**** BEGIN LOGGING AT Mon Dec 11 02:59:57 2006 Dec 11 07:44:46 hello Dec 11 08:09:58 good morning all Dec 11 08:16:11 good morning Dec 11 08:16:28 hey likewise Dec 11 08:32:10 good morning all Dec 11 08:34:02 good morning chris144, koen Dec 11 08:34:19 hi likewise Dec 11 08:34:23 hi koen Dec 11 08:34:25 morning sir ): Dec 11 08:34:33 hi Genesis Dec 11 08:47:46 mornin' everyone Dec 11 08:49:30 morning Dec 11 08:57:18 who is Rolf Leggewie? Dec 11 08:57:23 morning btw Dec 11 08:59:25 hey XorA Dec 11 09:00:55 hey Jin^eLD Dec 11 09:02:06 morning all Dec 11 09:02:12 hi Dec 11 09:02:17 XorA: Rolf = Laibsch Dec 11 09:02:36 RP: cheers Dec 11 09:03:39 Laibsch: next checkin will fix 1669, I sent wrong patch last night, it missed a change Dec 11 09:04:12 ~lart quilt Dec 11 09:04:12 * ibot tries to shut quilt up Dec 11 09:14:50 03xora 07org.oe.dev * rb564b6d6... 10/ (1 packages/mplayer/files/makefile-nostrip.patch): Dec 11 09:14:50 mplayer_0.0+1.0rc1.bb : update makefile-nostrip, I missed a change last Dec 11 09:14:50 night. #1669 is fixed now. Dec 11 09:17:04 ~lart forms that need MS excell to be completecd Dec 11 09:17:04 * ibot crushes forms that need MS excell to be completecd with a full height scsi disk Dec 11 09:24:33 any of you guys familiar with Monta Vista Pro stuff? I am just wondering if it has something special that would be really *needed* or if I could purge it and only keep the kernel patches for the vendor hardware Dec 11 09:25:52 * XorA shudders at MVP Dec 11 09:26:03 never use mvista stuff Dec 11 09:26:04 XorA: what is it anyway? Dec 11 09:26:12 ( you should never ... =) Dec 11 09:26:30 Genesis: well, so my question - if it is avoidable :) Dec 11 09:26:45 oki :) Dec 11 09:27:22 the thing is - http://www.ti.com/corp/docs/landing/davinci/index.html Dec 11 09:27:33 I may get my hands on TI DaVinci, and it comes with MVP Dec 11 09:27:46 Jin^eLD: MVP tends to be old version of 2.4.X kernels fscked to hell with low quality code, but sold to managers who dont have a clue as being fast to work with Dec 11 09:28:10 about TI , i read that they use gst framework on their DSP to make develop them easier Dec 11 09:29:12 ( perharps could interest some hardware prescriptor that camp here ) Dec 11 09:30:11 XorA: from what I read the TI comes with 2.6.10 MVP kernel Dec 11 09:30:27 RP: http://gitweb.freedesktop.org/?p=users/behdad/cairo.git;a=commit;h=d338ee97d697aade399e2b0cd6cf067011a7516e Dec 11 09:30:49 MV claims about some realtime patches, whatever Dec 11 09:30:50 hmm Dec 11 09:30:50 Jin^eLD: well at least thats only 10 revisions behind :-) Dec 11 09:30:54 :) Dec 11 09:31:43 well, the question is what exactly is there and if I have to use it, or if I could safely purge it; I surely would prefer to use OE Dec 11 09:31:43 Jin^eLD: likewise has a davinci board near his desk Dec 11 09:31:44 Jin^eLD: at wolfson we only support our drivers in 2.6.XX where XX is recent, we leave the montavista guys to backport stuff themselves Dec 11 09:32:37 koen: ever improving :) Dec 11 09:33:03 Jin^eLD: we have that board here. Dec 11 09:33:19 that's nice to hear Dec 11 09:33:34 koen: cool Dec 11 09:33:59 likewise: so what are you using? what does it require? Dec 11 09:34:34 likewise: I assume you got it running with OE? Dec 11 09:34:44 Jin^eLD: not tried yet. Dec 11 09:34:52 Jin^eLD: the board is collecting dust. Dec 11 09:34:57 Jin^eLD: http://focus.ti.com/dsp/docs/dspsplash.tsp?contentId=2552 Dec 11 09:35:22 Jin^eLD: MV is providing the toolkit and kernel - slowly but gradually Dec 11 09:36:21 Jin^eLD: yes, my aim is OE on this board, but only if we can use the link to the DSP, after which we hit the proprietary DSP codec stuff etc. Dec 11 09:36:40 * koen received an interesting email Dec 11 09:36:44 "I have made contact with TomTom, and they are willing to consider letting us be the first commercial Linux distribution of navigator IF we can get it working on our device and prove it to them without their help." Dec 11 09:37:10 likewise: do you know how this DSP link is done? Dec 11 09:37:38 Jin^eLD: no, lots of reading to do still Dec 11 09:38:16 likewise: let's see when I get the board... Dec 11 09:41:29 Dmnd, I wish I could replicate h/w as easily as source code... Dec 11 09:45:35 :) Dec 11 09:47:25 koen: Which distribution? Dec 11 09:47:42 koen: btw, what hardware is TomTom using? Dec 11 10:11:41 koen: Since you were the last to comment on bug 301 and since it seems to be easy to either fix or set to wontfix... Can you be bothered to deal with this one? Dec 11 10:21:06 morning all Dec 11 10:24:41 hello lrg Dec 11 10:26:18 RP: no idea which distro yet Dec 11 10:27:43 i've a compiled gnash , but it doesn't play flash :X Dec 11 10:31:13 Genesis: :-D Dec 11 10:34:49 RP: FULL_OPTIMIZATION_pn-perl = "-foo -bar" should work, right? Dec 11 10:43:25 koen: yes Dec 11 10:47:24 * koen mtn commit Dec 11 10:47:56 03koen 07org.oe.dev * r244049db... 10/ (1 conf/distro/include/angstrom-glibc.inc): angstrom glibc.inc: use -O1 for perl to fix #1616 Dec 11 11:09:03 Jin^eLD, I should have one of these in January Dec 11 11:09:03 http://www.lyrtech.com/DSP-development/dsp_fpga/sffsdrdevelopmentplatform.php Dec 11 11:10:00 Crofton: interesting! Dec 11 11:20:06 cute, 600MHz c64x Dec 11 11:20:40 they are giving us green hills integrity with it Dec 11 11:20:48 I want to fix that :) Dec 11 11:24:52 Crofton: how's openomap doing? Dec 11 11:26:06 fine Dec 11 11:26:20 my current headache is the USRP interface with usb1.1 Dec 11 11:27:00 which isn't an OE issue, so that is a good thing Dec 11 11:27:31 I need to build the dsp gw andplay with it also Dec 11 11:27:40 but I also need to go into writing mode Dec 11 11:43:21 has anyone ever played with the Nino pda's from Philips ? Dec 11 11:43:53 the run win ce, mips processor something around the 70Mhz range Dec 11 11:44:43 zecke, thanks for pushing directfb.. I will work on getting the other packages (debug..) built for it as well. Dec 11 11:44:48 Ifaistos: hey, obviously I did not start with 'packing' code Dec 11 11:44:55 Gerrath: thank you :) Dec 11 11:45:15 zecke : Hi ! Well i made some progress :) Dec 11 11:45:37 zecke : it takes sometime as my python charming skills are not that good Dec 11 11:46:08 hi all Dec 11 11:46:15 zecke : i managed to get all the components packed, with the exception of the ld lib.... no idea how to find that from OE Dec 11 11:46:32 ld lib is not required Dec 11 11:47:16 Ifaistos: you will need it if you run the toolchain on a system with a smaller glibc (2.1. vs. 2.5) Dec 11 11:48:01 zecke : any idea how to find it the ld lib from OE ? Dec 11 11:48:22 Ifaistos: using the hosts objdump Dec 11 11:48:35 Ifaistos: or ldd on the gcc, cc1, cc1plus and add these libs :) Dec 11 11:48:42 Ifaistos: but I would understand not to do this... Dec 11 11:50:16 zecke : why ? Dec 11 11:51:00 Ifaistos: you will need to package ld-linux.so, libc.so, libm.so, linux-gate (or how this is called) Dec 11 11:52:27 zecke : the other problem i have is how to determine from python if i am trying to copy a link or the actuall file... Dec 11 11:52:46 os.path should help you Dec 11 11:52:52 zecke : this is advanced pythonese :) Dec 11 11:53:54 http://docs.python.org/lib/module-os.path.html Dec 11 11:54:00 Ifaistos: os.path.islink(path) Dec 11 11:55:16 zecke: I ordered a usb2 dvb-t unit today :) Dec 11 11:55:29 zecke : thanks. i'll try to fix a couple more issues with the bbclass and then push it to the rep, to see how it works for other also Dec 11 11:55:35 I should be close enough to germany to get a decent number of channels Dec 11 11:55:56 awesome :) Dec 11 11:56:30 koen: good! dou you have dvb-t at all? Dec 11 11:56:32 koen: hehe Dec 11 11:56:45 koen: build yourself an antenna Dec 11 11:56:47 zecke: yes, the goverment stopped analog broadcasts last week Dec 11 11:57:03 zecke: and by law they are required to broadcast the public nets Dec 11 11:57:25 zecke : btw icecc --build-native runn from within a bbclass uses the hosts native settings unfortunatelly :( Dec 11 11:57:49 so you can watch the public channels everywhere in .nl through dvb-t Dec 11 11:58:31 2010 or 2008 is the final date when they need to switch all analog channels to dvd-t ? Dec 11 12:00:29 in austria they are turning analog off in a couple of months and I am very happy about that, since I am not bying dvb-t stuff I finally will no longer have to pay for this crap; here you are forced to pay if you have a tuner at home Dec 11 12:02:27 well in Greece you pay for the public tv depending on how your energy consumption is.... Dec 11 12:03:05 so you end up having factories paying enourmous sums to the goverment channels :) although they don't have even a single tv set Dec 11 12:03:29 odd Dec 11 12:03:44 well, I do not like this whole public tv policy at all Dec 11 12:04:04 in Germany they even want money if you have a PC - they are arguing that you theoretically could watch public TV on your PC Dec 11 12:04:13 that's just a ripoff Dec 11 12:04:25 if it has a cable....tax it :) Dec 11 12:05:14 :) Dec 11 12:08:22 Ifaistos: an if it has no cable, tax it as well! Dec 11 12:09:05 heh... it sounds like the principles of Japanesse marketing.... Dec 11 12:09:13 The UK works on receiver circuits. Own any receiver circuit and you're eligible for a TV licence. It neatly includes PC's capable of watching TV Dec 11 12:09:20 rule number one . The boss is always right Dec 11 12:09:42 rule numner two. In case the boss is wrong follow rule number one :) Dec 11 12:09:45 RP: and point out that includes watching via internet streaming Dec 11 12:12:05 have to go. cu later Dec 11 12:40:10 mickeyl: morning dude Dec 11 12:40:16 heya Dec 11 12:41:43 I have a problem with init: http://pastebin.ca/275324 ... glibc incompatibility ? why is it looking for glibc 2.4, my image is built with glibc 2.5 ? Dec 11 12:41:53 hey mickeyl Dec 11 12:44:33 morning mickeyl Dec 11 13:23:24 my angstrom build has glibc problem when running init on my mx31ADS: http://pastebin.ca/275324 someone has a clue what's wrong ? Dec 11 13:25:31 vlo: I saw this on lrg screen, it looks to me like you have a glibc2.4/2.5 init trying to run with a glibc 2.3.5 libc6 Dec 11 13:28:42 XorA: ok ... but my /lib/libc6.so seems to be glibc2.5 : Dec 11 13:28:42 valentin@lsa1pc7:~/EPFL/iMXBoard/linux/openembedded/build/mx31ADS/tmp/rootfs$ ll lib/libc.so.6 Dec 11 13:28:42 lrwxrwxrwx 1 valentin valentin 11 2006-12-07 22:43 lib/libc.so.6 -> libc-2.5.so Dec 11 13:29:02 vlo: wierd Dec 11 13:30:33 XorA: now I have received my JTAG probe, I could try to debug such things ... buuuut ... (I'm ashamed) I don't know how to use it efficiently ... Dec 11 13:34:01 anyone has some experience with macraigor JTAG probes ? Dec 11 13:34:15 vlo: lrg :-) Dec 11 13:36:30 03koen 07org.oe.dev * r5213367b... 10/ (3 files in 2 dirs): libeventdb: use the new GPE svn repo Dec 11 13:46:58 03awilcox 07org.oe.dev * rad4e848a... 10/ (3 files in 3 dirs): linux-storcenter: missing depend on uboot-utils for mkimage Dec 11 13:47:29 hey HopsNBarley Dec 11 13:47:43 hi koen, good trip this weekend? Dec 11 13:47:52 yep Dec 11 13:48:58 thanks for pointing out my non-staging area oops in my first commit. long fixed. Dec 11 14:23:32 RP: http://blogs.gnome.org/view/xan/2006/12/11/0 Dec 11 14:27:53 morning Dec 11 14:28:56 hey chouimat Dec 11 14:32:34 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Dec 11 14:33:28 hehe Dec 11 15:00:34 does anybody knows if there is an rdesktop port which runs without X? Dec 11 15:00:48 tmbinc: opie-rdesktop did Dec 11 15:01:01 tmbinc: it is basicly the normal rdesktop client built for QtE Dec 11 15:01:08 tmbinc: rdesktop client though :) Dec 11 15:01:37 tmbinc: http://rdesktop.sourceforge.net/ Dec 11 15:02:29 yes that's what i've searched (the opie version) Dec 11 15:02:30 thanks Dec 11 15:04:12 is there any reason not to convert x86.conf to task-base? Dec 11 15:04:32 tmbinc: there is a opie fork of rdesktop and there is native Qt support in rdesktop itself Dec 11 15:04:42 tmbinc: I don't know which is better... Dec 11 15:07:46 cbrake: Not that I'm aware of Dec 11 15:09:03 koen: cool :) Dec 11 15:09:15 RP: I've been doing task-base x86 builds and they seem to be working. Should I go ahead and make the change, or submit a bug? Dec 11 15:09:44 cbrake: Feel free to commit that change Dec 11 15:09:50 * koen is surprised noone has raised valid concerns about removing task-bootstrap tomorrow Dec 11 15:10:11 koen: has everything converted yet? Dec 11 15:10:16 03awilcox 07org.oe.dev * r7168bc90... 10/ (1 packages/apache2/apache2-native_2.2.3.bb): apache2-native: doesn't build if native apr is found. Dec 11 15:10:20 03cbrake 07org.oe.dev * re2e35578... 10/ (1 conf/machine/compulab-pxa270.conf): compulab-pxa270.conf: convert to task-base Dec 11 15:10:25 03cbrake 07org.oe.dev * r7ccd0151... 10/ (1 conf/machine/logicpd-pxa270.conf): logicpd-pxa270.conf: convert to task-base Dec 11 15:10:46 RP: everything maintained Dec 11 15:12:17 koen: I have misgivings with breaking things just for the sake of breaking them Dec 11 15:14:10 RP: it isn't breaking stuff Dec 11 15:14:24 03cbrake 07org.oe.dev * rbe66fa1e... 10/ (1 conf/machine/x86.conf): x86.conf: switch to task-base Dec 11 15:14:35 it's removing a deprecated part of OE that has been on the removal list for over a month Dec 11 15:14:53 koen: Will any currently working unmaintained machines break if we remove task-bootstrap? Dec 11 15:15:54 the unmaintained machine that were converted will still work Dec 11 15:16:12 I doubt there are unmaintained machines that work with task-bootstrap as-is Dec 11 15:37:09 * mrz80 is back (gone 01:04:36) Dec 11 15:47:54 03koen 07org.oe.dev * rf676cd28... 10/ (1 conf/machine/efika.conf): efika: add MACHINE_FEATURES Dec 11 16:09:51 grr, still no email from fedex Dec 11 16:10:25 koen, how well do we support headless efika's and ones with graphics cards? Dec 11 16:13:37 morning all Dec 11 16:16:19 hey do13 Dec 11 16:16:45 Crofton|oe: the default kernel supports ati and sis/xgi cards Dec 11 16:16:47 hi koen Dec 11 16:20:24 I will run without a card, if I ever get an efika Dec 11 16:26:04 hrw|gone: Did you have time to look at my RFC re sourceforge in the mailing list? Dec 11 16:42:47 'night Dec 11 16:49:11 ~kill selinux for stopping eclipse opening the gtk file selector Dec 11 16:49:21 * ibot shoots a hyper-charged tachyon gun at selinux for stopping eclipse opening the gtk file selector Dec 11 17:06:16 03tmbinc 07org.oe.dreambox * r0001554e... 10/ (1 classes/base.bbclass): base.bbclass: add source mirror Dec 11 17:14:45 hey Dirk Dec 11 17:14:53 tmbinc: ping Dec 11 17:18:47 koen: yes? Dec 11 17:19:27 tmbinc: are there dvb-c capable dreamboxen? Dec 11 17:19:33 yes Dec 11 17:19:38 the DM7025 has pluggable tuners Dec 11 17:19:44 and the DM500 is available as -C Dec 11 17:21:14 cool Dec 11 17:26:34 hey Liam Dec 11 17:29:00 * lrg -> train Dec 11 17:29:02 cu Dec 11 17:41:06 hey zecke Dec 11 17:44:45 zecke: Moin :) I have someting a bit OT. Dec 11 17:45:03 sure Dec 11 17:45:41 zecke: In your QA research, have you come across any good resources for testing on computer controlled equipment? Dec 11 17:46:17 no :) Dec 11 17:46:25 My company makes robotic equipment and we are trying to get more automated testing done... the problem is that things can move around on you :) Dec 11 17:46:39 zecke: Ok, just curious :) Dec 11 17:47:08 hvontres|home: hmm we did not to system tests at FU Fighters Dec 11 17:47:22 hvontres|home: we tested the parts separately with captured input data Dec 11 17:47:32 and looked which commands where sent to the driving engine Dec 11 17:47:43 or which objects where identified. So mostly baseline tests Dec 11 17:48:34 Good evening Dec 11 17:48:42 hey sirfred Dec 11 17:49:22 Hi, koen Dec 11 17:49:48 hvontres|home: Where is this company located? Dec 11 17:50:07 VoodooZ: Irvine, CA Dec 11 17:51:58 VoodooZ: But, be warned, currently we us Windblows for the UI and VxWorks for our Real Time stuff Dec 11 17:52:13 morning fooks Dec 11 17:52:20 * VoodooZ runs away screaming... Dec 11 17:52:49 Cool. I was just hoping you'd be closer to Canada so I could try to get a job there. :) Dec 11 17:53:38 VoodooZ: you moving? Dec 11 17:53:47 no, can't Dec 11 17:54:00 hey likewise & VoodooZ Dec 11 17:54:04 I'm in Canada (Ottawa, Ontario) so real far. Dec 11 17:54:06 koen: hi Dec 11 17:54:20 * VoodooZ is reporting for duty. Dec 11 17:54:26 :) Dec 11 17:54:42 VoodooZ: But just think: no more shoveling the drive in the morning :) Dec 11 17:54:55 * chouimat|away decides to talk the 5km trip to the grocery store since Ottawa is freaking huge Dec 11 17:55:04 s/talk/take Dec 11 17:55:31 hi likewise, thanks for the storcenter kernel build bug report yesterday. Dec 11 17:55:50 yeah, but in the last few years we get less and less snow so ... Dec 11 17:55:56 hehe Dec 11 17:56:15 I like skiing, but I love motorcycling so I get the best of both world really. Dec 11 17:56:25 you get used to the cold. Dec 11 17:56:35 VoodooZ: not for long ... the Gulf Stream will change is route soon so ... welcome back the cold ;) Dec 11 17:57:05 hehehe. I think I'm ready for it. -40degrees celsius, bring it on! Dec 11 17:57:16 bye folks Dec 11 17:58:21 ok back in 1.5 hours ... Dec 11 17:58:25 VoodooZ: you work@videotron ? How many embedded s/w engineers there? Dec 11 17:58:49 likewise: nope. That's the ISP I'm stuck with at home. Dec 11 17:59:02 I work for a net security firm Dec 11 17:59:13 HopsNBarley: no problem, I tried to bake u-boot but failed a few times. Dec 11 17:59:21 I'm trying to get back in the embedded field. Dec 11 17:59:30 Hello Dec 11 17:59:41 VoodooZ: Ehm, I mistake the ISP name with something similar, but cannot find the name now. Dec 11 17:59:46 VoodooZ: QNX? Dec 11 18:00:00 VoodooZ: are they in ottawa? Dec 11 18:00:01 hey koen. Heard anything about the slug guys converting to angstrom yet? Dec 11 18:00:12 hu Dec 11 18:00:17 QNX? possibly. I'm not sure. Dec 11 18:00:21 VoodooZ: rwhitby sent an RFC iirc Dec 11 18:00:21 hi even. Dec 11 18:00:25 :) Dec 11 18:00:33 likewise: close enough they are in Kanata :) Dec 11 18:00:44 I think there's a WindRiver office here though Dec 11 18:00:59 koen: really? where? I didn't see it on the yahoo groups. Dec 11 18:01:09 VoodooZ: to the OE list Dec 11 18:01:38 hi, all! Dec 11 18:01:52 ah. that's why I missed it. Dec 11 18:02:12 * VoodooZ will have to become a member of that list. Dec 11 18:02:13 is it good style to leave files unpackaged or it's better to explicitly delete them? Dec 11 18:03:25 chouimat|away: that's 10 miles? (I dunno sh*t about Canada, excuse the topographic errors :-) ) Dec 11 18:04:58 Any idea how QNX is doing, after all this time being Harman owned? Dec 11 18:05:41 hey mallum! Dec 11 18:05:48 hey koen Dec 11 18:07:24 likewise: from where I am in Ottawa Kanata is about 10km ... (and Kanata is also part of ottawa) ... and I'm about 17KM from downtown Dec 11 18:10:31 chouimat|away: thanks for the education. Dec 11 18:11:45 I have made an init script for icewm (the start|stop|restart kind), but it always takes up my console Dec 11 18:12:11 Is there a way I can make it become a baground process without resorting to xdm? Dec 11 18:12:23 background Dec 11 18:13:01 The only way I can make it stop using the tty, is starting it with >/dev/null Dec 11 18:13:10 But that is not a very good solution Dec 11 18:13:29 And I'm probably asking in the wrong channel Dec 11 18:13:32 :| Dec 11 18:14:27 B_Lizzard: take a look at mickeyl's daemonize patch for QtE Dec 11 18:14:46 Allright... Dec 11 18:14:51 B_Lizzard: fork, play with the process groups, exit the parent process... Dec 11 18:15:00 Aha Dec 11 18:15:14 Is that appliable to a bash script? Dec 11 18:17:39 no :) Dec 11 18:17:43 :) Dec 11 18:17:54 well I don't know Dec 11 18:18:15 I was wondering, GPE and OPIE don't have such problems cause they use their respective login programs? Dec 11 18:18:29 mickeyl: wb Dec 11 18:18:33 hi Dec 11 18:18:41 I tried to find some example as to how to free up the terminal there, but nothing relevant came up Dec 11 18:18:45 hello Dec 11 18:18:47 B_Lizzard: mickeyl's daemonize patch with neat ascii art in the Opie case Dec 11 18:18:55 I saw it :) Dec 11 18:19:06 "we are a daemon now" Dec 11 18:19:21 B_Lizzard: it might be possible with bash, but we don't have bash installed and I'm bad at shell programming Dec 11 18:19:33 Uh huh Dec 11 18:19:43 Well, searching the net found me nothing Dec 11 18:19:59 Maybe bash has something like a --quiet switch Dec 11 18:20:07 @echo off Dec 11 18:20:17 http://www.google.com/codesearch?q=%22we+are+a+daemon+now%22&btnG=Suche&hl=en&lr= Dec 11 18:20:40 hahaha Dec 11 18:23:25 splitnode! Dec 11 18:23:48 haha, directFB's keyboard.c has a bug :) Dec 11 18:23:59 :) Dec 11 18:24:23 0.9.24 Dec 11 18:26:33 oh, OE's version of my kernel-keymap handler has it too... Dec 11 18:27:20 hmm only qt-2.3.12 gets the old patch applied... Dec 11 18:28:09 I'm starting to hate kdrive Dec 11 18:28:32 sirfred: welcome to the club Dec 11 18:28:48 This offscreen management... Dec 11 18:29:34 It's hardwired everywhere that the offscreen zones are *after* the visible screen. I don't see a simple way to shorcircuit their management to use the one in libw100 Dec 11 18:31:15 you guys had some easy to remember diff options for patches Dec 11 18:31:32 apparently they aren't that easy ... -nerd maybe? Dec 11 18:31:43 What about directfb?I read somewhere there's an x server based on directfb Dec 11 18:33:09 Crofton: I think it was -Nurd.... kinda like nerd :) Dec 11 18:33:28 Crofton: -Nurd :) Dec 11 18:34:12 sirfred: there is Xdirectfb Dec 11 18:34:21 XorA|gone: Hello Dec 11 18:34:42 sirfred: and if you got accellerated directfb on w100 working, you wouldnt need to work out Xv Dec 11 18:34:55 ah Dec 11 18:35:07 -Nacr also Dec 11 18:35:08 :) Dec 11 18:35:14 XorA|gone: I didn't try. I thought kdrive was the best option. Dec 11 18:35:38 XorA|gone: The problem is that no solution seems to be optimal. Dec 11 18:35:48 sirfred: always the way Dec 11 18:36:20 -qte2 seems to be abandoned. Dec 11 18:36:44 -kdrive needs to be hardly modified to accept management of the internal imageon memory. Dec 11 18:36:47 http://en.wikipedia.org/wiki/Diff Dec 11 18:37:16 ~lart libusb Dec 11 18:37:16 * ibot wallops libusb with a main rotation server that needs rehubbing. It won't take long Dec 11 18:37:33 And the worse is that for example SDL on X seems not to use hardware surfaces. Dec 11 18:38:23 And also seems that kdrive does not expose acceleration for things like stretch+blitting, that should be useful for SDL Dec 11 18:38:38 KAA is that limited? Dec 11 18:38:57 koen: About offscreen management? Dec 11 18:39:07 acceleration Dec 11 18:39:37 koen: Well, stretching seems to be not supported, I think. Dec 11 18:39:42 ~seen gerwinin Dec 11 18:40:30 WTF is going on with mplayer on collie Dec 11 18:40:48 gerwinin was last seen on IRC in channel #oe, 1d 4h 51m 58s ago, saying: 'Zecke: I am planning to attent some more events as well , but do no want to stand there in an oe booth without having this thing discussed with yiu guys '. Dec 11 18:41:01 koen: There's no acceleration for lines or colour expansion neither Dec 11 18:41:17 They say this is not hardly used in modern desktop. Dec 11 18:42:30 does anyone know if there is anything special you must do to get recording support for the UCB1x00 codec and the PXA270 or if the native codec kernel drivers already support audio recording... Dec 11 18:42:38 XorA|gone: Have you tested that Xdirectfb ? Dec 11 18:43:06 sirfred: no, I never get round to making a .bb for it Dec 11 18:43:18 sirfred: something always seems to come up Dec 11 18:43:35 I think there's also a gtk port working on directfb. Dec 11 18:43:55 I hardly remember to have seen some gimp screenshot running on directfb Dec 11 18:45:52 heh Dec 11 18:46:05 Sony is promoting its evilness Dec 11 18:46:29 sirfred: these screenshots are old Dec 11 18:46:41 sirfred: but the debian etch installer is using gtk/gdk on directfb Dec 11 18:47:26 zecke: Interesting Dec 11 18:48:54 zecke: yup, and I think that approach might be nice for embedded devices Dec 11 18:49:13 java vm for arm: http://www.k-embedded-java.com/mika/ Dec 11 18:52:13 It seems there's also a SDL driver on directfb. Dec 11 18:53:16 XorA|gone: Did you finally saw the mplayer libw100 based driver in action? Dec 11 18:54:30 sirfred, I had tried to compile gtk for directfb a couple of months ago, but ran in to issues with getting cario to compile, I was working on debugging it but got pulled off onto something else and then we changed directions and went away from gtk but I think it should not be too much work to get it working.. Dec 11 18:54:50 sirfred: I did Dec 11 18:55:04 XorA|gone: How about performance? Dec 11 18:55:04 sirfred, I might play around with it some this week(end) Dec 11 18:55:13 Gerrath: Interesting. Dec 11 18:55:33 I'm just bitbaking directfb, to see how the code looks like. Dec 11 18:55:34 likewise: been there :) Dec 11 18:55:52 likewise: convergence, set-top box, Gtk+ on a set-top box :) Dec 11 18:55:58 zecke: where? gtk on directfb? Dec 11 18:56:03 gobject was just too slow :) Dec 11 18:56:07 sirfred: I tried playing a divx and it was dead slow Dec 11 18:56:15 XorA|gone: What res? Dec 11 18:56:21 sirfred: gimme a sec Dec 11 18:56:22 likewise: well I have never worked there, but convergence tried the gtk+ approach :) Dec 11 18:56:29 ah, ok. Dec 11 18:56:42 zecke: and then they threw it out? Dec 11 18:58:11 likewise: they were out of money before throwing it out :) Dec 11 18:58:17 sirfred: 608x336 24bpp 23.976 gotta go Dec 11 18:58:22 XorA|gone: jeje Dec 11 18:58:34 XorA|gone: You don't have enought decoding power, I'm afraid. Dec 11 18:58:56 XorA|gone: Your decoding counter should be over 100% Dec 11 19:00:43 XorA|gone: I tried with a recoded movie, at 320x204, at fullscreen , without rotating the source, and it didn't played bad, well, the sound was distorted, and the time in the audio decoding was more than 20% I remember Dec 11 19:01:13 XorA|gone: The video_out percent was at 2-3% only Dec 11 19:02:54 XorA|gone: I think that the only way we could play a video of the size you're testing, is performing the IDCT in the card, so the amount of data to transfer to the card will be reduced dramatically. Dec 11 19:11:37 sirfred: I actually thought that video was smaller, thats a HDTV rip, oops Dec 11 19:11:48 sirfred: I shall try again with a more sensible size sometime later Dec 11 19:12:15 sirfred: also try mplayer 1.0rc1, and maybe svn as they have been talking a lot about new arm optimisations Dec 11 19:19:44 03koen 07org.oe.dev * rd3a4f128... 10/ (1 site/powerpc-linux): powerpc-linux site: add some more entries Dec 11 19:20:52 XorA|gone: Thanks, I will try. Dec 11 19:26:53 sirfred: I ran the alphablend on hx4700 and got the following result. What do you suggest? http://handhelds.org/scap/port.20730.png Dec 11 19:27:46 mreimer: Buf Dec 11 19:27:51 mreimer: What a distorted result Dec 11 19:28:11 sirfred: makes me wonder if the lib thinks the screen is turned the wrong way? Dec 11 19:28:16 mreimer: Perhaps you could try to change the ac_init() call. Dec 11 19:28:29 mreimer: Now, it should be: ac_init(WIDTH, HEIGHT, AC_ROT90) Dec 11 19:28:31 koen: which tests did you run? Dec 11 19:28:44 sirfred: I'll try that Dec 11 19:28:59 mreimer: Perhaps with ac_init(480, 640, AC_ROT0) Dec 11 19:28:59 zecke: the libiberty configure on the efika itself Dec 11 19:31:30 mreimer: I'm thinking about adding some initialization function that relies on the framebuffer configuration. Dec 11 19:31:37 sirfred: good idea Dec 11 19:31:53 mreimer: The problem is the rotation info. Dec 11 19:32:13 sirfred: how does libw100 interact with w100fb and Xw100? Is it ok to use both? Dec 11 19:32:24 mreimer: No, sadly, it's not. Dec 11 19:32:51 sirfred: w100fb + Xfbdev + libw100 is ok, but not Xw100 + libw100? Dec 11 19:33:25 mreimer: Well, if you set the accel knob file of the w100fb to 0, the first combination shouldn't be dangerous. Dec 11 19:34:14 sirfred: ok. (I'm trying to figure out how to safely try the imageon mplayer driver.) Dec 11 19:34:50 'evening all Dec 11 19:35:05 mreimer: If you're using Xfbdev and set the /sys/bus/platform/devices/w100fb/accel to zero, it should work. Dec 11 19:35:14 ok Dec 11 19:35:16 hi mreimer Dec 11 19:35:17 hi zecke Dec 11 19:35:19 hi pb_ Dec 11 19:35:31 pb_: master :) Dec 11 19:35:50 pb_: time for _one_ m4/autoconf question? Dec 11 19:35:55 zecke: always Dec 11 19:36:44 sirfred: maybe you could use the FBIOGET_VSCREENINFO ioctl to get info about the screen size and rotation Dec 11 19:37:03 mreimer: I tried that, but the rotation seems to be not initialized. Dec 11 19:37:18 sirfred: I wonder if the w100fb driver needs to set that Dec 11 19:37:36 mreimer: Not sure, perhaps it's my fault. Dec 11 19:37:56 mreimer: I tried something like this: Dec 11 19:37:59 pb_: We have these shiny site files. The results are from ./configure runs. My aim would be to collect the test functions/m4 files (source), store them in a file for each vendor (e.g. samba.m4, gnu.m4, gpe.m4). To allow to (re)create the site file for a new/changed platform (man I'm tired...) Dec 11 19:38:46 memset((void *)&vinfo, 0, sizeof(vinfo)); vinfo.bits_per_pixel = 16; vinfo.xres = vinfo.virtual_xres = 640; ... Dec 11 19:38:53 vinfo.activate = FB_ACTIVATE_TEST; Dec 11 19:39:07 ioctl(fd, FBIOPUT_VSCREENINFO, &vinfo) Dec 11 19:39:13 pb_: how would you approach this? my current approach is in http://www.openembedded.org/repo/org.openembedded.dev/contrib/site-conf/ but I don't like it Dec 11 19:39:21 mreimer: And the rotation info I got was a very large number. Dec 11 19:39:35 sirfred: hmm Dec 11 19:39:48 pb_: I don#t like the m4_inclue in aclocal.m4. is there a way I can avoid this? Dec 11 19:40:03 mreimer: I suppose this should be set by w100fb, I will look into the code, perhaps it's just not set. Dec 11 19:40:28 sirfred: have you played with the 'rotate' sysfs attribute? looks like that should work too. Dec 11 19:40:42 sirfred: looks like the driver should look at the rotate setting in its activate function Dec 11 19:41:11 mreimer: Perhaps it's just not filled into the vinfo structure. Dec 11 19:42:39 zecke: I'll take a look Dec 11 19:43:10 pb_: my aim is clear though? Dec 11 19:43:23 I think so Dec 11 19:43:55 sirfred: that's better: http://handhelds.org/scap/port.21637.png Dec 11 19:44:39 mreimer: Much better. Dec 11 19:45:03 mreimer: Some strange offsets in some of the opie icons, have you changed the WIDTH and HEIGHT macros? Dec 11 19:45:35 sirfred: yes, I changed the macros to WIDTH 480 HEIGHT 640 Dec 11 19:45:45 zecke: why don't you like the m4_include in aclocal.m4? Dec 11 19:45:47 mreimer: I wonder why some of the icons are overlapped. Dec 11 19:45:54 sirfred: should I have just changed the ac_init args? Dec 11 19:46:16 mreimer: No, that is fine. Perhaps some bug in the demo itself. Anyway, this one is not longer present in the cvs. Dec 11 19:46:22 ok Dec 11 19:46:51 mreimer: It seems alphablending is working for your card, and it's not a w100, isn't it? Dec 11 19:46:54 zecke: or, alternatively, why aren't you using aclocal to build aclocal.m4? Dec 11 19:47:01 sirfred: it's a w3220 Dec 11 19:47:21 mreimer: I think that somebody with a w3220 said me that the scaling demos didn't work. Dec 11 19:47:21 pb_: because I'm a moron. I some failed to make it look at tme m4 dir Dec 11 19:47:31 ah Dec 11 19:47:38 does "aclocal -I m4" not do the trick? Dec 11 19:47:41 sirfred: I'll work my way through the other tests, starting with the scaling demos. Is there one you'd like me to try? Dec 11 19:47:58 mreimer: I'm curious about the benchmark one, and see the numbers you're getting Dec 11 19:48:04 sirfred: ok Dec 11 19:48:18 pb_: hmm looks better now Dec 11 19:48:34 autoreconf calls aclocal as well? Dec 11 19:48:35 mreimer: Can lead to hangs, well, any demo can. ;) Dec 11 19:48:39 zecke: yeah Dec 11 19:48:59 pb_: hmm, I blame it on BSD Dec 11 19:49:08 good thinking Dec 11 19:49:13 sirfred: I did see a hang or two when I first tried to run the tests Dec 11 19:49:42 mreimer: Perhaps some interaction with the w100fb acceleration. You should better turn off framebuffer acceleration Dec 11 19:49:51 sirfred: good idea Dec 11 19:49:56 mreimer: If you get hungs with the framebuffer accel off, it's a bug. Dec 11 19:49:59 pb_: hmm autoreconf -I m4 does not regenerate an aclocal Dec 11 19:50:20 do I need to instantiate it from configure.ac? Dec 11 19:50:21 sirfred: benchmark seems to have hung testing ext to ext memory BLT Dec 11 19:50:23 zecke: does --force help? Dec 11 19:50:30 pb_: I used -f Dec 11 19:50:32 ah Dec 11 19:50:35 pb_: and it didn't help Dec 11 19:50:37 odd Dec 11 19:50:46 mreimer: Sorry. :( Dec 11 19:50:55 mreimer: And what number before that? Dec 11 19:50:57 sirfred: but I'd guess that's because I'm running the imgblt patch and hadn't disabled accel yet Dec 11 19:51:00 mreimer: And what numbers before that? Dec 11 19:51:04 zecke: does --verbose give you any clues? Dec 11 19:51:08 pb_: this can't be blamed on BSD, but Ubuntu is happy to be blamed Dec 11 19:51:30 autoreconf is just a perl script, it should be fairly easy to figure out what it's doing. Dec 11 19:51:37 mreimer: Yes, that is probably the cause. Newer w100fb are using accel even for drawing the blinking cursor. So conflicts are really probably Dec 11 19:51:52 pb_: yes. It claims configure.ac: not using aclocal Dec 11 19:52:08 pastebin is slow. is there another similar service you all are using? Dec 11 19:52:22 zecke: ah Dec 11 19:52:33 koen: is it senseful to give ixp4xxbe/angstrom/uclibc a run here? Dec 11 19:52:33 zecke: I think it usually does that if it thinks your aclocal.m4 was not generated by aclocal Dec 11 19:52:46 mreimer: #oe@freenode.irc.net, people hates me. :) Dec 11 19:53:03 likewise: only with uclibc svn or 0.9.29 Dec 11 19:53:14 likewise: Jacques was trying something similar iirc Dec 11 19:53:25 pb_: thanks :) Dec 11 19:53:25 sirfred: why? because w100 is too fast now? :-) Dec 11 19:53:33 mreimer: rafb.net/paste Dec 11 19:53:33 pb_: so I will leave the GNU header in this file Dec 11 19:53:46 pb_: I tried with none aclocal.m4 and a touched one Dec 11 19:53:55 odd Dec 11 19:53:58 thanks koen. (and hi!) Dec 11 19:54:00 mreimer: That would be nice. :) Dec 11 19:54:13 hey mreimer Dec 11 19:54:16 zecke: that is odd. a touched one would upset it, but I would expect that it would generate aclocal.m4 if it didn't exist at all Dec 11 19:54:17 autoreconf2.50: running: aclocal --force Dec 11 19:54:20 oh well, crazy gnu Dec 11 19:54:30 sirfred: http://rafb.net/paste/results/hyaBus83.html Dec 11 19:54:30 pb_: and it didn't update it... Dec 11 19:54:44 pb_: I'm too tired to understand this :) Dec 11 19:54:58 oh, that is weird Dec 11 19:55:00 mreimer: Thanks, let's compare with my c7x0 Dec 11 19:55:36 pb_: hmm I assume -I m4 is not passed to aclocal Dec 11 19:55:44 zecke: it should be Dec 11 19:56:46 mreimer: The speed is really better for your PDA Dec 11 19:56:57 RP: Hello Dec 11 19:57:02 sirfred: this has a 624MHz xscale pxa270 Dec 11 19:57:48 mreimer: It seems your speeds are doubling mine, more or less. Dec 11 19:58:05 sirfred: wow Dec 11 19:58:17 sirfred: I think the memory bus is 208MHz Dec 11 19:58:24 sirfred: that might explain it Dec 11 19:58:26 pb_: probably my configure.ac is just too bad Dec 11 19:58:51 mreimer: I have 13.409945, 12.697497, 6.391711 and 6.394025 for your 24.712544, 24.762773, 11.552360 and 11.561250 Dec 11 19:59:47 Even the internal to internal gave me 126.198889 for your 179.154962 Dec 11 20:01:10 RP: Do you know at what speed runs the bus of the c7x0 ? Dec 11 20:01:33 sirfred: ~100MHz Dec 11 20:01:43 koen: Thanks. That makes sense. Dec 11 20:02:12 ~lart c7x0 bus Dec 11 20:02:12 * ibot takes a rusty axe and swings it violently, taking c7x0 bus's head off Dec 11 20:02:14 sirfred: full benchmark http://rafb.net/paste/results/0O1g2u82.html Dec 11 20:02:23 morning Dec 11 20:02:26 zecke: it seems Dec 11 20:02:31 hi hrw Dec 11 20:02:32 ~lart RHEL for 2.6.9 kernel Dec 11 20:02:33 * ibot moos at RHEL for 2.6.9 kernel Dec 11 20:02:40 [hrw@misc1 ~]$ sudo chroot X86rootfs/ bin/sh Dec 11 20:02:40 FATAL: kernel too old Dec 11 20:03:05 argh Dec 11 20:03:06 mreimer: Buf, your external to external speed is really awesome. Dec 11 20:03:22 * zecke just filed a bug report at directfb Dec 11 20:03:35 sirfred: that's good to hear. what test should I run next? Dec 11 20:03:36 mreimer: Sure you could get a really better mplayer performance than with the c7x0. Dec 11 20:03:50 koen: seen codesourcery's 3MB patch to uclibc? http://www.codesourcery.com/public/uClibc-0.9.28-csl-nptl-9.patch.gz Dec 11 20:03:54 mreimer: Do you have a test called videotest or something so? Dec 11 20:03:58 sirfred: yes Dec 11 20:04:12 sirfred: I'll recompile with the right rotation setting Dec 11 20:04:21 mreimer: That one tried to mimic the things the mplayer driver does Dec 11 20:04:35 mreimer: So, if that one works for you, the mplayer driver should work also Dec 11 20:04:39 ok Dec 11 20:05:33 mreimer: You should see the image of a hedgehog, scaled to fullscreen (aprox) rendered 100 times. You will get a frame rate at the end. Dec 11 20:06:28 mreimer: If you want to render more frames, you can just specify an integer as argument, for example: Dec 11 20:06:40 root@c7x0:/media/card/libw100/bin$ ./videotest 1000 Dec 11 20:06:40 Elapsed time for 1000 images: 3 seconds, 699092 microseconds. 270.337 fps Dec 11 20:06:43 sirfred: I have only a blank screen and the following line on the console: dstpoint x = 26(320 - 303)/2 y = 8(240 - 187)/2 Dec 11 20:07:17 mreimer: It hanged? Dec 11 20:07:23 sirfred: either that, or it hasn't finished rendering the first frame yet ;-) Dec 11 20:07:34 mreimer: Bad news. Dec 11 20:07:38 likewise: yes Dec 11 20:08:04 koen: you're running videotest too? Dec 11 20:08:04 hrw: that's why I said chroot sucks and you'd want qemu/vmware Dec 11 20:08:29 koen: nm. lol. I thought you were reporting the same result I was. "likewise" just like my result Dec 11 20:08:45 koen: on remote box? Dec 11 20:08:57 sirfred: do you have any suggestions for debugging this hang? Dec 11 20:09:19 mreimer: You will have to try simpler demos. Dec 11 20:09:20 hrw: yes, since you can put a qemu bin on your homedir Dec 11 20:09:21 * likewise did consider another nick once, but thought nah... Dec 11 20:09:26 sirfred: ok Dec 11 20:09:30 mreimer: To see what is really not working. Try the rotation ones. Dec 11 20:10:02 pb_: next question: Dec 11 20:10:16 mreimer: I'm afraid the way to perform rotations, specially the register ENG_CNTL is not compatible with the IMAGEON 100 Dec 11 20:10:24 pb_: I define the function in the m4 but need to touch the configure.ac to invoke/call/expand it Dec 11 20:10:36 pb_: is there an easy way to invoke these methods directly? Dec 11 20:10:52 mreimer: Another thing you can try, is running changing the debug level, you can launch the demos like this: Dec 11 20:10:57 AC_DEBUGLEVEL=0 ./demo bla bla Dec 11 20:11:04 sirfred: maybe I can figure that out by rotating the screen in PPC and reading registers with haret Dec 11 20:11:05 sirfred: ok Dec 11 20:11:55 mreimer: No, because the ENG_CNTL controls the trajectory of a blit operation. The whole screen is rotated changing other registers. Dec 11 20:12:09 sirfred: bummer Dec 11 20:22:38 zecke: not that I can think of, but you could make a script to programatically generate configure.ac from the .m4 files. Dec 11 20:24:20 pb_: okay Dec 11 20:25:30 sirfred: hostyuv: Elapsed time for 1000 frames: 2 seconds, 821592 microseconds Dec 11 20:25:58 Elapsed time for 1000 frames: 5 seconds, 105351 microseconds here Dec 11 20:26:23 mreimer: So, overlay is working for you. Your hang problem is probably caused by the rotation. Dec 11 20:26:31 mreimer: Try the rotblt demo. Dec 11 20:26:35 sirfred: rotblt 100 100 0 gives me an image of a hedgehog? Dec 11 20:26:49 mreimer: It should. And what about rotblt 100 100 1 ? Dec 11 20:27:16 mreimer: With 0 you request no rotation at all, so no problem, it's just like a ordinary blt. Dec 11 20:27:19 sirfred: all I get is a thin black line down the middle of the previous image Dec 11 20:27:46 mreimer: It seems that this is the problem. Dec 11 20:27:58 sirfred: it didn't hang though, which is good Dec 11 20:28:14 mreimer: Well, perhaps trying it more times it finish hanging. Dec 11 20:28:19 sirfred: http://handhelds.org/scap/port.23275.png Dec 11 20:28:52 sirfred: the opie icons are probably from the hostcopyrot tests Dec 11 20:29:26 mreimer: Where they here before running the rotblt test? Dec 11 20:29:32 sirfred: yes Dec 11 20:29:35 mreimer: ok Dec 11 20:29:39 sirfred: I didn't know how to clear the screen Dec 11 20:30:00 sirfred: does this alphablend2 result look right: http://handhelds.org/scap/port.23319.png Dec 11 20:31:29 sirfred: or this brush result: http://handhelds.org/scap/port.23375.png Dec 11 20:34:12 sirfred: font:Elapsed time: 0 seconds, 90607 microseconds Dec 11 20:36:01 Could someone explain to me, what scan[column].dc means if scan[] is an array ? Dec 11 20:36:32 mreimer: The alphablend2 result is not good, the colours should be rgb Dec 11 20:36:51 mreimer: The brush result is fine. Dec 11 20:37:29 sirfred: ovrotate puts bits on the screen which seem to be scrambled versions of the hedgehog picture Dec 11 20:37:50 sirfred: ovstretch shows a black box Dec 11 20:38:00 mreimer: It seems that the problem is the rotation, all the time. Dec 11 20:38:04 mreimer: Sorry, I have to leave now. Dec 11 20:38:12 sirfred: bye. thanks for your help Dec 11 20:38:18 mreimer: Thanks for testing. Dec 11 20:38:28 sirfred: I'm glad to do any testing you need. Dec 11 20:38:45 Laibsch: do not assign bugs to me Dec 11 20:39:06 Laibsch: if you want me to be interested in bug - add me to Cc: in it Dec 11 20:48:48 Had anyone ever had any problems with dbus starting up? Dec 11 20:49:00 I've been having problems these last 2 days Dec 11 20:50:01 B_Lizzard: like #1567? Dec 11 20:50:05 Starting gnumeric states: "Client failed to connect to the D-BUS daemon: Failed to execute dbus-launch to autolaunch D-Bus session" Dec 11 20:50:45 Even though dbus-launch exists, seems to run well, and dbus-daemon + friends are running Dec 11 20:50:48 Wait... Dec 11 20:51:44 I got that error message, yes Dec 11 20:51:49 With dbus 0.92 Dec 11 20:52:20 I moved up to 1.0.1 to see if that fixed anything, but nothing Dec 11 20:52:30 Instaid, I got a different error Dec 11 20:53:15 env doesn't give me the required thingies Dec 11 20:53:26 As per the bug report Dec 11 20:54:40 !oebug 1567 Dec 11 20:54:41 * * Bug 1567, Status: NEW, Created: 2006-11-10 06:39 Dec 11 20:54:42 Would exporting the variables manually work for now? Dec 11 20:54:43 * * koen(AT)dominion.kabel.utwente.nl: Dbus fails on waitpid() in /etc/X11/Xsession Dec 11 20:54:44 * * http://bugs.openembedded.org/show_bug.cgi?id=1567 Dec 11 20:54:46 hi all ... can someone point me to a doc about adding some files in a final image ??? touching only image-xxx.bb file ... are some config files Dec 11 20:54:52 Yes, I saw that Dec 11 20:54:53 cu Dec 11 20:55:43 koen: Any leads on the bug? Dec 11 20:55:55 B_Lizzard: could you add your experiences to that bug? Dec 11 20:55:59 Aye Dec 11 20:56:05 B_Lizzard: no, as not all my machines have it Dec 11 21:02:49 Added comments Dec 11 21:05:12 thanks Dec 11 21:05:39 No Problem, I want to see the bug fixed as much as you do Dec 11 21:07:38 As a note, I do not seem to have a /etc/X11/Xsession file Dec 11 21:08:40 I have a /etc/X11/Xsession.d/69gconfd-dbus, which is a shell script that runs /usr/libexec/gconfd-2 & Dec 11 21:10:18 This should be a problem with gconf-dbus or gnome-vfs-dbus Dec 11 21:10:34 Since those are the only packages that get rebuilt every day Dec 11 21:10:40 And can introduce problems Dec 11 21:10:48 no, it's a dbus problem Dec 11 21:11:04 Then why did it pop up just now? Dec 11 21:11:10 I have the srcdate locked down for those Dec 11 21:11:18 Ah, ok Dec 11 21:11:32 have a look at https://bugs.freedesktop.org/show_bug.cgi?id=7675 Dec 11 21:11:38 dbus is 'magic' Dec 11 21:12:29 That bug's been there since August Dec 11 21:12:29 :| Dec 11 21:15:01 03freyther 07org.oe.dev * rf9592acb... 10/ (1 contrib/site-conf/aclocal.m4): Use aclocal -I m4 to generate the aclocal.m4. Somehow autoreconf refuses to do this Dec 11 21:15:07 03freyther 07org.oe.dev * rdc5138de... 10/ (4 files in 3 dirs): Dec 11 21:15:07 Add gnu.m4 which has a copy of the libiberty calls from binutils.m4 Dec 11 21:15:07 Make sure we can regenerate the values for binutils which Dec 11 21:15:07 koen just added to OpenEmbedded. Dec 11 21:16:27 If you need me for anything in order to fix the bug, tell me Dec 11 21:16:35 And I can probably deliver Dec 11 21:16:38 I need patches :) Dec 11 21:17:12 If the bug is in mainline, well... Dec 11 21:19:26 If we export the variables manually, would they work? Dec 11 21:19:44 Scratch that Dec 11 21:19:51 no, since dbus-launch fails to launch Dec 11 21:20:08 Yeah Dec 11 21:20:29 ok for me to just jump in with questions ? Dec 11 21:20:34 I'll check dbus mainline, see if they have any similar problems Dec 11 21:23:32 nite Dec 11 21:23:59 nite Dec 11 21:24:26 RP, ping Dec 11 21:25:43 dfoley: sure Dec 11 21:27:02 armv4t eabi angstrom and uclibc doesn't work for me Dec 11 21:27:02 is it known to build ? Dec 11 21:27:20 dfoley: you need uclibc svn for that Dec 11 21:27:35 dfoley: and somehow build the kernel before uclibc Dec 11 21:27:47 dfoley: my masterplan is to wait for the .29 release Dec 11 21:28:08 right i was doing that and saw that. I can wait for .29 Dec 11 21:28:19 what I really really want is a thumb system Dec 11 21:29:40 a pure thumb system? Dec 11 21:29:51 or the eabi "we might use thumb" system Dec 11 21:30:05 well I guess the kernel can be arm, but yes. Dec 11 21:30:47 a pure thumb system would be very neat Dec 11 21:31:02 yeah 16bit rocks :) Dec 11 21:31:48 I actully use linux as a first boot instead of a bootloader. Dec 11 21:31:48 well I still have a bootloader it just doesn't do much. Dec 11 21:31:48 so i have a 2M linux on nor flash that loads another linux for development. Dec 11 21:31:48 a thumb system would hopefully be quite a bit smaller. Dec 11 21:32:24 iirc the openslug people said it was 70% of the size of a regular system Dec 11 21:32:25 off topic. anyone use wvdial or alternative ? Dec 11 21:32:42 since you still need regular instructions Dec 11 21:32:46 70% would be good Dec 11 21:54:04 can anyone think of an oe package that uses autotools and builds a python extension? Dec 11 21:54:39 I'm having trouble building enigma2 on the dreambox branch on a 64 bit userspace and thinking another package might help Dec 11 21:58:08 Good night Dec 11 21:59:19 thaytan: gnome-python might use it Dec 11 22:00:32 thaytan: http://ewi546.ewi.utwente.nl/tmp/hx/python-pygnome_2.16.2.bb Dec 11 22:01:38 python-pygnome itself has some problems, which is why that recipe isn't in the repository yet Dec 11 22:01:49 oh, pygtk would mebbe Dec 11 22:16:15 sweet Dec 11 22:16:18 koen: good call Dec 11 22:16:31 slurping the acinclude.m4 in for pygtk2 fixed it Dec 11 22:16:51 that keep me busy for 3 days Dec 11 22:17:01 glad you figured it out in less time :) Dec 11 22:17:07 kept* Dec 11 22:17:13 me too! Dec 11 22:20:04 keep that acinclude.m4 handy - you'll need it again Dec 11 22:21:49 :) Dec 11 22:33:55 good nite all Dec 11 22:36:36 gnight Dec 11 22:40:13 I am stuck at making a bb file Dec 11 22:40:31 I am getting a cross compiler badness :( Dec 11 22:53:38 kergoth: ping **** ENDING LOGGING AT Tue Dec 12 02:59:57 2006