**** BEGIN LOGGING AT Wed Jun 15 02:59:58 2011 Jun 15 03:43:33 timeless: ? Jun 15 03:43:47 timeless: I've only been connecting over IPv6 for like 2 years now…? Jun 15 04:35:15 MohammadAG: shall I kick your alter ego? Jun 15 04:38:06 Sazpaimon, considering his salary is $20 mil, he damn well better be. :rolleyes: Jun 15 04:38:27 Sazpaimon, otherwise I'd say you guys are getting ripped off. Jun 15 04:46:06 GeneralAntilles, sallary doesn't mean much in terms of performance Jun 15 04:47:07 how a player jives with a team means a lot more Jun 15 04:47:56 look at jayson werth Jun 15 04:48:16 he's a great player, but $126 millon? Jun 15 04:50:50 he also can't singlehandedly hold up a terrible team like the nats Jun 15 04:53:08 where's optonline.net located? Jun 15 05:04:18 Sazpaimon, that's why the Rays are so awesome. :) Jun 15 05:05:34 GAN900: are you running a ZNC? Jun 15 05:06:13 * DocScrutinizer wonders who's faking MohammadAG, in connect-i-cut Jun 15 05:11:01 Anyone know where easy debian mounts it's chroot to by default? Jun 15 05:11:25 irssi proxy Jun 15 05:11:35 hi guys Jun 15 05:12:05 I have a weird problem on maemo on N900, I was playing Doom and it crashed and now my keyboard layout is messed up Jun 15 05:12:10 everywhere, not just in the game Jun 15 05:12:38 . works as left, left works as down and ä works as up Jun 15 05:12:51 reboot, or start doom and stop it Jun 15 05:12:55 at a guess Jun 15 05:13:01 I've got the Finnish unit so I've got a scandinavian layout Jun 15 05:13:05 rebooted, didn't fix it Jun 15 05:13:10 restarted the game and closed it Jun 15 05:13:13 without it crashing Jun 15 05:13:15 didn't fix it Jun 15 05:13:51 is there some command for resetting the keyboard layout or something? Jun 15 05:15:17 try settings - keyboard - layout Jun 15 05:15:31 Frampis: sounds like you have the intl or US layout now Jun 15 05:15:40 yup Jun 15 05:15:58 and obviously doom is doing something *NASTY* Jun 15 05:16:12 yes, same thought. sounds like a bug Jun 15 05:16:23 like storing current keymap and restoring it on normal quit Jun 15 05:16:39 that's a bug for sure, but a fubar bug Jun 15 05:17:30 thanks guys Jun 15 05:17:39 restarting doom won't help as it backup-stores his own mess of last segfault Jun 15 05:17:50 yeah I was overthinking this, forgot to check the actual GUI settings menu :D Jun 15 05:18:27 I found the setting and had to change it back Jun 15 05:18:36 insane idiocy to load a new keymap to deal with key meanings Jun 15 05:19:36 Frampis: so you did what doom is supposed to do on quit: restore the keymap that's been there on doom-start Jun 15 05:19:48 seems that way, yes Jun 15 05:20:02 ~lart doom and its devels Jun 15 05:20:02 * infobot drops a truckload of VAXen on doom and its devels Jun 15 05:20:28 that should cure it Jun 15 05:20:41 :-P Jun 15 05:21:02 the port is actually called PrBoom Jun 15 05:21:13 I was trying to get a doom 2 wad to work but it just crashed Jun 15 05:21:42 whatever, I wasn't gonna play it anyway. Jun 15 05:27:28 why does the N900 need a custom doom port Jun 15 05:27:37 couldnt someone just compile boom or zdoom? Jun 15 05:28:45 unless you wanted to turn using the touchscreen and shoot by tapping Jun 15 05:28:50 but that'd be awkward Jun 15 06:16:35 Hey guys. Does anyone know if we can modify the usb-connection screen? The one with "mass storage" and "PC suite" modes? Jun 15 06:20:18 what do you want to do with it? Jun 15 06:20:23 host mode gui or something? Jun 15 06:22:09 I was thinking in terms of adding a couple extra buttons, allowing for only the SD card OR MyDocs to be chosen for mass storage Jun 15 06:22:39 and also, see about a mode that would allow for *both* mass storage mode as well as USB networking Jun 15 07:19:50 i am very new with using N900 . i have just run simpletest program written using ARToolkit APIs on my PC ubuntu10.10 . i want to port this program on my N900 mobile.Can anybody help me in process of installing and running this program or can provide me some useful link Jun 15 07:20:08 please help... I am really stuck at this point Jun 15 07:22:33 please provide me some useful link regarding above stated problem. Jun 15 07:22:42 http://maemo.org/development/ Guest12410 Jun 15 07:24:50 @appiah - thank you for helping me out. I just wanted to know that if i want to port a sample porgram written using ARToolkit APIs on my N900 then i have to do learn maemo 5 also Jun 15 07:25:11 http://www.reghardware.com/2011/06/15/review_asus_eee_pad_transformer/ <-- hmm Jun 15 07:26:04 I have no idea on what ARToolkit APIs is so I wouldn't know Jun 15 07:26:31 I've tried it...I don't think there is an arm port of it but not sure Jun 15 07:26:50 Tried just compiling your software for arm and just running it on n900? Jun 15 07:27:19 @joga-- Can you tell me which are the libraries i need to port on my N900 mobile so that my porgram runs fine Jun 15 07:27:42 O_o Jun 15 07:27:48 Guest12410, no, I don't know sorry, you need to find it out on your own and if you feel capable of porting it, finding it out wouldn't be hard in comparison Jun 15 07:29:12 ok thank you both of you. Jun 15 07:29:21 thank you very much. Jun 15 07:37:10 Jun 15 07:37:17 HELO ladies and gentlemen Jun 15 07:55:30 ~seen khertan Jun 15 07:55:36 khertan <~khertan@AAmiens-553-1-180-120.w92-142.abo.wanadoo.fr> was last seen on IRC in channel #maemo, 18h 7m 13s ago, saying: 'cloudyLights, what your app does ?'. Jun 15 08:16:28 hello maemonians Jun 15 08:17:06 <_trine> mece, we are a dying race Jun 15 08:36:07 * RST38h moos at crashanddie Jun 15 08:36:16 moo Jun 15 08:36:21 baaaa Jun 15 08:36:34 crash: I have done the impossible by making sound work in Android! Jun 15 08:36:37 (yes, what am I doing getting into a cownversation?) Jun 15 08:36:46 khertan: hi there Jun 15 08:36:56 khertan: remember me from yesterday? Jun 15 08:36:59 RST38h, sound didn't work in Android? Jun 15 08:37:11 RST38h, I figured that was kindof a useful feature for a phone Jun 15 08:37:16 Oh, you mean on the N900? Jun 15 08:37:30 crash: Oh it does work Jun 15 08:37:37 cloudyLights, yep ... did you successfully package your app ? Jun 15 08:37:58 khertan: in http://khertan.net/articles/maemo/pypackager_howto I see the line: Jun 15 08:38:01 crash: But the native API is OpenSL, brought to you by the same mentally ill people ad OpenGL Jun 15 08:38:13 p["/usr/lib/python2.5/site-packages"] = files Jun 15 08:38:21 Ok, not entirely mentally ill, just crazy =) Jun 15 08:38:36 khertan: what does that line mean and waht should I change it to? Jun 15 08:39:01 is there a command to post a file from the N900 to a pastebin? Jun 15 08:39:03 The 800-page pdf never explain how to play normal, pcm sound Jun 15 08:39:13 cloudyLights, hum ... i most of the time put my python package in /usr/lib/python2.5/site-packages Jun 15 08:39:31 and files is what os.walk found in the folder Jun 15 08:39:43 p is a dictionnary Jun 15 08:40:12 with install path as key, and current local path as value Jun 15 08:40:50 exemple : you want to put script1.py script2.py and script3.py in /opt/myuberscript Jun 15 08:41:09 script1.py are located in the same folder as your pypackager script Jun 15 08:41:38 ok Jun 15 08:41:41 p['/opt/myubserscript'] = ['script1.py','script2.py', 'script3.py'] Jun 15 08:42:22 mhow can I pastebin my make.py from my N900 now? Jun 15 08:42:27 I want to show you Jun 15 08:42:48 use your n900 browser :) Jun 15 08:49:09 khertan: http://paste.ubuntu.com/627182/ Jun 15 08:49:36 I didnt put any files in /usr/lib/python2.5/site-packages as I didnt understand why Jun 15 08:49:56 my files are at /usr/lib/switchProfByMeeting Jun 15 08:50:04 RST38h, yeah, I have a bit of experience with massively big pdf files "explaining" how to implement something -- I feel for you Jun 15 08:50:18 crash: Oh, you cannot imagine... Jun 15 08:50:49 Well, I spent last night trying to understand the g723, g726, g727 and g729 codecs, so yeah, I can imagine :P> Jun 15 08:50:57 crash: These maniacs have redefined all the basic types, created their own COM-like framework, added support for media streaming, media metadata, special effects, etc Jun 15 08:51:00 cloudyLights, so you can comment also the things below #src Jun 15 08:51:15 cloudyLights, and you should enable binary build at last line Jun 15 08:51:26 crash: All that in order to implement an audio API. Jun 15 08:51:39 ok, that is pretty bad. Jun 15 08:51:51 crash; And the PDF is not really explaining anything, it is just listing endless interfaces Jun 15 08:52:23 cloudyLights, there will be only one file writen, as you overwrite the dict key value on line 41 and 42 Jun 15 08:52:49 ps ubuntu pastebin suck you can't correct and reply Jun 15 08:54:52 cloudyLights, http://pastebin.com/Hj1gvimx Jun 15 08:55:44 http://paste.ubuntu.com/627186/ ? Jun 15 08:56:18 o ok Jun 15 08:56:22 let me see Jun 15 08:59:26 RST38h: OpenGL is not *that* bad Jun 15 09:00:51 * Arpa-MST is away (Arpa Systems) Jun 15 09:21:45 khertan: ok your lines helped and now I see switchProfByMeeting_0.0.1-1_armel.deb Jun 15 09:22:09 how can I ask dpkg to list the files without installing it Jun 15 09:22:09 ? Jun 15 09:22:59 cloudyLights, dpkg -c {.Deb package} Jun 15 09:24:19 want to try? Jun 15 09:24:41 its very simple and shoulndt kill your N900 Jun 15 09:26:22 cloudyLights, i'm currently trying to implement theme to qt in meego de Jun 15 09:26:41 so didn't have too much time for this :) Jun 15 09:26:42 sorry Jun 15 09:26:51 np Jun 15 09:29:01 khertan: thanks a lot Jun 15 09:29:11 can I also run it on a pc? Jun 15 09:29:19 I assume I can Jun 15 09:30:43 yep you should be able to install it with the python setup tools Jun 15 09:30:51 (easy_install install pypackager) Jun 15 09:31:00 as it s availble on the pypi index Jun 15 09:31:06 http://pypi.python.org/pypi/pypackager/3.0.20 Jun 15 09:40:54 khertan, yay, a Qt theme Jun 15 09:41:59 shut Jun 15 09:41:59 :) Jun 15 09:42:26 i ve got so many problem that i didn't think it ll work one day Jun 15 09:47:38 So, I want to watch this episode, the built-in player is completely ignorant and tells me the file has an invalid type or something, while mplayer reports that it's an 848x480 H264, 900kbps avi with an mp3 audio track. If I want to play it I hear the first half second of audio and the then audio disappears, while the video is stuttery but close to smooth. How do I get it to play smoothly? Jun 15 09:48:16 Shapeshifter: Find it in a different format on the torrent? Jun 15 09:48:17 tried with mplayer? Jun 15 09:48:37 robbiethe1st: I'm saying mplayer above. Jun 15 09:48:45 So use it? Jun 15 09:48:53 Or.. Jun 15 09:48:59 hm Jun 15 09:49:14 That's odd, then... Jun 15 09:53:24 "Unable to play media. Media format not supported" is what the built-in player says. That player is terrible anyway. Jun 15 09:53:43 I even have decoders-support Jun 15 09:55:12 Is there a navigation software for N900 that would have voice guiding? Jun 15 09:55:13 hm... Try with kmplayer Jun 15 09:55:36 It seems to be good at wrapping mplayer and selecting good options to play something Jun 15 09:55:47 mappero sortof has it Jun 15 09:56:30 sort of? Jun 15 09:56:43 well its there, but takes some manual setup to get working Jun 15 09:57:02 my biggest gripe with the built in media player is lack of gapless mp3 playback. Jun 15 09:57:38 i always expect i might have to do a bit of transcoding of video Jun 15 09:57:55 and that actually happens pretty raarely Jun 15 09:58:21 nid0, Ok, care to give some pointers, a good howto perhaps, if you know. (I can google too, of course :)) Jun 15 09:58:36 HRH_H_Crab, have you tried the community media player? Jun 15 09:58:51 HRH_H_Crab, it's a drop-in replacement for the built-in media player, graphically identical, with some extra features Jun 15 09:59:10 hmm i wasnt aware of that Jun 15 09:59:25 ill have to check it out Jun 15 09:59:35 thanks for the tip Jun 15 09:59:38 plus, it's open source, so we can add features Jun 15 09:59:53 crash: does it support themes already? Jun 15 09:59:55 T-Co_: your best bet would be the mammoth mappero thread on TMO, it should all be documented there. Sygic is another option, but isnt free Jun 15 10:00:10 RST38h, no idea, I haven't looked at it in months Jun 15 10:00:13 nid0, Thanks Jun 15 10:01:22 robbiethe1st: thanks. kmplayer is an improvement, the sound at least plays constantly. too stuttery though. Jun 15 10:01:42 nid0, I tried Marble and it works great, expect that it wont play the voices... Jun 15 10:02:02 Hm, there ought to be a way to play it. you may need to, perhaps, use a converter to switch the container format Jun 15 10:02:19 to, say, mp4 or maby mkv Jun 15 10:03:33 robbiethe1st: the cpu is at 100%, I would assume the n900 is simply to slow for this Jun 15 10:03:59 Shapeshifter: Yes. But we've got hardware-decoding for a couple of formats, if you can get the right codecs to decode it Jun 15 10:04:09 h264 is one of them Jun 15 10:04:21 crash: does it support themes already? Jun 15 10:04:27 If you can get the stock player to play it, it'll work, I think Jun 15 10:04:31 it supported themes from the start since it loads the same images Jun 15 10:04:51 robbiethe1st: but not mplayer iirc. mplayer cannot use the hw-acceleration is what I thought Jun 15 10:04:54 ! Jun 15 10:05:02 crashanddie, gapless playback is something to be implemented in mafw-gst-renderer Jun 15 10:05:05 * RST38h wonders if the volume control bug is gone then =) Jun 15 10:05:13 Correct, though it may be possible with some very convoluted options. Jun 15 10:05:27 RST38h, what bug? Jun 15 10:05:52 crashanddie: what's the name of that community player? Jun 15 10:05:54 Mohammad: the volume control in the original media player always shows up with the gray background of fixed color Jun 15 10:06:07 Mohammad: they forgot to make it themeable Jun 15 10:06:43 Shapeshifter: http://talk.maemo.org/showthread.php?t=63554 Jun 15 10:07:47 Shapeshifter: gi Jun 15 10:07:51 Shapeshifter: hi Jun 15 10:08:05 cloudyLights: hi there Jun 15 10:08:09 robbiethe1st: thanks Jun 15 10:08:22 I wonder... http://wiki.meego.com/ARM/N900/Tips_and_Tricks/DSP Jun 15 10:08:58 if anyone here care to see this app I wrote Jun 15 10:09:13 it switchs the profile according to the calender Jun 15 10:09:17 is it this one: http://www.reghardware.com/2011/06/15/pin_spy_app_pulled/ ;) Jun 15 10:10:11 I hope to get someone else try it before I post it in the forum Jun 15 10:10:28 RST38h, oh, haven't seen that in OMP Jun 15 10:11:30 so who has the C skills to implement gapless playback? Jun 15 10:11:35 Shapeshifter: Hm, we might be able to use dsp decoding with gstreamer - "gstreamer0.10-dsp" in the repos Jun 15 10:11:49 I do Jun 15 10:11:52 But... cba Jun 15 10:11:57 let me ask: does anyone need an app to silence the phone when in a meeting then bring it back to general? Jun 15 10:12:09 Erhm Jun 15 10:12:13 Isn't there "profiles" for that? Jun 15 10:12:27 right Jun 15 10:12:46 I'd like something that modifies profiles based on GPS Jun 15 10:12:46 but the profile doesnt change back to general after the meeting ends Jun 15 10:12:53 me too Jun 15 10:12:57 (well, if GPS were a viable long-term alternative) Jun 15 10:12:58 will soon offer it Jun 15 10:13:12 You could make a "conference joiner" Jun 15 10:13:15 first I need it to alert me when I am near my station Jun 15 10:13:17 Or try to add it to the calendar Jun 15 10:13:47 so that the phone automatically joins a phone conference, waits and types the codes required to enter the conference room Jun 15 10:13:53 GPS indoors is going to be poor, and battery drain :-( Jun 15 10:14:09 Jaffa, which is what I said with "if it were a viable long-term alternative) Jun 15 10:14:17 I do <-- and the gstreamer skills? Jun 15 10:14:29 gstreamer isn't that hard. Jun 15 10:15:11 gstreamer is a terrible mess Jun 15 10:16:11 Shapeshifter, you've never worked in my company I take it ;) Jun 15 10:16:19 Compared to some of our products, gstreamer is a work of art. Jun 15 10:17:31 crashanddie: I used gstreamer once for a project of mine and I've never been more frustrated with a library. A fraction of it is decent, a large part of it is badly documented and awkward, and a good lump of it is just absolutely terrible. Jun 15 10:17:56 especially the documentation is incredibly bad. the whole API... Jun 15 10:18:20 I will use tower ID to now I am near my bus station Jun 15 10:18:55 core and base are well documented, but it goes steeply downhill from there Jun 15 10:22:12 cloudyLights: Chat to lardman - he's planning a framework (called "Proximus") which would make writing this app easy Jun 15 10:22:27 cloudyLights: http://wiki.maemo.org/Proximus Jun 15 10:22:43 thanks mate Jun 15 10:23:06 as I am not working with maemo and I am at work Jun 15 10:24:34 I got so excited with khertan help and having to finally package my program Jun 15 10:24:58 then I had a work amte ask so many cifs questions Jun 15 10:25:04 back to work Jun 15 10:25:16 wish I could do these chats at night Jun 15 10:32:54 what the heck is this? http://30.media.tumblr.com/tumblr_lmd8j77siG1qcakwko1_500.jpg srsly? Jun 15 10:34:01 MohammadAG, is gapless playback need to be implemented in MAFW or in gstreamer or in the media player? Jun 15 10:37:14 mafw Jun 15 10:37:21 gstreamer already has an about to finish signal Jun 15 10:39:10 LOL @ the image Venemo Jun 15 10:39:21 MohammadAG, yeah LOL indeed Jun 15 10:39:27 MohammadAG, maybe https://bugzilla.gnome.org/show_bug.cgi?id=440952 could give a clue Jun 15 10:39:29 04Bug 440952: was not found. Jun 15 10:40:03 MohammadAG, or this: https://bugzilla.gnome.org/show_bug.cgi?id=130426 Jun 15 10:40:04 04Bug 130426: was not found. Jun 15 10:40:08 povbot, you are stupid Jun 15 10:40:08 Venemo: Error: "you" is not a valid command. Jun 15 10:40:13 xD Jun 15 10:41:56 Venemo, I could do a proof of concept with gstreamer Jun 15 10:42:01 I think I could at least Jun 15 10:42:14 just start playing another file when the old one is about to finish Jun 15 10:42:27 the PS3 does gapless playback best imo Jun 15 10:42:40 MohammadAG, start _decoding_ the new file when the old one is about to finish - then you can start playing immediately Jun 15 10:43:50 I just hope that in case the N9 does come out I can see it in the subject here... Jun 15 10:44:15 cloudyLights, the news will be full of it Jun 15 10:44:42 yet I dont check the news any more Jun 15 10:44:49 prefer to make news.... Jun 15 10:45:22 cloudyLights, for example? Jun 15 10:45:32 write good applications Jun 15 10:45:38 MohammadAG, banshee's gapless playback does indeed work Jun 15 10:45:41 cloudyLights :) Jun 15 10:46:38 I mean there is too much text regarding meego/N9 and yet some usefull stuff are missing Jun 15 10:46:44 MohammadAG, why don't we just code a mobile GUI for banshee and forget about mafw? Jun 15 10:46:52 I still think most of thesmart phones are not so smart Jun 15 10:47:06 cloudyLights, N9 is mostly rumors Jun 15 10:47:20 but most ppl NEED some usefull apps Jun 15 10:47:26 not browser and games Jun 15 10:47:54 cloudyLights, I agree Jun 15 10:48:31 a good calender for example doesnt make a phone smart Jun 15 10:48:49 but it is missing in my N900 and I had such in the palm E3 Jun 15 10:50:19 cloudyLights, well, N900's calendar sort of works Jun 15 10:50:53 for example that program that told what train to take from muenchen was usefull Jun 15 10:51:20 usefull app: how to get to that book store from here now Jun 15 10:51:57 cloudyLights, yeah, mapping apps are useful Jun 15 10:52:10 Venemo, banshee isn't the best player imo Jun 15 10:52:29 I recognizes songs by extension rather than mime type for example Jun 15 10:52:34 it* Jun 15 10:52:55 MohammadAG, do you know that for a fact? Jun 15 10:53:09 yes Jun 15 10:56:09 a phone is smart if it can tell me where I left it Jun 15 10:56:30 or where I left my music on it Jun 15 10:56:51 assuming you use 32Gb and dont know where the album is at Jun 15 11:45:02 cloudyLights: :| Jun 15 11:45:23 are you the guy that help ppl in mythtv? Jun 15 11:45:33 that mysql wizard? Jun 15 11:45:45 cloudyLights: huh, not at all. I dislike relational databases. Jun 15 11:46:05 o ok Jun 15 11:53:17 00 Jun 15 11:53:20 ] Jun 15 12:08:17 DocScrutinizer, using the initrd image to install MeeGo on the eMMC, apparently it doesn't have the mtd blocks at all Jun 15 12:18:04 MohammadAG, or you could go to GimpNet and join #rhythmbox or #banshee to see how they implemented gapless playback Jun 15 12:19:33 Venemo, cross fade engine afaik Jun 15 12:20:12 that Jun 15 12:20:13 would Jun 15 12:20:14 be Jun 15 12:20:18 great! Jun 15 12:20:20 MohammadAG, it is not crossfading, definitely Jun 15 12:20:31 only thing I dont like about the music player Jun 15 12:20:46 that gap between songs which ruins some albums Jun 15 12:22:18 Venemo, it is in ryhtmbox Jun 15 12:22:26 the option says use cross fading Jun 15 12:22:48 MohammadAG, yes, rhythmbox uses crossfading, banshee doesn not (seem to) Jun 15 12:23:17 I'm not sure if banshee uses it Jun 15 12:23:19 * MohammadAG checks Jun 15 12:23:52 their UI contains an "enable gapless playback" option Jun 15 12:24:43 MohammadAG, I tested it with some Mike Oldfield albums, and it indeed works Jun 15 12:24:51 nope, there's still a gap for me Jun 15 12:24:58 weird. Jun 15 12:25:10 none of the mediaplayers I've used do it like the PS3 Jun 15 12:25:13 what version is that? Jun 15 12:25:23 Banshee 2.0 (2.0.0) Jun 15 12:25:29 hm Jun 15 12:25:35 the gap is small, but it's there Jun 15 12:25:40 I have 2.0.1, and there is no gap at all Jun 15 12:25:55 the PS3 lowers the volume of the first track, merges the second track and raises the volume of that Jun 15 12:25:56 MohammadAG: heya Jun 15 12:26:06 note that some songs do have a gap at their ends Jun 15 12:26:09 in effect, the two songs are merged together without any noticeable gap Jun 15 12:26:15 achipa, hi Jun 15 12:26:32 MohammadAG: what's the final word on that Qt 4.7.2+ stuff ? did you manage to get a fix for the raster/rendering bug ? Jun 15 12:26:33 Venemo, nope, works fine on my PS3 Jun 15 12:26:45 MohammadAG, that is called crossfading Jun 15 12:26:45 achipa, no, but I checked 4.8 and it doesn't have the bug Jun 15 12:26:59 MohammadAG: oooh, even better Jun 15 12:27:01 the CSSU reverted back to 4.7.0 with some patches Jun 15 12:27:08 anyway, look at rhythmbox's inspiration Jun 15 12:27:09 achipa, 4.7.4 has the bug otoh Jun 15 12:27:32 if 4.8 works, we should take that one Jun 15 12:27:40 achipa, however 4.8 is a release candidate for now (or a beta, not sure) Jun 15 12:27:48 it lacks hildonization Jun 15 12:27:50 achipa, it requires additional dependencies which are not available on Maemo Jun 15 12:28:02 yeah, that too Jun 15 12:28:09 Venemo: that was the "if"... Jun 15 12:28:09 it needs a newer glib to build QtDBus Jun 15 12:28:25 does it really need it (find it hard to believe) Jun 15 12:28:28 MohammadAG, does anything stop us from upgrading glib as well? Jun 15 12:28:39 achipa, it needs one call Jun 15 12:28:58 Venemo, no, but I'd ask javispedro since I have no idea how glib ties with the rest of the system Jun 15 12:29:21 MohammadAG, basically everything that begins with g depends on glib Jun 15 12:29:24 I would file a bug against it if it's not abysmally difficult to ifdef it Jun 15 12:29:37 MohammadAG, but it's worth a try Jun 15 12:30:38 latest MeeGo daily image is quite fast btw Jun 15 12:31:17 MohammadAG: there has been a lot to improve Jun 15 12:31:22 Can it already be installed to N900? Jun 15 12:31:54 meh, Kasvopus hanged on it Jun 15 12:33:39 k, I'm actually impressed with MeeGo, if only they would settle on a standard UI toolkit (QML isn't standard) Jun 15 12:34:10 MohammadAG do you have MeeGo installed? Jun 15 12:34:36 * SpeedEvil wants athena widgets. Jun 15 12:34:37 can you please check if proximity sensor info is in the same location? Jun 15 12:34:45 Xaw?:D Jun 15 12:34:46 A nice standard toolkit, that everyone knows. Jun 15 12:35:08 blood pressure doubled Jun 15 12:35:08 Let's put the backwards in backwards compatibility! Jun 15 12:35:40 basically, they're pushing QML everywhere Jun 15 12:35:45 I'm not sure that's the right choice Jun 15 12:35:52 trx, same location? Jun 15 12:35:54 MohammadAG, QML isn't (won't be?) a bad thing. Jun 15 12:36:12 mobilemotif Jun 15 12:36:17 qml won't be a bad thing if it achieves market success. Jun 15 12:36:40 all OSs were doing it right at one point Jun 15 12:36:47 Maemo 5, iOS, Android, Symbian Jun 15 12:36:53 iOS did it best though Jun 15 12:36:58 MohammadAG, like on maemo there is "sys/devices/platform/gpio-switch/proximity/state" Jun 15 12:37:10 my SD card is dead Jun 15 12:37:18 so i cant check myself Jun 15 12:37:20 trx: QML isn't getting it right either - for example - the accellerometer driver polls. Jun 15 12:37:29 then came QML, and each application has a new UI that needs time to get used to Jun 15 12:37:33 hi *, I am trying out the omvoiceserver for maps, it seems to recoginize the existance of espeak (which works from cli), but the test button doesn't do anything. Any hints? Jun 15 12:38:29 SpeedEvil it does? Jun 15 12:38:48 trx, no SD card in my N900 :p Jun 15 12:39:04 SpeedEvil, that would be Qt Mobility Jun 15 12:39:10 SpeedEvil, accelerometer driver isn't related to QML.... Jun 15 12:39:11 oh cool Jun 15 12:39:13 Sorry - I'm barely awake Jun 15 12:39:22 My sleep cycle is broken. Jun 15 12:39:24 yeah, qt mobility Jun 15 12:39:32 the new MeeGo terminal has pinch to zoom, remind me how one can pinch on a single touch screen? Jun 15 12:39:33 MohammadAG ok, nevermind then :) Jun 15 12:39:56 trx, point is, I have MeeGo on the eMMC :p Jun 15 12:40:00 SpeedEvil, MohammadAG, MeeGo's Qt Mobility seems to do the right thing (only MeeGo kernel's accelerometer driver sucks) Jun 15 12:40:13 Venemo: probably Jun 15 12:40:28 Venemo: I shoul dwake up properly, and ivnestigate things before commenting. Jun 15 12:40:36 MohammadAG oh :D Jun 15 12:40:37 SpeedEvil, I know this for a fact ( DocScrutinizer and myself investigated the matter) Jun 15 12:41:46 MohammadAG: iOS is not doing that good regarding toolking... cocoa is starting to show it's age and now they have to revamp without breaking stuff... and that's always the most difficult part Jun 15 12:42:22 MohammadAG: and of course, the remark that QML is not a UI toolkit Jun 15 12:42:38 achipa, point is, most apps look the same on iOS Jun 15 12:42:45 and on Maemo/Android etc Jun 15 12:42:51 QML is breaking that Jun 15 12:42:58 no it's not Jun 15 12:43:02 MohammadAG, not QML, but how people use QML Jun 15 12:43:11 the (yet) unavailability of components does Jun 15 12:43:18 but that's soon (TM) to change Jun 15 12:43:22 Venemo, I don't expect everyone to rewrite hildon for maemo 5 Jun 15 12:43:49 MohammadAG, they don't need to. Jun 15 12:43:50 you can see plenty of dumb iOS/Android 'apps' wrapped from web stuff that just looks awful Jun 15 12:44:20 anyway, I think we all agree that QML itself is a good direction, but it isn't yet mature enough Jun 15 12:44:37 sorry, I don't think it's a good direction Jun 15 12:45:05 at least it's a better direction than web and what Android does :) Jun 15 12:45:08 why not? Jun 15 12:45:30 what's wrong with QWidgets again? Jun 15 12:45:35 why deprecate those in favor of QML Jun 15 12:47:55 MohammadAG, they will be working still Jun 15 12:48:11 i don't like the fact that they will be deprecated in the first place Jun 15 12:48:18 Venemo, I don't like working with deprecated APIs Jun 15 12:48:55 I prefer C++ than some JS-based language, if I wanted that I would've gone for WebOS Jun 15 12:49:48 MohammadAG, think of QML as a declarative GUI language and forget the fact that it's JS based Jun 15 12:50:31 how can you forget it's JS based when the syntax is that of JS Jun 15 12:55:23 MohammadAG, do you really program in C++ though? :P Jun 15 12:55:34 MohammadAG, I thought you were one of those cool Qt kids. Jun 15 12:57:19 trx, http://pastebin.com/raw.php?i=EGNW5gZK Jun 15 12:57:28 crashanddie, that's C++ :p Jun 15 12:57:52 Find is written in c++? Jun 15 12:58:15 MohammadAG thanks Jun 15 12:58:15 no, that was a reply to what he said :p Jun 15 13:00:18 MohammadAG, Qt is far removed from C++ ;) Jun 15 13:00:35 re Jun 15 13:33:38 MohammadAG: the biggest problem with QWidgets is that they were essentially a lowest-common-denominator wrappers Jun 15 13:34:14 and QML has nothing to do with QWidgets, you could have had QML+QWidgets (and it would suck of course :) ) Jun 15 13:34:47 achipa, load average: 15.70, 10.79, 5.41 <-- isn't something that should happen :p Jun 15 13:34:48 plus QWidgets are not *deprecated*, bud *done*, blah, blah Jun 15 13:35:02 I always get that they're deprecated on #meego Jun 15 13:36:05 MohammadAG: yes, because nobody maintains them on #meego, but that you can keep on using them on other platforms that DO have an interest in them Jun 15 13:36:48 (wrapper story -> MeeGo doesn't have a native 'way' or 'style' which could be wrapped in QWidgets, so you would actually have to implement stuff, hence the lack of appeal) Jun 15 13:37:30 The Maemo5 QWidgets would be also major suckage if they did not plug-in/reuse/integrate stuff from Hildon Jun 15 13:37:39 (i.e. again, it's not QWidgets that you are missing) Jun 15 13:37:54 that's what MeeGo should do then, have a meego touch theme Jun 15 13:38:30 point is, I have to reimplement stuff whether I go with QWidgets or QML on MeeGo Jun 15 13:38:30 MohammadAG: welcome to MTF Jun 15 13:38:40 achipa, MTF is deprecated Jun 15 13:39:21 MohammadAG: yes, been there, done that :) Jun 15 13:39:27 but eventually somebody realized that reinventing each wheel separately is not a good idea Jun 15 13:39:48 and QML isn't reinventing the wheel? :P Jun 15 13:40:04 MTF would have been (and sort of is, if you look at Harmattan) a good solution, but it would solve the problem ONLY for MeeGo/Harmattan Jun 15 13:40:35 QML is a foundation that can (at least theoretically ;) ) solve it for all platforms, and/or make it easier to change Jun 15 13:40:41 MTF is a good widget set, except it breaks the whole code once deploy everywhere concept Jun 15 13:41:01 exactly my point from 2 lines ago :) Jun 15 13:41:29 having a theme for each system would've been easier :p Jun 15 13:41:59 I would disagree, but hey, everyone can make an estimate :) Jun 15 13:42:36 And *this* is the problem that Qt Quick Components is supposed to address Jun 15 13:42:44 Except on MeeGo. Unless you mean Harmattan. Jun 15 13:42:45 also, most of the QML hatred (from C++, python, etc) comes from JS, not QML per se Jun 15 13:42:47 Jun 15 13:43:18 achipa, isn't QML a glorified JS anyway? Jun 15 13:43:20 if QML is Qt Quick, QWidgets must be Qt Slow? Jun 15 13:43:34 Jun 15 13:43:38 ;P Jun 15 13:43:39 javispedro: but true Jun 15 13:43:41 ;) Jun 15 13:43:53 QWidgets is way quicker than QML :P Jun 15 13:43:57 especially at long lists Jun 15 13:44:18 MohammadAG: not really, has surprisingly little to do with JS actually Jun 15 13:44:51 I wonder how long it'd take me to make a QML version of Sociality Jun 15 13:45:06 as a UI only Jun 15 13:45:20 the whole app took me 3 days to make it functional Jun 15 13:45:38 MohammadAG, long lists of how long Jun 15 13:45:40 MohammadAG: It's like saying HTML is a glorified JS... they are linked, but just because of historical reasons and inertia Jun 15 13:45:41 in QWidgets/C++ of course Jun 15 13:45:48 lcuk, 100+ I'd say Jun 15 13:45:49 are we talking 10s 100s 1000s Jun 15 13:45:51 right Jun 15 13:46:18 i heard someone complaining that a list with a couple of photos and a label were sluggish and non optimal beyond 6 items Jun 15 13:46:28 MohammadAG: I can learn to play a piano piece in 3 days, I bet learning it on a flute would take me 100... Jun 15 13:46:35 because I don't play a flute Jun 15 13:46:47 that doesn't mean a piano is easier to play than a flute Jun 15 13:47:06 lcuk: rubbish or special case Jun 15 13:47:07 my cat can play the piano Jun 15 13:47:14 but he would not be able to play a flute Jun 15 13:47:26 i never said he was in tune or anything Jun 15 13:47:34 but his mass and inertia can press the keys :P Jun 15 13:48:00 I guess this is the other way round... you can click around something in QML... but not with QWidgets... Jun 15 13:48:04 you should have seen him jump the first time he did it on jakes toy piano Jun 15 13:48:14 click around? Jun 15 13:48:26 QML is the best toolkit for making WP7-style panorama applications =) Jun 15 13:48:36 is it? Jun 15 13:48:40 :) Jun 15 13:48:47 lcuk: well, that's all I know to do with it! Jun 15 13:49:06 javispedro, I have said this before: Jun 15 13:49:20 it is not the toolkit, but the IDE which differentiates utility Jun 15 13:49:25 lcuk: there is a little bit of thinking involved in how you make your model - even QGV is fast enough for a 6 item list Jun 15 13:49:29 because the IDE turns your ideas into code Jun 15 13:49:42 lcuk: but if you make a sh!+ model then yes, it will be jerky Jun 15 13:49:51 i agree achipa Jun 15 13:49:52 well, 254 QWidgets each showing an image scrolls through and works well Jun 15 13:50:01 I don't want to imagine that in QML Jun 15 13:50:08 it would be nice I expect Jun 15 13:50:09 lcuk: you have a point, as the qt creator qml designer resembles the ms wp7 designer much Jun 15 13:50:29 MohammadAG, qml is rapid (like other things) if you keep the nuimber of layers minimal Jun 15 13:50:36 even gtk is Jun 15 13:50:45 lcuk: the fact that I was constantly thinking about wp7 while plastering controls on the form probably is the reason I unconsciously made a panorama UI Jun 15 13:50:48 Qt works for long lists Jun 15 13:50:51 * lcuk wonders why his fish are swimming backwards Jun 15 13:50:52 QML jerks a bit Jun 15 13:51:02 lcuk, Guinness in the pool? Jun 15 13:51:52 MohammadAG, http://liqbase.net/20110614_003.mp4 (2.5mb tiny) Jun 15 13:52:03 meh, I think I'll try QML, but only as a UI Jun 15 13:52:04 sketches of fish swimming in a liqflow pool :) Jun 15 13:52:10 MohammadAG: when you say list, are you comparing the same base (i.e. same C++ model) or just stuff thrown in a scrollarea ? Jun 15 13:52:17 if I like it, I'll back it up with some C++ and make the switch Jun 15 13:52:32 achipa, latter Jun 15 13:52:54 MohammadAG: well yes, of course QML will be slower, because you are not doing the same thing Jun 15 13:53:08 achipa, hm Jun 15 13:53:15 how come scrollarea is slow? Jun 15 13:53:19 MohammadAG: what people don't realize is that QML is essentially the *SAME* c++ widget code Jun 15 13:53:21 achipa, shouldn't a QScrollArea behave the same way as a ScrollArea? Jun 15 13:53:27 technically a single moving backpane and showing the items onscreen? Jun 15 13:53:57 seeing as we are resolution limited, those items will be a fraction of all items Jun 15 13:54:04 or does qml/qt do extra to render? Jun 15 13:55:05 lcuk: the QML ListView stuff is done with models Jun 15 13:55:22 which is actually the more sane way to do it Jun 15 13:55:34 hrr Jun 15 13:55:38 so that your memory/resource consumption does not depend on the size of the list, etc Jun 15 13:55:42 so's QListView, but I admit I never used that Jun 15 13:55:54 achipa, but each visible node on the list has visibilityu check and nothing rendered if it is offscreen etc Jun 15 13:56:09 achipa, hm Jun 15 13:56:21 * lcuk just initialises/loads as things come onto screen sometimes Jun 15 13:56:32 QListWidget is saner imo :p Jun 15 13:56:52 this is code. there is no sanity. Jun 15 13:56:55 lcuk: correct, and listview should do the same, in fact it should recycle the elements (so if your list has a million items, you don't get a list with a million c++ objects) Jun 15 13:57:21 you will only have as many objects as visible and the rest will be created/cached from the model Jun 15 13:57:22 achipa, what happens if each item is a different size? Jun 15 13:57:46 you need to store your data, then a million references to the size of each element then rejig everything anyway to know where in the list you are Jun 15 13:57:55 lcuk: same story - the point being it get killed if it goes too far off screen Jun 15 13:58:11 lcuk: exactly, and that's why you need a fast model Jun 15 13:58:19 but to know whether it is onscreen needs to know the history of all proceeding items Jun 15 13:58:21 achipa, QML is exceptionally slow, and it uses 10 times the memory the equivalent QWidget solution uses Jun 15 13:59:13 Venemo, ignore the memory construct issue Jun 15 13:59:16 achipa, part of this could be fixed by translating QML into C++ code (at build time) and compiling it Jun 15 13:59:19 slowness is more worrysome Jun 15 13:59:22 which aspect is slow Jun 15 13:59:23 Venemo: I would disagree but there is a high chance we don't interpret the word 'equivalent' the same way Jun 15 13:59:35 Venemo: say hello to qtquick 2.0 Jun 15 13:59:46 achipa, QML 2.0 isn't compiled either Jun 15 13:59:57 achipa, it just uses a different JS engine Jun 15 14:00:03 (AFAIK) Jun 15 14:00:21 there ws some discussion about this on the SF MeeGo summit and compiled QML was on the table Jun 15 14:00:22 Venemo, QML and HTML are similar DOM models Jun 15 14:00:35 and nobody advocates compiling those to c++ first Jun 15 14:00:45 Venemo: the compilation won't help you much per se, you ARE getting C++ objects Jun 15 14:00:53 achipa, yeah Jun 15 14:00:54 shitload of time needed to start a QML app :/ Jun 15 14:00:55 Venemo: what you save on is the initial parsing Jun 15 14:01:02 right MohammadAG Jun 15 14:01:10 did you see the qml performance tool Jun 15 14:01:17 achipa, sure, but QML itself is interpreted still. fortunately the new JS engine will have JIT compilation. Jun 15 14:01:18 that one I don't dispute at all :) Jun 15 14:01:18 initialisation goes through so many layers Jun 15 14:01:24 and so many thousands of times Jun 15 14:01:27 Venemo: QML is not. JS is Jun 15 14:01:33 no, but this reminds me of Python Jun 15 14:01:37 Venemo, HTML is interpreted Jun 15 14:01:41 same way Jun 15 14:01:45 same level of detail Jun 15 14:01:51 hmm Jun 15 14:01:57 a descriptive language, "Interpreted"? Jun 15 14:02:02 achipa, will I be able to use the backends of QML from C++? eg. in Qt 5 can I write a C++ app that uses the same things as QML? Jun 15 14:02:21 Venemo: you sort-of already can Jun 15 14:02:38 4343 4332 root S 22676 9.2 0.0 /opt/socialityqml/bin/socialityqml Jun 15 14:02:39 Venemo: it's a very high price to pay just to be able to say 'yay, I did not use JS or QML !' Jun 15 14:02:41 what the f Jun 15 14:02:48 9.2MBs for a Hello World?! Jun 15 14:03:01 MohammadAG: AKA debug build Jun 15 14:03:02 nice memory usage Jun 15 14:03:14 achipa, Release build Jun 15 14:03:19 MohammadAG, your qt app uses memory too Jun 15 14:03:31 also - the memory usage is somewhat linux-style, we like preloading and preinitializing a lot Jun 15 14:03:32 lcuk, well duh, but not 9.2MBs for "Hello world" Jun 15 14:03:45 I am sure microsoft can top that Jun 15 14:03:55 I don't care about Microsoft Jun 15 14:04:09 I'll see how much memory a XAML hello world uses Jun 15 14:04:12 a QMainWindow with a QLabel saying Hello World takes 4MBs Jun 15 14:04:17 (estimation) Jun 15 14:04:44 bah, just double? Jun 15 14:04:45 as usual, the question is whether this is incremental or not Jun 15 14:04:50 what the f. 4mb for a Hello World?! Jun 15 14:04:50 that's a few years of hw only ;) Jun 15 14:05:02 if at 100MB the difference is 100 vs 105, I don't give a s&&t Jun 15 14:05:32 that's like saying python uses 10000x more memory than C Jun 15 14:05:40 it does. Jun 15 14:05:43 :D Jun 15 14:06:14 my question is why? Jun 15 14:06:24 the point being that for print "hello world" python will fare bad Jun 15 14:06:28 as achipa I understood it was just a way of expressing Qt constructors in a shorter way Jun 15 14:06:47 so it would also use the same qobjects, same paint() methods, Jun 15 14:06:48 etc. Jun 15 14:06:57 I imagine the next device should have 1GB of HW RAM for this kind of stuff Jun 15 14:07:10 MohammadAG: otherwise it would look bland when compared to recent devices Jun 15 14:07:23 javispedro, most apps have a work part Jun 15 14:07:25 javispedro, no, it would suck from all the QML apps Jun 15 14:07:28 the bit that is more than glue. Jun 15 14:07:35 most of MeeGo seems QML by now Jun 15 14:07:35 and that work part is where slowness is Jun 15 14:07:37 lcuk: not hello world =) Jun 15 14:08:04 MohammadAG: not really... the N900s 256 is small, it works just fine with 512... with 1gb it fliesssss Jun 15 14:08:26 achipa, are you referring to the rumoured harmattan devices? Jun 15 14:08:30 javispedro: that is what happens Jun 15 14:08:30 :D more ceiling Jun 15 14:08:54 achipa, I'd imagine 512 is way too low Jun 15 14:08:56 achipa: but why does it seemingly use more RAM? Jun 15 14:09:23 hildon-desktop, home, status-menu etc probably take up 8MBs of RAM together Jun 15 14:09:32 pffffffffffffffffffffffffffffffffffffffffffffft Jun 15 14:09:34 only 8? Jun 15 14:09:39 now I'd compare the Intel QML UI to that, but cba Jun 15 14:09:44 Venemo: I have all sorts of hardware running all sorts of Qt versions if that's you mean... the exopc for example has 1gb, beagleboard 512mb... no ieda what harmittan stuff you're talking about ;) Jun 15 14:09:52 hehehe Jun 15 14:10:01 achipa :) Jun 15 14:10:43 MohammadAG, right now, hildon-desktop uses 3.5 MB and puzzle-master uses 9.8 MB Jun 15 14:10:54 javispedro: yes it does, but much of that overhead is simply from (at this point) focusing more on features than saving memory Jun 15 14:11:06 javispedro: you will never ever see intel stuff below 1gb Jun 15 14:11:26 let's hope that QML will be better in Qt 5 Jun 15 14:11:27 Venemo, puzzle master is a game Jun 15 14:11:28 Which will mean that caring about memory gets left behind. Jun 15 14:11:28 i have a 486 that proves you wrong. Jun 15 14:11:36 when you compare, do it with apps that do the same Jun 15 14:11:37 achipa, ^ Jun 15 14:11:42 so IMHO it's understandable that most of the time is spent on making stuff work, esp with regard to scenegraph, than battling QGV and QtScript memory usage Jun 15 14:11:47 i.e compare hildon-desktop with whatever desktop meego uses Jun 15 14:11:54 Also - RSS is a lie! Jun 15 14:12:02 lcuk: RECENT Intel stuff, as in Atoms, etc, etc Jun 15 14:12:03 :) Jun 15 14:12:14 I am sure the 486 is made from atoms too Jun 15 14:12:20 RSS counts - for example - the address space libc is mapped into Jun 15 14:12:20 but not Atoms Jun 15 14:12:22 MohammadAG, well yeah, but p-m shouldn't use 9.8 MB either... Jun 15 14:12:23 :) Jun 15 14:12:24 http://www.selenic.com/smem/ Jun 15 14:12:25 lcuk, :):) Jun 15 14:12:28 lol Jun 15 14:12:51 Venemo, Qt is loaded into memory Jun 15 14:13:11 compare Qt apps together to cancel out differences Jun 15 14:13:23 what is the memory stuff in top as Jun 15 14:13:30 what does VIRT/RES/SHR mean Jun 15 14:13:35 for example, puzzle master takes up 9MBs, my mediaplayer takes up 7-ish Jun 15 14:13:40 top is not the best way for gauging memory usage Jun 15 14:13:48 just roughly Jun 15 14:13:48 use htop Jun 15 14:13:54 do I add up all those? Jun 15 14:13:58 virt is pointless, forget it Jun 15 14:14:07 res is kinda it Jun 15 14:14:08 what is it though? Jun 15 14:14:11 SHR - shared with other apps (such as shared libraries) Jun 15 14:14:14 shr is shared Jun 15 14:14:19 VIRT - swapped out to disk Jun 15 14:14:30 not necessarily swapped out... allocated Jun 15 14:14:33 RES - actual memory amount currenty owned by the app Jun 15 14:14:33 btw, KDE's way of showing memory usage is nice Jun 15 14:14:41 doesn't mean it's used or swapped... Jun 15 14:14:46 MohammadAG, memory usage is irrelevant when VirtualBox uses 103% of my CPU Jun 15 14:14:55 as we know our beloved linux loves to overcommit :) Jun 15 14:14:59 according to top Jun 15 14:15:07 103%? Jun 15 14:15:10 helo RST38h Jun 15 14:15:16 ehlo javispedro Jun 15 14:15:26 2439 lcuk 20 0 86852 18m 3988 S 3.3 1.9 0:05.58 liqbase-playgro Jun 15 14:15:34 that's just top's silly way of saying 'more than one core/hyperthread' Jun 15 14:15:37 Venemo whats wrong with that Jun 15 14:15:43 nid0 :) Jun 15 14:15:51 to be fair to it, it does have every sketch ever created since day1 available and prelisted Jun 15 14:15:54 genuine question Jun 15 14:16:05 nid0, how can you use more than 100% of a CPU? Jun 15 14:16:12 Venemo: top cpu % is / by number of cores/threads available Jun 15 14:16:19 Venemo: by having multiple cores/threads Jun 15 14:16:24 mhmm Jun 15 14:16:30 if you're using a quad core hyperthreaded cpu, thats 13% Jun 15 14:16:47 I use a dual-core non hyperthreaded cpu Jun 15 14:17:14 it LOOKS better to say 103% than to divide all values b the number of cores/HTs Jun 15 14:17:21 imagine the server folks Jun 15 14:17:23 achipa, looks better indeed :) Jun 15 14:17:26 hehe Jun 15 14:17:32 cringing why is my system hardly going over 10% ! Jun 15 14:18:20 achipa, because you don't run windows in VirtualBox Jun 15 14:19:38 * MohammadAG doesn't get the concept of widgets in QML Jun 15 14:20:36 anyway, I will go and get some food now Jun 15 14:20:40 bye-bye guys :) Jun 15 14:20:43 bbl Jun 15 14:29:39 khertan: how can I add a sudo line in my package? Jun 15 14:29:52 MohammadAG: think about QML as .ui files if you have to Jun 15 14:30:09 MohammadAG: what does it matter where QLabel comes from Jun 15 14:30:28 achipa, sorry, I can't do that, ui files are compiled, QML files aren't Jun 15 14:30:47 let's pretend for a moment that they are Jun 15 14:31:18 k, so instead of a QPushButton I need to implement a Button.qml file Jun 15 14:31:33 MohammadAG: no, you get that from qt components Jun 15 14:31:44 achipa, is Qt components on maemo Jun 15 14:31:47 yes Jun 15 14:32:03 and on all systems with 4.7.2+? Jun 15 14:32:38 yes, though some might lack color styling or lack particular components (say, 20footDatePicker) Jun 15 14:32:51 but the idea is that they would be omnipresent Jun 15 14:33:23 does anyone knows how to add a sudo line in a package I write Jun 15 14:33:23 ? Jun 15 14:33:29 and yes, it's a shame the recent work on them is not (yet) available, but soon TM that obstacle will go away too Jun 15 14:35:38 MohammadAG: and where QML components kick ass compared to QWidgets is that most of the time they have very little native code, so you can swap them around Jun 15 14:36:29 achipa, how do I get Qt Components? do I just copy the qml files into my project? Jun 15 14:36:51 * javispedro ponders creating the wp7 pivot control in qml for a start Jun 15 14:36:56 MohammadAG: that is the ugliest way (but nonetheless works) Jun 15 14:37:23 on a linux system that hasn't forsaken dependencies you would depend on them like any other package Jun 15 14:37:57 but they are certainly more mobile (in terms of code) than the binary libQtGui stuff Jun 15 14:38:53 a big question: qml is not directly reliant upon qt is it? Jun 15 14:39:05 GML+ ;P Jun 15 14:39:16 well, like we have various browsers Jun 15 14:39:19 each parsing html Jun 15 14:39:20 lcuk: define directly... Jun 15 14:39:48 it would be feasible (for instance) for google to produce qml parser Jun 15 14:39:56 for direct use in android Jun 15 14:39:56 there is nothing preventing you from making your own declarative engine if that's what you mean Jun 15 14:40:38 but that would be sort of like what gnash is trying to do with SWF, that sort of doesn't really work in practice Jun 15 14:40:45 (sort of) Jun 15 14:40:52 i understand Jun 15 14:41:37 are there any other declarative languages? Jun 15 14:41:42 but then again, Google got away by snatching Java and violating it into a form that is nowadays more popular than the original :) Jun 15 14:41:44 is html5 and css3 close? Jun 15 14:42:49 close... maybe, better... IMHO not (and the gap is widening and HTML standardization is snail-paced) Jun 15 14:43:07 NOK at 6.03 Jun 15 14:43:31 buy buy buy! Jun 15 14:43:39 (please! ;P ) Jun 15 14:44:06 last one out at nok please shutdown the lights Jun 15 14:45:50 the light are already off, it's summer time (it does get a bit darker around 1AM on the south) Jun 15 14:46:17 pft, same as here, no money left even for A.C Jun 15 14:46:21 s/A.C/AC Jun 15 14:46:46 it's also the end of public universities here =) Jun 15 14:47:54 but I have to wonder how Nokia investors are going to congratulate Elop for chopping share value in half Jun 15 14:52:35 javispedro: They'll be "happy" waiting to see the reaction to the WP devices :-/ Jun 15 14:53:26 what's all the talk about the 21st being something big Jun 15 14:54:43 according to J. random blogger invitations have been distributed for a nokia event Jun 15 14:54:54 which would map to the fcc expiration date nicely Jun 15 14:55:09 MohammadAG: June 21st is the biggest day in 2011 Jun 15 14:55:27 quite literally, it is the longest day in the northern hemisphere Jun 15 14:55:32 hey again Jun 15 14:55:35 * achipa ducks for cover Jun 15 14:58:04 javispedro, well, yeah, the 24th, though I'm not sure if the new "disruptive" device is the N9 Jun 15 14:58:12 achipa, really? this is the longest day? Jun 15 14:58:43 Venemo_N900: yessiree, in 2011 it's the 21st Jun 15 14:58:56 MohammadAG, I'm not sure how disruptive it will be, if at all Jun 15 14:59:16 Venemo_N900, it will lower stocks by 90% Jun 15 14:59:17 achipa :) Jun 15 14:59:24 MohammadAG, why? Jun 15 14:59:25 Mohammandag: who knows. I am putting my finger on this one though. Jun 15 14:59:56 just saying, that's how disruptive it's going to be :p Jun 15 15:00:06 lol Jun 15 15:00:26 maybe whats going to be disruptive is elop and investors commiting collective suicide Jun 15 15:01:02 that would be disturbing, not disruptive! :p Jun 15 15:01:06 maybe Jun 15 15:01:20 disturbing?? Jun 15 15:01:22 it wouldn't disturb me at all Jun 15 15:01:29 I'd bring in the champagne! Jun 15 15:01:41 xD Jun 15 15:01:53 no fine wine? Jun 15 15:01:59 javispedro, deal! Jun 15 15:02:10 achipa, this weekend I am heading to London Jun 15 15:02:54 well, I am if I have been paid Jun 15 15:03:01 to an event called Hacktivate Jun 15 15:03:14 where idea is to describe how you would use technology to change the world Jun 15 15:03:22 and to try and build prototypes Jun 15 15:04:32 lcuk, I would use nuclear power plants to save the atmosphere Jun 15 15:05:07 unfortunately, one of them blew up and now the stupid masses think that nuclear energy is bad Jun 15 15:05:18 i would burn the atmosphere to save the planet Jun 15 15:05:48 * javispedro --> away Jun 15 15:08:19 heh Jun 15 15:08:23 bitcoin fail Jun 15 15:08:24 http://gawker.com/5811868/a-500000-geek-cyberheist Jun 15 15:14:10 user fail* Jun 15 15:14:48 who keeps a wallet with $500,000 in it somewhere nickable, regardless of whether its cash, travellers cheques, or bitcoins Jun 15 15:19:16 nid0: But isn't that part of the problem? Most people aren't geared up to store any form of money worth $500,000 Jun 15 15:25:34 lcuk: actually I should have been there myself Jun 15 15:26:47 lcuk: but there are two countries in the whole of the European continent that require paperwork for me to get in, and the UK is one of those :) thus I passed... Jun 15 15:27:35 achipa, cool Jun 15 15:27:41 do you know anyone else that is going? Jun 15 15:28:33 I know there will be a couple of FN guys there Jun 15 15:28:39 Jaffa: most people arent, because they arent storing $500,000 Jun 15 15:28:44 if you are, you take care of it Jun 15 15:28:45 dont you? Jun 15 15:28:52 or DN or whatever we are called :) Jun 15 15:30:15 Venemo_N900: wait until an earthquake brings down the hoover dam or sorts, a hurricane kills the wind plants and some will start advocating that we should not use electricity at all :) Jun 15 15:30:39 achipa, heheh Jun 15 15:31:24 achipa, by that time we will need electricity more than ever Jun 15 15:31:44 mostly because all fossil fuel will run out Jun 15 15:31:45 Nokia bicycle chargers FTW ! :D Jun 15 15:39:45 * user0 is back! Jun 15 15:40:26 DocScrutinizer MohammadAG , Jun 15 15:40:29 any last suggestions ? Jun 15 15:40:48 before i send this device back to the store :P Jun 15 15:41:00 nuclear cars ftw Jun 15 15:41:28 huh? Jun 15 16:10:09 so much for the topic :/ Jun 15 16:10:29 http://www.nokian900applications.com/nokia-n900-compatible-with-android-market/ Jun 15 16:10:47 anyone know if there is an easy way to make the easy debian image size larger? Jun 15 16:13:44 Jef91, ask qole Jun 15 16:19:20 http://www.youtube.com/watch?v=MqD9UaDpA_0 <<< nice MeeGo Netbook Bug Jun 15 16:19:21 :) Jun 15 16:26:51 khertan: why do the file I install from the package are forgetting their prmitions? Jun 15 16:27:18 khertan: how to make all my .py files have 755? Jun 15 16:30:48 it s didn't care of permission as it didn't exist on fat drive Jun 15 16:31:08 so i set the permission in the post install script Jun 15 16:31:16 ok Jun 15 16:31:38 need I also add a script for uninstall? Jun 15 16:32:01 if you need to do other action than removing file no Jun 15 16:32:12 update-sudoers Jun 15 16:32:19 if you didn't need to do other action than removing file no Jun 15 16:32:34 update-sudoers ? Jun 15 16:32:43 you modify the sudo file prefs ? Jun 15 16:32:46 so yes Jun 15 16:34:02 can I use p to find any file ending with .py and chmod 755 it? Jun 15 16:34:08 hmmm Jun 15 16:34:18 I only have a few files.. Jun 15 16:37:36 cloudyLights, yep it s possible Jun 15 16:39:46 "#!/bin/sh\nchmod 755 "+"\nchmod 755".join(p['/usr/...'] should works ... not tested Jun 15 16:51:16 khertan: help pls Jun 15 16:52:07 err. min Jun 15 16:53:36 ok got my mistake Jun 15 16:55:20 crashanddie: still there? im home now and im looking for that community player you mentioned earlier. Jun 15 16:55:27 cant seem to find it in my repositories Jun 15 16:55:36 do i need to add one? Jun 15 16:55:46 HRH_H_Crab, which player are you referring to? Jun 15 16:55:50 ~cssu Jun 15 16:55:50 somebody said cssu was http://wiki.maemo.org/Community_SSU Jun 15 16:56:03 crashanddie: thanks. Jun 15 17:10:07 Well. Moo. Jun 15 17:15:14 hmmm. is the thing im looking for called "someplayer" ? Jun 15 18:07:12 crashanddie: hmmm. im guessing that the media player you are talking about is the "standard" one that is replaced when you install the cssu stuff? Jun 15 18:07:24 if so, sadly it still isnt gapless. :( Jun 15 18:07:54 (or ive done something wrong) Jun 15 18:08:24 does anyone need a app to silence the phone by the calender? Jun 15 18:08:32 I am looking for testers Jun 15 18:08:56 i suppose the implication is that now i at least have one that is potentially still being improved so might have that feature at some point in the future though, which is a good thing. Jun 15 18:09:12 my app will switch the N900 to silent when a meeting starts , and then switch it back to regular Jun 15 18:20:23 HRH_H_Crab: just had a discussion in cssu chan about why cssu is enforcing new "replacement" apps that can't switch back to the original ones. It's indeed a flaw in CSSU concept in my book Jun 15 18:22:21 DocScrutinizer: i havent noticed any problems per se. Jun 15 18:22:52 its just that in this case the replacement app hasnt fixed the issue with the original one that i hoped it would. Jun 15 18:23:33 anyway, the cssu stuff makes it quite clear that its for powerusers / testers etc. Jun 15 18:23:45 and in a worse case scenario one could just reflash i guess. Jun 15 18:24:07 for all subsystems where that's possible, such replacements should either come as quite normal package in extras(-devel) and provide an interface to user where she can select what's the default "system" app handling e.g. mp3. Or, for sake of conserving space, you get an *optional* package from CSSU-repo that you can install whenever you like, and that package would replace your standard mediaplayer by the new one permanently and free the Jun 15 18:24:08 storage eaten by the meamo5-stock component Jun 15 18:26:15 i dunno. im ot a developer. just a somewhat "geeky" end user. i think their approach seems o.k. in as far as i understand it, but i can see your point of view too. Jun 15 18:27:01 there's no good rationale for CSSU to permanently replace stock system bits like mediaplayer, camera, whatnot - other than to keep storage footprint small Jun 15 18:27:14 but as i see it, Jun 15 18:27:50 just replace rogue things that kill your system - make the rest optional Jun 15 18:27:57 the whole point of installing cssu would be because a user wasnt happy with the status quo (closed source original apps that are no longer being developed / getting bug fixes) Jun 15 18:29:23 not sufficiently a rationale to *force* users of CSSU to permanently irreversibly replace all their stock system components Jun 15 18:29:59 as long as there's no dependency to the new stuff, make it optional Jun 15 18:31:13 if it needs the special hacks to allow removal of parts of maemo5-metapackage aka system aka MP, get it into CSSU that is about doing exactly that Jun 15 18:31:44 but still you can make it an optional package to install via HAM from cssu-repo Jun 15 18:32:36 well, speaking as an end user, i'll just be happy if one day the "stock" media player has gapless playback. Jun 15 18:32:39 :P Jun 15 18:32:55 i have way too many dj mix cds that sound like shit on my n900 Jun 15 18:33:01 CSSU isn't meant to be the mutually exclusive alternative to stock PR1.3 Jun 15 18:33:30 and even albums like "live at folsom prison" by johnny cash have segues between songs that sound crappy when it transitions between them. Jun 15 18:33:40 HRH_H_Crab: completely on same page with you regarding gapless Jun 15 18:47:11 Mohammad-, mind iding yourself or something... Jun 15 18:48:00 lol >_> Jun 15 18:48:21 yeah, makes tabcompleting your name confusing :P Jun 15 18:48:44 it's not me Jun 15 18:48:58 unless I have another XChat instance up, but that would show my IP Jun 15 18:49:36 yeah Jun 15 18:49:54 HRH_H_Crab: also, you could just... install a different media player and use that Jun 15 18:50:05 i really liked mediabox, not sure about gapless but it's open, soo Jun 15 18:50:36 ZNC is clean, no duplicate users Jun 15 18:53:19 rm_work: ive tried several of the other ones but i don't like them much. Jun 15 18:54:48 iirc some (none?) of them would pause correctly if there was an incoming call. Jun 15 18:55:02 maybe i should try trawling through the various options again. Jun 15 18:59:29 yeah dunno Jun 15 18:59:42 ono my home internet appears to have hiccuped Jun 15 18:59:45 >_> Jun 15 19:25:37 Tick tick tick Jun 15 19:27:26 boom Jun 15 19:49:01 yo Jun 15 19:49:28 howto launch an command in osso-xterm from osso-xterm. Jun 15 19:49:46 i.e. osso-xterm -e ping google.com? Jun 15 19:49:58 any o' you bro's know? Jun 15 19:50:32 MohammadAG: for example? Jun 15 19:54:40 osso-xterm "ping google.com"? Jun 15 19:57:40 MohammadAG: that Mohammad- is a ZNC instance running on ool-44c62415.dyn.optonline.net == 68.198.36.21 using secure connection. I recall you used this "account" quite some time ago. I don't think pinging it and asking to "id" does help - you should contact the sysop of the server at 68.198.36.21 and ask hin to get your account out of that ZNC for good Jun 15 19:59:00 seems the server is located in Connecticut, USA - if that helps you remember who might be owner of this ZNC Jun 15 19:59:26 Only server I know of in the US is in New York Jun 15 19:59:34 maybe Jun 15 19:59:44 locateIP isn't reliable Jun 15 19:59:48 ZNC can only run once Jun 15 20:00:08 ?? Jun 15 20:00:15 MohammadAG: 'osso-xterm ping google.com' causes an instance of osso-xterm to start, however it crashes immediatley Jun 15 20:00:25 crashes or stops Jun 15 20:00:31 ZNC can only have one instance with the same config Jun 15 20:00:42 vi__, probably cause ping needs root privs Jun 15 20:00:47 n one server, yes Jun 15 20:00:59 yeah, that's the only server with the config Jun 15 20:01:07 probably you're right about that Jun 15 20:01:48 as there can't be two ZNC listening to same port for clients connecting Jun 15 20:02:13 just checked a full list of processes, only one Jun 15 20:02:31 anyway evidently that gone-mad ZNC is on 68.198.36.21 Jun 15 20:02:42 * MohammadAG stops ZNC Jun 15 20:02:51 LOL Jun 15 20:02:55 woop Jun 15 20:02:55 byebye mohammad- Jun 15 20:03:30 heh Jun 15 20:03:33 hi again Mohammad- Jun 15 20:03:43 it's definitely yours, it took over your name :P Jun 15 20:03:47 no, no program will run when invoked this way Jun 15 20:03:51 funnily though Jun 15 20:03:55 if i type Jun 15 20:04:20 grr Jun 15 20:04:34 hi MohammadAG51 Jun 15 20:04:55 what is the run-standalone.sh script for? Jun 15 20:05:13 vi__: to allow root to access user session dbus Jun 15 20:05:15 vi__: IIRC running things without having to deal with connecting to dbus Jun 15 20:05:36 because otherwise the system will kill whatever you run after a few seconds Jun 15 20:05:51 thank you Jun 15 20:06:41 is there a way to find out all users who have *@ool-44c62415.dyn.optonline.net ? Jun 15 20:06:54 Yes Jun 15 20:06:55 alas not Jun 15 20:07:00 on my client Jun 15 20:07:04 Hack OptOnline, get their customer records Jun 15 20:07:10 irssi users know of ways to do Jun 15 20:07:16 echo "ls" | osso-xterm does not work either Jun 15 20:07:39 Something like /who * with a filter will do Jun 15 20:07:48 vi__: osso-xterm doesn't take stdin Jun 15 20:08:31 MohammadAG: you know about chat to ZNC by /querying *status ? Jun 15 20:08:35 then how do I invoke an instance of osso-xterm running a particular program? Jun 15 20:08:50 DocScrutinizer, yes, but it's not the ZNC I'm on Jun 15 20:09:13 vi__, add a read at the end to see why it exists Jun 15 20:09:23 err Jun 15 20:09:27 ? Jun 15 20:09:31 using osso-xterm obviously doesn't spawn a shell, so when the process exits, the terminal closes Jun 15 20:09:42 vi__: you might have a look into mc.desktop Jun 15 20:09:54 MohammadAG: exactly Jun 15 20:09:57 to the point Jun 15 20:10:26 osso-xterm ash -c "ls -l" Jun 15 20:10:45 or osso-xterm "ls -l && read foo" Jun 15 20:10:52 yep Jun 15 20:11:52 MohammadAG: netstat is your friend, regarding weird phantom ZNC instances ;-) Jun 15 20:12:00 that's not working btw Jun 15 20:12:03 for some reason Jun 15 20:12:15 MohammadAG: && is shell syntax Jun 15 20:12:25 osso-xterm "ash -c 'ls -l && read foo'" Jun 15 20:12:56 oh and use ; not && Jun 15 20:13:00 xterm can't spawn "ls -l && read foo" : file not found Jun 15 20:13:24 bah Jun 15 20:13:32 why is this so annoying? Jun 15 20:13:55 vi__, osso-xterm "ash -c 'ping google.com; read foo'" Jun 15 20:14:33 even osso-xterm "ash -c 'ping google.com; read foo; exit'" ;-D Jun 15 20:15:46 that does work, however I am HUNGRY to understand why Jun 15 20:15:59 what is the parameter -c do? Jun 15 20:16:10 what is the 'read foo' for? Jun 15 20:16:17 xterm just loads and executes ONE binary Jun 15 20:16:49 obviously -c = command, read read some stuff from stdin to a environment variable, but here it just prevents the closing of osso-xterm Jun 15 20:16:52 I see the ping google.com is essentially a 1 line script (terminated by ;) Jun 15 20:16:53 -c is telling this binary which happens to be shell to execute the -commmand Jun 15 20:17:09 aah Jun 15 20:17:12 read foo reads the keypress into the variable $foo Jun 15 20:17:22 of course, pressing anything will return Jun 15 20:17:58 so something at the end of a script like echo "Press any key to continue or press any other key to exit" && read keypress would work Jun 15 20:18:02 then exit closes the shell which in turn makes xterm quit Jun 15 20:18:13 there's no need for that afaik Jun 15 20:18:24 I love you guys, teach me your ways. Jun 15 20:18:26 could be wrong, but my battery ran out Jun 15 20:18:41 isn't -c supposed to exit the shell when all commands return? Jun 15 20:18:48 maybe Jun 15 20:19:57 o/ Jun 15 20:20:02 bbl Jun 15 20:20:26 nah just a moment Jun 15 20:21:01 MohammadAG: what did you do then to make Mohammad- quit and re-appear, a few minutes ago? Jun 15 20:21:43 [2011-06-15 22:01:36] byebye mohammad- Jun 15 20:21:45 [2011-06-15 22:02:12] heh Jun 15 20:21:46 [2011-06-15 22:02:14] hi again Mohammad- Jun 15 20:22:31 MohammadAG: you obviously have control over that ZNC instance Jun 15 20:22:56 DocScrutinizer, nah, it connected faster than my proper ZNC Jun 15 20:23:03 hmm Jun 15 20:23:22 Mohammad- didn't quit Jun 15 20:23:31 when I quit, it immediately changed nicks, that's how ZNC works Jun 15 20:23:47 err Jun 15 20:24:13 netstat is your friend, find the processes that keep connections to freenode Jun 15 20:24:21 here's an example Jun 15 20:24:37 actually sec, need to do something first Jun 15 20:25:26 MohammadAG: I could do that for you, using netstat ;-) Jun 15 20:25:42 MohammadAG: I could do that for you, using nmap ;-) Jun 15 20:25:58 but I guess netstat is faster and more informative Jun 15 20:28:03 http://mg.pov.lt/maemo-irclog/latest.log.html#t2011-06-15T23:01:23 Jun 15 20:29:11 hi javispedro Jun 15 20:29:18 mo Jun 15 20:29:28 * DocScrutinizer waves Jun 15 20:29:35 * DocScrutinizer heads out Jun 15 20:30:15 moo javispedro Jun 15 20:30:34 MohammadAG: clear your nicknames module in ZNC Jun 15 20:30:44 * javispedro hits N810 Jun 15 20:30:59 DocScrutinizer, my ZNC has this IP Jun 15 20:31:03 today the bme rebooting bug is specially ... damnable Jun 15 20:31:03 err, hostmask Jun 15 20:31:09 as I'm trying to copy a few files and it keeps rebooting Jun 15 20:31:13 pool-96-250-166-104.nycmny.fios.verizon.net Jun 15 20:31:19 that's the only one I have root access to Jun 15 20:32:17 * javispedro curses Jun 15 20:32:40 anyone remembers if a fixed n810 bme was released by nokia? I kinda remember the story.. Jun 15 20:33:09 MohammadAG: *** Mohammad- is now known as MohammadAG [23:01] Jun 15 20:33:41 DocScrutinizer, I don't have access to that Jun 15 20:33:57 I see, find owner of that server Jun 15 20:34:07 javispedro: I don't think so Jun 15 20:34:15 Though I haven't been paying that close attention Jun 15 20:34:20 somebody who granted you ZNC some 12 months ago Jun 15 20:37:32 my only ZNCs were yours and the one I'm currently on Jun 15 20:37:46 so unless someone's moving servers, I have no idea where that's from Jun 15 20:48:29 funny Jun 15 21:07:41 there sure are a lot of people on tmo with very little grasp of english lately Jun 15 21:09:16 Sazpaimon: politeness and courtesy too most likely ;) Jun 15 21:12:32 maemo.org seems geared toward n900, is n8x0 still supported? Jun 15 21:13:14 there's a few dedicated n8x0 users on the forum Jun 15 21:14:04 good to hear Jun 15 21:14:18 just picked my first n800 up off ebay Jun 15 21:15:03 * lcuk ponders requesting maintainers for n8x0 Jun 15 21:16:04 looking for a website or instructions which will bring me quickly up to speed. Jun 15 21:16:28 looking for maintainers is hard when dealing with old hardware and volunteer projects Jun 15 21:16:41 http://wiki.meego.com/ARM/N8x0 Jun 15 21:16:55 appears like it may have stalled... Jun 15 21:17:36 was hoping being a linux device the support might be long lived Jun 15 21:17:55 i thought meego has a 256MB minimum Jun 15 21:18:03 or is that for x86 Jun 15 21:18:05 for ram that is Jun 15 21:18:32 hello :) Jun 15 21:18:33 hows everyone today? Jun 15 21:18:54 ggoebel: it is linux based, but if those parts that are closed source don't receive updates by its owners those parts are stuck. Jun 15 21:18:58 "a few" is, like, 3. Jun 15 21:19:23 javispedro, more than a couple Jun 15 21:19:29 andre__: Jun 15 21:19:32 unless you swing that way Jun 15 21:19:54 MohammadAG, DocScrutinizer , how are you guys doing today? Jun 15 21:20:40 just sent my n900 back to the shop for them to deal with the problem Jun 15 21:21:50 ggoebel: linux has shorter support than windows Jun 15 21:21:53 lcuk: yeah, i *hope* ABL is still functioning properly and availble from extras Jun 15 21:21:59 ggoebel: lot shorter Jun 15 21:22:17 ggoebel: unless you can do it yourself Jun 15 21:23:14 yeah, usually when a new version of a linux OS comes out the old one is quickly abandoned Jun 15 21:23:24 windows at least has a long term support cycle Jun 15 21:23:50 Windows also has a much longer release cycle. Jun 15 21:24:20 nice... found http://www.gronmayer.com/it/ Jun 15 21:24:23 windows also has fairly tight-fisted control over what happens at the OS level :P Jun 15 21:24:36 They've already dropped support for the version two versions ago, it's just that was released in 2001. Jun 15 21:24:50 jacekowski: depends on whether you're talking about driver support or programming apis Jun 15 21:25:09 derf, windows xp was a special exception Jun 15 21:25:18 ggoebel: both Jun 15 21:25:42 in microsoft's defense they created a new OS, canned it, and then made another while XP was out Jun 15 21:26:11 jacekowski: beg to differ... but suppose we beg off the usual back and forth and agree to disagree? Jun 15 21:26:18 talking about longhorn -> vista Jun 15 21:26:20 ggoebel: also http://ageofikon.com/packrat/ Jun 15 21:26:30 rm_work: thanks Jun 15 21:27:13 ggoebel: get any linux app that was made 10 years ago, and try to run it now Jun 15 21:27:16 rm_work, :) https://github.com/lcuk/liqbase-playground/blob/master/sub_published/liqcontrolpanel/liqbrightness.c#L48 Jun 15 21:27:22 still needs wiring up correctly! Jun 15 21:27:25 heh Jun 15 21:27:28 yeah Jun 15 21:27:37 it's the wiring up that was annoying as @#&$ originally Jun 15 21:27:43 which is why i had to do it :P Jun 15 21:27:48 otherwise i feel like it would have been done Jun 15 21:28:05 oh, and figuring out how to make a GTK statusbar applet actually LOOK right Jun 15 21:28:07 >_< Jun 15 21:28:27 heh yeah Jun 15 21:28:41 i think it took several weeks to find a single define function Jun 15 21:28:55 yerp Jun 15 21:29:06 and garage.maemo.org's first successfully paid bounty :P Jun 15 21:29:13 * rm_work bought jott a bear Jun 15 21:29:15 *beer! Jun 15 21:29:18 NOT A BEAR Jun 15 21:29:19 OH GOD Jun 15 21:29:22 * rm_work is mauled Jun 15 21:29:30 lcuk: why not link it at compile time? Jun 15 21:30:02 jacekowski, complex Jun 15 21:30:12 just needs a switch() statement really Jun 15 21:30:21 lcuk: ROFL: "// from the bowels of abl" Jun 15 21:30:26 but had other things work related sidetracking Jun 15 21:30:52 rm_work, ;) Jun 15 21:32:21 was gonna say you needed to set the scaling factors properly, but then saw your comment at the top when looking for the defines :P Jun 15 21:32:36 because the system method is like... WTF Jun 15 21:32:44 the formula for that number is retarded Jun 15 21:32:45 T_T Jun 15 21:33:11 have you looked at autodetection for choosing the method yet? Jun 15 21:33:25 the way we did it was based on the hardware revision >_> which is... bleh Jun 15 21:33:28 but works Jun 15 21:34:07 rm_work, not looked in that file for ages Jun 15 21:34:13 i tend to write hacky code to get the job done at first and then go back to fix things once it actually WORKS end to end, doing regression testing as I go Jun 15 21:34:16 I did however find all the sketches relating to its design Jun 15 21:34:23 that's one of those things we never really found a better method for Jun 15 21:34:27 lol yeah Jun 15 21:34:39 i remember, you linked them :P Jun 15 21:34:50 would like to see the end product at some point Jun 15 21:34:56 it sounds like it'd work nicely Jun 15 21:35:27 i'm imagining putting my finger at the top, dragging left/right and also down, then getting to the next level and repeating, and again Jun 15 21:35:42 wonder if it actually works that smoothly in reality Jun 15 21:35:59 though you have to meter the requests because otherwise you spam DSME and it goes whacko Jun 15 21:37:22 http://liqbase.net/liq.20110615_223401.brightness.scr.png Jun 15 21:37:36 it would work today if it was wired up Jun 15 21:37:45 I however dunno which to do, send me a patch Jun 15 21:38:35 rm_work, disco brightness ;) Jun 15 21:38:42 should have noise whilst you wiggle Jun 15 21:38:53 lol yeah Jun 15 21:39:04 lcuk: what do you need a patch for? your file looked good, besides the scaling Jun 15 21:39:15 which is only really necessary for the sys method, i believe Jun 15 21:39:18 which is only on 770 <_< Jun 15 21:39:28 i am currently setup on ideapad Jun 15 21:39:43 ah Jun 15 21:39:44 yeah Jun 15 21:39:48 my n900 is in user mode and scratchbox isnt running Jun 15 21:39:50 i never got HAL support functioning Jun 15 21:39:58 does Meego use HAL? Jun 15 21:40:00 or DSME still Jun 15 21:40:06 if it still uses DSME... well, you have the code Jun 15 21:41:53 ABL was super useful for getting a real job BTW... I got to say things like "have experience with translation of applications using l10n :P Jun 15 21:41:55 " Jun 15 21:42:12 hehe indeed Jun 15 21:43:01 I am heading to London this weekend Jun 15 21:43:29 cool Jun 15 21:43:31 I'm not :( Jun 15 21:43:32 I wish I was Jun 15 21:43:34 heh Jun 15 21:43:35 that'd be awesome Jun 15 21:43:40 going to this: http://www.guardian.co.uk/media/blog/2011/jun/01/hactivate Jun 15 21:43:40 also a long and expensive flight <_< Jun 15 21:44:06 you would fall asleep in the mcdonalds :P Jun 15 21:44:06 LOL Jun 15 21:44:10 partnered with Nokia :P Jun 15 21:44:12 lcuk: lol Jun 15 21:44:19 at least I didn't VOMIT in the mcdonalds Jun 15 21:44:28 yes, true Jun 15 21:44:32 i'm holding on to that :P it's all I've got Jun 15 21:44:43 :D Jun 15 21:45:01 the collective will of every McDonalds employee is also pleased Jun 15 21:45:17 rm_work, Nokia are supporting part of it sure Jun 15 21:45:20 but they say: Jun 15 21:46:11 i just think thats funny Jun 15 21:46:23 i was really impressed with them before, for supporting the community the way they did Jun 15 21:46:33 but now with Elop and all the crap thats going on, it's just "wtf" Jun 15 21:46:43 like, are they supporting us, or dropkicking us into the nether? Jun 15 21:46:49 I can't tell <_< Jun 15 21:47:42 rm_work, meh Jun 15 21:48:17 lcuk: yes, "meh" describes how I feel right now about that Jun 15 21:48:32 but i guess for now i'll say "it's still cool that they're supporting the community like that" Jun 15 21:48:49 not sure where the balance is coming from Jun 15 21:49:12 but the first revisions of that and the paperwork I have talks about being platform agnostic Jun 15 21:49:18 but does not specify a toolkit Jun 15 21:49:35 so I can stand up and talk about liqbase bits Jun 15 21:49:41 in context of liqbase Jun 15 21:49:42 :) Jun 15 21:52:00 something just came to mind Jun 15 21:52:02 * lcuk joins the symbian channel Jun 15 21:52:07 lol Jun 15 21:52:10 user0-: OH NO Jun 15 21:52:20 quickly, forget everything! Jun 15 21:52:35 * rm_work is in a jovial mood today Jun 15 21:52:43 hmm they've done it already Jun 15 21:52:59 "they"? Jun 15 21:53:26 https://elektranox.org/n900/screenshots.html Jun 15 21:54:12 its just not everything works Jun 15 21:56:52 what? that's obviously SHR Jun 15 21:56:59 of course we did it Jun 15 21:57:12 like 9 months ago Jun 15 21:57:46 DocScrutinizer : i sent the damn phone back to the shop Jun 15 21:58:27 http://www.youtube.com/watch?v=ywkWbb_BetI&feature=related Jun 15 21:58:43 user0-: probably best yu can do Jun 15 21:59:52 DocScrutinizer : i could have installed meego on an sd card and got the code Jun 15 22:00:20 ah, bought device with lock code? :-) Jun 15 22:00:21 but this shouldn't happen so let them deal with the problem Jun 15 22:00:29 sure you could, but that doesn't explain why it is acting up like it did Jun 15 22:00:36 exactly Jun 15 22:00:59 ShadowJK : it started asking for one after flashing it Jun 15 22:01:06 ShadowJK: yep, plus it wasn't resettable as it's supposed to be, according to MohammadAG Jun 15 22:02:25 mmm supposed to be trivial to crack, but it takes effort i guess Jun 15 22:02:43 DocScrutinizer : i was expressing how useless the rescue initrd was moments ago at #meego :P Jun 15 22:02:49 not trivial when system refuses to boot Jun 15 22:03:00 hehe Jun 15 22:03:10 I guess reply was ..... Jun 15 22:03:37 that i was right and that the wiki page is missing a lot :P Jun 15 22:04:09 they were asking each other what was the point of usb networking Jun 15 22:04:18 ShadowJK: if you got the opportunity eventually, you'd not want to miss to test enabling lockcode prior to rootfs reflash, to see if it gets reset Jun 15 22:04:44 lol @ user0- Jun 15 22:05:06 Uh, i'm pretty sure it doesn't get reset? Jun 15 22:05:12 the info online regarding that was mixed Jun 15 22:06:39 i ended up translating sites from Arabic and Chinese to know the answer yesterday Jun 15 22:07:10 isn't lock code stored with CAL and stuff.. Jun 15 22:07:23 didn't even get it, worked for some people, didn't for others Jun 15 22:09:07 ShadowJK: MohammadAG claims the fact it shall get queried is in gconf, and gconf gets reset on flashing rootfs Jun 15 22:09:35 a big hoax, but comes handy when overriding a forgotten lockcode Jun 15 22:10:13 was it MohammadAG or someone else who stared at disassembly of NOLO and stuff.. Jun 15 22:10:15 gconf should be stored on ~ imo Jun 15 22:10:22 alas I never heard anybody actually *testing* it, that's why I asked if you... eventually... (I'm rarely ever flashing rootfs) Jun 15 22:10:27 what does NOLO have to do with this? Jun 15 22:10:37 and no, it was probably jacekowski Jun 15 22:10:49 ShadowJK: jacekowski Jun 15 22:11:40 iirc he found the lock code in CAL, and figured out the default code is stored in plain text, and when setting code it gets stored as hash(code+salt), where salt is constant Jun 15 22:11:52 but my memory isn't that reliable :P Jun 15 22:11:59 MohammadAG: (should be stored in ~) maybe it is, partially, and that's why you need to nuke both, rootfs AND eMMC Jun 15 22:12:09 would make a lot of sense Jun 15 22:12:23 no need to nuke eMMC to clear gconf actually :p Jun 15 22:12:26 just rootfs Jun 15 22:12:36 and setting a gconf value as root or user does the same Jun 15 22:12:40 ShadowJK: (figured default) that's been me :-) Jun 15 22:12:57 ah ok Jun 15 22:13:16 MohammadAG: please TRST it Jun 15 22:13:19 :-D Jun 15 22:13:31 test what? Jun 15 22:13:42 it's 12345 on the other N900 here Jun 15 22:13:42 we all know how to *simulate* it with gconftool Jun 15 22:13:46 plain text Jun 15 22:13:53 I don't want to change that :P Jun 15 22:13:56 yes, that means you never changed it Jun 15 22:14:09 yeah, not my personal N900 so no reason to do so Jun 15 22:14:29 DocScrutinizer : you remember the flashers output that was something like ( x kb of y kb ) yesterday ? Jun 15 22:15:03 you made a geekish joke back then with something regarding the speed of light.. Jun 15 22:15:05 (simulate) and I believe in your tests with gconftool regarding the lockcode disabling. Nevertheless there *might* be more to it Jun 15 22:15:15 y was 0 on windows Jun 15 22:15:16 user0-: yes Jun 15 22:15:33 on linux x=y Jun 15 22:15:51 transmitting arbitrary amount of data in 0 seconds is speed-of-light for flashing Jun 15 22:15:53 how it should be Jun 15 22:15:59 I wonder if the gconf key is written to on boot Jun 15 22:16:28 ShadowJK, DocScrutinizer qwerty12 lost his key and had to recover, he found out flashing recovers access to the device Jun 15 22:16:44 well, if *anybody* dares to enable lockcode prior to next flash... Jun 15 22:16:50 we'd know for sure Jun 15 22:16:51 :) Jun 15 22:17:14 i'll be flashing it if it comes back with an older firmware tomorrow Jun 15 22:17:46 no problems as long as you know your own lockcode Jun 15 22:17:54 so i'll do whatever testing you have in mind Jun 15 22:18:11 thats for sure :) Jun 15 22:18:18 first post here http://talk.maemo.org/showpost.php?p=434946&postcount=1 Jun 15 22:18:43 MohammadAG: ok, I'll look into it, but it might have changed, no? Jun 15 22:19:10 I can tell it's an old post, just from the URL ;-P Jun 15 22:19:20 nah Jun 15 22:19:27 12-19-2009 Jun 15 22:20:23 perhaps i'll get SHR running on an sd card before flashing tomorrow Jun 15 22:25:04 MohammadAG: also he doesn't state WHAT he had to flash, maybe >>let me remind you that I had to reflash the phone and blasted all of the files I had on here along with all my settings and applications<< implies he flashed eMMC too? Jun 15 22:25:39 DocScrutinizer, extra research shows http://talk.maemo.org/showpost.php?p=538825&postcount=2 Jun 15 22:26:35 :nod: Jun 15 22:27:01 still I'm worried nobody succeded in actually doing it recently Jun 15 22:27:45 MohammadAG : i read his/her posts yesterday Jun 15 22:28:07 I'd feel a WHOLE LOT better when talking users like user0- thru all of it, when I got some recent confirmation of the process working with PR1.3 Jun 15 22:28:45 his, saying her in the maemo community wouldn't be sexism imo :p Jun 15 22:28:49 err, him* Jun 15 22:28:52 it's just so simple to change the cruft that checks lockcode on boot Jun 15 22:29:16 so nobody knows if Nokia fixed that vulnerability in PR1.3 or not Jun 15 22:31:12 if I were to implement that, I'd place a cookie to ~user/ as well, and only ignore lockcode if both rootfs and ~user got nuked Jun 15 22:32:02 so all your data is safe, and you recover by lockcode loss via FULL reflash (root&eMMC) Jun 15 22:32:12 clean solution - and so easy Jun 15 22:32:32 DocScrutinizer : have you been getting many users with the problem? Jun 15 22:32:44 dunno Jun 15 22:33:09 this* Jun 15 22:36:36 DocScrutinizer : if i find the time tomorrow i'll be flashing it a few times to see what was causing it Jun 15 22:36:50 cool, thanks Jun 15 22:37:05 don't forget your lockcode ;-) Jun 15 22:37:15 :D Jun 15 22:37:21 just in case Jun 15 22:46:11 btw, *which* process actually deals with lockcode? Jun 15 22:46:23 mce? Jun 15 22:46:28 wait for it, wait for it, mce Jun 15 22:46:37 hehe Jun 15 22:46:40 lol Jun 15 22:46:40 stop mce; sleep 1; start mce prompts for the code Jun 15 22:46:50 gonna strace the sucker ;-) Jun 15 22:47:19 see where it looks for secret bits about asking lockcode or not Jun 15 22:47:30 gconf :p Jun 15 22:48:00 just string the library with the UI in control panel Jun 15 22:48:01 why not ~user/.gnome/letslockit as well? Jun 15 22:48:08 it's libcpdevlock I think Jun 15 22:48:13 it's in /usr/lib/hildon-control-panel Jun 15 22:48:34 cause Nokia didn't think an average crook can flash rootfs Jun 15 22:48:42 I'll jus see if it touches eMMC when *enabling* lockcode Jun 15 22:49:37 nope, sure it doesn't :p Jun 15 22:49:44 as it would need to flip both toggles if there's one in rootfs *and* one somewhere hiding on eMMC Jun 15 22:50:01 MohammadAG: how do you know? Jun 15 22:51:43 given I had source of mce, I'd "fix" this vulnerability in ~5min Jun 15 22:52:21 refer what they did with cherry ;-) Jun 15 22:53:49 I dug around a lot when looking for something for chem|st Jun 15 22:53:52 not sure what it was Jun 15 22:56:38 -- if get-gconf(GCONFKEY_USE_LOCK) Jun 15 22:56:40 ++ if get-gconf(GCONFKEY_USE_LOCK) or fstat(PATH_TO_LOCKCOOKIE_IN_HOME, void) Jun 15 22:57:26 can go to mce any time, for any past or future mce update Nokia ships Jun 15 22:59:02 unless somebody compares md5 for mce in PR1.0 with what we got today and oesn't find a diff, we don't know for sure Jun 15 23:00:04 or somebody *demonstrates* flashing rootfs is sufficient for PR1.3 Jun 15 23:01:17 arguably the option to boot meego or rescue-initrd and access eMMC data this way somewhat completely defeats the purpose of a lockcode Jun 15 23:02:27 but you never know if they didn't think they at least should fix the unprotected MyDocs from simple rootfs flashing which is what every noob would try first to unlock the device Jun 15 23:04:34 Nokia could even fix NOLO and completely protect the device this way Jun 15 23:10:27 NOLO: if I'M_FLASHING_ROOTFS and check-cal-key(MCE-HAS-SET-LOCKCODE, "yes") then {set-cal-key(LOCK-CODE, "invalid); set-calkey(MCE-HAS-SET-LOCKCODE, "reflash-emmc-too-to-unlock")}; Jun 15 23:11:06 nolo means ackward in Finnish Jun 15 23:11:16 yo Jun 15 23:11:28 and FIASCO means... errhm Jun 15 23:11:46 I guess the developers of Nolo also knew it ;) Jun 15 23:15:02 btw my approach as sketched above still has vulnerabilities, but definitely it would spoil the day for joe random-lifter interested in your data Jun 15 23:21:11 joe random-lifter will say "hrm, this looks like a cellphone. sell to pawn-shop." Jun 15 23:21:22 i think you may be overestimating the majority of the criminal element Jun 15 23:21:57 "but you never know if they didn't think they at least should fix the unprotected MyDocs from simple rootfs flashing which is what every noob would try first to unlock the device" Jun 15 23:21:58 rofl Jun 15 23:22:05 every noob wouldn't even know what maemo is Jun 15 23:30:40 this doesn't really matter here, does it? Jun 15 23:31:19 just saying Jun 15 23:32:33 you make it sound like any random person on the street might pick up my cell phone and immediately think "AH, I know, I'll just flash the rootfs and that way i'll have easy access to his data" >_> Jun 15 23:33:05 90% of people on the street will be like "durrr, cellphone" and the rest will be like "cool, smartphone... wait, this isn't android. LAME." Jun 15 23:33:30 as long as the average service shop finds the MyDocs data gets exposed despite lock code, on a regular flashing of "firmware", this might just be sth Nokia feels the should change Jun 15 23:33:34 your point is still valid though if you're legitimately concerned with security Jun 15 23:34:20 hopefully you understand my point tho :) Jun 15 23:34:25 yes Jun 15 23:34:43 I simply didn't bother to elaborate in it Jun 15 23:35:23 i'm just poking fun mostly, as it seems you're the pedantic person in this channel most often :P Jun 15 23:35:35 as it's about what Nokia might wnat to do or have done already, not about actually implementing sth community driven to improve security Jun 15 23:38:04 I have lost my N900 -actually stolen- and after using another phone for a month I am dying to get an N900 again, so I am checking ebay for used ones. what is a good price for a used N900 in good-condition? Jun 15 23:39:16 smhar: I paid $300CAD recently Jun 15 23:39:33 and watched a brand new phone sold for the same price Jun 15 23:41:06 on a random shop I've seen two used ones sold for 240EUR Jun 15 23:41:17 I think it's too much though Jun 15 23:41:50 s/sold/on sale/. Jun 15 23:42:34 tried local shops here, no new and no used. I felt like I am the only one who is -was- using it :-) Jun 15 23:43:49 smhar: where are you? Jun 15 23:44:07 I've been amazed when I've seen that pair of N900. Been the first time I've seen one in a shop Jun 15 23:44:12 SpeedEvil, Bahrain Jun 15 23:45:03 I found I'm getting ads from nokia as baner ads, advertising the n900 as still for sale on contract. UK only. And I don't know if it's real. Jun 15 23:46:01 they're still on sale new in canada Jun 15 23:46:55 SpeedEvil, Nokia and their local agent here never ran an ad for N900. it was a very silent comer to the market Jun 15 23:47:14 stores aren't stocking them though, as far as I have seen - it's special order only Jun 15 23:48:20 price is rediculous though Jun 15 23:51:13 I think Nokia discontinued it Jun 15 23:58:14 http://share.ovi.com/media/joerg900.public/joerg900.10130 Jun 15 23:58:22 (2 N900) Jun 16 00:36:59 I wonder how long share.ovi.com will persist - wasn't it discontinued? Jun 16 01:36:14 hi when i play certain videos on n900 after i play it once it says file not supported? Jun 16 01:46:58 are you overclocking? Jun 16 01:48:06 or undervolting Jun 16 02:13:21 iluminator101: there's a known issue with some video encoded in some particular format makes the GFX hw accel hang. AFAIK you have to reboot after that to recover. Shouldn't say "file not supported" for just one file then, but doesn't playback any video Jun 16 02:14:02 the effect is reproducable with particular "offending" video files Jun 16 02:14:48 probably a bug in the hw accel decoder code Jun 16 02:16:51 actually OC and messing with core voltages can also cause a lockup, and in case of OC this can be permanent flaw that causes frequent lockups even when clock is reverted to normal Jun 16 02:19:31 Corrupt files can also wedge the decoder in a state that needs a reboot Jun 16 02:31:43 I can reproduce this undervolting. The "gpu" really doesn't like it. Jun 16 02:32:42 sure, otherwise Nokia had chosen lower voltage to make for longer battery endurance Jun 16 02:33:26 Some people are just lucky. But then again they probably have dodgy usb ports Jun 16 02:33:57 and if OC was without risk and damage, Nokia for sure would not run the device underclocked to give nerds sth they can "add 2 eggs and a glass of milk to bake the cake" Jun 16 02:35:40 DocScrutinizer, exactly have to reboot Jun 16 02:35:41 warranty is max 2 years afaik. So expect device to break in less than that when you overclock. Of course that's worst case scenario Jun 16 02:35:45 Then mine runs quite ok slightly overclocked and undervolted. Just enough to get better battery and play some videos it wouldn't otherwise. Jun 16 02:35:46 DocScrutinizer, how can i fix this Jun 16 02:36:00 don't play back the killing file Jun 16 02:36:23 iluminator101: so are you undervolting or not? Jun 16 02:36:51 GonzoTheGreat, no stock didnt make any changes Jun 16 02:37:25 did you pause the video? Jun 16 02:37:28 find which file kills the gfx engine. delete the file Jun 16 02:37:58 get same movie in another encoding or rate or whatever Jun 16 02:38:01 GonzoTheGreat, yea sometimes Jun 16 02:38:21 Seem to remember it did that for me. It couldn't resume from those points. Jun 16 02:38:50 I can't remember where those "bookmarks" are stored. Jun 16 02:39:34 Try moving it somewhere and rename it and moving it back. It think it is handled by tracker. Jun 16 02:39:58 * GonzoTheGreat waits for DocScrutinizer to get warmed up Jun 16 02:40:28 err wut? Jun 16 02:40:42 I mentioned the "t" word Jun 16 02:41:40 it's some particular encoding sequence occasionally occuring in random arbitrary video files. If you run into such a file that has the killer sequence, you can do really little about it Jun 16 02:41:51 meh Jun 16 02:42:11 the t word gets eaten for breakfast Jun 16 02:42:19 You can add another media player - mplayer for example - and watch it using that. Jun 16 02:42:24 and actually is unrelated here Jun 16 02:42:36 SpeedEvil: indeed Jun 16 02:42:42 good point Jun 16 02:47:46 DocScrutinizer: could it be that seeking to a "bookmarked point" in certain files could trigger this but not playing "across" this point? Jun 16 02:48:22 might, but unlikely Jun 16 02:49:16 well, maybe there are special "hints" for searching in such video files, that aren't used for normal playback Jun 16 02:49:32 so if such a hint is flawed... X-[ Jun 16 02:50:16 hints like a linked list of forward pointers to next second, minute, etc Jun 16 02:52:09 you'd not need that for normal playback, but for searching to a particular point it's useful, and can break things if such pointer is defect Jun 16 02:56:00 usually the distance between 2 i-frames (full picture) is limited, i.e. there must not be more than x differential frames between 2 i-frames. Sometimes encoders violate that rule. Results may vary depending on decoder **** ENDING LOGGING AT Thu Jun 16 02:59:57 2011