar 26 02:29:14 hello Mar 26 02:30:26 jhass840: I read above. Sure we would be glad to finaly have a good internationalized predictive keyboard Mar 26 02:31:16 jhass840: if you need some inspiration for the building of a mmap-suitable file structure, there is the compiler in gobject-introspection that does that Mar 26 02:31:47 ptitjes: nice, thanks Mar 26 02:31:53 you need to define blob structures Mar 26 02:32:11 a type of blob for each of your node types Mar 26 02:33:05 say a blob contains your character data, and indexes to reference other blobs Mar 26 02:33:13 I've been working this out with raster at great length and I have learned a lot about what I need to do, I'm pretty excited about it Mar 26 02:33:24 nice Mar 26 02:33:29 sorry, go on Mar 26 02:33:55 jhass840: will you compact branches on the trie Mar 26 02:37:52 anyway men nice night Mar 26 02:37:54 bye **** ENDING LOGGING AT Thu Mar 26 02:59:57 2009