IRC log for #openmoko on 20100106

13:49.05*** join/#openmoko infobot (i=ibot@rikers.org)
13:49.05*** topic/#openmoko is Users of Openmoko Mobile Phone Hardware | WikiReader users, please join #openmoko-wikireader | Developers, please join #openmoko-cdevel | New Users read http://wiki.openmoko.org + http://www.catb.org/~esr/faqs/smart-questions.html | Make sure to check http://wiki.openmoko.org/wiki/Om_2009 | SHR images: http://build.shr-project.org/shr-testing/images/om-gta02/ | hackable:1 rev5 is out ! http://blog.hackable1.org/2009/12/hackable1-rev5-is-out.
13:49.46PaulFertserI had a nightdream today that for some unknown reason i had two extra SCSI disks in my desktop with noisy coolers in front of both. O_o
13:50.11DocScrutinizer51~botsnack
13:50.11infobotDocScrutinizer51: thanks
13:50.12PaulFertser(and of course i actually have no scsi controller on desktop)
13:50.25*** mode/#openmoko [-o DocScrutinizer51] by ChanServ
13:50.27gena2xnasty dreams
13:50.36gena2x~botsnack
13:50.36infobotthanks, gena2x
13:50.49DocScrutinizer51-logs
13:50.57DocScrutinizer51~logs
13:50.58infobotAll conversations are logged to http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged.
13:51.42PaulFertsergena2x: well, at least i wasn't chased by no evil guys like in my friend's dreams :)
14:01.57*** join/#openmoko valos (n=valos@coquelicot-a.easter-eggs.com)
14:03.34*** join/#openmoko emacsian (n=ram@122.167.246.129)
14:07.26*** join/#openmoko Tupeco (i=tuomom@xob.kapsi.fi)
14:08.23*** join/#openmoko Gabrys (n=Gabrys@158.75.12.65)
14:10.42*** join/#openmoko Ghiottone (n=ale@87.13.245.31)
14:24.34gena2xPaulFertser: Cool. without debugging almost 2x in important areas like context switch, tcp connection, pipe, 10x in signal handling
14:29.47gena2xPaulFertser: Paul ;)
14:31.33*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
14:32.12lindi-gena2x: hmm, what are you talking about?
14:32.59gena2xlindi-: I am bencmarking kernel
14:33.20lindi-gena2x: yes but what are you changing?
14:33.27gena2xlindi-: with different kernel options disabled
14:33.51lindi-gena2x: ok, what would affect context switches?
14:33.58gena2xlindi-: now it disabled whole 'debug' under 'kernel nacking'
14:34.40gena2xdebian-gt Linux 2.6.29d 191.5 378.2 626. 846.3       1126.       3222
14:34.41gena2xdefconfig Linux 2.6.29w 312.5 751.9 1209 1477.       1920.       5823
14:34.41gena2xdefconfig Linux 2.6.29w 313.8 1069. 1414 1496.       1921.       5839
14:35.06gena2xnow, better wait, I'll push all results to web
14:35.23lindi-gena2x: if you have test scripts it would be interesting to run those
14:35.38gena2xlindi-: i am using lmbench
14:35.55gena2xlindi-: just good microbencmark
14:36.42lindi-ok
14:40.22DocScrutinizer51aaah. tweaking the shit out of scheduler once more? ;-P
14:40.48DocScrutinizer51BSFH
14:41.06gena2x2x is tweak?
14:41.30DocScrutinizer51aka cooperative multitasking, or windows 3,x
14:42.08gena2xi am not even started with FSCE
14:42.14gena2xOr how it is?
14:42.32gena2xthis is just plain disable of debug :)
14:42.44DocScrutinizer51double speed of task switching? nice! just if the taskswitching is only 0.1% of cpu total load... so what?
14:43.49gena2xnot only task switching... for example pipe have 15% more bandwidth
14:43.57gena2xand 2x better latency
14:44.17DocScrutinizer51scheduler tweaks to speed up desktop experience vanished with tans100MHz cpus for a good reason
14:44.34DocScrutinizer51so now you're speaking
14:44.40gena2xDoc, I am not tweaking :)
14:44.41DocScrutinizer51ok I'll hush
14:45.25DocScrutinizer51pip +15 is a decent number
14:45.38gena2xopen/close timing - 10x
14:45.45DocScrutinizer51woooow
14:46.24DocScrutinizer51looking forward to your collected results anxiously
14:46.57gena2xloopback communication bu tcp 19.5 from 16 (in MB/s)
14:48.35gena2xthis is not tweaking... this is essestinal things..
14:48.41*** join/#openmoko alexxy (n=alexxy@gentoo/developer/alexxy)
14:54.36DocScrutinizer51ack
14:55.25*** join/#openmoko retracile_ (n=eli@adsl-99-188-14-94.dsl.rcsntx.sbcglobal.net)
15:00.12*** join/#openmoko mnt_real (n=sinan@bas12-montrealak-1167974851.dsl.bell.ca)
15:06.41*** join/#openmoko gena2x (n=gena2x@bsdmn.com)
15:09.34*** join/#openmoko retracile__ (n=eli@76-204-213-211.lightspeed.allntx.sbcglobal.net)
15:11.58*** join/#openmoko Gabrys (n=Gabrys@158.75.12.65)
15:15.54*** join/#openmoko naeg (n=naeg@194.208.239.170)
15:16.42*** join/#openmoko rhaeder (n=rhaeder@188.109.251.228)
15:20.54*** join/#openmoko ssweeny (n=ssweeny@li124-16.members.linode.com)
15:34.43gena2xfor example, simple top in system without debugging eats 1.6% cpu, with debugging - 3.2% cpu average
15:36.35rhaederXserver bootup takes already 6 minutes here :/
15:37.01rhaederand that auto-suspend is back after updating packages :(
15:38.16*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
15:38.23rhaeder"/etc/init.d/xserver-nodm restart" did the trick here
15:38.29rhaederstill auto-suspending
15:46.16*** join/#openmoko raster (n=raster@enlightenment/developer/raster)
15:49.17*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
15:55.14*** join/#openmoko Sup3rkiddo (n=sudharsh@unaffiliated/sudharsh)
15:59.46elopiohello
15:59.59elopiowhat calendar application do you use on shr?
16:01.18*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
16:07.23*** join/#openmoko Openfree|H (n=Openfree@222.65.62.165)
16:12.16PaulFertsergena2x: without debugging we won't get meaningful backtraces at all?
16:12.39gena2xPaulFertser: I didn't check yet.
16:12.55PaulFertserfell asleep but now's awake again :)
16:13.00*** join/#openmoko pH5 (n=ph5@e178218022.adsl.alicedsl.de)
16:13.04gena2xPaulFertser: Morning :)
16:13.21*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
16:13.24rhaederthe "Enter PIN" dialog is no longer showing up :(
16:14.58gena2xPaulFertser: But i guess performance impove worth it even if we compleely loose backtrace.
16:15.33PaulFertsergena2x: well, i'd prefer to have both ;)
16:16.03PaulFertsergena2x: i'll push mtd-related defconfig amendment shortly, as soon as i feel fully woken up.
16:16.27gena2xPaulFertser: Climb down from Olimp of kernel development and thing about poor users
16:17.27gena2xPaulFertser: They suffer :)
16:17.31PaulFertsergena2x: microtests are good but will overall performance differ measureably? That's the question we need to answer.
16:17.42gena2xPaulFertser: Yes.
16:17.49PaulFertsergena2x: i suffer too. I still use debian with full-python stack and slow old shr apps.
16:17.59gena2xPaulFertser: You can feel it.
16:18.22PaulFertsergena2x: as my only phone. I suffer every time i want to write or read a message.
16:19.03gena2xPaulFertser: i told, that top with debug - 3,2%, without - 1,6% of cpu. how do you think can this be noticed?
16:19.13PaulFertsergena2x: i think disabling rcu debugging and things like that is beneficial.
16:19.30PaulFertsergena2x: i just would very much want to have meaningful backtraces included.
16:19.55gena2xPaulFertser: now to provoke backtrace?
16:20.30gena2xPaulFertser: In running kernel?
16:20.42gena2xPaulFertser: I think function names are kept.
16:21.21gena2xPaulFertser: But have to check it.
16:21.35gena2xPaulFertser: lmbench is so sloooowwww.
16:21.57gena2xPaulFertser: it makes report for several minutes now.
16:22.28gena2xPaulFertser: i also tried to disable PREEPTION
16:22.36gena2xPaulFertser: PREEMPTION
16:22.57gena2xPaulFertser: this has significal (but less) impact also
16:23.12gena2xPaulFertser: and didn't try alltogether
16:23.34gena2xPaulFertser: What do you think about preemption?
16:23.36rhaederok, fixed it by myself: phonefsod (or so?) needs to be restarted ("Enter SIM code" was not popping up)
16:24.08rhaedergena2x: may help :) and other timer
16:24.36*** join/#openmoko Guest36361 (n=leon@fco.xs4all.nl)
16:25.25rhaederI mean IO scheduler :)
16:25.52rhaederI compiled in Deadline and Anticipatory/CFQ are modules
16:25.53gena2xIO sheduler changing is real tweak.
16:26.05rhaederit works :)
16:27.26rhaederwhen I use a default Debian kernel (host system, not the FR), it comes with Anticipatory, I have more sound hicks (framedrops) than with Deadline
16:27.34gena2xI'm not surprised, but to talk to include it to defconfig, need to prove that. this can be done doing some reliable testing.
16:28.20rhaederrun "heavy IO apps" like Quake4 (demo) or so in full details (max texture quality)
16:29.19gena2xYou need to prove that on FR :)
16:29.30gena2xif we are speaking about FR
16:29.31rhaederFR is no server ;)
16:30.39gena2xFR is just another Linux computer
16:30.47*** join/#openmoko everslick (n=clemens@91-113-62-59.adsl.highway.telekom.at)
16:30.59nullixif you manage to get Quake4 working on FR i would promote you as the genius developer of the deceny...
16:31.51rhaedernullix: won't work. Quake4 is non-free and X86
16:31.55rhaeder:(
16:32.20nullixyes i know, it is sad.
16:32.21gena2xrhaeder: np, we have qemu
16:32.29*** join/#openmoko holycow (n=new@69.67.174.130)
16:33.16rhaederdo you run the FR seriously as a web-server??? it is a "desktop system" and no server system. so choose the right pill ;)
16:33.20rhaederblue or red? ;)
16:33.33nullixjust having 3D on FR would be a very good surprise
16:34.18gena2xnullix: Use mesa :)
16:35.15nullixwell, with 1FPS ?
16:35.36gena2xlindi-: here?
16:36.35rhaeder9 frames in 5.2 seconds =  1.733 FPS <-- glxgears :/
16:36.51rhaedermore a slide-show than performance ;)
16:37.26rhaederOpenGL renderer string: Software Rasterizer <-- that is the reason :)
16:37.28gena2xrhaeder: make you window smaller :)
16:37.50nullixwith 8x8 pixel it might be better...
16:37.56rhaederLOL
16:37.58gena2xnullix: sure
16:38.23gena2xnullix: mb even quake will have decent 10^-3 fps.
16:42.34*** join/#openmoko emacsian (n=ram@122.167.97.135)
16:42.47rhaederunplayable
16:42.54rhaederunless you play in god mode
16:43.34*** join/#openmoko dos1 (n=dos@unaffiliated/dos1)
16:59.37PaulFertsergena2x: preemption is very important for percieved responsiveness (and latency).
17:00.26gena2xgena2x: for desktop I think
17:00.35gena2xPaulFertser: for desktop I think
17:00.51gena2xPaulFertser: Why do you think we need it on FR?
17:00.53rhaederhttp://www.opkg.org/package_202.html#write_comment <-- (Quix0r, last comment) happens with many packages from opkg.org :(
17:01.13PaulFertsergena2x: (provoke backtrace) there's some file in proc for magic sysrq key
17:01.42PaulFertsergena2x: do we need responsiveness or raw computing power? I think it's the first for sure.
17:02.06gena2xPaulFertser: I think second.
17:02.37gena2xPaulFertser: we dont need responsiveness at any cost
17:02.44rhaederusers will disagree with SHR/FR when latency is slow and even more when they find out, you *COULD* but did not enable latency improvements
17:03.07rhaederno blaming or offensives here :)
17:03.12gena2xPaulFertser: In fact latency in descreased that preemption is off.
17:03.23gena2xs/that/than/
17:03.37rhaederlike I wrote above: what do we need? a server OS or desktop OS? I vote for 2nd
17:03.38PaulFertsergena2x: i doubt that
17:03.41rhaederfor sure
17:03.52gena2xPaulFertser: see test results
17:04.14gena2xrhaeder: we need fast responsive system, for sure
17:04.21rhaederyes :)
17:05.00rhaederon a server you don't need fast responses, better high IO throughput :)
17:05.27gena2xPaulFertser: preemption should complicate all 'preetible' kernel functions, this compication can be neglibible on fast systems and very visible on slow.
17:06.37*** join/#openmoko radekp (n=radek@88.146.48.30)
17:06.48gena2xPaulFertser: in fact, we have nothing with low latency. recall our ui.
17:07.14gena2xPaulFertser: everyone will be very happy if that window in shr will appear in 3 seconds.
17:07.17*** join/#openmoko peter-b (n=peter_b@mnch-5d85c78e.pool.mediaWays.net)
17:07.38*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
17:07.39gena2xPaulFertser: we have now low-latency user iteraction.
17:07.43gena2xs/now/no/
17:07.53lindi-gena2x: who is we?
17:08.16gena2xlindi-: hi, please that do you think?
17:08.33lindi-gena2x: not enough data
17:09.14*** part/#openmoko nullix (n=philippe@92.103.174.253)
17:10.22gena2xdata: http://www.bsdmn.com/openmoko/summary1.txt
17:10.27gena2xlindi-: now?
17:11.05lindi-gena2x: s/Mhz/MHz/
17:11.22gena2xto authors of lmbench
17:11.31gena2xthis is autogenerated :)
17:11.37gena2xcomments:
17:11.56gena2xdone several runs of each test
17:12.41gena2xeach test if deconfig + some 1 option set (remove debug, remove preempt, -O2 insteal -Os).
17:13.02gena2xFile tests should ignored
17:13.57gena2xoverall comment - preemption should be disabled, debug must be removed, speed is neligble, but I think we can remove -Os too.
17:14.43gena2xas you can see performance impact is huge.
17:14.48lindi-gena2x: speed?
17:15.08gena2x-O2 vs -Os optimization
17:15.31gena2xread my comments here ;)
17:16.40gena2xPaulFertser: can you take a look too?
17:18.23gena2xany comments?
17:21.02PaulFertsergena2x: synthetic tests are not enough to convince me preemption should be turned off ;)
17:21.21*** join/#openmoko radekp (n=radek@88.146.48.30)
17:22.04gena2xPaulFertser: so, that will be enoght?
17:22.34PaulFertsergena2x: a community consensus
17:23.10gena2xPaulFertser: so, you is neutral :)
17:24.19gena2xPaulFertser: in fact I'll post results now. and tomorrow we'll how about consensus
17:25.49gena2xlindi-: that do you think?
17:26.33lindi-gena2x: not enough time to understand all those right now :)
17:26.54gena2xlindi-: I like you method ;)
17:27.17PaulFertsergena2x: done: http://git.openmoko.org/?p=kernel.git;a=commit;h=a15608f241a40b41fed5bffe511355c2067c4e88
17:27.55lindi-gena2x: i'm nostly worried about stability, fr is fast enough for me
17:28.03gena2xPaulFertser: wow
17:28.33gena2xlindi-: You worry it will be so fast that became unstable?
17:29.06gena2xlindi-: I/we can do some basic checks with major distibutions
17:29.25lindi-gena2x: you never know what happens when you change parameters :)
17:29.51gena2xlindi-: this can't be reason to slow down own device.
17:30.00PaulFertsergena2x: preemption is what gives you context switches when you need them. Somehow i always thought that's beneficial for an interactively used device.
17:30.37gena2xPaulFertser: you also have to modify some functions to be preemtible
17:31.10gena2xPaulFertser: this makes that functions slower
17:31.59gena2xPaulFertser: they should store/restore the state i guess.
17:32.57gena2xPaulFertser: I can investigate it in better details - I'm not a kernel pro dev, just read a bit.
17:34.25gena2xPaulFertser: http://kerneltrap.org/node/2702
17:37.13gena2xPaulFertser: and it is marked as 'EXPERIMENTAL'
17:37.41*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
17:38.44*** join/#openmoko der-lehmann (n=stefan@188-192-133-169-dynip.superkabel.de)
17:39.43*** join/#openmoko _phil (i=crypt@shell.bawue.net)
17:40.52rhaedermaybe something that should be fixed: OGGs as ringtones will result in playing the file as raw data (treated as WAV?)
17:42.11rhaederdon't get me wrong: I prefer muted phones :)
17:43.06rhaederand the om200x.x where able to play oggs :)
17:43.17rhaederI think they used a wrapper app or so
17:43.32rhaeder"play" ?
17:44.14rhaederplaysound from libsdl?
17:44.27PaulFertserrhaeder: rather gstreamer
17:44.45rhaederplaysound works here on my host :)
17:45.04rhaederwhat do I need to alter?
17:46.17rhaederresponse of mplayer is slow, it needs some time to initialize
17:46.32rhaederI think the same applies to gstreamer
17:46.54PaulFertserrhaeder: well, fso used some tricks like pre-creating pipeline etc.
17:47.16rhaederthrough oggdec?
17:47.30PaulFertserrhaeder: i'm not sure
17:48.03rhaederno vorbis-tools package :(
17:48.23rhaederbut that would limit it to OGG (do we like MP3? ;-) )
17:48.50rhaederwith "we" I mean FR community. and could be harmful to the project
17:49.07rhaederbecause of software patents :(((
17:49.49rhaedernow tell that the "customer" or "consumer" that he/she needs to convert all MP3 to OGG before playback is possible on FR...
17:50.27rhaederin realistic view, the most have (still and sadly) a lot MP3, with some exceptions like myself
17:51.02rhaedersuch a feeling far fetched... ;)
17:51.47rhaederin Half-Life 2, a Valgraunt said "such a theory far fetched" to Alyx Vaince (or so)
17:52.54PaulFertserrhaeder: we do not care about software patents, we do not sell anything, especially in the usa :)
17:53.26rhaederso OGG/WAV are the choices?
17:53.45PaulFertserrhaeder: anything gstreamer can play. I used to have an mp3 ringtone.
17:54.17rhaederconvert it? free your music/ring tones? ;)
17:56.10PaulFertserrhaeder: i now use arkanoid.sid again
17:57.23rhaederwhere are the modules/classes located for playback?
17:57.32rhaederI mean ring tone
17:58.35*** join/#openmoko toi (n=toi@d54C2A96D.access.telenet.be)
18:00.07*** join/#openmoko playya (n=playya@unaffiliated/playya)
18:05.39*** join/#openmoko Gabrys (n=Gabrys@77-253-139-158.adsl.inetia.pl)
18:06.39rhaederself.AudioDbusObj.PlaySound(play_tone, 0, duration, reply_handler=dbus_ok, error_handler=dbus_err) <--- that is the line for playing files
18:06.41rhaederhmmm
18:06.52rhaederI'm out of ideas
18:08.12*** join/#openmoko playya_ (n=playya@unaffiliated/playya)
18:09.06*** join/#openmoko pespin (n=Sharwin_@90.163.53.166)
18:19.36*** join/#openmoko redarrow (n=thomas@unaffiliated/redarrow)
18:20.02rafahi, several days ago I was talking about a openmoko SD distribution for games only.. Some brave person with balls to test?
18:21.13tomcatekrafa: what do You provide with?
18:21.38tomcatekor what is the purpose of this disro?
18:22.37groszekhm, btw. anyone tried to run windows mobile on fr? ;-)
18:22.42rafatomcatek: a rootfs for SD and kernel. So you need a SD card with fat and ext2. The distro has a minimal rootfs and several games (some of them known in openmoko, like pingus or mokomaze; and other new games). The idea is onvert the Freerunner open mobile phone in a Linux-based handheld game console.
18:22.55rafaconvert*
18:23.29rafatomcatek: we also re worked some games. For example, current pingus for openmoko was okey just without sound. Now you can play with sound and music and good performance
18:23.33tomcatekrafa: do You know, if hardware acceleration is working there?
18:23.48gena2xok, I posted my measures with comments to list
18:24.08SpeedEvilhardware accelleration is very iffy.
18:24.22rafatomcatek: What do you mean with hardware acceleration? THe graphics issues are well known. We re work every game until we are happy how it looks.
18:24.26SpeedEvilas in the bus the graphics accel is connected to is so slow.
18:24.35SpeedEvilfair enough.
18:24.39rafatomcatek: SpeedEvil is right
18:25.21rafatomcatek: but, whatever.. we modify every game if we need better performance.
18:25.36DocScrutinizer51let me test...
18:25.44DocScrutinizer51~glamo
18:25.45infobotglamo is probably http://www.google.com/search?q=site%3Alists.openmoko.org+glamo+raster
18:25.57DocScrutinizer51yup
18:26.05DocScrutinizer51no more comment
18:26.14tomcatekrafa: is the phone usability still there?
18:28.40rafatomcatek: No. "Gamerunner is an Openmoko Linux distribution. It aims convert the Freerunner open mobile phone in a Linux-based handheld game console."
18:28.43DocScrutinizer51some ghostwriter should make a book of raster's lessons about glamo and gfx-accel in general
18:28.53rafatomcatek: just games
18:30.29rafaDocScrutinizer51: make several books using raster's mails
18:30.53*** join/#openmoko holycow (n=new@69.67.174.130)
18:30.53rafaNo balls enough here,... argh :)
18:33.48*** join/#openmoko CVirus (n=Satan@41.233.140.245)
18:35.09WeissI'm kind of cautious about raster's comments in general though
18:35.37*** join/#openmoko DocScrutinizer-8 (n=jr-N810@openmoko/engineers/joerg)
18:36.03Weissslow bandwidth and not being able to accelerate "everything" is a problem for ALL hardware, it's just that raster's code (E) works particularly badly in that case
18:36.27Weissif we could actually USE the acceleration in the manner it's meant to be used, we'd be flying along.  that's what all the other phones out there are doing..
18:37.36rafaWeiss: you can write books and other could not like your ideas. But yes, E is for other kind of hardware, better hardware.
18:37.52Weissthe way X currently does acceleration is also very bad for (say) GTK
18:38.04WeissI did a post on SHR-Devel about this problem quite recently
18:38.09SpeedEvilFor 99% of games also - 320*240 is what you want to do
18:38.15SpeedEvilOr even less
18:38.23rafaWeiss: I always feel that a great GUI could be made using just SDL for example. But that is my feeling, and I have never tried to start.
18:38.44rafaWeiss: yeah, GTK2 is not for fr.
18:39.04rafaSpeedEvil: that is that we are trying to do ;)
18:39.07*** join/#openmoko borg_ (n=olaf@80.149.17.21)
18:39.40holycowit needs flash
18:39.47holycowwhatever you are doing you need to include flash
18:39.54holycowwaits for the kickban
18:39.55holycow:)
18:40.04rafaholycow: which thing? the game distro?
18:40.07PaulFertserholycow: i'm in the mood for that, wanna try?
18:40.14holycowlol!
18:40.17Weisshehe
18:40.18rafa:)
18:40.48holycowhaha, man that comment totally deserves a kickban tho, just the thought of flash ...
18:40.49holycowheh
18:40.52Weissthe problem which raster (rightly, I guess) emphasises is that to get decent performance we need to write highly platform-specific code, and it's fair not to want to do that
18:41.09holycowas an aside, the cowon s9 media player uses an all flash interface, most retarded idea ever
18:41.39holycowwell the other way to tackle this problem is lower our expecations and wait out the cpu/chipset wars
18:41.42SpeedEvilFlash is on the n900.
18:41.49SpeedEvilFlash 'HD' gets 4fps.
18:41.53holycowi just saw a 2.5" sbc with an intel atom cpu on
18:41.53SpeedEvilon youtube
18:42.03SpeedEvilmplayer gets 60fps.
18:42.50rafaWeiss: well, about platform-specific code : If 10.000 mobile phones are around without decent performance on GUI, then I would support the idea to write platform-specific code.
18:43.06*** join/#openmoko wd (n=wd@82.119.147.190)
18:45.27Weissindeed
18:47.08*** join/#openmoko Batchy (n=batchy@AVelizy-155-1-75-108.w90-2.abo.wanadoo.fr)
18:48.12*** join/#openmoko alabd (n=alabd@unaffiliated/alabd)
18:49.24*** join/#openmoko emacsian (n=ram@122.167.246.129)
18:49.41DocScrutinizer51SpeedEvil: don't mention flash and N900 please >:-(
18:50.10SpeedEvil<PROTECTED>
18:50.15SpeedEvilIt's currently dissassembled.
18:50.19DocScrutinizer51nah' simply don't mention flash at all
18:50.47SpeedEvilI am planning to use it for something silly.
18:50.53Weisshttp://lists.shr-project.org/pipermail/shr-devel/2009-December/001702.html
18:51.00Weiss(r.e. acceleration)
18:51.12SpeedEvilReplacing the controller board ina microwave oven.
18:51.29SpeedEvilIt's a huge stainless thing with lots of features - but the board is mostly broken.
18:51.34DocScrutinizer51SpeedEvil: still pending project? OMG
18:55.39*** join/#openmoko fredrin (n=fredrik|@ti0035a340-dhcp0191.bb.online.no)
18:56.24*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
18:57.03*** join/#openmoko Sargun (n=Sargun@atarack/Staff/Sargun)
18:59.02*** join/#openmoko fsteinel (n=fsteinel@fedora/fsteinel)
19:02.26*** join/#openmoko CVirus (n=Satan@41.233.140.245)
19:03.05*** join/#openmoko Docscrutemp (n=Joerg_rw@openmoko/engineers/joerg)
19:04.02*** join/#openmoko fredrin_ (n=fredrik|@ti0035a340-dhcp0191.bb.online.no)
19:09.21*** join/#openmoko playya (n=playya@unaffiliated/playya)
19:17.59*** join/#openmoko Amorphous (i=jan@unaffiliated/amorphous)
19:19.32alabdit's Time-consuming to listen free rintones one by one   , does anyone know a simple one  ?
19:21.33*** join/#openmoko gargoyle-grin (n=randerso@gentoo/contributor/gargoyle-grin)
19:24.23*** join/#openmoko eolson (n=eolson@maze.mcs.anl.gov)
19:38.52alabdCan sattelite maps be used for showing location maps in gps.. ?
19:39.08*** join/#openmoko ericthefish (n=ivan@94-192-230-157.zone6.bethere.co.uk)
19:49.01*** join/#openmoko tg (i=irc@x-net.hu)
19:51.16lindi-alabd: what?
19:51.48alabdlindi-:  ho thanks for gps problem idea
19:52.09alabdCan sattelite maps be used as gps maps
19:53.17tomcatekalabd: You mean raster maps?
19:53.52alabdyea
19:54.09tomcatektangogps uses them
19:54.22alabdhow about navit ? tomcatek
19:55.33tomcatekalabd: see navit wiki. I'm not sure
19:55.51alabdmethink didn't mentioned there
19:55.58alabdanyway how to use them for tango ?
19:56.05Undrwateralabd: supported maps are mentioned in the wiki
19:56.06alabdwhere to downnload
19:57.07*** join/#openmoko Laibsch (n=Laibsch@p5B3B2A17.dip.t-dialin.net)
20:03.52*** topic/#openmoko by Laibsch -> Users of Openmoko Mobile Phone Hardware | WikiReader users, please join #openmoko-wikireader | Developers, please join #openmoko-cdevel | New Users read http://wiki.openmoko.org + http://www.catb.org/~esr/faqs/smart-questions.html | Make sure to check http://wiki.openmoko.org/wiki/Om_2009 | SHR images: http://build.shr-project.org/shr-testing/images/om-gta02/ | hackable:1 rev5 is out !
20:04.03LaibschURL was 404
20:04.39alabdUndrwater: tomcatek in wiki nothing said about sattelite raster maps http://wiki.openmoko.org/wiki/Navit#Set_up_the_maps_you_want
20:05.13Undrwateralabd: look at navit wiki
20:05.46alabdis not this navit wiki  ? http://wiki.openmoko.org/wiki/Navit#Set_up_the_maps_you_want
20:05.48alabdUndrwater:
20:06.05rhaederhi. which package do I install when I want to build own opkg packages?
20:06.17UndrwaterThat's what I"m seeing in the profile settings window.  Phone does not ring or vibrate.
20:06.18UndrwaterHere's possibly relevant output of phoneuid.log:
20:06.18Undrwater2010.01.05 21:33:38.079593 [phoneuid]MESSAGE: Using log level 'INFO'
20:06.18Undrwater2010.01.05 21:33:38.118154 [libphone-ui]MESSAGE: Loading phoneuid
20:06.18Undrwater2010.01.05 21:33:43.928516 [libphone-ui]MESSAGE: No speaker value for idle found, using none
20:06.18Undrwater2010.01.05 21:33:43.928922 [libphone-ui]MESSAGE: No microphone value for idle found, using none
20:06.20Undrwater2010.01.05 21:33:43.979228 [libphone-ui]MESSAGE: no vibrator configured - turning vibration off
20:06.22rhaederhmm, that can be find out with opkg --file?
20:06.22Undrwater2010.01.05 21:33:45.085501 [libphone-ui-shr]WARNING: No such file or directory
20:06.24UndrwaterI'm up to date on upgrades, and I've tried reinstalling dbus, fso, libphone-ui, and a few others I don't remember without luck.  
20:06.25*** part/#openmoko Laibsch (n=Laibsch@p5B3B2A17.dip.t-dialin.net)
20:06.27Undrwateroops
20:06.30Undrwaterhttp://wiki.navit-project.org/index.php/Main_Page
20:07.05Undrwateralabd: http://wiki.navit-project.org/index.php/Main_Page
20:07.30rhaederopkg search ipkg-build :( no luck
20:08.02alabdUndrwater: thanks
20:08.14Undrwaterrhaeder: maybe more for desktop?
20:08.36rhaedersomewhat. :) a shell script with desktop icon
20:08.44rhaederand vala-terminal output ;)
20:08.55rhaedersorry, I cannot code python :(
20:10.07rhaederopkg-dev?
20:11.11Undrwaternot sure what the -dev packages provide extra (maybe some degugging stuff?), but they're basically the same as the non -dev ones
20:12.14rhaederdebug packages should have -dbg suffix
20:12.27alabdUndrwater: if here http://maps.navit-project.org/download/ shows a country and we press get map will downloaded map conatin each city streets ?
20:13.02Undrwatersome may be missing...but that's a better question to ask on #navit
20:13.33rhaedernope, only header and .pc file :(
20:14.55alabdUndrwater: what does OSM mapnik mean ?
20:15.19Undrwateralabd: no idea
20:21.40*** join/#openmoko retracile (n=eli@76-204-213-211.lightspeed.allntx.sbcglobal.net)
20:24.46alabdtomcatek: are you sure about sattelite maps or you mean those reqular maps ?
20:25.02*** join/#openmoko fredrin|fail (n=fredrik|@ti0035a340-dhcp0191.bb.online.no)
20:25.33*** join/#openmoko redarrow_ (n=thomas@unaffiliated/redarrow)
20:25.45*** join/#openmoko batti (n=batti@dslb-088-078-016-225.pools.arcor-ip.net)
20:28.01*** join/#openmoko retracile_ (n=eli@76-204-213-211.lightspeed.allntx.sbcglobal.net)
20:33.02rhaeder:(
20:42.39*** join/#openmoko CVirus (n=Satan@41.233.140.245)
20:43.24*** join/#openmoko sthuebne_ (n=sthuebne@84.189.199.74)
20:43.45*** join/#openmoko onen|openBmap (n=quassel@vbo91-1-89-87-201-85.dsl.club-internet.fr)
20:43.48*** join/#openmoko mjr (i=mjr@aulis.sange.fi)
20:45.17*** join/#openmoko spydon_ (n=spydon@nl103-146-152.student.uu.se)
20:49.50*** join/#openmoko batti (n=batti@dslb-088-078-016-225.pools.arcor-ip.net)
20:50.42*** join/#openmoko tim_abell (n=tim@78-86-198-4.zone2.bethere.co.uk)
20:51.44*** join/#openmoko tomcatek (n=tomcatek@078088161149.elblag.vectranet.pl)
20:55.16tim_abellcan anyone tell me how what happens between git and the shr-unstable feeds being updated? or point me to some documentation?
20:55.24*** join/#openmoko Elleo (n=Elleo@mikeasoft.com)
20:56.08tim_abelli'm wondering which branch of git://git.openembedded.net/openembedded shr-unstable and the feeds are built from, and how long it takes
21:00.40*** join/#openmoko jas4711 (n=jas@c80-216-24-211.bredband.comhem.se)
21:03.20*** join/#openmoko jas4711 (n=jas@c80-216-24-211.bredband.comhem.se)
21:05.54*** join/#openmoko tjikkun (n=tjikkun@2001:7b8:356:0:204:bff:fe80:8080)
21:17.01*** join/#openmoko jonkristian (n=jonkrist@174.84-48-204.nextgentel.com)
21:23.44*** join/#openmoko Ghiottone (n=ale@host171-143-dynamic.51-82-r.retail.telecomitalia.it)
21:24.41*** join/#openmoko valos (n=valos@mar92-6-82-227-150-198.fbx.proxad.net)
21:28.11*** join/#openmoko dvermd (n=roudoudo@78.234.93.192)
21:29.09*** join/#openmoko bmm (n=bram@195-240-186-6.ip.telfort.nl)
21:30.42*** join/#openmoko objorn (n=objorn@unaffiliated/objorn)
21:35.13*** join/#openmoko demitar (n=demitar@vaxjo8.235.cust.blixtvik.net)
21:39.37*** join/#openmoko Zorkman (n=Man@78-22-123-127.access.telenet.be)
21:49.42*** join/#openmoko tomcatek (n=tomcatek@078088161149.elblag.vectranet.pl)
21:52.04*** join/#openmoko jserv-- (n=jserv@61.57.131.211)
21:53.27*** join/#openmoko KoFish (n=kofish@pr0nography.org)
22:06.20*** join/#openmoko dns53 (n=dns@202.124.88.226)
22:14.54*** join/#openmoko aloril (n=aloril@84.249.126.153)
22:47.05Undrwaterhow do i find out what package installed a particular file?  whatprovides isn't working for me
22:48.36Undrwaternvm, figured it out
22:49.20*** join/#openmoko CVirus (n=Satan@41.233.140.245)
22:55.24*** join/#openmoko raster (n=raster@enlightenment/developer/raster)
22:56.39*** join/#openmoko Fox_Muldr (i=quakeman@f050234213.adsl.alicedsl.de)
22:57.00*** join/#openmoko noxchat_tester (n=noxchat_@balticom-140-70.balticom.lv)
22:57.15*** part/#openmoko noxchat_tester (n=noxchat_@balticom-140-70.balticom.lv)
23:44.00*** join/#openmoko Gabrys1 (n=Gabrys@77-253-139-158.adsl.inetia.pl)
23:44.51*** join/#openmoko Gabrys (n=Gabrys@77-253-139-158.adsl.inetia.pl)
23:47.32*** join/#openmoko Gabrys (n=Gabrys@77-253-139-158.adsl.inetia.pl)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.