| 01:45.07 | lkcl | i think the clue in the bitbake gpe-image build is in "parsing finished. 0 cached, 2868 parsed, 72 skipped, 0 masked". working out which ones are "skipped" and why would help... |
| 01:45.45 | lkcl | nope. no mention of "soundcore" in the tmp/cache. |
| 01:45.48 | lkcl | odd. |
| 02:16.14 | *** join/#htc-blueangel Killer3k (n=Phonics3@82-71-29-232.dsl.in-addr.zen.co.uk) |
| 08:00.07 | *** join/#htc-blueangel rob_w-cgi (n=txuser@dd1010.kasserver.com) |
| 08:00.08 | *** mode/#htc-blueangel [+o rob_w-cgi] by ChanServ |
| 08:20.42 | rob_w-cgi | morning room |
| 09:53.46 | lkcl | moornin |
| 10:07.07 | *** join/#htc-blueangel ben017f (n=ben@219-88-216-1.jetstream.xtra.co.nz) |
| 10:50.19 | *** join/#htc-blueangel rob_w-cgi (n=txuser@81.209.148.105) |
| 10:50.19 | *** mode/#htc-blueangel [+o rob_w-cgi] by ChanServ |
| 13:34.06 | *** join/#htc-blueangel rob_w-cgi (n=txuser@dd1010.kasserver.com) |
| 13:34.06 | *** mode/#htc-blueangel [+o rob_w-cgi] by ChanServ |
| 13:39.53 | Funklord | hmm.. seems BAs are running short here, maybe I'll get a universal |
| 14:06.05 | *** join/#htc-blueangel cp15_ (n=txuser@dd1010.kasserver.com) |
| 14:06.16 | cp15_ | Good Morning |
| 14:13.34 | Funklord | morning =) |
| 14:35.37 | ben017f | morning |
| 15:31.23 | *** part/#htc-blueangel oho (n=phm4@myrtle.ukc.ac.uk) |
| 16:32.59 | *** join/#htc-blueangel cp15 (i=tducrm@p549A2EAF.dip0.t-ipconnect.de) |
| 16:32.59 | *** mode/#htc-blueangel [+o cp15] by ChanServ |
| 16:33.16 | cp15 | Good morning |
| 17:49.34 | *** join/#htc-blueangel _cronos_ (n=cronos@ron34-1-82-224-149-102.fbx.proxad.net) |
| 18:00.47 | *** join/#htc-blueangel rob_w (n=bob@p85.212.194.64.tisdip.tiscali.de) |
| 18:00.47 | *** mode/#htc-blueangel [+o rob_w] by ChanServ |
| 20:03.22 | immolo | heya |
| 20:18.22 | *** join/#htc-blueangel iclem (n=chatzill@iclem.net) |
| 20:19.27 | iclem | good evening all |
| 20:32.35 | cp15 | Hi iclem |
| 20:56.35 | lkcl | _christ_ this gpe-image kernel thing is pissing me off. |
| 20:57.23 | lkcl | the modules are listed in the blueangel.conf file as dependencies but half of the ipkg kernel module packages aren't being copied onto gpe-image. |
| 20:57.34 | lkcl | kernel-module-soundcore for example. |
| 20:57.53 | cp15 | Hmmm... Immolo, are you here? |
| 20:58.22 | lkcl | even if i _explicitly_ mention kernel-module-soundcore as a dependency, it _still_ doesn't get packaged up. |
| 20:59.12 | cp15 | Have you already tried a find . -type f | xargs grep blueangel_audio |
| 20:59.18 | cp15 | Maybe it gets cached somewhere |
| 20:59.24 | immolo | cp15, between two things but yes |
| 20:59.46 | cp15 | Probably you can help: Is there a special trick to get the kernel modules in the gpe image? |
| 21:00.18 | immolo | what do you mean? |
| 21:00.48 | cp15 | lkcl has extended the BA_MODULES_kernel26 list (like you before), but the new modules are not in the gpe image |
| 21:00.55 | immolo | oh |
| 21:01.41 | lkcl | soundcore doesn't get included on gpe-image, for example, even though it's listed in BA_MODULES_kernel26. |
| 21:01.41 | lkcl | which is really weird. |
| 21:02.27 | immolo | by gpe-image we are talking the rootfs after gpe-image has baked? |
| 21:02.38 | lkcl | bitbake gpe-image. |
| 21:03.22 | lkcl | off it goes, builds the image.... and despite soundcore (and others) being in BA_MODULES_kernel26, half of the modules listed are missing from the final .tar.bz2 |
| 21:03.27 | immolo | did you remove all stages etc of the xanadux kernel after you changed the blueangel config |
| 21:03.43 | lkcl | at one point, yep. |
| 21:04.02 | lkcl | i deleted pretty much everything in the tmp/ directory. |
| 21:04.11 | immolo | can I see your machine.conf to check against mine? |
| 21:04.34 | lkcl | a version (earlier version) of the monodoc diff is at.... |
| 21:05.10 | lkcl | http://hands.com/~lkcl/gpe.blueangel.patch |
| 21:05.25 | immolo | ok |
| 21:05.55 | immolo | let me just check where bb puts files |
| 21:06.01 | lkcl | which is pretty straightforward: add some extra modules is about it. |
| 21:06.04 | immolo | been a few weeks |
| 21:06.21 | immolo | yeah I had to delete something to get it added just trying to see what |
| 21:07.50 | lkcl | thanks. |
| 21:08.07 | immolo | no problem |
| 21:09.22 | immolo | is the stages/xanadux and stamps\xanadux stuff deleted? |
| 21:09.41 | lkcl | will try that.... |
| 21:09.49 | immolo | there is some more |
| 21:10.26 | immolo | you also need to delete the gpe-image stuff from those files as well if I remember right |
| 21:11.07 | immolo | then try bitbake gpe-image and it should bake xanadux something else then gpe-image |
| 21:11.43 | lkcl | okay - where's stages/xanadux? there's a staging directory... |
| 21:12.04 | immolo | is it in tmp/ ? |
| 21:12.12 | lkcl | no. |
| 21:12.23 | lkcl | cache cross deploy rootfs staging stamps work |
| 21:12.36 | immolo | ah staging |
| 21:12.46 | immolo | knew it had stage in it :P |
| 21:13.00 | immolo | sorry it's been a while since I've had oe in front of me |
| 21:14.30 | lkcl | there's nothing called xanadux in staging (find staging | grep xanadux) ... |
| 21:15.11 | immolo | damn what does it get saved under, is it ba something? |
| 21:15.33 | lkcl | work/xanadux-ba-2.6-2.6.12-r0 |
| 21:16.12 | immolo | let me see if my stuff is on backup and find out where it alll goes |
| 21:16.18 | lkcl | ack |
| 21:21.23 | lkcl | okay - rebuild just for fun, kernel-module-soundcore has been created, for example... |
| 21:21.40 | lkcl | unpacking gpe-image.bb to tmp/work/gpe-image-1.0-r19... |
| 21:22.10 | lkcl | ah _ha_. here's something new - do_populate_staging... |
| 21:24.45 | lkcl | nope. FRAK. |
| 21:26.56 | cp15 | Hmm... do a find . -name run.do_rootfs.* |
| 21:27.02 | cp15 | And then send me the file |
| 21:27.18 | cp15 | But before do a ls -l on it to see whether its timestamp is current |
| 21:28.28 | cp15 | Ah... Another thing comes into mind |
| 21:28.41 | lkcl | yeh? :) |
| 21:28.52 | cp15 | Take a look into stamps and see if there is something is there with bootstrap in its name |
| 21:28.59 | lkcl | ack.. |
| 21:29.08 | lkcl | gotta find them first... |
| 21:31.00 | lkcl | task-bootstrap-1.0-r25 |
| 21:31.43 | cp15 | Yeah, delete the stamps |
| 21:31.53 | cp15 | This is the package which holds the kernel modules |
| 21:32.02 | lkcl | ack. |
| 21:33.06 | lkcl | hey, phil blundell wrote the task-bootstrap bb file. |
| 21:33.14 | lkcl | where do i remember his name from.. |
| 21:34.31 | cp15 | He is responsible for most of gpe... But I also had thought of from where I know him else |
| 21:35.48 | lkcl | FRAK. bastard. it was the task-bootstramp thing. |
| 21:35.50 | lkcl | got it. |
| 21:39.12 | cp15 | So you now have a Image with all modules? |
| 21:39.53 | lkcl | yep. |
| 21:40.02 | lkcl | friggin stupid friggin thing. |
| 21:40.30 | cp15 | Can you check whether the modules have kernel version 2.4.12-hh2-xda0-ba0? |
| 21:40.42 | lkcl | sure. |
| 21:40.42 | cp15 | 2.6.12 of course |
| 21:41.16 | lkcl | vermagic 2.6.12-hh2 |
| 21:41.40 | cp15 | Ok, then you have to redo the kernel build |
| 21:41.47 | lkcl | xanadux-ba-2.6-2.6.12-r0 |
| 21:41.49 | lkcl | burble? :) |
| 21:41.52 | cp15 | Yes |
| 21:42.17 | lkcl | uhhnnn... errr... i thought i just did! |
| 21:42.39 | lkcl | but then again, i didn't do a monocrock update... |
| 21:43.25 | lkcl | when you say redo the kernel build, do you mean with bitbake? |
| 21:43.27 | cp15 | You probably have to remove the sources to force a new cvs checkout |
| 21:43.43 | immolo | you do |
| 21:43.45 | lkcl | i got rid of... oh, yes, yes that'd be right. |
| 21:44.01 | cp15 | Yes, remove sources/linux-2.6-xda_cvs.sourceforge.net_K2-6-12-hh2-xda0-ba0_.tar.gz |
| 21:44.40 | cp15 | I whish there would be a better way |
| 21:44.47 | lkcl | off we go again.... |
| 21:46.13 | cp15 | How have you fixed quilt and other packages referring to savannah? |
| 21:46.26 | lkcl | lemme see... |
| 21:47.20 | lkcl | 1-line patch: use cvs://anonymous@cvs.sv.nongnu.org, method=pserver |
| 21:47.27 | lkcl | bit of a pain, that. |
| 21:48.26 | lkcl | cvs update, hurrah.... tag=K2-6-12-hh2-xad0-ba0 |
| 21:49.25 | lkcl | brb, build time & tea time... |
| 22:19.10 | lkcl | okay, built... time to retry. i'm going to need to drag the kernel off of there, aren't i? |
| 22:21.21 | lkcl | muhahahah. off we go... |
| 22:21.31 | cp15 | The kernel and initrd from gnulinux.biz should be ok |
| 22:22.00 | lkcl | oops. could not load /lib/modules/2.6.12-hh-xda0-ba0/modules.dep |
| 22:22.02 | lkcl | oops :) |
| 22:22.05 | lkcl | hmmm |
| 22:24.08 | cp15 | Yes, you need the current initrd from gnulinux.biz |
| 22:24.22 | lkcl | oh. right... |
| 22:25.16 | cp15 | I have changed the kernel postfix from -hh2 to -hh2-xda0-ba0 recently, which gives incompatible modules |
| 22:25.40 | lkcl | ack. |
| 22:36.40 | cp15 | Any success? |
| 22:37.53 | lkcl | gimme 2secs... |
| 22:38.23 | lkcl | gpe image "build" date from bitbake says 200512132233 |
| 22:38.39 | lkcl | file that gets created is image 20051213210706... |
| 22:39.41 | cp15 | Don't think this is a huge problem :-) |
| 22:39.59 | lkcl | yehh, well as long as it _is_ getting recreated... |
| 22:40.32 | lkcl | boot... |
| 22:40.46 | lkcl | i mean recreated not copied from some unknown cache location i don't know about |
| 22:40.51 | lkcl | okay. |
| 22:40.54 | lkcl | seems okay so far |
| 22:40.58 | lkcl | running /sbin/init... |
| 22:41.45 | lkcl | gpe 2.7 package stuff... |
| 22:45.45 | cp15 | Now tell me that it works :-) |
| 22:45.54 | lkcl | damn. have to go back to that lsmod trick. |
| 22:45.59 | lkcl | nope! :) |
| 22:46.06 | cp15 | No touchscreen? |
| 22:46.15 | lkcl | nope :) *lol* |
| 22:46.45 | cp15 | Hmm... Can you upload it again? I will try to test it tomorrow |
| 22:47.03 | lkcl | willdo... gimme a while. |
| 22:54.03 | lkcl | hurrah, i have ssh access. |
| 22:54.48 | cp15 | Is the touchscreen module loaded? |
| 22:55.24 | lkcl | himalaya_tsc2200 |
| 22:55.29 | lkcl | himalaya_ts3 |
| 22:55.37 | lkcl | evdev |
| 22:55.38 | lkcl | tsdev |
| 22:55.43 | cp15 | is there a /dev/input/event0? |
| 22:55.45 | cp15 | Probably |
| 22:56.09 | lkcl | yep. |
| 22:56.15 | cp15 | echo $TSLIB_TSDEVICE |
| 22:56.21 | cp15 | echo $TSLIB_CONFFILE |
| 22:56.51 | cp15 | Ah... I think I know |
| 22:56.51 | lkcl | <PROTECTED> |
| 22:57.11 | lkcl | <PROTECTED> |
| 22:57.19 | cp15 | Probably the load order of kbd and ts is wrong, so ts is at /dev/input/event1 |
| 22:57.27 | lkcl | evdev wasn't loaded |
| 22:57.29 | lkcl | ah. |
| 22:57.47 | cp15 | Try od /dev/input/event1 and see if it reacts on the touchscreen |
| 22:58.01 | lkcl | yep. |
| 22:58.28 | lkcl | plenty. |
| 22:58.51 | cp15 | Ok mv /etc/modutils/blueangel_kp /etc/modutils/zblueangel_kp |
| 22:59.02 | cp15 | update-modules |
| 22:59.12 | lkcl | suggestions?acp |
| 22:59.13 | lkcl | acp |
| 22:59.14 | lkcl | ack |
| 22:59.20 | lkcl | burble... |
| 22:59.25 | lkcl | done |
| 22:59.52 | lkcl | you're not going to say "okay now reboot it" are you? |
| 23:00.05 | cp15 | Maybe you can fix it online |
| 23:00.11 | cp15 | /etc/init.d/gpe-dm stop |
| 23:00.17 | cp15 | rmmod blueangel_kp |
| 23:00.19 | lkcl | gpe-dm not running |
| 23:00.21 | cp15 | rmmod himalaya_ts3 |
| 23:00.33 | cp15 | modprobe himalaya_ts3 |
| 23:00.36 | cp15 | modprobe blueangel_kp |
| 23:00.46 | cp15 | /etc/init.d/gpe-dm start |
| 23:01.17 | cp15 | In blueangel.conf replace module_autoload_blueangel_kp with module_autoload_zblueangel_kp |
| 23:01.40 | lkcl | xxxxingA |
| 23:01.48 | cp15 | And also remove the wrong # pxa2xx_udc is built in to the kernel comment |
| 23:03.42 | lkcl | zblueangel_kp? in conf/machines/blueangel.conf? it's as simple as that - the z will be detected as "load this last"? |
| 23:04.02 | cp15 | Yes |
| 23:04.05 | lkcl | cool. |
| 23:04.41 | cp15 | The list gets sorted alphabetically by the shell |
| 23:05.38 | cp15 | Did usb work on the first try or was a reboot required? |
| 23:05.45 | lkcl | reboot required. |
| 23:05.49 | lkcl | worked 2nd time. |
| 23:07.12 | lkcl | ok. |
| 23:07.14 | lkcl | where am i. |
| 23:07.17 | lkcl | gomunicator. |
| 23:07.28 | cp15 | There seems to be a strange misordering. The modules get configured after the network |
| 23:07.53 | cp15 | Is it there? |
| 23:08.12 | lkcl | something wrong with sourceforge cvs - i had to take gomunicator out for a mo. |
| 23:08.25 | lkcl | now this is happy, will retry putting gomunicator back in. |
| 23:08.48 | cp15 | Ok, before you start check whether the following ipkgs are installed: |
| 23:09.48 | cp15 | alsa-dev |
| 23:09.51 | cp15 | alsa-conf |
| 23:10.03 | cp15 | libasound2 |
| 23:10.39 | cp15 | If not, add them to gomunicator DEPENDS |
| 23:10.55 | lkcl | ack. |
| 23:11.54 | lkcl | nope, they not there. added now. |
| 23:13.01 | lkcl | handling bitbake files.... zzznnnn... |
| 23:13.09 | lkcl | 88%... |
| 23:13.13 | lkcl | 90%... |
| 23:13.16 | lkcl | zzz... |
| 23:13.26 | lkcl | 95%... |
| 23:14.22 | cp15 | Hmm... I am afraid to tell you, but you have to add them also to RDEPENDS |
| 23:14.27 | lkcl | hmmm... libasound - doesn't exist - you mean alsa-lib? |
| 23:15.06 | lkcl | no problem |
| 23:15.24 | cp15 | Just a moment |
| 23:18.03 | cp15 | It gets built by alsa-lib, so hopefully you don't need to specify it on its own |
| 23:18.19 | lkcl | what provides alsa-conf? |
| 23:18.31 | cp15 | Also alsa-lib |
| 23:18.59 | cp15 | BTW: What would be your preferred ringtone format? mp3 or ogg? |
| 23:19.17 | lkcl | haven't _got_ a clue - don't care :) |
| 23:19.19 | immolo | ogg |
| 23:19.38 | immolo | better quailty smaller size |
| 23:20.13 | immolo | did you ever do anymore work on video ringtones? |
| 23:21.58 | cp15 | Hmm... libmad0 is a lot smaller than libogg + libvorbis... But it may keep the balance with a few installed ringtones |
| 23:22.22 | *** join/#htc-blueangel JuNuKN (n=JuNuKN@Kbd55.k.pppool.de) |
| 23:22.27 | cp15 | No, not yet... I have dumped the idea to use gstreamer for now, since it is simply too slow |
| 23:23.31 | cp15 | A ringtone which comes with a delay of 2 seconds is not acceptable IMO |
| 23:23.44 | immolo | no :P |
| 23:24.28 | cp15 | Maybe we get it faster when we move to PXA flash/DOC, but for now I think I will implement ogg (and maybe mp3) directly into gomunicator |
| 23:25.17 | immolo | well I say the best idea is to get it working then fiddle with gstreamer after |
| 23:25.25 | lkcl | what's the plan for reading/writing sim card contacts? |
| 23:25.41 | cp15 | None yet :-) |
| 23:25.47 | lkcl | plus, any plans for splitting gomunicator into a client/server thing? |
| 23:26.05 | lkcl | so you can run a daemon that will fire up the user-interface on-demand? |
| 23:26.10 | cp15 | Do you think we should implement a external program like win or integrate it into gomunicator? |
| 23:26.52 | cp15 | I don't want to split gomunicator right now, because it requires a clean interface which is hard to design without knowing every required feature in advance |
| 23:26.53 | immolo | client/server seems the best then it is using as little ram as needed |
| 23:27.05 | lkcl | definitely split gomunicator into client/server, then make the server bit the only way to access /dev/gsm |
| 23:27.18 | lkcl | nehhh, bugger that :) hack iiiiit :) |
| 23:27.53 | lkcl | sorry. automatic reaction there. what were we talking about? |
| 23:28.29 | lkcl | with a client-server thing, you can have a separate program accessing the sim card for contacts etc. |
| 23:28.49 | lkcl | or provide a library plugin for other contact-based programs. etc. |
| 23:28.54 | cp15 | And: The ram issue isn't really an issue. It has always to be enough ram left to activate the gui. You don't want to miss a call because all ram is occupied |
| 23:29.45 | cp15 | It doesn't make sense to access the sim card for more than copying the contacts to the address book |
| 23:30.10 | lkcl | or writing. |
| 23:30.11 | cp15 | SIM access is extremely slow with the HTC GSM modem, because they "forgot" to implement the usual commands to access it |
| 23:30.20 | lkcl | ah that would explain it. |
| 23:30.55 | lkcl | okay - so what's going to arbitrate sim card reading and incoming phone call answering? |
| 23:31.29 | cp15 | But OTOH they didn't implement direct SIM access, so it is not possible to run the gsm algorithm and "export" the SIM |
| 23:32.19 | lkcl | urk. |
| 23:32.55 | cp15 | Don't think the arbitration is that hard. We simply forbid to queue commands here, one command can only read one entry, and so there is enough space left for a RING |
| 23:33.02 | lkcl | wtf going on with gomunicator build: one of the dependencies is libgpg-error0 apparently. |
| 23:33.08 | immolo | lkcl, you building the distro with gcc 3.4.4 or 4? |
| 23:33.20 | immolo | to change the suject a little :P |
| 23:33.26 | lkcl | gcc 3.4.4 |
| 23:33.39 | lkcl | you think i'm mad? gcc 4? muhahahah |
| 23:34.12 | immolo | you try telling the rest of the linux world a high version number doesn't mean better |
| 23:34.19 | immolo | and apple for some reason |
| 23:34.28 | lkcl | it doesn't mean better _yet_... |
| 23:35.00 | cp15 | It breaks compile of a lot of packages because it is much stricter in error checking |
| 23:35.14 | immolo | 4.1 is suppose to be the first speed boosted version |
| 23:35.21 | immolo | 4.0 is just framework |
| 23:36.06 | lkcl | tell me about it... i've got a system i can't upgrade because unicorn-source (pci adsl modem) doesn't compile with gcc 4.0 |
| 23:36.15 | immolo | cp15, yeah it's taken ages to get gcc4 fully working on gentoo |
| 23:37.03 | immolo | and that was x86 |
| 23:37.04 | lkcl | that's because gentoo mix up the build dependencies with the development dependencies - freedce can't be compiled unless you install a _separate_ version of flex 2.5.34 in /usr/local ... |
| 23:38.00 | immolo | lkcl, never followed it really just saw how much bloat gcc4 added to my system and fdisked :P |
| 23:38.09 | lkcl | :) |
| 23:38.29 | lkcl | btw - how am i meant to add gomunicator to the gpe-image stuff? |
| 23:38.43 | immolo | I added it to meta-gpe.bb |
| 23:39.01 | cp15 | I would add it to blueangel.conf |
| 23:39.16 | cp15 | Not everybody who builds a gpe image wants gomunicator |
| 23:39.35 | lkcl | onto blueangel.conf ain't being a happy bunny. |
| 23:39.52 | cp15 | Why? |
| 23:40.25 | lkcl | i'm getting build errors - that don't exist - in the do_rootfs logfile |
| 23:40.36 | cp15 | Show me |
| 23:40.42 | immolo | my idea was to get it working and then work out how to get it to play nice |
| 23:40.59 | lkcl | ikpkg: cannot create directory "/usr/lib/ipkg': permission denied |
| 23:41.00 | lkcl | ??? |
| 23:41.48 | lkcl | cannot satisfy the following dependencies for gomunicator: |
| 23:42.18 | lkcl | -f /home/lib ... rootfs ... libgpg-error0.postinst ... |
| 23:42.41 | cp15 | Oops, looks terrible |
| 23:42.48 | lkcl | yep :) |
| 23:43.51 | lkcl | hmmm... i never did get that tea... |
| 23:43.52 | lkcl | brb |
| 23:44.50 | cp15 | I think I will do some mattress listening |
| 23:45.15 | cp15 | Good night |
| 23:45.38 | immolo | night |
| 23:52.11 | lkcl | he he, he's off :) |
| 23:52.35 | lkcl | i like that phrase. |
| 23:55.14 | lkcl | hmmmm... immolo - what did you do about the dependency stuff in gomunicator? |
| 23:56.03 | immolo | built alsa-lib |
| 23:56.19 | immolo | that was the only one I had a problem with |