IRC log for #htc-linux on 20070426

01:01.32*** join/#htc-linux elysion_ (n=elysion@k203c.lanhovi.tontut.fi)
04:35.12*** part/#htc-linux _law_ (n=_law_@213.173.86.202)
05:01.36*** join/#htc-linux sadeness_ (n=LamersIn@nat1.rlan.ru)
05:15.07*** join/#htc-linux drno (n=drno@eki.student.utwente.nl)
05:20.48*** join/#htc-linux ellisway (n=ellis@host-87-74-241-174.bulldogdsl.com)
05:21.42*** join/#htc-linux _law_ (n=law@mail.stiftadmont.at)
06:05.48*** join/#htc-linux rob_w|mis (n=rob_w@p549B9FB6.dip0.t-ipconnect.de)
06:25.11*** join/#htc-linux goxboxlive (n=goxboxli@195.159.97.196)
06:25.33goxboxlivegood morning
06:27.53*** join/#htc-linux pleemans (n=peter@d5152D19B.access.telenet.be)
07:13.22*** join/#htc-linux Fossi (n=cmewes@195.177.48.115)
07:44.07*** join/#htc-linux cyrill62 (n=cyrill62@130.252.146.195.dynamic.adsl.abo.nordnet.fr)
07:59.11*** join/#htc-linux cyrill62 (n=cyrill62@130.252.146.195.dynamic.adsl.abo.nordnet.fr)
08:00.22*** join/#htc-linux cyrill62 (n=cyrill62@130.252.146.195.dynamic.adsl.abo.nordnet.fr)
08:32.19*** join/#htc-linux cyrill62 (n=cyrill62@130.252.146.195.dynamic.adsl.abo.nordnet.fr)
08:34.57*** join/#htc-linux rmoravcik (n=rmoravci@pc-3s0zt5w2e4y0vzmhnrzq3a21zqajzfw.users.student.utc.sk)
08:38.09*** join/#htc-linux psokolovsky_ (n=psokolov@empty-crossing.volia.net)
08:55.14*** join/#htc-linux BabelO (n=root@lun34-2-82-238-28-28.fbx.proxad.net)
09:03.29cr2hi BabelO
09:05.55BabelOhi cr2 :)
09:06.08cr2WARNING: Qtopia does not support x86_64. Using -platform linux-g++-32
09:06.08cr2WARNING: Qtopia does not support x86_64. Using -xplatform linux-generic-g++-32
09:06.38cr2i'm not interested in -device zaurusclamshell
09:06.48cr2because it does not support the phone.
09:07.14cr2is greenphone the only feasible target ?
09:07.18BabelOdo you take device/htcphone from xanadux cvs tree ?
09:07.30BabelOi put a configure with right command line
09:07.41cr2no. i think there were some tslib issues.
09:08.12*** join/#htc-linux rob_w|mis (n=rob_w@p549b9fb6.dip0.t-ipconnect.de)
09:08.42BabelOhttp://xanadux.cvs.sourceforge.net/xanadux/opie2/devices/htcphone/
09:08.57cr2ok.
09:09.20BabelOtslib is working with this, the only thing is to point to your OE include directory and OE lib directory
09:10.05cr2hm. i think we should talk to ljp so it will be merged into the main tree.
09:10.42BabelOyes, but have to try with latest opieII sources, because they merged with qtopia 4.2.1
09:11.01cr2ok. trying to build now.
09:12.25BabelOif you get htcphone, you have only to specify -device htcphone to configure
09:12.35BabelOand change some line in configure files
09:15.16cr2ok. i need to become acquainted with the directory structure.
09:16.50BabelOit is simple, all directory in devices/<mobilesdevices>/* can surcharge a class in qtopia tree
09:17.29BabelOonly two special thing are for brigtness and apm that is in custom.cpp files
09:18.14goxboxliveBabelO: Where exactly are the include and libs in OE. I want to try that myself.
09:18.49elliswaymorning all
09:19.44goxboxlivehi ellisway
09:20.44elliswayhi goxboxlive hows it going m8 did u get anywhere the other day
09:21.39*** join/#htc-linux cyrill62 (n=cyrill62@130.252.146.195.dynamic.adsl.abo.nordnet.fr)
09:22.00goxboxlivewith gprs? No i didnt try any firther
09:22.07BabelOgoxboxlive: last time i try OE, i see that sources are decompressed in a directory tree and libs are in imaage directroy tree before building bz2 file
09:22.49BabelOgoxboxlive: do you read log about gprs yesturday ? cr2 say that maybe you miss parameter vendor passed to modprobe usbserial vendor=0xXX
09:23.34elliswayi cant remember what you were trying to build but it was failing during te compile wit yyparse if i remember correctly
09:23.36goxboxliveBabelO: Yes i saw that, i have to try it out when i get home.
09:24.17goxboxliveBabelO: Bu back to OPie2. At Xanadux there is no files in the gnokii folder, there is only empty directories.
09:28.07BabelOgoxboxlive: yes it is normal ;) now with line discipline from pH5 we only need a c file specific to HTC, i don't comit mine because i do  lot of really bad code :)
09:30.06BabelOif you want i can send you my C file for gnokii 0.6.13
09:30.46goxboxliveBabelO: yes thank you
09:34.13*** join/#htc-linux twostupiddogs (n=towstupi@84.18.144.39)
09:34.43twostupiddogsI'm trying to build opie-image
09:35.13twostupiddogsI've created a new OE installation, and some times it stops with the following error: ERROR: Can not check RPATH scanelf not found
09:35.17BabelOmost thing different from AT used by gnokii is that on HTC it seems answer is not on same line
09:35.33twostupiddogsthen when I run bitbake again it goes on. What's wrong here?
09:35.53twostupiddogshi there BTW
09:37.03twostupiddogsbuild system is Fedora Core 6 x86_64
09:40.31cr2BabelO: if it's completely separate directory, why don't we merge it right now ?
09:41.05cr2goxboxlive: /sbin/modinfo usbserial
09:41.21cr2parm:           product:User specified USB idProduct (ushort)
09:41.21cr2parm:           vendor:User specified USB idVendor (ushort)
09:41.21cr2parm:           debug:Debug enabled or not (bool)
09:42.40cr2twostupiddogs: i build it on a dual opteron, and it works.
09:43.24cr2builds a lot of junk at the same time too. did you get the host.conf from xanadux CVS ?
09:43.26BabelOcr2: yes, it is a separate directory, but need to be sure it compile fine before
09:43.35cr2BabelO: ok.
09:43.58twostupiddogsyou updated the cvs with local.conf last time
09:44.28cr2yes.
09:44.54twostupiddogsthen if it't the local.conf you talking about it's the one from xanadux
09:45.14cr2ok.
09:45.18twostupiddogsit seems it is looking for an scanelf program which is not in my system
09:45.59cr2don't have it too.
09:46.27twostupiddogsexacly, maybe some recent changes...
09:46.57twostupiddogsAnyway it seems to be going on with the process, let's see what happens. tnx cr2
09:56.08cr2qbootsourceaccessory.cpp
09:56.22cr2i386 compilation failed here.
09:57.42cr2WARNING: Qtopia does not support x86_64. Using -platform linux-g++-32
09:57.57twostupiddogsyou talking to me?
09:58.14cr2no.
09:58.16cr2:)
09:58.28cr2it's opie2
09:58.37twostupiddogsok, :)
09:58.51cr2opie has zero phone support.
09:59.12cr2opie2 has some hooks, so we will just extend them.
10:00.16twostupiddogscr2: I know, but was just a way to make a start. opie-image is complex enough and somebody else has done it before with success. Then I'll start experimenting...
10:00.36cr2yes, i agree.
10:01.54twostupiddogsby the way, gave me the same error on util-macros_X11R7.1-1.1.5.bb, maybe it's bitbake... what version you using? mine is 1.8.0
10:02.13cr2hm.
10:02.24cr2this is just insane.
10:02.58cr21.6.3
10:03.05twostupiddogsI know those things shouldn't even be in the building. qtembedded does not use X11 right?
10:03.16cr2no.
10:03.33cr2i think it's all because of bluez-utils
10:03.47twostupiddogsho, ok.
10:03.48cr2of maybe some angstrom conspiracy.
10:03.57twostupiddogslol.ù
10:04.07twostupiddogss/ù//
10:04.43cr2:) if you compile gnome-something when you want a qt-based system, i call it a conspiracy.
10:04.59twostupiddogsagreed.
10:05.15BabelO:)
10:06.06cr2i'd be happy to run qt on x11, though.
10:06.07BabelOcr2: what command line do you use to compile qtopia ?
10:06.34cr2BabelO: just playing.
10:06.52cr2what is the real command line ?
10:06.52BabelOlook here it is full command line for arm http://xanadux.cvs.sourceforge.net/xanadux/opie2/devices/htcphone/configure?revision=1.1&view=markup
10:07.05BabelOi like the -D SYSINFO_GEEK_MODE :)
10:07.17cr2:)
10:07.19twostupiddogscr2: anyway, rerunning bitbake goes on... I'll investigate a little. don't want to run bitbake 1000 times :( everytime I build something...
10:07.21twostupiddogs:)
10:07.57cr2BabelO: -extra-qtopiacore-config "-I/usr/local/arm/include"
10:08.06cr2imho this should not be so.
10:08.11BabelOyes this is where i put all my include for arm
10:08.15sadenesstest
10:08.19sadenesss/es//
10:08.23sadenesscoool!
10:08.30cr2-extra-qtopiacore-config "-depths 8,16"
10:08.48cr28 not really needed.
10:09.18BabelOhum maybe
10:09.35cr2BabelO: is tslib included into opie2  ?
10:09.55BabelOsupport is included but ot library
10:10.21cr2headers ?
10:10.36BabelOthat's why there is some -I and -L for bluetooth, dbus and tslib header and library
10:11.13cr2ok, then it's a good idea to integrate it with the oe tree.
10:11.57BabelOi m interesting about helix support
10:12.48cr2BabelO: what is helix, i always forget-
10:12.58cr2some media playaer
10:13.00cr2?
10:13.11BabelOyes
10:13.25BabelOand it support rtsp protocol
10:13.43cr2ok.
10:14.08cr2it is a good idea to sort out the tslib header/lib path
10:14.24cr2so the opie2 compilation will look like 'make opie2'
10:14.45cr2and put the Makefile into CVS.
10:15.10cr2i have it in ~/stuff/tmp/work/armv5te-angstrom-linux-gnueabi/tslib-1.0-r11/tslib-1.0
10:16.03BabelOyes but you need multiple -I line, ine for each module ( dbus, ts, bluetooth)
10:17.03BabelOcr2: https://helixcommunity.org/
10:19.07BabelOlunch time, bbl.
10:19.20cr2BabelO: such file can be autogenerated.
10:28.17twostupiddogscr2: should I worry about a line with: ERROR: libtdl.la failed sanity check ?
10:29.15twostupiddogsBTW I found the scanelf, it's in pax-utils-native
10:30.06cr2afaik, no.
10:30.27twostupiddogss/libtdl.la/libltdl.la/
10:30.37twostupiddogsok. continuing the build ... :)
10:32.31elliswaycr2 the helix community develops the code that was made opensource by realnetworks so its basicly real player real encoder and the real server
10:33.21cr2ellisway: ok.
10:57.05twostupiddogsopie-image: gtk-doc-native.bb...building ... :)
10:59.50twostupiddogsopie-image: glib-2.0-native-2.12.4 .. conspiracy confirmed :)
11:01.59goxboxlivecr2: OE refuse to accept on eof our bug: How do we ommit this? Is there anything that can be done in linux kernel to make it work without having our own /etc/iniit.d/opie file:http://bugs.openembedded.org/show_bug.cgi?id=1733
11:02.29*** join/#htc-linux RoEn_PC (n=roen@p54a671d2.dip.t-dialin.net)
11:06.31cr2goxboxlive: i don't know how opie handles /dev/input/*
11:06.42cr2but your patch looks ugly anyway :)
11:40.25twostupiddogsgoxboxlive: I'm not expert here, but it seems that for the touchscreen the detect-stylus package is not included into your opie-image
11:41.20*** part/#htc-linux rmoravcik (n=rmoravci@pc-3s0zt5w2e4y0vzmhnrzq3a21zqajzfw.users.student.utc.sk)
11:44.59goxboxlivecr2: It is how opie has done it as long as i can remember.
11:45.52goxboxliveBut anyway, i dont know how to make a udev rule so i guess we wouldnt have keyboard support in Opie anymore without any local hacks.
11:47.23twostupiddogsudev is pretty simple to write. I'm looking it up know, though I think that the OE guys are right, there is something in the kernel/driver that should be fixed (this is behind my knowings though)
11:48.05twostupiddogsby the way the keyboard module, is not compiled as a module right?
12:03.47twostupiddogsgoxboxlive: this is an awful hack but it work. put it in your /etc/udev/rules.d/local.rules (at the end is ok)
12:03.59twostupiddogs# Switch the inputs as opie wants them to be
12:04.03twostupiddogsSUBSYSTEM=="input", KERNEL=="event1",   NAME="input/event3"SUBSYSTEM=="input", KERNEL=="event3",   NAME="input/event1"
12:04.09twostupiddogsSUBSYSTEM=="input", KERNEL=="event1",   NAME="input/event3"SUBSYSTEM=="input", KERNEL=="event3",   NAME="input/event1"
12:04.16twostupiddogsSUBSYSTEM=="input", KERNEL=="event1",   NAME="input/event3"
12:04.18twostupiddogsSUBSYSTEM=="input", KERNEL=="event3",   NAME="input/event1"
12:04.37twostupiddogss/SUBSYSTEM=="input", KERNEL=="event1",   NAME="input/event3"SUBSYSTEM=="input", KERNEL=="event3",   NAME="input/event1"//
12:05.02twostupiddogsretring a better post ... :)
12:05.10twostupiddogsput this...
12:05.16twostupiddogs# Switch the inputs as opie wants them to be
12:05.23twostupiddogsSUBSYSTEM=="input", KERNEL=="event1",   NAME="input/event3"
12:05.28twostupiddogsSUBSYSTEM=="input", KERNEL=="event3",   NAME="input/event1"
12:06.16twostupiddogsand of course you have to change the opie start script ... export QWS_KEYBOARD=USB:/dev/input/event1
12:06.24twostupiddogsgoxboxlive: you there?
12:12.27BabelOcr2: compiling ?
12:13.39*** join/#htc-linux Foxdie (i=Foxdie@77-98-48-193.cable.ubr05.harb.blueyonder.co.uk)
12:32.06*** join/#htc-linux pof (n=pof@173.red-62-57-1.user.auna.net)
12:38.10goxboxliveI just dont get it why it is so difficult to have own files in OE for universal. I cant see that hey have removed any files for zaurus, seems like they also need their own files
12:38.14*** join/#htc-linux BabelO (n=root@lun34-2-82-238-28-28.fbx.proxad.net)
12:46.54cr2BabelO: need to do some real work too, so moved it to the evening :)
12:47.28cr2goxboxlive: zaurus is a broken platform to such extent, that you should not take it as an example ;)
12:48.06cr2there is a good reason they have finally decided to drop openzaurus in favour of angstrom.
12:49.28cr2imho, opie should open _all_ files in /dev/input/* to gather events. the "kernel" part is finished and will not be changed.
12:50.26cr2but the device sequence can be completely random, depending on the order of inserted modules. and if they are inserted at all.
12:52.34BabelOcr2: yes i understand, same here, i ve a lot of bug tracker todo today ;)
14:38.08*** join/#htc-linux rmoravcik (n=rmoravci@adsl-dyn49.91-127-60.t-com.sk)
15:49.39*** join/#htc-linux goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
15:51.17goxboxlivei never remember how to read private messages
15:53.51goxboxlivegot it
15:59.50twostupiddogsgoxboxlive: maybe you already dropped this 'cause of cr2 post. For the keyboard/touchscreen issue, maybe OE would accept a patch that modifies the opie start script ...
16:00.01twostupiddogswith something different like distros do.
16:00.32twostupiddogsExample to include a file which is in /etc/sysconfig with init script variables initialization...
16:00.44goxboxlive<twostupiddogs> i dint manage to read what you wrote, but i saw there where something about the keyboard stuff. I have the read the logs first.
16:00.53*** join/#htc-linux ellisway (n=ellis@host-87-74-241-174.bulldogdsl.com)
16:00.55goxboxlivei where at work
16:01.10twostupiddogsok
16:01.14twostupiddogsI'll wait...
16:11.30goxboxlivecr2: Are yo able to build Opie2 with tslib and gnokii?
16:16.15twostupiddogsgoxboxlive: you busy?
16:16.35goxboxlivetwostupiddogs> are you saying that i should add this to the opie startup script after adding the stuff you wrote to udev? export QWS_KEYBOARD=USB:/dev/input/event1
16:17.21twostupiddogsthat was the patch you sent to oe, so I suppose that was the original opie script.
16:17.23twostupiddogsyes
16:17.49twostupiddogsnot add, but use the original script
16:18.01twostupiddogsanyway I got a better solution...
16:18.30goxboxlive<twostupiddogs> But the problme is that they wouldt us to have such machine specific stuff in the file.SO actually, the want us to have it removed, like for the touchscreen
16:18.32twostupiddogsI'd do as many distributions do to change the init scripts withouth touching them.
16:18.55twostupiddogsYes, and I propose this:
16:19.01goxboxliveWe had in the opie startup script export TSLIB blabla /dev/dev/input0 but
16:19.27twostupiddogsdo a patch like this
16:19.29twostupiddogs# Source config
16:19.29twostupiddogsif [ -f /etc/sysconfig/opie ] ; then
16:19.29twostupiddogs<PROTECTED>
16:19.29twostupiddogselse
16:19.29twostupiddogs<PROTECTED>
16:19.30twostupiddogs<PROTECTED>
16:19.33twostupiddogsfi
16:19.34twostupiddogsexport QWS_MOUSE_PROTO
16:19.36twostupiddogsexport QWS_KEYBOARD
16:19.38twostupiddogsthen put this into /etc/sysconfig/opie for htc_universal.
16:19.40twostupiddogsQWS_MOUSE_PROTO=TPanel:/dev/input/touchscreen0
16:19.42twostupiddogsQWS_KEYBOARD=USB:/dev/input/event3
16:20.07twostupiddogswhat is the original opie startup script? Still trying to build opie-image here... :(
16:20.21goxboxliveok i c
16:21.08goxboxliveThe TS isent an issue anymore as it has been fixed with some udev stuff, but the Keyboard still needs some export.
16:22.01twostupiddogsI have not much experience with opie, but the original one has the export or should detect it?
16:22.06goxboxliveBut i dont think that psokolovsky_ will allow me to patch our own file in OE. We should have it common with the other devices
16:22.17twostupiddogsbbl
16:22.22goxboxliveok
16:23.46twostupiddogsback. That's why we propose to not have a specific patch, but a more common one that let's the distro/user decide what options opie should start with
16:24.06twostupiddogsthe one above is just an example, I should check the original one.
16:26.55twostupiddogsbtw the udev hack I wrote is really awful...should not be used. As cr2 has pointed out.
16:27.52goxboxlivehttp://www.openembedded.org/repo/org.openembedded.dev/packages/opie-init/opie-init/opie
16:28.10goxboxlivethis is the original one.
16:28.25twostupiddogsok checking ...
16:29.33goxboxliveWell i dont have any idea how we should do it. It's like, our kernel develpper says that our keybaord stuff is OK and the opie maintainer is saying that we have to fix it. And me, i have no idea, i just gather all the stuff and make a image out of it.
16:30.17twostupiddogsok, the patch you sent was a wrong one, the export of the keyboard isn't there in the original startup script...
16:30.36twostupiddogsI understand, you doing a great work (tnx) :)
16:30.51goxboxlive<twostupiddogs> lesson
16:30.55goxboxlivelisten i mean
16:31.04twostupiddogshere, tell me
16:31.34goxboxliveI know that there was no export of keyboard, that's why i've added it so the HTC Universal could also have keyboard support in OPIE.
16:32.11goxboxliveSecond: The Opie startup script has been changed since my patch where placed as a bug.
16:32.34goxboxliveHOld on an i will make you patch to see how diffrent it is now.
16:32.34twostupiddogsok. now I get it
16:32.36goxboxlivegood
16:34.46goxboxlivejrs@linuxserver:~/cvs/oe/org.openembedded.dev/packages/opie-init/opie-init$ diff  opie htcuniversal/opie
16:34.47goxboxlive11c11
16:34.47goxboxlive<
16:34.47goxboxlive---
16:34.47goxboxlive> export QWS_KEYBOARD=USB:/dev/input/event3
16:35.32goxboxliveSo, we have to figure out how we can make Opie understand where our keyboard is. Like it does with TS
16:35.52twostupiddogsperfect, so the problem here is that opie should get the keyboard automagically, though this is not the case in HTC, thing that it didn't do in the previous own
16:36.09twostupiddogsyou know where it detects the TS?
16:36.22goxboxliveyes in the same file
16:37.11goxboxliveno, actually it doesnt anymore
16:37.23cr2goxboxlive: it is a fundamentally wrong idea to hardcode any device names.
16:37.38twostupiddogsyep, that's what they changed. yes agreed.
16:38.01twostupiddogsBut probably it was for testing, and now they fixed it and wrote some code to discover them out.
16:38.02cr2i have given you an example. i can create a device entry per key in /dev/input/*
16:38.03goxboxliveAll i know is that UDEV is creating a symlink to /dev/input/touchscreen0 form the TS input
16:38.23cr2so UDEV is not a solution too.
16:38.33twostupiddogsyes, but this is in the local.rules... which I don't know from which package it comes from
16:38.41twostupiddogsNope, defenatly
16:38.41goxboxliveAnd after that we didnt need to export TS in the opie script
16:38.50twostupiddogsyes.
16:39.02cr2imagine, you have /dev/input/event0, /deV/input/event1, ... , /dev/input/event70.
16:39.09cr2what are you going to do ?
16:39.14goxboxlivewell as i said, i dont know udev or not. But we cant have it the we have it today
16:39.40twostupiddogsusually a cat /proc/bus/input/devices should give the kind of device they are.
16:40.16twostupiddogsYou already tried with the new opie script?
16:40.34goxboxliveI have no idea how Opie detects a keyboard like it detects the ts. But if we figured that out we wouldnt had this problem anymore.
16:40.44goxboxliveyes
16:41.11twostupiddogsok, I'll do some debugging on this, and get back to you later...
16:41.12*** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be)
16:41.30goxboxliveIt's just to uncomment the export of event3 and the you will see if the KB works or not
16:41.34goxboxlivethx
16:41.36goxboxlivebbl
16:57.34*** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be)
17:01.34*** join/#htc-linux rob_w (n=bob@p85.212.40.40.tisdip.tiscali.de)
17:03.34*** join/#htc-linux asylumed (n=insanity@196.30.118.75)
17:59.21*** join/#htc-linux scorpio16v (n=scorpio1@ACB4D80B.ipt.aol.com)
18:19.19*** join/#htc-linux BabelO (n=kvirc@lun34-2-82-238-28-28.fbx.proxad.net)
18:43.41*** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
18:45.42twostupiddogsgoxboxlive: you here?
18:46.47twostupiddogsit seems not, then BBL
19:15.50*** join/#htc-linux psokolovsky__ (n=psokolov@82.193.98.1)
19:19.05*** join/#htc-linux twostupiddogs (n=towstupi@62.123.180.130)
19:21.28*** join/#htc-linux goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
19:23.39*** join/#htc-linux psokolovsky (n=psokolov@82.193.98.1)
19:25.02goxboxlivepsokolovsky: How is opie figuring out where the touchscreen is?
19:25.04*** join/#htc-linux asylumed (n=insanity@196.30.118.76)
19:25.11BabelOcr2: what thing do we broke today ? ;)
19:25.45goxboxliveBabelO:Do you have wlan worimg now?
19:26.28BabelOgoxboxlive: no wlan :(
19:26.49goxboxliveok
19:26.51twostupiddogsgoxboxlive: got time?
19:27.07goxboxlive<twostupiddogs> yes
19:27.24twostupiddogsok,I found out the workaround
19:28.01twostupiddogsthe so called fix for the touchscreen is only a script that gets called before opie is started
19:28.29goxboxliveok
19:28.37twostupiddogsit simply sets the TSLIB_TSDEVICE variable to /dev/input/touchscreen0
19:28.40goxboxliveand how do we do the same for the keyboard
19:29.23twostupiddogsI have written a udev rule to get a /dev/input/keyboard0
19:29.24goxboxliveYes, and if we could make a similar device for keyboard we could have it working
19:29.24goxboxliveyou go
19:29.30twostupiddogsand the a script that gets called before opie starts
19:29.40twostupiddogsonly on universal
19:29.54twostupiddogsthis way we can set QWS_KEYBOARD in there...
19:30.16twostupiddogspasing scripts ...
19:30.17goxboxliveWhere is TSLIB_TSDEVICE set?
19:30.31twostupiddogsin /etc/profile.d/tslib.sh
19:30.36goxboxliveright
19:30.45goxboxlivehold on.
19:30.56goxboxlivepsokolovsky: Are you here?
19:30.58twostupiddogsso I'v created a /etc/profile.d/htc-univ.sh
19:31.32twostupiddogs#!/bin/sh
19:31.32twostupiddogsHW=`cat /proc/cpuinfo | grep "HTC Universal"`
19:31.32twostupiddogsif [ "$HW" != "" ]; then
19:31.32twostupiddogs<PROTECTED>
19:31.32twostupiddogsfi
19:31.50psokolovskyHi!
19:31.53psokolovskywazzup?
19:32.04goxboxliveok, great, now we need to ask psokolovsky if we can use it
19:32.15twostupiddogsfinishing ...
19:32.29goxboxlivetwostupiddogs explain for psok how we can solve the keyboard stuff in OPie
19:32.30twostupiddogsthe udev rule is in /etc/udev/rules/local.rules
19:32.38twostupiddogsok
19:32.57psokolovskywell, TS(es) and keyboard(s) are different things
19:33.00twostupiddogson Universal we need to set the QTS_KEYBOARD for qt to recognize it
19:33.29psokolovskykernel doesn't multiplex TS so far (one obvious reason is that few devices have >1)
19:34.09psokolovskybut kernel does multiplex all input event sources into /dev/tty0, etc, so input stuff should work without any dirty magic
19:34.10twostupiddogsok, but still qt needs the QWS_KEYBOARD
19:35.06twostupiddogsok, but it seemd that the keyboard input if not handled by opie, it goes wrong
19:35.18psokolovskytwostupiddogs: no, you don't
19:35.34psokolovskytwostupiddogs: what makes you think it's opie at fault?
19:36.02goxboxlivepsokolovsky: Beacuse it works just fine in x11
19:36.39twostupiddogsqte loads the inputs, but then nothing input goes to console, with strange chars. ..
19:36.59twostupiddogsStill this is a workaround...
19:37.40twostupiddogswe use a udev rule to make a /dev/input/keyboard0 link
19:37.41psokolovskywell, it's not a reason, unfortunately. maybe X does that magic which you try beneath the hood. good for X. but opie manages to work without it on many devices, so if uni has issues, the first place to suspect is itself
19:38.16psokolovskyok, first thing you'd want to tell me if keyboard works in plain console
19:38.26twostupiddogswell
19:38.49twostupiddogsunless you start opie without setting the QWS_KEYBOARD
19:39.15twostupiddogsthen it simply writes junk.
19:40.01goxboxlivepsokolovsky: Yes tkeybaord works in plain console.
19:40.06twostupiddogshaven't tryed switching consoles though
19:40.14psokolovskyboot plain kernel with busybox. hit keys on keyboard tell me if anything shows on screen.
19:40.17goxboxliveI have it works
19:40.25twostupiddogsI mean when opie is not recognizing the keyboard
19:41.02goxboxliveYes, when Opie is not recognizing the keyboard it still works in plain console.
19:41.31twostupiddogswill you be there in a 30mins, my wife is waiting for me to dinner :)
19:41.37twostupiddogsgone
19:41.41psokolovskywell, that's strange.
19:42.21goxboxlivepsokolovsky: One person told me something once, it happends because of some USB stuff
19:42.21psokolovskyok, please submit stuff you did to OE bugtracker, I'll work on adding ability to add per-device overrides later
19:43.08twostupiddogsbbl
19:43.13goxboxliveok
19:44.23goxboxliveThis is the only line i need in the opie startup script: export QWS_KEYBOARD=USB:/dev/input/event3 But as cr2 said. It is dirty because sometime it the eventX can be changed and then we have to edit the patch again
19:45.20goxboxlivebbl
19:49.07cr2BabelO: let's start with opie2 compilation and document all the steps.
19:49.36BabelOcr2: ok
19:49.52cr2i've got the svn tree .
19:50.23cr2opie/trunk/*
19:51.00cr2now i want to checkout your htcphone tree from sf.net
19:51.25BabelOi ve to do same step as you :)
19:51.49cr2to opie/trunk/devices
19:51.56BabelOyes
19:53.19*** join/#htc-linux RoEn_PC (n=roen@p54A671D2.dip.t-dialin.net)
19:53.27cr2cvs -d:pserver:anonymous@xanadux.cvs.sourceforge.net:/cvsroot/xanadux login
19:53.53cr2ok, slow but it works.
19:54.24*** join/#htc-linux goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
19:55.40cr2got the whole opie2 dir and need to set the correct links.
19:56.09cr2ln -sf opie2/devices/htcphone .
19:56.21cr2done.
19:56.27cr2what's next ?
19:56.36*** part/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
19:57.53BabelOnext step
19:58.21BabelOedit devices/htcphone/configure
19:58.51cr2-sound-system oss ?
19:58.54BabelOto set up correctly -I and -L to point to your arm library and include file of tslib.h dbus.h and bluetooth
19:59.09BabelOcr2: yes, the most easier to compile
19:59.21BabelOi ve not alsa source tree :(
19:59.31cr2wait :) let's do it step by step, so the description stays in the log.
20:00.26cr2hm. i assume that the whole oe tree is available for the compilation.
20:00.46cr2ok. let's start with oss, although we don't have it.
20:01.01cr2displaysize is settable ?
20:01.14BabelOi wrote a README of our test an then comit it
20:01.35BabelOdisplay size is settable in opie startup script too
20:01.56BabelOwe have alsa oss emulation for the moment
20:02.00cr2ok.
20:02.12cr2-dbus -dbuspath /home/fcr/armlinux/opie/usr
20:02.21cr2what should be in this path ?
20:03.24BabelOthis path is where is your arm dbus installation, so for me it is path in my opie image where dbus is ;)
20:03.31cr2~/stuff/tmp/work/armv5te-angstrom-linux-gnueabi/dbus-1.0.2-r2/dbus-1.0.2/
20:03.46BabelOyou must found usr/bin/dbus and usr/lib/libdbus.so under
20:04.30cr2already compiled ?
20:04.42cr2the source is here ~/stuff/tmp/work/armv5te-angstrom-linux-gnueabi/dbus-1.0.2-r2/dbus-1.0.2/dbus/
20:04.56BabelOyes already compiled
20:05.08BabelOthey do a link directly to it
20:05.44cr2~/stuff/tmp/work/armv5te-angstrom-linux-gnueabi/dbus-1.0.2-r2/install/dbus
20:05.55BabelOyes :)
20:06.13cr2yes, usr/ lib/ and so on.
20:06.33cr2do i need full paths of can use ~ ?
20:06.39cr2s/of/or/
20:06.55BabelOfull path
20:07.29cr2ok.
20:08.02cr2-I and -L what should be here ?
20:08.25cr2also -extra-qtopiacore-config
20:09.12BabelOthe two -I and -L is setup to find all .h file needed and .so
20:10.01cr2but which files are needed ?
20:10.27cr2is it the compiler path ?
20:10.53BabelOno
20:11.01BabelOwait i list my file
20:11.31BabelO-L is for libts.so
20:12.04BabelO-I include is for dbus  libmng.h  libmng_conf.h  libmng_types.h  png.h  pngconf.h  tslib.h  zlib.h
20:12.41cr2that's a lot of packages.
20:13.00cr2greenphone uses
20:13.03cr2-extra-target-qt-config "-qt-libjpeg"
20:13.03cr2-extra-target-qt-config "-qt-libmng"
20:13.03cr2-extra-target-qt-config "-qt-libpng"
20:13.04cr2-extra-target-qt-config "-qt-zlib"
20:13.48BabelOyes it is autodetected
20:13.58*** join/#htc-linux rmoravcik (n=rmoravci@adsl-dyn49.91-127-60.t-com.sk)
20:15.27cr2ok, i'll remove these options and will add something when the compilation will break.
20:15.43BabelOok :)
20:15.45cr2done.
20:16.09BabelOfirst step is to pass configure -device htcphone corerctly
20:16.24cr2i'm in opie/trunk now.
20:17.00BabelOok we will see if your computer is fast :)
20:17.05cr2WARNING: Qtopia does not support x86_64. Using -platform linux-g++-32
20:17.05cr2Checking the compiler (host): OK (GCC 4)
20:17.05cr2Checking the compiler (target): FAIL
20:18.15BabelO:(
20:18.15cr2how does it know where is my arm-gnueabi compiler in the oe tree ?
20:18.15BabelOyou have arm-linux-* in your path ?
20:18.15cr2no.
20:18.15BabelO:)
20:18.15cr2ok.
20:18.35cr2PATH=~/stuff/tmp/cross/bin/:$PATH ./configure -device htcphone
20:18.41cr2the same result
20:19.09BabelOhumm
20:19.38cr2looking into ./configure
20:20.43cr2oh. it's a perl script.
20:20.54cr2is there something like 'bash -x' in perl ?
20:21.42BabelOhumm i don't know :(
20:21.59goxboxlivecr2: Could yo pastebin your configure file when you have it running? I will compare it with mine.
20:22.26cr2ok.
20:22.37cr2it obviously can't find the compiler.
20:22.54BabelOi f i remember i have same problem
20:23.23BabelOi link all my arm-linux-gnueabi-* to arm-linux-*
20:23.36cr2konqueror ./doc/html/buildsystem-user-guide.html
20:23.37BabelObecause i don't understand how he can find it
20:25.42cr2<PROTECTED>
20:25.42cr2<PROTECTED>
20:26.25cr2devices/greenphone/mkspecs/qws/linux-greenphone-g++/qmake.conf
20:26.59cr2you have linux-blueangel-g++
20:27.28cr2and your CC is arm-linux-gcc
20:27.28BabelOhumm yes and no
20:27.48BabelOyes CC is arm-linux-gcc
20:27.54cr2arm-linux-g++
20:28.13cr2AR and strip
20:28.33cr2i suggest to rename blueangel to htcphone
20:28.42*** join/#htc-linux twostupiddogs (n=twostupi@62.123.180.130)
20:28.45cr2don't know how it searches the path
20:29.12BabelOyes it was a try, we can keep linux-arm-g++ for this
20:30.29cr2arm-angstrom-linux-gnueabi-gcc
20:30.39cr2arm-angstrom-linux-gnueabi-
20:30.45cr2is the prefix
20:31.02BabelOok
20:32.17cr2edited.
20:32.36cr2FAIL
20:32.54cr2WARNING: Qtopia does not support x86_64. Using -platform linux-g++-32
20:33.02cr2i don't like this silly message.
20:33.49cr2maybe it's bad to have -xplantform arm ?
20:34.37cr2-xplatform htcphone ?
20:34.39BabelOi m trying same step as you actually
20:35.48cr2yes. hit some bugs, but it goes further.
20:36.28cr2Build type:    linux-g++-32
20:36.29cr2Architecture:  i386
20:36.32cr2???
20:36.54cr2Configuring Qtopia Core (target)
20:36.54cr2configure -embedded arm -platform linux-g++-32 -xplatform
20:37.16cr2stopped it now.
20:37.52goxboxlivewhy? The fist it listing is your host, then it lists your target
20:38.11cr2Configuring Qt (host)
20:38.11cr2configure -platform linux-g++-32 -prefix
20:39.04cr2why does it compile something in a 32bit compatibility mode ?
20:40.08BabelOcr2: it is qt base because of designer and other tools
20:40.20cr2# Qt assumes non-absolute paths to -platform and -xplatform and some tests (eg. x86_64
20:40.20cr2# support) rely on this. If the values are in the Qt depot path then make them relative
20:40.37cr2the x86_64 is supported, so we are doing something wrong.
20:40.54BabelOstrange i have same problem on my computer now :(
20:40.59BabelOi don't understand
20:41.36cr2opie/trunk/qtopiacore/qt/mkspecs/x86_64 does not exist, check -platform parameter is correct
20:42.24cr2linux-g++-64
20:43.28BabelOok i forgot to export PATH to my compiler :)
20:43.38cr2ok, gave up. let's the -m32 work.
20:44.12BabelOnow ./configure -device htcphone is working for me
20:44.17*** join/#htc-linux dcsmith (n=dcsmith@ool-4351bc6d.dyn.optonline.net)
20:44.34cr2-depths 16,32
20:44.51cr2got added to the target somehow.
20:45.42cr2-DQT_EXTERNAL_SOUND_SERVER
20:45.46cr2whatever that it.
20:45.57cr2-no-qt3support ?
20:46.26cr2The tslib functionality test failed!
20:48.28BabelOwhat is your -m32 option ?
20:48.30cr2it does not want to work in x86_64 mode
20:48.30cr2-L ~/stuff/tmp/work/armv5te-angstrom-linux-gnueabi/tslib-1.0-r11/install/tslib
20:48.30cr2added
20:51.07BabelOhumm i m not happy, dbus is desactivated again :(
20:54.27BabelOcr2: new message :) WARNING: Failure to find: qbootsourceaccessory.cpp
20:58.12twostupiddogsgoxboxlive: you there?
20:59.11goxboxlive<twostupiddogs> yes, but i am following the conversation between babalO and cr2 now. They are trying to figure out how to build opie2. I am going to try my self
20:59.47twostupiddogsgoxboxlive: ok then we will see this another time.
21:00.28cr2BabelO: compilation fails because it tries to check for some tslib on the host system
21:00.43BabelOhum
21:00.59goxboxliveyes, i guess this it's best to take the keyboard stuff with psokolovsky. He is the one who wouldnt let us have our own start script.
21:01.19BabelOcr2: i restart configure to have dbus activated
21:06.12cr2tslib.o: ../../../../qt/config.tests/unix/tslib/tslib.cpp
21:06.12cr2<PROTECTED>
21:06.57cr2so the INCPATH should include the tslib headers
21:07.15BabelOyes only tslib.h
21:07.54BabelOhe he i m compiling with dbus :) but not bluetooth
21:08.10BabelOconfigure search for dbus-1.so
21:08.14BabelOno other file
21:10.23cr2but it's only opie/trunk/devices/opie2/devices/htcphone/mkspecs/qws/linux-htcphone-g++
21:10.28cr2and .
21:10.36cr2no -I
21:10.46cr2hm. DEFINES
21:13.54BabelOcr2: i ve started compiling
21:13.58BabelOwaiting for error
21:16.58cr2i've changed the makefile, but it's autogenerated and was overwritten.
21:17.32psokolovskygoxboxlive: well, I don't have aim "to not let", I just want to find real cause of problem.
21:18.11goxboxlivepsokolovsky: Yes of course, we will try to figure something out
21:18.31BabelOcr2: change the .pro file instead
21:20.26twostupiddogspsokolovsky: you got any suggestion on how to debug this out?
21:22.12psokolovskytwostupiddogs: initial idea is apparently to describe the problem fully (for example, keyboard doesn't work - do other buttons work?). then, ultimate way to debug this is to add debugging output to opie's input handler and see what it gets.
21:24.01twostupiddogspsokolovsky: actually I have seen some opie code, OInputSystem, and it says the input devices are good. Though I have no idea from where to go from there.
21:24.07cr2BabelO: which .pro file ? the -L and -I clearly do not progagate thought this build system.
21:24.43twostupiddogsI know this might not be an opie problem, but it's easyer for me to see where it fails so to check the rest
21:24.48goxboxlivecr2: i have the same problem as . It complaines about missing L and I switchs
21:26.12BabelOcr2: it depend -L -I is for opie else you need -L -I for extra config
21:29.59cr2these switches are not included in the generated Makefile
21:30.21cr2i've added an echo to the .configure file where this error message comes out.
21:41.48*** join/#htc-linux MrVanx (n=chatzill@82-69-60-250.dsl.in-addr.zen.co.uk)
21:42.23BabelOtoo long time to compile :(
21:48.06BabelOcr2: you found ?
21:51.53cr2no.
21:52.06goxboxliveis this you bug cr2?
21:52.19goxboxlivehmm hold on
21:52.21cr2?
21:52.47goxboxlivei'll pastebin, but had to reconfigure
21:55.02goxboxlivehttp://pastebin.ca/460165
21:56.00goxboxliveAdn my path to L and I is in the configure file
21:57.27BabelOgoxboxlive: -extra-qtopiacore-config "-I/usr/local/arm/include"
21:57.31BabelOthis line is ok ?
21:57.42*** join/#htc-linux TripleDES (n=sergio@158.pool85-50-141.dynamic.orange.es)
21:58.37cr2yes, it's the same place.
21:58.41goxboxliveyes
21:58.44goxboxliveok
21:59.14goxboxliveBabelO: Are you sure there isent any other thing you have updated since you uploaded to xanadux?
21:59.18cr2BabelO: but it's a configure flag ??
21:59.44BabelOgoxboxlive: no i restart from scrtch like you
21:59.44goxboxliveBecause this is the same problem i have had since i updated your htcphone files
22:00.15goxboxliveohh, then there has to be something in your lib and include that we miss.
22:01.02goxboxliveThis is my configure file: http://pastebin.ca/460173
22:02.36goxboxlivein /usr/locael/arm/include i have only tslib.h
22:02.47BabelOok i look into mine
22:03.16goxboxlivein /usr/locael/arm/lib i have : libts-0.0.so.0  libts-0.0.so.0.1.1
22:03.35BabelOlinuxbox:/usr/local/arm/lib # ls
22:03.35BabelOlibts-0.0.so.0  libts-0.0.so.0.1.1  libts.la  libts.so  pkgconfig  ts
22:03.45BabelOsame :)
22:04.03BabelOcr2: yes all configure flags
22:04.05cr2why didn#t they just include tslib into opie2.
22:04.31BabelOdon't know :( tslib is used as a plugin by qtopiacore
22:04.45BabelOgoxboxlive: and why -I/usr/include
22:05.21cr2i'll edit ./qtopiacore/qt/config.tests/unix/tslib/tslib.cpp
22:05.48goxboxliveBabelO: Could  i just dele that line?
22:06.05*** join/#htc-linux cyrill62 (n=cyrill62@68.250.146.195.dynamic.adsl.abo.nordnet.fr)
22:06.09cr2so it wil always succeed.
22:06.16BabelO;)
22:07.03BabelOgoxboxlive:  you can point the two -I to same directory
22:07.36goxboxliveok
22:08.56cr2did not help.
22:10.41cr2removed tslib
22:10.44BabelOcr2: in which directory are you when it fail, i ll look at my Makefile
22:11.15goxboxliveBabelO: I am shadow building. Are you too?
22:11.33BabelOno
22:11.42goxboxlivecr2: are you?
22:11.57cr2WARNING: Failure to find: qbootsourceaccessory.h
22:12.05cr2WARNING: Failure to find: qbootsourceaccessory.cpp
22:12.06BabelOyes same for me
22:12.14cr2what is shadow building ?
22:12.16BabelOit is new feature :)
22:12.32goxboxliveyou build form another place than the source
22:12.32cr2Qtopia has been configured. You can build Qtopia by running make.
22:12.40BabelOall obj file are in another directory
22:12.59cr2./doc/html/qbootsourceaccessory.html
22:13.02goxboxlivecr2: How did you manage it?
22:14.25cr2without the tslib
22:14.32cr2no touchscreen
22:14.37cr2g++ -c -m32
22:14.38goxboxliveok
22:15.00cr2BabelO: is it only for the qdesigner ?
22:15.25BabelOno, linguistic, and other tools
22:16.51BabelOuic to generate forms from .ui file
22:17.13cr2saw alsa.c somewhere in the process.
22:17.44BabelOi don't understand your problem with tslib :(
22:18.24cr2somewhere deep in the openssl now.
22:18.42cr2maybe it wants the tslib for the host system ?
22:19.16twostupiddogsgoxboxlive: I know this is not as exiciting as building opie2, but I think I found where the keyboard problem is...
22:20.33goxboxlive<twostupiddogs> wow that's great, what was the problem?
22:20.42twostupiddogsThe thing is simple: "we" start opie on tty3, but there is already a getty process on that console
22:20.59twostupiddogschanging the line in inittab did the trick
22:21.32twostupiddogs3:2345:off:/sbin/getty 115200 tty3
22:21.59twostupiddogsIt'd be better to start it somewhere else, like tty5 and the have some special key to get there...
22:22.17cr2goxboxlive: it's your invention :)
22:22.23goxboxliveohh
22:22.42goxboxliveso the tty3 was busy
22:22.56cr2BabelO: it could have used my local qt4.2 installation.
22:22.56goxboxlivehmm that's my fault
22:23.05twostupiddogsyes, so keyboard was busy on the tty
22:23.31twostupiddogsalso if you did get some errors they got to the console .... (try pushing the camera button)
22:24.20goxboxlive<twostupiddogs> have you tried it out with the original initab and not mine, actually it's not mine it's milan who made it. I thought it where clever.
22:24.46twostupiddogsyou mean the one in OE?
22:24.55goxboxlivetwostupiddogs yes
22:25.27twostupiddogsI'm trying ...
22:25.30goxboxlivegood
22:26.11twostupiddogsMNCI?
22:26.19cr2BabelO: still crosscompiling to i386
22:26.33goxboxlivetwostupiddogs hold on i 'll check
22:26.41twostupiddogsor miniinit?
22:27.31goxboxlivehttp://www.openembedded.org/repo/org.openembedded.dev/packages/sysvinit/sysvinit/inittab
22:28.10twostupiddogsok not mnci, the one in mainline.
22:28.15goxboxliveyes
22:30.00goxboxlive<twostupiddogs> DO you also have the symbols working in Opie now? Like / \ , . and so?
22:30.03twostupiddogsok rebooting just to be sure... This one simply opens a shell to the console
22:30.21twostupiddogsNot tryed yet.
22:30.27*** part/#htc-linux MrVanx (n=chatzill@82-69-60-250.dsl.in-addr.zen.co.uk)
22:30.27cr2BabelO: the qt3compat is enabled ?
22:30.28twostupiddogsdon't think so though
22:31.15twostupiddogsok booting up...
22:31.20goxboxlive<twostupiddogs> Ok, i know psokolovsky have added support for kernel input, it doesnt read my kernel-keymap file.
22:32.18twostupiddogsok with that one opie works...
22:32.29twostupiddogswe have not vt1 console though
22:32.32goxboxlivewithout export QWS.....
22:32.48twostupiddogsThe FN key seems not to work
22:33.14goxboxliveI know, but somone told me to just install 'screen' and the we would have multiple terminals
22:33.17goxboxliveok
22:33.36goxboxlivetwostupiddogs great work btw, thx for helping out
22:34.02twostupiddogsyou welcome :)
22:34.05goxboxlivewell i t's getting late (again). good night all.
22:34.30goxboxliveIs there any other way we could have vt1+4?
22:34.39goxboxlivebeside installing screen?
22:34.46twostupiddogsttys are there with messages tty1 kernel tty2 opie output
22:35.30goxboxliveok, so when you are in the plain shell you are able to switch between the tty's?
22:35.41twostupiddogsI think the way you setted this up, is ok, only remove the tty3 from inittab, then you still have tty1, tty2 and tty4
22:35.50twostupiddogsyes
22:36.01goxboxliveok
22:36.16twostupiddogstty4 is not there though now. Was before
22:36.47goxboxliveok
22:36.58goxboxlivewell, i'll try it out tomorrow.
22:37.00goxboxlivenight
22:37.07twostupiddogsnight
22:37.32twostupiddogscr2: you still trying for opie2?
22:38.57cr2twostupiddogs: it's compiling.
22:39.17twostupiddogsI have a problem with opie-image (still) :(
22:39.45twostupiddogsit failes during opie-dagger telling me that SWModule is not defined
22:40.51twostupiddogsactually opientextdlg.cpp:66 undefined type struct sword::SWModule
22:41.23twostupiddogsand then swmgr.h forward declaration of swmgr.h
22:41.35twostupiddogsis this some kind of gcc incompatility?
22:43.12cr2don#t know.
22:43.17cr2failed here too.
22:43.20cr2make[5]: *** No rule to make target `qbootsourceaccessory.cpp', needed by `.obj/release-shared/qbootsourceaccessory.o'.  Stop.
22:44.02twostupiddogsI think I saw something out there talking about this issue or something like that. It was from x86_64
22:44.24twostupiddogsI read a post somewhere that qtopia wasn't compiling on x86_64 :(
22:44.35twostupiddogsBut don't remember actually where
22:44.55cr2created it with 'touch' moving firther now :)
22:45.14twostupiddogs:)
22:45.25*** join/#htc-linux BabelO (n=pocketir@193.253.141.90)
22:45.50BabelOsorry internet issue
22:45.54twostupiddogsIs the cross compiler using my distro compiler?
22:46.38cr2?
22:46.53cr2BabelO: have you compiled opie2 ?
22:48.11twostupiddogsit's version 4.1.2 (the cross one)
22:48.34cr2yes.
22:48.53cr2arm-angstrom-linux-gnueabi-g++
22:49.36twostupiddogsok then this is correct. I'll check the sources to see what exactly is wrong here...
22:50.47BabelOno same problem as you with qboot
22:50.57cr2it does not like make -j 3
22:51.36cr2simple make is too slow :)
22:51.49BabelOwhat is this file qboot
22:52.17BabelOi use make j 2but computer cpu is full
22:53.15cr2i think they did not make dependencies right.
22:54.32cr2in phoneapp now :)
22:55.01BabelOyes it continue with touch
22:56.58BabelOthere is lot of change since my last try
22:57.52cr2it's something to ask ljp about.
22:58.04cr2~ping ljp
22:58.07aptpong ljp
22:59.49BabelOnot here
23:00.18cr2htcphonehardware.cpp:48:18: error: misc.h:
23:00.27cr2htcphonehardware.cpp:49:20: error: gnokii.h: No such
23:00.28BabelOyour computer is to fast
23:00.42cr2LOL
23:01.10BabelOyes maybe have to comment for the moment
23:01.11cr2how do i fix it ?
23:02.51cr2if ((toclean[i]!=0x02) && (toclean[i]!=0x16))
23:03.02cr2this ir redundant now.
23:03.08BabelOyes
23:03.44BabelOlot of thing like this
23:04.24BabelOyou can comment content of readgsmdata
23:04.36cr2what about gnokii ?
23:05.16BabelOi use gnokii 0.6.13
23:05.27cr2what is the current version ?
23:06.17cr2Current gnokii version: 0.6.14
23:06.19BabelODon t know i try lot of version and only one compile for arm
23:07.22cr2we can contact them about fixes to the CVS version.
23:07.49cr2how is gnokii binded into the opie2 build process ?
23:08.22BabelOwe put it in libraries
23:08.37BabelOand add it to pro files include line
23:08.58cr2so i need to link it to the 3rdparty ?
23:09.50BabelOyes
23:09.55cr2like dbus
23:10.12cr2what is library/gsm ?
23:10.31twostupiddogscr2: ok bypassed this opie-dagger..
23:10.41cr2ok. it's voice codec.
23:10.49cr2twostupiddogs: good.
23:11.13twostupiddogsI think it's a bug. Had to add an include inside the opentextdlg.h (???)
23:11.17BabelOyes provided into qtopia
23:12.04cr2ok, which pro do i need to edit ?
23:12.15BabelOcr2 it is late and internet didn t come back
23:12.25cr2twostupiddogs: contact psokolovsky
23:12.29cr2BabelO: ok.
23:13.35cr2we have moved forward. that's good.
23:13.41cr2good night :)
23:13.57BabelOit is in src server directory
23:14.45twostupiddogsi will. night
23:14.52BabelOgood night

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