**** BEGIN LOGGING AT Thu Jan 11 02:59:58 2007 Jan 11 05:48:13 Morning Jan 11 06:21:39 03pfalcon 07org.oe.dev * re3e82037... 10/ (1 packages/qte/qte-mt_2.3.10.bb packages/qte/qte_2.3.10.bb): qte* 2.3.10: Bump PR for latest changes. Jan 11 06:21:43 03pfalcon 07org.oe.dev * r570b67d0... 10/ (1 packages/opie-pics/opie-pics.inc): opie-pics: Use @base_conditional() instead of doubel indirection for GUI_MACHINE_CLASS testing. Jan 11 06:21:48 03pfalcon 07org.oe.dev * r6aa011cd... 10/ (1 packages/gpe-bootsplash/files/angstrom/splash-p.svg): Jan 11 06:21:48 gpe-bootsplash: Add missing portrait bootsplash for Angstrom. Jan 11 06:21:48 * Based on splash-l.svg plus GPE logo (mostly to make it "portrait" orientation Jan 11 06:21:48 so gpe-bootsplash render it correctly). Jan 11 06:21:48 * Yeah, it's understood it's all temporary, etc. But let's not sure users Jan 11 06:21:50 something badly rendered/inconsistent anyway. Jan 11 06:21:52 03pfalcon 07org.oe.dev * r1c580900... 10/ (1 packages/gpe-bootsplash/files/angstrom/splash-l.svg): Jan 11 06:21:54 gpe-bootsplash: Fix up Angstrom's splash-l.svg, move graphics a bit up, so Jan 11 06:21:56 it wasn't overwritten by psplash's progress indicator. Jan 11 06:22:00 03pfalcon 07org.oe.dev * r786a0913... 10/ (1 packages/gpe-bootsplash/gpe-bootsplash_1.15.bb): gpe-bootsplash 1.15: Bump PR for Ansgtrom bootsplash changes. Jan 11 06:22:03 03pfalcon 07org.oe.dev * r242e744e... 10/ (1 packages/opie-taskbar/opie-taskbar/opie): Jan 11 06:22:05 opie-taskbar: OPIE startup: if touchscreen device is not available, use mouse. Jan 11 06:22:07 * Tested on qemuarm vs h3900. Jan 11 06:22:09 03pfalcon 07org.oe.dev * r26996adc... 10/ (3 files in 2 dirs): opie-taskbar: Bump PR for mouse support in startup script. Jan 11 06:22:14 03pfalcon 07org.oe.dev * r27be13cf... 10/ (1 packages/opie-taskbar/opie-taskbar/opie): Jan 11 06:22:16 opie-taskbar: OPIE startup: Run OPIE on VT 3, as a workaround for psplash issues. Jan 11 06:22:18 * psplash is finished properly before starting OPIE, it just leaves VTs in weird Jan 11 06:22:20 state. It seems that actually OPIE cannot change VT to 2 after psplash, and instead Jan 11 06:22:22 starts on VT 1, with console input activated, etc (i.e. user can type over Jan 11 06:22:26 graphical framebuffer). Using VT 3 works ok. Jan 11 07:15:48 can anybody tell why this is happening http://pastebin.ca/312847 Jan 11 07:17:07 i have written the defconfig there Jan 11 07:20:48 03hrw 07org.oe.dev * r207db565... 10/ (1 packages/perl/perl-5.8.7/config.sh-x86_64-linux): perl: cleaned config.sh-x86-64 from paths of my build Jan 11 08:21:30 hi all Jan 11 08:22:17 i should hnow if the kernel whicth used by angstrom distro is build by bitbake or just download an image ? Jan 11 08:32:18 morning Jan 11 08:33:06 ~lart apt for general crappiness Jan 11 08:33:07 * ibot readies the nuke launcher and fires some rounds at apt for general crappiness Jan 11 08:43:05 XorA: Hi! Jan 11 08:43:59 XorA: yesterday I tried the method you said. Jan 11 08:44:14 No success:-( Jan 11 08:45:11 sza2: Im puzzled something insane is happening that I cant figure Jan 11 08:45:15 XorA: After I made an update there was a big recompilation. Jan 11 08:45:38 XorA: even the cross-compiling tools Jan 11 08:46:01 XorA: then I start with a clean install Jan 11 08:46:26 XorA: download OE.mtn, pull, etc. Jan 11 08:47:00 XorA: I set up local.conf Jan 11 08:47:29 XorA: and there result is same with a fresh install. Jan 11 08:48:36 XorA: is it possible that something wrong on the host system? Jan 11 08:50:54 sza2: OE is supposed to protect you from the host system in 90% of cases Jan 11 08:51:17 sza2: the rest of the cases show up in compile failures, not by building the wrong packages Jan 11 08:52:01 XorA: I thougt the same, however I'm a newbie in this business. Jan 11 08:53:34 XorA: maybe I post the message earlier, but: Jan 11 08:53:48 XorA: checking for GCONF_DBUS... configure: error: You need either ORBit or D-BUS for IPC Jan 11 08:54:00 XorA: FATAL: oe_runconf failed Jan 11 08:55:05 sza2: what gconf is it trying to build Jan 11 08:55:28 XorA: gconf-osso-2.6.4-3.1osso13-r3 Jan 11 08:55:53 sza2: you see, it should have built gconf-dbus as thats what DISTRO = "generic" asks for Jan 11 08:56:32 XorA : I get this error http://pastebin.ca/312847 Jan 11 08:57:37 i have written the defconfig http://pastebin.ca/312883 and have copied it there Jan 11 08:57:37 anushsh: read the log it tells you Jan 11 08:57:50 XorA: Ok, I' double check, but I think it is built already. Jan 11 08:58:44 Xora : The log is defconfig file not fouund. How can it be when I already have the defconfig there Jan 11 08:58:46 XorA: And gconf-dbus is not built automatically as a dependency? Jan 11 09:00:56 anushsh: examine carefully linux-rp_*.bb and linux-rp.inc and make sure you do the right things Jan 11 09:01:15 ok Jan 11 09:04:55 will the FIC phone be avail to buy at FOSDM? Jan 11 09:12:59 topic Jan 11 09:37:17 03koen 07org.oe.dev * r9bf06067... 10/ (1 packages/esound/esound-gpe_20070109.bb): esound gpe: use revision instead of date, fix path Jan 11 09:46:52 good morning Jan 11 09:47:03 good morning Jan 11 09:49:58 Bonjour Jan 11 09:50:44 I just updated OE, and xserver-kdrive fails on compilation Jan 11 09:50:57 Undoubtly by an update to the .bb file Jan 11 09:52:02 Might thou be interested in my output? Jan 11 09:52:19 Maybe someone can help me solve the issue Jan 11 09:52:35 Here: http://pastebin.ca/312916 Jan 11 09:56:39 hrw|gone: I saw your entry about the linking to .la files problem being solved, but was unable to test it as I've left it building and glibc has been trying to generate the es_* locales since tuesday. I'll try today to disable binary locale generation and see if it finally builds Jan 11 10:08:04 re Jan 11 10:08:43 hrw|work: do you expect to be at FOSDM? Jan 11 10:08:43 hi Jan 11 10:08:43 hi hrw|work Jan 11 10:09:06 niv_one_three: I will be Jan 11 10:11:09 good morning Jan 11 10:11:54 * Bernardo would love to have mythfrontend on his simpad... Jan 11 10:12:12 me too :-( Jan 11 10:15:47 niv_one_three: then we would have to find a 802.11g pcmcia card with memory, probably... Jan 11 10:17:40 802.11g wouldnt be supported by the pcmcia bandwidth Jan 11 10:22:30 are you sure? there are a couple 802.11g cf cards out there Jan 11 10:26:04 they are CF type II Jan 11 10:26:16 CF -> pcmcia are only type I Jan 11 10:26:51 hi zecke Jan 11 10:27:24 niv_one_three: do you know what differencies are between CF I and CF II? Jan 11 10:28:28 * koen looks at the cf II -> pcmcia converter in his hands Jan 11 10:28:49 16 bit vs. 32 bit ? Jan 11 10:28:55 * koen looks at the cf II -> pcmcia converter in his netbook pro Jan 11 10:29:04 s/his/lrgs/ Jan 11 10:29:07 well , I admit I just googles for a week on this Jan 11 10:29:20 the answer is 'about 2 mm' Jan 11 10:29:39 and finnaly understand that: pc card =! pcmcia Jan 11 10:30:11 someone want cf II -> pcmcia converter? I got one and it does not accept CF I type cards ;( Jan 11 10:30:23 and all 802.11g chipsets "talk" 32 bit Jan 11 10:30:23 that was what I thought, that CF I and CF II difered on size only Jan 11 10:30:37 32bit PCMCIA is called cardbus and is actually hotplug PCI Jan 11 10:30:46 hrw|work: I can send you one of my sandisk adapters Jan 11 10:30:49 hrw|work: I suspect that the postage fee would be more that the price of a new one Jan 11 10:31:11 * XorA has the other way, an adapter to plug PCMCIA into zaurus Jan 11 10:31:14 I'll have to check, but I think the sandisk plus is CFI Jan 11 10:31:25 http://fotokonijnenberg.nl/product_details.php?id_product=16296 Jan 11 10:31:51 Bernardo: snadisk Plus IS CF type I Jan 11 10:31:51 koen: no, I can take it to fosdem and give away for someone Jan 11 10:32:32 hrw|work: if you need one CFI -> PCMCIA adapter, just say Jan 11 11:04:30 Hm, you've done something funny to the kernel... make-kpkg complains that "The upstream version .. does not contain a digit" Jan 11 11:06:46 ~lart apt for being totally stupid Jan 11 11:06:47 * ibot accelerates a free AOL cd to 50,000 rpm and lets apt feel it for being totally stupid Jan 11 11:06:50 Was able to fix that by altering the debian/changelog file generated by make-kpkg Jan 11 11:07:31 yo zecke Jan 11 11:16:37 There seem to be some missing config dependencies... At least APM and Corgi BL are possible to turn into modules, generating unresolved symbols from the build process. Jan 11 11:24:20 DataBeaver: those things are not written to be in modules Jan 11 11:26:06 how do i fix this http://pastebin.ca/312970 Jan 11 11:26:10 Shouldn't they be bool instead of tristate then? Jan 11 11:27:46 DataBeaver: patches welcome Jan 11 11:28:06 anushsh: paste complete logs Jan 11 11:30:31 On another note, I have all the OE patches now but am getting the same error ("Slave PCM not usable") as yesterday. Jan 11 11:30:51 Don't have RULES_DEBUG enabled so I can't be 100% sure if it's exactly the same problem. Jan 11 11:31:27 03florian 07org.oe.dev * r2537e8f4... 10/ (1 packages/gpe-conf/gpe-conf_0.2.3.bb): gpe-conf: Add new release 0.2.3 Jan 11 11:31:31 03florian 07org.oe.dev * r960d0a30... 10/ (1 packages/blueprobe/blueprobe_0.18.bb): blueprobe: Add 0.18 Jan 11 11:33:22 Any ideas what to do now or shall I just muck around in pcm_native.c again until it works? Jan 11 11:33:55 DataBeaver: I am using CONFIG_BACKLIGHT_CORGI=m since quite a while without problems. Jan 11 11:34:06 florian_kc: MAINTAINER is forbidden in .bbs nowadays Jan 11 11:35:51 pH5: I'm getting arch/arm/mach-pxa/built-in.o:(.data+0x228c): undefined reference to `corgibl_limit_intensity' Jan 11 11:37:29 hrw|work : that was the complete log Jan 11 11:38:16 it breaks down at kernel_do_compile Jan 11 11:38:41 DataBeaver: ah, so it has to be compiled in on zauruses because corgi/spitz_pm reference it but are always compiled in. CONFIG_PXA_SHARP_C7xx and CONFIG_PXA_SHARP_Cxx00 should depend on BACKLIGHT_CORGI then Jan 11 11:40:01 florian_kc: could you remove that from the .bbs please and add it to your entry in MAINTAINERS? Jan 11 11:44:02 koen: ah right, i forgot Jan 11 11:44:09 morning all Jan 11 11:45:23 hi mickeyl Jan 11 11:45:55 hi ade|desk Jan 11 11:46:16 hi mickeyl Jan 11 11:47:11 koen: fixed Jan 11 11:48:16 brb, rebooting Jan 11 11:49:27 Hm, this time it's 4 -> 0 so it might not be any quantum effect... Jan 11 11:49:59 Need to debug this a bit... Jan 11 11:52:41 Okay, mask = 0004, m = ffff and this results in 0000 ? Jan 11 11:53:44 Not good. Jan 11 11:55:49 hey pb_ Jan 11 11:55:51 morning guys Jan 11 11:55:59 hey mickeyl Jan 11 11:58:55 hi mickeyl Jan 11 11:59:00 hi mickeyl Jan 11 11:59:01 hey anushsh Jan 11 11:59:11 anushsh: got any further with your kernel? Jan 11 11:59:13 yo Bernardo, long time no see. what are you up to? Jan 11 11:59:40 me? nothing much, lots of work lately Jan 11 11:59:45 mickeyl : this is the new http://pastebin.ca/312970 :p Jan 11 11:59:46 sounds familiar to me Jan 11 11:59:56 lol Jan 11 12:00:14 d Jan 11 12:00:16 d'oh Jan 11 12:00:30 anushsh: you sure S points to the right directory? Jan 11 12:00:38 S = "${WORKDIR}/path-to-my-linux-kernel-dir" Jan 11 12:00:50 and got a reprimand from my boss for too much internet use Jan 11 12:00:56 really? sucks Jan 11 12:01:02 mickeyl : lemme check Jan 11 12:01:02 bosses suck Jan 11 12:01:06 that's why I'm a freelancer ;) Jan 11 12:01:09 banks suck Jan 11 12:01:13 even more so Jan 11 12:01:14 ya Jan 11 12:01:15 Bernardo: you slowed down his pr0n downloads? Jan 11 12:01:19 hehe Jan 11 12:01:23 lol Jan 11 12:01:47 mickeyl: hurry up and become a billionaire so you can hire us all Jan 11 12:01:54 working on it :D Jan 11 12:02:09 alas, my stocks are falling though Jan 11 12:02:10 heh Jan 11 12:02:31 my stocks are worth -ive :-( Jan 11 12:02:37 looks like the only thing to make money is to actually work *shrug* Jan 11 12:02:57 (for me, that is) Jan 11 12:08:17 hrw|work, mickeyl: http://www.openembedded.org/~koen/index.php?path=amare/ Jan 11 12:08:25 if I could afford it, I'd have left one week after I got hired Jan 11 12:09:14 koen: looks cool to me Jan 11 12:09:48 koen: nice but Amare had search input entry in header which should be available Jan 11 12:10:08 hrw|work: right, I guess I didn't enable that Jan 11 12:10:24 yeah php. do we have enable_global_foo on? Jan 11 12:10:43 hrw|work: just a quick wget, untar, select Jan 11 12:10:52 is that a fixed width theme? Jan 11 12:10:56 mickeyl: but when I have some time I still try to build stuff using OE... and bug hrw when I can't get it to work (sorry hrw, and thanks for the patience) Jan 11 12:11:22 cool. keep us posted Jan 11 12:13:38 hi all Jan 11 12:14:25 when i build my system with angstrom, bitbake rebuild an kernel or just download an image ? Jan 11 12:15:08 because i have started sthe system ine the hard disk ide but the system don't start in my compact flash emulate in ATA Jan 11 12:15:36 i think that the kernel have not the module to start the ATA Jan 11 12:23:09 koen: website looks nice on shots Jan 11 12:23:14 mickeyl : do i need to downlaod the source and give the path to it in S ? Jan 11 12:24:11 no idea ? Jan 11 12:25:10 disaster: configure kernel, enable proper ATA driver, build, change kernel, boot Jan 11 12:30:26 hrw|work, i know compil a kernel in standart distro, it' the same operation for oe ? Jan 11 12:30:57 in my debian source , i download the kernel source, makemenuconfig ... Jan 11 12:31:09 debian box Jan 11 12:34:18 bbiab Jan 11 12:37:28 what is bbiab ? Jan 11 12:37:44 be back in a bit Jan 11 12:38:17 kk Jan 11 12:38:50 Crofton|home, you know how recompil a kernel for the oe ? Jan 11 12:39:08 bitbake -c clean linux; bitbake linux Jan 11 12:39:14 or bitbake -c rebuild linux Jan 11 12:39:15 is the standart operation or different for oe ? Jan 11 12:39:43 but where i must edit the option kernel ? Jan 11 12:40:46 hmmm Jan 11 12:41:29 You want to build a kernel with your own config Jan 11 12:42:31 yes i must enabled the ATA driver Jan 11 12:42:53 you know whcih kernel you are building in oe? Jan 11 12:43:26 which bb file? Jan 11 12:46:14 yes Jan 11 12:46:27 kernel-epia-2.6.12 Jan 11 12:47:10 ia have found files conf ine org.openembedded.dev/packages/linux/kernel-epia-2.6.12 Jan 11 12:47:52 but in ATA/ATAPI section Jan 11 12:48:00 no line for the ATA Jan 11 12:48:26 you knows what is the right line to add to use a ATA Jan 11 12:49:53 Can anyone help me with Xorg and the touchscreen? Jan 11 12:51:52 Got a few problems with it: (1) it's awfully noisy, (2) can't seem to generate clicks and (3) how to calibrate it Jan 11 12:53:11 use kdrive Jan 11 12:54:28 No way to make it work with regular Xorg? Jan 11 12:54:38 DataBeaver: use the source, luke Jan 11 12:57:01 hmh... Jan 11 13:02:24 Is it possible to incorporate the required changes into an input driver for a regular Xorg server? Jan 11 13:03:05 disaster, you make xconfig in the directory where the kerenl was built and fix the config options, save the new config Jan 11 13:03:44 the copy .config over the appropraite file in the epia directory and rebuild the kernel Jan 11 13:04:18 if this is something of braod interest to epia users, find the maintainer and ask them to make the change in the metadata Jan 11 13:13:56 crofton : how be sure of a number of distribution Jan 11 13:14:02 ? Jan 11 13:15:36 not sure what you mean? Jan 11 13:17:56 Crofton|home : our disro is epia, Mean you that a kernel epia? Jan 11 13:18:21 Crofton|home : what package can install for xconfig? Jan 11 13:19:43 koen: kdrive seems to address at least the calibration issue and the inability to click may be due to the noise, but it seems that the noise is coming from the kernel side. Any patch for that? I've seen lcd noise patches for tosa, but not for spitz. Jan 11 13:22:15 I determined this by doing cat /dev/input/mouse0 and holding the stylus at one point on the screen. This resulted in lots of spam from the device. Jan 11 13:24:13 DataBeaver: learn about tslib Jan 11 13:25:10 k, I'll check that Jan 11 13:30:08 the kernel is specified in the machine file (I think ..) Jan 11 13:47:13 koen: added Pingwinaria into events Jan 11 13:49:43 hrw|work: I'll add bossa as soon as I get confirmation Jan 11 13:50:32 koen, you have any idea if putting a usb2.0 card in EFIKA really works Jan 11 13:50:37 it sort of works ... Jan 11 13:50:50 but I don't know if the data is crap Jan 11 13:52:21 Crofton|home: it should Jan 11 13:52:31 Crofton|home: but remember that it's a bigendian cpu Jan 11 13:52:36 yeah Jan 11 13:52:47 I wonder how that screws things up Jan 11 13:57:04 Hmph, why is tslib only supported in kdrive and not the normal x server... Jan 11 13:57:31 DataBeaver: because you have not yet submited your patch for Xorg? Jan 11 13:57:38 or XSun, whatever is normal x server Jan 11 13:57:48 Xorg in my case Jan 11 13:57:58 re Jan 11 14:00:35 ~hail wireshark Jan 11 14:00:37 * ibot bows down to wireshark and chants, "I'M NOT WORTHY!!" Jan 11 14:01:24 XorA: can you get it to work properly? I was never able to use it decently with my rt2500 card Jan 11 14:02:55 Bernardo: seems to work on cards in my Dell desktop/laptop Jan 11 14:03:30 koen: does utwente has ircnet server? Jan 11 14:03:47 I can only get decent captures on wired cards, not on 802.11 cards Jan 11 14:04:12 Bernardo: wireless cards cant enter promiscuous mode Jan 11 14:04:36 hrw|work: yes, irc.snt.utwente.nl Jan 11 14:04:45 thx Jan 11 14:05:14 XorA: I know, but even so I had problems... probably PEBCAK Jan 11 14:08:25 ha the system start good on my compact flash Jan 11 14:08:44 the problem was the timout with dma Jan 11 14:09:10 just add the param in the kernel boot Jan 11 14:09:10 ide=nodma Jan 11 14:09:26 but now to add vlc fo example, how i do ? Jan 11 14:10:50 bitbake vlc, find ipk and copy to machine, ipkg install vlc maybe? Jan 11 14:10:59 very roughly Jan 11 14:15:10 thx Jan 11 14:17:14 thw a lot Jan 11 14:17:18 thw*x a lot Jan 11 14:17:38 disaster, hopefully things are starting to make sense ... Jan 11 14:18:48 okey Jan 11 14:18:51 musty go Jan 11 14:18:53 bye Jan 11 14:25:36 ~seen likewise Jan 11 14:26:08 likewise was last seen on IRC in channel #oe, 2d 17h 56m 55s ago, saying: 'cu guys'. Jan 11 14:34:02 mickeyl: make OpenMoko INC. create a ridicoulous NDA/License as well. E.g. "If any intellectual property is transfered all copies of the source has to be returned to Luke Skywalker" Jan 11 14:38:02 any cheap Pentium M based development board to recommend for a mobile application? Jan 11 14:45:22 ls -la Jan 11 14:45:26 oops Jan 11 14:45:28 :( Jan 11 14:46:17 wow Pentium M ... thats some mobile application Jan 11 14:56:13 I'm planning on setting up a demonstration of OE for local companies in my area (Pittsburgh, PA, United States). Other than the wealth of informations and presentations on the OE wiki is there any other presentations or demos any one would recommend for me to reference while I'm putting my material together? Jan 11 14:56:32 I based on those one Jan 11 14:57:49 well, it's for my next robotics project. I want to have lots of power for opencv stuff. Jan 11 14:58:01 I got it to compile under OE Jan 11 15:02:22 heh, i'd transport the data from the local robot to a bigger system that is fixed and heavy duty, making the robot lighter and cheaper to make Jan 11 15:02:25 Gerrath: that wiki should list every presentation we know off Jan 11 15:05:45 Gerrath, good uck, we need more oe'rs in this time zone :) Jan 11 15:07:51 Hm, need to figure out where to get the xkb layout for my Z... Jan 11 15:08:05 * cbrake is an EST OE'er :-) Jan 11 15:08:21 any svg experts in da house? Jan 11 15:08:42 is there a tool to remove proprietary chunks? Jan 11 15:08:42 * koen only gets as far as "OMG vector!!" Jan 11 15:08:42 (scriptable) Jan 11 15:08:54 mickeyl: sed Jan 11 15:09:00 hmm Jan 11 15:09:03 that goes over my caps Jan 11 15:09:09 anything readymade? Jan 11 15:09:26 mickeyl: peruse libxml2/expat/gmarkup? Jan 11 15:09:42 i thought more along the lines of "svgclean *.svg" :) Jan 11 15:09:55 i can code this, sure Jan 11 15:10:03 but i'd like not if i don't need to Jan 11 15:10:13 1) code it 2).... 3) profit Jan 11 15:12:03 mickeyl: SVG is XML so python-xml? Jan 11 15:12:32 hmm i rather tell the designer to save everything again in adobe illustrator than to waste an hour coding that :/ Jan 11 15:13:24 inkscape Jan 11 15:15:06 Shouldn't a generic SVG parser just skip those? Jan 11 15:17:00 mickeyl: I'm taking a look at adobe illustrator... Jan 11 15:17:44 DataBeaver: it's not about skipping (yes, that works), it's about me being reluctant adding lots of icons each worth of 5kb of data being 190Kb to SVN Jan 11 15:17:45 mickeyl: ask him to export into TinySVG 1.1? Jan 11 15:17:59 mickeyl: to embed fonts Jan 11 15:18:01 ah. Jan 11 15:18:15 mickeyl: or to convert fonts... Jan 11 15:18:30 zecke: will do, thanks Jan 11 15:18:57 no promise though Jan 11 15:19:56 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND Jan 11 15:19:57 23899 hrw 25 0 1101m 984m 8976 R 99.1 77.8 35:23.05 firefox-bin Jan 11 15:20:21 sure Jan 11 15:20:21 someone still want firefox on embedded? Jan 11 15:20:32 no, thanks :) Jan 11 15:20:44 my bets are on gdk-webkit atm. Jan 11 15:20:48 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND Jan 11 15:20:48 1778 tdb 15 0 57588 28m 14m S 0.0 46.7 0:53.05 firefox-bin Jan 11 15:20:52 I thought that had died? Jan 11 15:20:54 yeah :) Jan 11 15:21:09 (That's running on my Z) Jan 11 15:22:00 mickeyl: yeah :) WebKit rocks! iPhone is using it as well Jan 11 15:22:19 hrw|work: Is it possible to rotate a qpf font without having access to the original ttf font (which I cannot find anywhere on the net)? Jan 11 15:22:38 Laibsch: rather not Jan 11 15:22:44 Laibsch: qpf is one-way converting Jan 11 15:22:52 I see. Jan 11 15:23:01 Laibsch: unismall? Jan 11 15:23:42 yes, of course. Jan 11 15:23:50 I cannot find the original ttf Jan 11 15:23:56 Even from the zaurus-ja page. Jan 11 15:24:23 mickeyl: I've been looking at the buildsystem for webkit (javascriptcore/webcore/webkit) and I haven't figured it out yet Jan 11 15:24:51 hrw|work: Might have got it: http://prdownloads.sourceforge.jp/zaurus-ja/773/unismall-1.0.0.tar.gz Jan 11 15:24:58 koen: there is no code to build the gdk stuff Jan 11 15:25:15 zecke: there are patches in bugzilla Jan 11 15:25:24 koen: we still need a browser for openmoko and that Alp Toker guy seemed to have success in getting it to compile Jan 11 15:25:33 I have a GtkAligments that embeds WebKit Jan 11 15:25:34 zecke: but the docs only mention xcode, while cmake also seems present Jan 11 15:25:42 koen: and qmake Jan 11 15:25:47 mickeyl: btw, Alp Toker is also a pdf/cairo god Jan 11 15:25:52 ah Jan 11 15:25:55 so he knows his stuff Jan 11 15:26:09 zecke: needing Qt? Jan 11 15:26:14 mickeyl: yes ;) Jan 11 15:26:21 mmh Jan 11 15:26:53 Laibsch: its qpf inside Jan 11 15:27:41 damn Jan 11 15:27:50 mickeyl: http://lists.freedesktop.org/archives/cairo/2007-January/009059.html Jan 11 15:28:13 hrw|work: But there is a README Jan 11 15:28:20 and it points to http://hp.vector.co.jp/authors/VA013391/fonts/ Jan 11 15:28:21 Laibsch: in japanese Jan 11 15:28:25 koen: nice Jan 11 15:29:45 hrw|work: Yes, no problem. I think the author of the font is HP. I will verify distribution terms etc. Jan 11 15:30:00 naga10 is where unismall was derived from Jan 11 15:30:27 Laibsch: nice Jan 11 15:30:46 hrw|work: bdf files should be OK as well to create the qpf files, right? Jan 11 15:30:59 huhu Jan 11 15:31:18 Laibsch: iirc you should be able to generate them Jan 11 15:32:47 morning Jan 11 15:33:33 hi chouimat Jan 11 15:38:12 hey Bernardo, have you chance to try the subtitles? Jan 11 15:39:10 have you had that is Jan 11 15:39:17 * Jin^eLD is still not quite awake Jan 11 15:41:43 hi Jin^eLD Jan 11 15:41:53 not yet - got my media center pc last night Jan 11 15:41:55 :) Jan 11 15:42:19 but I'll be able to check this weekend Jan 11 15:43:20 aah, so DSM is going to be sold I guess? :) Jan 11 15:43:24 also, the last time I checked (before this last change) ff and rev no longer gave "invalid" when I pressed them on the remote, but didn't seem to work. Maybe now with the new manufacturer url they'll work Jan 11 15:43:47 hmm, I have not yet checked that Jan 11 15:43:48 I'm sure it will... With box, remote, manual, the works. Jan 11 15:44:09 I need to find some time to do a more extensive test with the DSM that we have at work Jan 11 15:44:55 I had hopes of keeping it until finding out that the main apps were closed source, and there was no documentation on how to get video out of it. That made me give up fast on the idea of adding it as a machine in OE Jan 11 15:45:35 well Jan 11 15:46:06 actually we would just need the specs of the chipset Jan 11 15:46:16 to control the decoders and stuff Jan 11 15:46:31 I am not sure, but D-Link did post the kernel? Jan 11 15:48:39 they posted the kernel and their patches to busybox Jan 11 15:49:04 some guys managed to extract the bootloader from the firmware image Jan 11 15:49:26 I read a post where some guy tried to do something with video, he was writing about ioctl calls Jan 11 15:49:28 so we have all the parts - and the sources posted by dlink have the script to build a image Jan 11 15:49:49 Jin^eLD: mediatomb builds again: http://www.angstrom-distribution.org/repo/?action=details&pnm=mediatomb :) Jan 11 15:50:15 cool :) Jan 11 15:51:03 koen: the NSLU2 people are not using OE, right? are there any reasons for it? Jan 11 15:51:58 Jin^eLD: they use OE Jan 11 15:52:07 oh, they do? Jan 11 15:52:19 I looked at unslug once I think Jan 11 15:52:22 and it looked very different Jan 11 15:52:40 Jin^eLD: they have a wrapper around OE ... Jan 11 15:52:52 uh.. what is that good or? Jan 11 15:52:56 s/or/for/ Jan 11 15:53:04 well Jan 11 15:53:10 Jin^eLD: you type make ;) Jan 11 15:53:19 ah, ok :) I get the idea Jan 11 15:54:03 I gotta check their channel, would be good to have a package of my server for the NSLU2 distro Jan 11 15:57:43 hrw|work: Just checked the README and the naga10 font is (as expected) free for non-commercial redistribution as long as the original README (in Japanese) is included. Jan 11 15:57:46 koen, Crofton|home thanks. Jan 11 15:58:05 hrw|work: free for redistribution in original and dervived form. Jan 11 16:01:01 crap Jan 11 16:01:05 did I miss anything? :> Jan 11 16:11:09 hrw|work: Do you have make-qpf for x86? Jan 11 16:12:20 Laibsch: removed some time ago but w8 - will check here Jan 11 16:12:30 Laibsch: have Jan 11 16:12:34 Nice. Jan 11 16:13:15 Laibsch: will put on web so you can get it Jan 11 16:13:20 Great Jan 11 16:14:22 Laibsch: http://ewi546.ewi.utwente.nl/tmp/hrw/t/makeqpf-x86.tar.bz2 Jan 11 16:14:32 Laibsch: 33ab6dc5933b434161e8182cd9cd8163 is md5 Jan 11 16:15:24 03florian 07org.oe.dev * r7ba28832... 10/ (1 packages/gpe-conf/gpe-conf_0.2.3.bb): gpe-conf: Remove obsolete MAINTAINERS field Jan 11 16:15:31 03florian 07org.oe.dev * r78963df9... 10/ (3 files in 3 dirs): libmimedir: Add patch by Jan Arne Petersen to support parsing of Google Calendar files. Jan 11 16:15:55 florian_kc: can you update gpe-conf in .oz354x? Jan 11 16:17:19 Is there an xkb rules file for spitz? Jan 11 16:17:26 DataBeaver: no Jan 11 16:17:35 DataBeaver: we do not use xkb Jan 11 16:17:37 hrw|work: i think so :-) Jan 11 16:17:39 iirc Jan 11 16:17:42 k... Jan 11 16:17:46 florian_kc: would be great Jan 11 16:18:10 hrw|work: Got it. Thanks. Jan 11 16:18:31 we should go through thh the gpe related packages and check for updates anyway. Jan 11 16:18:48 florian_kc: see my RFC :) Jan 11 16:19:20 Laibsch: np. inside you will find 'dofonts.sh' script - its useful Jan 11 16:19:41 was just looking at it. Jan 11 16:19:43 florian_kc: http://lists.linuxtogo.org/pipermail/gpe-list/2007-January/000160.html Jan 11 16:19:47 Seems very useful indeed. Jan 11 16:20:11 Hm, I'll need to look at getting the OE kdrive xserver built as a Debian package... Jan 11 16:20:14 Laibsch: you need to run qvfb first Jan 11 16:20:21 hrw|work: But the script does not seem to accept any command line switches? Jan 11 16:20:51 Laibsch: it does not need any Jan 11 16:21:27 koen: right.. i read this already but an useful answer consists of more than a single line... Jan 11 16:21:34 Laibsch: makeqpf parse QTDIR/lib/fonts/fontdir for fonts and generate binaries for them in one rotation Jan 11 16:22:42 oh right, libmimedir wants an update as well. Jan 11 16:25:06 it seems like I can bring about 175 euros of "gifts" to FOSDEM Jan 11 16:25:28 and likewise wants me to give him a few picoPSUs Jan 11 16:25:53 I suppose I should use up my entire "gift" allowance Jan 11 16:32:58 03koen 07org.oe.dev * r17ce0a15... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: prefer gtk 2.10.7 Jan 11 16:38:26 hrw|work: Do I absolutely need to have a GUI? My OE work is usually done on a headless machine but I can ssh into X. Jan 11 16:39:34 Laibsch: I do many^Wmost of builds on headless machines Jan 11 16:39:52 Laibsch: but makeqpf need access to framebuffer Jan 11 16:39:53 hrw|work: I meant for qvfb. Jan 11 16:39:58 OK Jan 11 16:40:08 and /dev/console Jan 11 16:40:32 hrw|work: Why is it "export LD_LIBRARY_PATH=`pwd`/../lib/" and not "export LD_LIBRARY_PATH=`pwd`/lib/"? Jan 11 16:40:48 Laibsch: because I have those libs in ../lib/ Jan 11 16:40:51 Do I call dofonts.sh from inside the directory where the bdf fonts are? Jan 11 16:41:08 hrw|work: I see. They are in ./ in your tar. Jan 11 16:41:13 Laibsch: bdf fonts needs to be in QTDIR/lib/fonts/ Jan 11 16:41:14 Laibsch: I know Jan 11 16:41:23 Laibsch: do not ask why I had them in ../lib/ Jan 11 16:41:32 OK, I won't ;-) Jan 11 16:46:21 koen, htcuniversal folks would like to upgrade kernel version for Angstrom to the latest. ok? Jan 11 16:47:20 koen: and this one builds Jan 11 16:48:25 Any idea when Richard will be back? Jan 11 16:48:45 no Jan 11 16:49:09 psokolovsky: have them attach a patch to the bugtracker Jan 11 16:49:33 Laibsch, I run Angstrom headless on the EFIKA Jan 11 16:49:42 psokolovsky: considering the fact they want a kernel that is not yet in OE Jan 11 16:50:39 koen, well, the version they want is 2.6.19-hh7. that's only change. they addressed to me, as person who looks after linux-handhelds-2.6 kernel ;-) Jan 11 16:50:55 and 2.6.19-hh7 is in OE Jan 11 16:51:08 "[17:38] goxboxlive: hrw|work: To have htc universal build for angstrom you could add 2.6.19-hh8 in to angstrom.conf.It where tagged today and has our lates drivers ." Jan 11 16:51:26 that's not hh7 Jan 11 16:51:32 and not in OE Jan 11 16:51:49 Packaged contents of kernel-modules into /a/home/hrw/devel/build/angstrom/tmp/deploy/glibc/htcuniversal//ipk/kernel-modules_2.6.19-hh7-r0_htcuniversal.ipk Jan 11 16:51:57 psokolovsky: We want hh8, not hh7 Jan 11 16:52:11 goxboxlive: then first hh8 needs to be added Jan 11 16:52:28 psokolovsky: that's why I told them to attach a patch to the bugtracker Jan 11 16:52:29 hrw|work: yes Jan 11 16:52:43 morning all Jan 11 16:53:03 hey RP & kergoth Jan 11 16:53:25 hi kergoth RP Jan 11 16:54:20 hey Jan 11 16:55:19 RP mornin Jan 11 16:55:44 Kristoffer: please subscribe to openembedded-devel mailinglist Jan 11 16:56:31 Kristoffer: that way you get mails like this: http://lists.linuxtogo.org/pipermail/openembedded-devel/2007-January/001228.html Jan 11 16:57:54 each devel with r/w to metadata need to be subscrbed to atleast oe-dev Jan 11 16:58:00 Im on the mailinglist, but got back from long holiday, So I've basicly done major cleaning Jan 11 16:58:09 koen thx, will push the patch tonight Jan 11 16:59:18 Btw, just for the fun of it I been working on getting armedslack to work on my spitz. Gotten everything up except proper mouse handling, it doesnt seem calibrated. Any suggestions? Would be wonderful to get this working using standard xorg but Im guessing that is out of the question Jan 11 17:03:19 http://jaaksi.blogspot.com/ Jan 11 17:09:18 hi..! does anyone here knows of a way to check whether the EABI patch is working on my system or not..? Jan 11 17:09:24 any kernel method/log/something? Jan 11 17:23:33 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Jan 11 17:28:02 I wonder what the purpose of bb files like gpe-beam is that have no SRC_URI? Jan 11 17:29:32 Laibsch: check gpe.bbclass and gnome.bbclass Jan 11 17:29:46 Dammnit Jan 11 17:30:30 I just updated OE (After a month), and I'm getting "undefined reference" problems with a lot of applications Jan 11 17:31:18 I know how to fix them, but I'm having difficulty finding the applications I have to rebuild in order to get rid of the problems Jan 11 17:31:36 For instance, udev gives me this: Jan 11 17:31:49 /usr/local/oe/tmp/work/sh3-linux/udev-100-r5/udev-100/extras/volume_id/vol_id.c:222: undefined reference to `volume_id_probe_adaptec_raid' Jan 11 17:32:30 I cannot understand which package I need to rebuild... Jan 11 17:32:58 B_Lizzard: after a month? I would rebuild from scratch Jan 11 17:33:36 I have almost 11000 packages in my deploy folder, plus a lot of bb files I made myself... Jan 11 17:33:46 big build Jan 11 17:33:50 which distro? Jan 11 17:33:54 JLime Jan 11 17:34:47 B_Lizzard: udev Jan 11 17:34:47 I could start a new tmp folder and replace the old one, but isn't that a monumental job? Jan 11 17:35:00 you prepare for release? Jan 11 17:35:07 Considering that I need OE to build images Jan 11 17:35:08 Yes Jan 11 17:35:20 B_Lizzard: preferably one that has a seperate libvolumeid package Jan 11 17:35:28 koen: It's udev that gives me that problem Jan 11 17:36:53 Well, I used udev-100 before, but this one got the "pr" bumped Jan 11 17:36:59 So I need to rebuild it Jan 11 17:37:23 100? Jan 11 17:37:30 I think that was buggy Jan 11 17:37:33 * koen still uses 0.92 Jan 11 17:37:34 Not for me Jan 11 17:37:46 It works great for me Jan 11 17:38:28 'Why it is worth to choose OE when there are other solutions? For me OE allows for quick start into embedded world without going through horror way of cross-compilation which is quite important in situation when we want to concentrate on out device development. After creating few simple config files we can build working system, use ready recipes to build needed tools (no need to search though internet for them as they are rather added into OE already) and Jan 11 17:38:43 :/ Jan 11 17:38:48 What was that? Jan 11 17:38:48 ~botmail for sirfred: Ximageon doesn't seem to work on w3220 :( Jan 11 17:38:55 hrw|work: cbrake? Jan 11 17:38:58 what do you think about such summary for OE related article? Jan 11 17:39:01 koen: me Jan 11 17:39:25 hrw|work: cbrake had a similar paragraph on his site a while back Jan 11 17:39:55 good to know Jan 11 17:40:46 this was realtime translation from Polish to English btw ;D Jan 11 17:41:18 hrw|work: http://bec-systems.com/web/content/view/57/9/ Jan 11 17:42:29 thx Jan 11 17:46:42 One more question: Does OSS exist for OE? Jan 11 17:47:10 We have a working /dev/dsp, but ALSA needs additional configuration in order to work Jan 11 17:48:39 Sound under OPIE works with the help of opie-qss, but what about X? Jan 11 17:52:30 esound Jan 11 17:52:47 Does esound use /dev/dsp natively? Jan 11 17:52:50 B_Lizzard: machines are required to provide OSS emulation Jan 11 17:53:14 Ah Jan 11 17:53:18 Thanks a lot Jan 11 17:55:09 hello all. I want to write some small custom apps to go into my rootfs, and I would like them to be integrated with OE and BitBake. What is the usual procedure for this? Ideally I would just like to point a .bb file at my source dir for the app, and have magic happen. Is sourcepkg the answer here? If so, is there some documents about it I can read? Jan 11 17:55:47 sorry, I mean "inheriting from sourcepkg" Jan 11 17:56:18 You could write a .bb file for the application and place it in org.openembedded.org/packages Jan 11 17:56:52 All the other .bb files can serve as examples, and there's some good documentation on the OE website Jan 11 17:57:00 bradbev: write recipe (.bb file), build it. if want to get it added into OE then submit it into bugtracker and pester us to add them Jan 11 17:57:38 B_Lizzard, thanks. I was under the impression that most .bb files do an unzip stage first. Are there any that start with a raw source dir that I can look at? Jan 11 17:57:53 Hmmm Jan 11 17:57:57 Haven't seen any Jan 11 17:58:03 You could "fake" it Jan 11 17:58:10 Hello Jan 11 17:58:18 hi sirfred Jan 11 17:58:20 "fake" it? :) Jan 11 17:58:31 sirfred: Ximageon does not work on w3220 (Xw100 works) Jan 11 17:58:38 By adding the source dir in a tar.gz file, and place it in your source folder Jan 11 17:58:43 hrw|work: Probably Jan 11 17:58:55 bradbev: SRC_URI = "file://source.tar.gz" can be used Jan 11 17:59:08 Ah Jan 11 17:59:13 hrw is tottaly right Jan 11 17:59:19 hrw|work: Basic acceleration should work, or could be fixed to work. Jan 11 17:59:40 hmm, I'm not making any sense here. I don't really want to do the zip or unzip stage at all. This is software that is in source form that I normally compile and just throw on the machine Jan 11 17:59:47 hrw|work: Unfortunately, xv relays in rotate/scale features that seems to be different in w3220 Jan 11 18:00:08 hrw|work: sounds good -- OE has really helped me doing embedded work. Jan 11 18:00:14 The "unzip stage" is pretty automated Jan 11 18:00:22 hrw|work: Perhaps mode setting for randr is also different. Jan 11 18:00:22 You don't actually need to do anything Jan 11 18:00:31 hrw|work: Do you know how do it fails exactly ? Jan 11 18:00:33 B_Lizzard, yes, but for me the "zip" stage isn't. Jan 11 18:01:01 cbrake: after Pingwinaria (where article will have premiere) I will translate it for English and publish on my website Jan 11 18:01:11 RP: Hello? Jan 11 18:01:11 Well, bitbake has to find some way to move the source files in it's tmp folder Jan 11 18:01:20 bradbev: you can write own do_unpack() which will copy sources from your tree Jan 11 18:01:20 hrw|work: great!\ Jan 11 18:01:30 hrw|work: hi..! is there a way by which i can check whether my eabi patch is working? Jan 11 18:01:38 sirfred: koen know more as he has device Jan 11 18:01:42 hrw|work, that sounds pretty good to me :) Jan 11 18:01:46 thanks guys Jan 11 18:02:17 shirour: no idea - if you have 2.6.16 then you should be able to boot eabi Jan 11 18:02:18 hrw|work: could it be that makeqpf takes eons? Jan 11 18:02:24 hrw|work: I think that the basic acceleration could be fixed to work on the w3220. xv, probably not. Jan 11 18:02:26 shirour: pastebin kernel config please Jan 11 18:02:37 hrw|work: We can always only enable the xv port for the w100. Jan 11 18:02:53 Laibsch: for one font? it was <30 minutes for all rotations, 20 sizes, 14 fonts Jan 11 18:02:55 hrw|work: I mean detecting it at runtime. Jan 11 18:03:25 hrw|work: ok, tough at the beginning when i only had "config_aeabi" i got errors with oabi, so i've added "config_oabi" and they disappeared, so i think the configuration is all right... Jan 11 18:04:12 sirfred: probably Jan 11 18:04:16 hrw|work: do normal executables work with eabi kernel? Jan 11 18:04:16 hrw|work: well, this is a slow machine, 4-5 fonts. Do I have to specify what size I want? I believe naga10 is always 10 point. Jan 11 18:04:28 Laibsch: naga10 is font name rather Jan 11 18:04:43 ttf-kochi-gothic-naga10 - Kochi Subst Gothic Japanese TrueType font with naga10 (non-free) Jan 11 18:05:02 Description: Kochi Subst Gothic Japanese TrueType font with naga10 (non-free) ttf-kochi-gothic-naga10 is high quality, Japanese gothic TrueType font. It includes naga10 font, so it's non-free. Jan 11 18:05:15 hmm, or I can point SRC_URI at a cvs server. Am I able to also point it at an SVN server? Jan 11 18:05:23 bradbev: yes you are Jan 11 18:05:34 woohoo - problem solved. Thanks a bunch Jan 11 18:05:36 bradbev: but user:pass is broken for svn Jan 11 18:06:10 oh, hmm. Shouldn't matter for me right now I don't think. Jan 11 18:06:51 <[cc]smart> I had write access to OE once. How can i find out if this is still the the case ? Jan 11 18:06:55 cbrake: I'm planning to start my company website soon Jan 11 18:07:26 [cc]smart: your key was? Jan 11 18:07:29 ccsmart@? Jan 11 18:08:09 [cc]smart: do 'mtn list keys' on OE database and check. if your key is not there then contact koen and mickeyl Jan 11 18:08:16 <[cc]smart> i was told there are two public keys floating around @....de and @ oe.org Jan 11 18:08:42 <[cc]smart> it is in mtn list keys, but afaik, this is only the public one Jan 11 18:09:22 hrw|work: excellent. Are you going to base it on wordpress or something else? Jan 11 18:09:22 [cc]smart: private one you have to keep Jan 11 18:09:22 <[cc]smart> if i'm correct mtn list keys cannot be taken for grant when it comes to write access Jan 11 18:09:22 <[cc]smart> yes Jan 11 18:09:22 cbrake: wordpress as I have now on blog Jan 11 18:09:35 cbrake: I know this quite good so why to change? :D Jan 11 18:09:42 hrw|work: nod Jan 11 18:10:13 cbrake: will move most of my OE related articles to it to have my blog less OE related, more private related Jan 11 18:10:24 time to split work from private D: Jan 11 18:11:40 cu Jan 11 18:11:44 hrw|work: site looks good. I subscribe to your feed. Jan 11 18:12:01 cbrake: I have BEC systems in feed reader for some time ;) Jan 11 18:12:10 hrw|gone: :-) Jan 11 18:12:42 cbrake: we need to create section on OE website where we will add links to misc articles about OE Jan 11 18:12:52 kind of 'press room' Jan 11 18:12:57 hrw: agreed. Jan 11 18:13:24 cbrake: I got to this idea after one of your articles ;D Jan 11 18:13:31 ok. cu for sure now Jan 11 18:13:39 hrw: my OE based work has really stepped up, so thats about all I write about now Jan 11 18:15:20 ~seen RP Jan 11 18:15:39 rp is currently on #gpe (2d 7h 38m 50s) #oe (2d 7h 38m 50s) #openzaurus (2d 7h 38m 50s) #handhelds.org (2d 7h 38m 50s). Has said a total of 16 messages. Is idling for 1h 22m 56s, last said: 'morning all'. Jan 11 18:17:21 03kristoffer 07org.oe.dev * rc4bca390... 10/ (3 files in 3 dirs): Jan 11 18:17:21 parted/parted-1.8.0/memory_allocation.patch : Addition of patch Jan 11 18:17:21 * Parted doesnt work on ARM when used with glibc 2.5 Jan 11 18:17:21 This patch although not fixing the issue adds Jan 11 18:17:21 some cleanups. Supplied by maintainers. Jan 11 18:30:58 Anyway to make a static kdrive?... Jan 11 18:31:49 Or whatever is needed to get working X with touschreen working properly Jan 11 18:33:29 * luke-jr ponders OpenWrt for CVS one-time-use cams Jan 11 18:33:57 hrw|gone, I plan to launch an aggressive promotion of OE in my area so the press room concept would really help out :-) The bottom line is I would be also willing to help out with the press room. Jan 11 18:34:42 Kristoffer: Xfbdev should work... . At least it's included in many distros for ipaq's. Jan 11 18:34:54 (if you don't need acceleration) Jan 11 18:36:49 mmp, yeah, but my main concern is getting the touchscreen working properly. When using Xfbdev it doesnt seem calibrated Jan 11 18:37:04 and no calibration tool exists in usual distributions Jan 11 18:37:20 Kristoffer: it exists - xtscal ... Jan 11 18:37:39 But touchscreen is accessed using tslib Jan 11 18:38:12 I know, my main goal atm is to compile a xserver with builtin tslib support (which armedslack doesnt have) Jan 11 18:38:16 but I don't know much about these things... . I'm just using them:) Jan 11 18:38:29 :) Jan 11 18:39:16 Kristoffer: but I believe that compiling xtscal for armedslack should not be big issue... . Jan 11 18:41:45 mmp, will look into it thx Jan 11 18:42:05 Kristoffer: np:) And good luck. Jan 11 18:42:13 mmp: I have angstrom-bootstrap running on my h5550 using 2.6 :) Jan 11 18:42:23 koen: :) great. Jan 11 18:42:36 koen: I believe GPE should run too... Jan 11 18:42:50 (at least at me it does, although +- few bugs:) Jan 11 18:43:00 but iirc you need PDA for locale generation:) Jan 11 18:43:06 mmp: my lcd is broken/loose connector Jan 11 18:43:12 :( Jan 11 18:43:13 03pfalcon 07org.oe.dev * rd1600a34... 10/ (3 files in 2 dirs): (log message trimmed) Jan 11 18:43:13 linux-handhelds-2.6 2.6.19-hh8: Add latest subrelease (drop -hh5). Jan 11 18:43:13 * Changes since -hh7 include: Jan 11 18:43:13 1. Lab revamped somewhat, h3900 support for using LAB as 1st stage Jan 11 18:43:13 bootloader completed (well, so far it was tested to boot at least from Jan 11 18:43:13 HaRET, not as 1st stage bootloader). Jan 11 18:43:14 goxboxlive, ok, -hh8 is in OE, submit patch to upgrade Angstrom to it. Jan 11 18:43:15 2. Further cleanup/restructuring of the tree. Jan 11 18:43:35 mmp: angstrom generates locales using qemu on the buildmachine Jan 11 18:44:37 I was thinking that you were once saying that it's a bit slow... . But I don't remember, not mentioning that my head is full of school at this time:) Jan 11 18:45:11 hrw|gone: if we switch website theme, we'd need a css for tinderbox, bonsai ,... , ... , etc as well Jan 11 18:46:23 koen: btw, if there will be some issues with it, let me know - I have never had PDA running for longer than few hours and only for debugging purposes - often rebooted. Jan 11 18:46:42 it's been up for a day know Jan 11 18:46:52 mmp: does SD and usbhost work? Jan 11 18:47:13 koen: Hello. Jan 11 18:47:27 koen: SD seems to work properly. At least I was able to work even with 2GB SD card... Jan 11 18:47:33 koen: I've seen your message about Ximageon not working on w3220. Jan 11 18:47:34 mmp: cool Jan 11 18:47:46 but the USB host (on which is hooked wifi) doesn't :( (yet) Jan 11 18:47:47 sirfred: I just get a blank screen Jan 11 18:47:57 sirfred: I might still have kernel accell turned on Jan 11 18:48:19 sirfred: didn't have time to debug (serial cable in use elsewhere), so just letting you know Jan 11 18:48:19 koen: Hmm, probably the cause is the mode setting I'm using. Jan 11 18:48:41 koen: I remember to have seen in the kernel that the display_cntl register is different for the w3220 Jan 11 18:48:48 So, perhaps that's the reason Jan 11 18:49:07 Anyway, I have to rework that part of the driver to be able to support 320x240 modes Jan 11 18:49:23 koen: Are you going to disconnect wifi, or there is some (theoretical) chance to get working the second port of USB host (ohci-hcd tells me there are two of them)? Jan 11 18:49:55 So, I was trying today to relay in framebuffer ioctls, but strange things happens that I would like to comment with RP Jan 11 18:50:16 koen: One question about w3220, does it has the same amount of memory than w100 ? Jan 11 18:51:02 Because it seems that into kernel w100fb.c all the cards are assumed to have the same amount of memory. Probably that's fine for the framebuffer layer, but not for a server that can use offscreen memory to accel. Jan 11 18:51:21 sirfred: I have no idea Jan 11 18:51:33 sirfred: if you can send me a test case I can run it in the weekend Jan 11 18:51:36 koen: OK. I will try to talk with RP about that. Jan 11 18:52:37 koen: The problem is that the registers I'm using to tell how much memory the card has, I've just seen that they're set into w100fb.c to a fixed value. And even worse, when the mode is changed to a 320x240 mode, the external memory value is not exposed. Jan 11 18:53:41 So, I would like to talk with Richard about the reasons of that, and also to think if it's a good thing to rely into the framebuffer ioctls to make the mode changes, or just copy the needed code to Ximageon Jan 11 18:54:51 Hi guys, sorry to butt in but I have a question about wifi on Zaurus C1000. Is this the right place to ask? Jan 11 19:18:52 sirfred: the w3220 in hx4700 and e800 have 384K internal RAM and 2M external RAM Jan 11 19:19:54 mreimer to the rescue! Jan 11 19:20:02 hey koen Jan 11 19:20:39 * koen tries to figure out 640x480x16bit Jan 11 19:20:49 koen|away: yeah, doesn't fit in 384K :-( Jan 11 19:30:47 03Laibsch 07org.oe.dev * r3d6c872a... 10/ (1 packages/ttf-fonts/ttf-sazanami_20040629.bb): ttf-sazanami: rprovide virtual/japanese-font, some clean-up Jan 11 19:30:51 03Laibsch 07org.oe.dev * r47970583... 10/ (7 files in 2 dirs): gpe-beam: unify Jan 11 19:53:28 mmp do you have a working angstrom-gpe-image for an h5550? Jan 11 19:53:56 koen, tells me that you have a fb patch needed to get X working ... (and the testing images are both bootstrap only) Jan 11 19:54:35 (I still have no idea why my builds do not boot, though) Jan 11 19:54:38 Any bright ideas about this line Jan 11 19:54:39 ./opiecore/device/odevice_jornada.cpp: model = cpu_info.mid( loc+2 ).simplifyWhiteSpace(); Jan 11 19:54:39 ./opiecore/device/odevice_jornada.cpp: model = cpu_info; Jan 11 19:54:51 Im understanding it as looking at the hardware line inside cpu_info Jan 11 19:54:55 correct? Jan 11 20:09:22 I hate ipkg Jan 11 20:20:12 I HATE IPKG Jan 11 20:20:38 * koen thinks of setting up Ipkg Anonymous Jan 11 20:21:07 #ipkg-anonymous ? Jan 11 20:21:48 heh Jan 11 20:22:34 koen, Our current feed size in JLime makes ipkg incredibly slow or simply freeze Jan 11 20:23:10 Kristoffer: wanna make it twice as fast? Jan 11 20:23:25 koen, I feel a joke coming Jan 11 20:23:32 for i in `find . -name Packages` ; do grep -v ^Source: $i|gzip -c9>$i.gz ;gunzip -c $i.gz>$i ; done Jan 11 20:23:49 haha nice Jan 11 20:24:03 I WILL DO THAT NOW Jan 11 20:24:10 reduces the Packages.* files by 50% Jan 11 20:24:19 and ergo ipkg memory use Jan 11 20:24:32 Yeah, but it still has to parse through the whole damn thing Jan 11 20:24:42 And, the overhead of compression Jan 11 20:24:53 it stores it uncompressed Jan 11 20:25:28 Well, worth the try Jan 11 20:25:37 I'll try that Jan 11 20:25:44 Do that, and upload so we can test it Jan 11 20:25:54 hrw and I talked about putting all info in a sqlite3 db and writing a frontend that just calls 'ipkg install http://foo1 http://foo2' Jan 11 20:25:58 But I don't have enough space on my disc for another backup Jan 11 20:26:14 koen: and? Jan 11 20:26:23 B_Lizzard: haven't had time Jan 11 20:26:30 :) Jan 11 20:26:30 Aha Jan 11 20:26:37 and o-hand has added .deb support to OE Jan 11 20:26:43 which is worth a try Jan 11 20:26:53 (even if it means using ipkg to install .debs) Jan 11 20:26:59 I'd have to rebuild everything Jan 11 20:27:16 Which is as bad as using multiple feeds Jan 11 20:27:22 mv *.ipk *.deb Jan 11 20:27:42 Yeah, but what's the difference? Jan 11 20:27:45 Using dpkg? Jan 11 20:28:07 more like for i in *.ipk; do mv $i `basename $i .ipk`.deb; done Jan 11 20:28:10 yes, dpkg and apt do things in a bit more reliable way Jan 11 20:28:23 I'll check that out then Jan 11 20:28:29 * koen needs more time Jan 11 20:43:49 mccarthy: sorry delay, I don't have image, just working Jan 11 20:43:57 ... PDA:) Jan 11 20:44:14 * koen dances the jig Jan 11 20:44:20 I can try to build one and put it somewhere ... Jan 11 20:44:50 mmp, that would be quite cool Jan 11 20:45:02 any people from brazil here? Jan 11 20:45:10 mccarthy: but basically it should not be a problem for anyone, you just need CVS HEAD kernel... Jan 11 20:45:40 mmp: do you have a list of modules that need to get installed (thinking of usbhost, mq lcd, etc) Jan 11 20:45:48 mmp, meaning what in terms of OE updating and bitbaking? Jan 11 20:45:52 koen: usbhost doesn't work yet... . Jan 11 20:46:19 I absolutely forgot, I guess some modules eve don't get packaged :( Jan 11 20:46:54 mccarthy: setting one variable so distro (i.e. angstrom) will use HEAD kernel instead of 2.6.16 one and putting few modules Jan 11 20:47:01 manually to the PDA Jan 11 20:47:23 but automatically it has few days of work before I'm afraid... . Jan 11 20:47:38 mmp: add kernel-module-foo to MACHINE_EXTRA_RDEPENDS in h5000.conf Jan 11 20:48:37 hmm, that could be a way, too... Problem actually is that I should rebuild whole oe tree, which was not updated for a month... . Jan 11 20:49:06 mmp: I added a correct /boot/params and fstab this week :) Jan 11 20:49:12 :-) Jan 11 20:49:30 hmm, I managed to boot with old fstab without problems, weird... . Jan 11 20:49:33 or... Jan 11 20:49:36 * mmp looks Jan 11 20:49:53 it boots actually into userspace now Jan 11 20:50:11 mmp: usbhost needs dma-bounce in the same buffer as sd? Jan 11 20:50:15 aha, so to fstab problbaby doesn't remount Jan 11 20:50:20 mmp, so I would add "PREFERRED_VERSION_linux_handhelds-2.6 = "HEAD""? Jan 11 20:51:02 "2.6.19-hh8+cvs${SRCDATE}" iirc Jan 11 20:51:03 (to my local.conf) Jan 11 20:51:11 or "2.6.19-hh8" Jan 11 20:51:15 koen: I don't know, the only thing I know about that bug is that either DMA transfers don't work correctly, or changes in USB code in last kernel releases were too dramatical Jan 11 20:51:37 mccarthy: don't know, I'm currently building kernel outside OE. maybe even removing preferred version could solve that... . Jan 11 20:51:46 or, as koen said:) Jan 11 20:51:55 that would be safer:) Jan 11 20:52:29 koen: I don't know how exactly works dmabounce, so I can't say. I have just rough image. Jan 11 20:52:45 mmp: I can't even code C :) Jan 11 20:53:12 can I put MACHINE_EXTRA_RDEPENDS in my local.conf too? (and what modules do I put there?) Jan 11 20:53:16 koen: :-)) Well, would believe quite a much, but not this:) Jan 11 20:53:44 mccarthy: you can, but h5000.conf will overwrite it anyway Jan 11 20:54:04 mccarthy: you should need h5400_lcd and mq1100fb Jan 11 20:54:15 and maybe some fbcon-things. Jan 11 20:54:45 what about wifi. isn't it some weird thing with an a and a bunch of numbers? Jan 11 20:54:49 mccarthy: but be warned, things don't work too reliably Jan 11 20:55:16 mccarthy: usbhost doesn't work, so I don't know how would the drivers work... Jan 11 20:56:25 The only thing I know that USB host either receives USB of zero length, or there is DMA error somewhere:) (thanks to Alan Stern, which told me what output from usbmon says:) Jan 11 20:57:17 On the other hand, bluetooth (at least the most low level part) seems to work quite well... . Jan 11 20:59:45 mmp: I fixed bluetooth detection as well :) Jan 11 20:59:49 :) Nice Jan 11 21:00:18 and florian_kc was nice enough to put that in the new release of blueprobe Jan 11 21:00:52 koen: btw, you mentioned some dmabounce && sd && ushbost - yes, they probably need to use the same DMA area, don't know if same buffers as well. Jan 11 21:01:05 mmp: http://projects.linuxtogo.org/plugins/scmsvn/viewcvs.php?rev=8947&root=gpe&view=rev Jan 11 21:01:08 (probably = I did not read SAMCOP documentation) Jan 11 21:01:35 mmp: h2200 needs the same for sd, but h5xxx shares the buffer with usbhost afaik Jan 11 21:01:39 * koen looks at mreimer Jan 11 21:01:59 (blueprobe) great:) Jan 11 21:02:36 mmp: koen's right Jan 11 21:02:45 I just know that "they both use the same SAMCOP SRAM area for DMA transfers" (from hho wiki) Jan 11 21:03:35 Hmm, that sounds quite weird... . What does the "same buffer" mean? Jan 11 21:04:04 Same DMA engine taking care about it? Shared structures allocated by something? Jan 11 21:04:08 buffer = sram Jan 11 21:04:11 mmp: same SRAM Jan 11 21:04:17 Aha, OK:) Jan 11 21:04:32 hamcop has 32k and samcop 64k Jan 11 21:04:35 (from memory) Jan 11 21:04:53 I guess samcop has less... but I'm not sure Jan 11 21:04:56 hamcop has 16k, samcopy probably double that Jan 11 21:04:59 gotta run Jan 11 21:05:01 bbiab Jan 11 21:05:03 that might be Jan 11 21:05:08 mreimer: bye:) Jan 11 21:05:21 Hmm, but that idea with sd controller is quite interesting... Jan 11 21:05:31 I'm going to have look on how it's implemented:) Jan 11 21:07:11 hi all Jan 11 21:07:41 koen,mreimer: _that_ code is much different from things I've seen, thanks for hint:) Jan 11 21:07:55 hey mr_nice Jan 11 21:08:07 koen: hi Jan 11 21:12:38 mmp: I think the main problem was to have a sane way to share the sram Jan 11 21:13:18 koen: well, but the equivalent might be (I did not read the code yet) to force ohci-hcd.c to use _our_ DMA thingies. Jan 11 21:13:54 I think there could be some way tho create DMA zone (or how it is called) for samcop devices... . Jan 11 21:13:55 you could just divide it in half Jan 11 21:14:23 Yes, but still, it's modification to code which should not be modified... . But maybe there's way to override it... . Jan 11 21:17:49 I guess there should be. I think that worked even in 2.4 without hacking the ohci code. Jan 11 21:42:34 03pfalcon 07org.oe.dev * rf37da99f... 10/ (1 packages/opie-sysinfo/opie-sysinfo_1.2.2.bb): Jan 11 21:42:34 opie-sysinfo: Make jornada-6xx-7xx.patch jlime-local. Jan 11 21:42:34 * It requires corresponding patch for libopie2, but that in turn fails to Jan 11 21:42:34 apply to CVS HEAD. These two patches should be submitted upstream. Jan 11 21:42:38 03pfalcon 07org.oe.dev * ra0e24570... 10/ (1 packages/libopie/libopie2_cvs.bb): Jan 11 21:42:38 libopie2 cvs: keyboardless-buttonmap.patch was accepted upstream, drop. Jan 11 21:42:39 * Recommended SRCDATE for this recipe: 20061220 Jan 11 21:42:42 03pfalcon 07org.oe.dev * r340d510d... 10/ (5 files in 3 dirs): (log message trimmed) Jan 11 21:42:43 opie-multikey: Rename buttons to add/remove custom keymap to be more clear. Jan 11 21:42:45 * opie-multikey needs usability improvements at all - it may take a user Jan 11 21:42:47 some effort to figure out that to change language layout, one needs to call Jan 11 21:42:50 config (which lives on a key marked as shift), select language, and tap Jan 11 21:42:51 OK at window top. Most people would tap "Add" and present with file selection Jan 11 21:42:53 dialog with unknown to them purpose. So, name buttons "Add custom..." and Jan 11 21:43:00 later guys Jan 11 21:43:02 mmp: look for dma_declare_coherent_memory in http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/drivers/soc/samcop_base.c.diff?r1=1.33&r2=1.34 Jan 11 21:43:06 thanks for the help koen and mmp Jan 11 21:43:13 no luck building myself (again). Jan 11 21:43:23 any images would be welcomed ;) Jan 11 21:49:30 morning Jan 11 21:49:47 (well, almost) Jan 11 21:50:44 hey likewise Jan 11 21:51:45 hey koen Jan 11 21:54:42 damn, flight are expensive these days Jan 11 22:05:55 hi Jan 11 22:06:14 psokolovsky_: why those patches are not on opie-devel? Jan 11 22:06:24 hrw, they are Jan 11 22:06:32 I have (yet another) thick question. Say I have 2 boards that run very similar sets of packages, is the normal way to build the images by having ./build/device1 and ./build/device2? Is there a way to share the /tmp directory so that I don't have scads of disk space used (not to mention recompiling the toolchain?) Jan 11 22:07:29 bradbev, yes, multi-machine builds are supported Jan 11 22:07:43 bradbev: add 'INHERIT += "multimachine"' into local.conf Jan 11 22:08:07 bradbev: it will need build from scratch but will give you what you want Jan 11 22:10:52 thanks Jan 11 22:17:09 mreimer: thanks, sorry delay:) Jan 11 22:17:16 np Jan 11 22:18:10 03mickeyl 07org.oe.dev * r864030c3... 10/ (3 files in 2 dirs): portmap: use DEBIAN_MIRROR instead of hardcoding one Jan 11 22:18:14 mreimer: hmm, but the SD driver uses some other magic to control the DMA... Jan 11 22:18:24 some samcop routines... . Jan 11 22:19:07 mmp: IIRC the OHCI stuff just writes to it, whereas the SD driver reads and writes to it Jan 11 22:19:16 mmp: have you tried it yet? it might just work Jan 11 22:19:34 mreimer: yes, I did, but usbmon says I'm receiving 0-length URBs Jan 11 22:19:41 mmp: ohci doesn't need to do DMA to that SRAM, just put usb descriptors or something like that in there Jan 11 22:19:58 mmp: can you read/write the SRAM with devmem2? Jan 11 22:20:10 mmp: on h2200 the nand clock has to be on for the SRAM to work Jan 11 22:20:18 SD controller works, so I should be able... Jan 11 22:20:21 * mmp checks. Jan 11 22:21:18 weird, I don't have SD controller enabled in actual setup? Jan 11 22:23:10 mmp: just in case the SD controller is stomping on OHCI's memory, try removing the SD driver from your config Jan 11 22:23:44 mreimer: it seems that there is something weird happening... . SD doesn't seem to work, but that's strange Jan 11 22:30:34 hmm, so SD driver freezes whole system when accessing SD card (at least as a mdule) Jan 11 22:30:56 koen: so it seems that neither SD works :( Jan 11 22:31:01 but with 2.6.16 it does Jan 11 22:31:05 (at least did) Jan 11 22:34:04 is the clock enabled? check CLKCON Jan 11 22:34:20 and do you have the dma driver built in too? Jan 11 22:34:23 and its clock enabled? Jan 11 22:34:28 koen: about website change - I know that it will need new themes for other sites Jan 11 22:35:37 mreimer: everything worked before, DMA should be enabled too. I guess clock code did not change - it even seems to read few sectors from SD, but freezes then Jan 11 22:36:06 mmp: maybe it's dropping an interrupt Jan 11 22:37:50 Weird, when hard-compiled to the kernel, it doesn't read anything from the card. Jan 11 22:38:00 Except for ID Jan 11 22:38:44 hmm, maybe... Jan 11 22:38:56 But I believe I fixed all clockings in samcop_base correctly... Jan 11 22:41:15 cu Jan 11 22:46:09 parted 1.8.0 doesn't have memory_allocation.patch, which is needed Jan 11 22:46:21 Bitbake fails with an error Jan 11 22:46:40 Ooops Jan 11 22:59:15 it's quite late; good night; and thanks to all:) Jan 11 23:00:28 PEOPLE OF THE NIGHT, I BID YOU FAREWELL Jan 11 23:09:42 hi Jan 11 23:09:54 is there a pdaxrom repo? Jan 11 23:11:49 what is pdaxrom? Jan 11 23:12:25 oleg, privet zemlyakam ;) Jan 11 23:12:57 Jin^eLD: translit is usually forbidden Jan 11 23:13:06 whatever, just saying hello :> Jan 11 23:13:16 who knows if you are running koi or cp Jan 11 23:13:17 koen_: in fact I need efltk and ede bb's Jan 11 23:13:33 koen_: pdaxrom seems to be a Zaurus project Jan 11 23:14:00 Jin^eLD: so you wanted to speak Russian on an international channel? lol Jan 11 23:14:10 olegfink: not at all Jan 11 23:14:32 but from time to time people "out" themselves, we had some russians who are .de now and so on Jan 11 23:14:40 olegfink: what is a zaurus? Jan 11 23:14:56 but infact it was never more then just "hello and how are ya" Jan 11 23:15:08 koen_: is it a sort of a bad humour? Jan 11 23:15:32 Jin^eLD: well it's ok ;) Jan 11 23:15:51 olegfink: I never assumed it wasnt ;) Jan 11 23:16:00 Jin^eLD: you're ecouraged to visit #ubuntu-russian then Jan 11 23:16:08 nah, not much of an ubuntu guy ;> Jan 11 23:16:31 intel is little endian and powerpc is bigendian? Jan 11 23:16:51 Jin^eLD: neither am I Jan 11 23:16:56 Crofton|home: which intel? Jan 11 23:17:03 x86 :) Jan 11 23:17:09 Crofton|home: yes, that's LE Jan 11 23:17:33 Jin^eLD: but it's more a community than a ubuntu club Jan 11 23:17:33 I think I need to big to little some data :) Jan 11 23:17:46 olegfink: i see... Jan 11 23:17:58 it's night now Jan 11 23:19:14 Crofton|home: so you found your garbled data problem? Jan 11 23:19:37 not sure Jan 11 23:19:40 * Jin^eLD had some garbled data with CIFS, bewaer... Jan 11 23:19:43 beware that is Jan 11 23:19:50 but I suspect the usrp assumes the data format is LE Jan 11 23:20:00 but thats a different topic :) Jan 11 23:20:09 and ultimately the data in a har buffer is treated as 16bit ints Jan 11 23:20:28 and what comes out is gibberish :) Jan 11 23:20:38 so I byte the shorts, and try again Jan 11 23:20:53 Jin^eLD: which handheld do you run? Jan 11 23:21:44 olegfink: none, I have custom hardware, mostly on ARM or MIPS Jan 11 23:21:46 Crofton|home: so you need to include a runtime check in the ossie stuff, right? Jan 11 23:22:04 Jin^eLD: lucky you Jan 11 23:22:17 compile time Jan 11 23:22:30 olegfink: I guess :) we have lots of different hardware at work Jan 11 23:22:39 or change firmware in the FPGA :) Jan 11 23:22:54 but for now I will do a compile check that byte swap if powerpc Jan 11 23:23:04 * olegfink only has an old HP Jornada 680, that's all Jan 11 23:23:09 since I suck at verilog Jan 11 23:23:15 Crofton|home: btw, arm can also be BE Jan 11 23:23:29 Crofton|home: the ixp425 board on my desk runs in BE mode now Jan 11 23:23:29 I know, but they are usually LE? Jan 11 23:23:56 Crofton|home: you could select it at boot Jan 11 23:24:29 I do not really care which way it is, only that I get intelligle data out of the USRP Jan 11 23:24:35 from the EFIKA Jan 11 23:24:48 right Jan 11 23:25:07 could you do byteswapping inside teh usrp? Jan 11 23:25:20 (not elegant, but cheaper for the receiving end) Jan 11 23:25:22 if I could change the verilog, I think I could Jan 11 23:25:46 I suspect the USRP assumes LE, and the data ands up goign to an ADC Jan 11 23:26:11 Jin^eLD: so what about some russian channels, eh? Jan 11 23:32:34 uhm.. :) I do not know any.. well, I know one on IRCNet but I do not hang around there too often Jan 11 23:33:26 Jin^eLD: so you are suggested, once more time, to visit #ubuntu-russian at freenode Jan 11 23:34:38 ok.. I'll drop by sometime Jan 11 23:35:13 ok Jan 11 23:45:00 for people who use multimachine, is the TMPDIR always set to the same place? Jan 11 23:45:48 bradbev: I think so, but I'm not sure Jan 11 23:46:04 thanks mreimer Jan 12 00:38:18 So, no one else has had sound driver problems with 2.6.19? Jan 12 00:43:30 I'm still having this quantum bug on my Z after applying the OE patches (including the dcache fix) Jan 12 00:47:51 Also it seems that after audio playback is successfully started, reloading the sound driver results in no sound until the machine is reset Jan 12 01:13:13 03lenehan 07org.oe.documentation * rf74db950... 10/ (6 files in 4 dirs): Jan 12 01:13:13 usermanual: Add a list of all the installation variables and there typical Jan 12 01:13:13 values to the reference section. Jan 12 01:21:35 DataBeaver: and why are you reloading the sound module? Jan 12 01:22:00 JustinP: To test chages I made to it? Jan 12 01:22:18 DataBeaver: ok, then ;-) Jan 12 01:22:44 Of course I wouldn't reload it in normal use, so this is a rather minor annoyance. Jan 12 02:20:02 03lenehan 07org.oe.documentation * rc805bc51... 10/ (3 files in 3 dirs): Jan 12 02:20:02 usermanual: Add a section describing the staging directories to the Jan 12 02:20:02 usermanul reference section. **** ENDING LOGGING AT Fri Jan 12 02:59:57 2007