**** BEGIN LOGGING AT Sun Feb 19 10:59:57 2006 Feb 19 12:59:52 hrm Feb 19 13:00:03 my OE tree has too much uncomitted stuff in it Feb 19 13:00:15 hey zecke Feb 19 13:02:40 hey Feb 19 13:02:52 koen: did you setup postgres already? Feb 19 13:03:23 koen: i have got an seperate dir where i place my patches, like hrw (http://www.hrw.one.pl/2005/10/31/my-openembedded-enviroment/en/) Feb 19 13:04:22 zecke|breakfast: I've complete step one Feb 19 13:04:43 zecke|breakfast: but I don't seem to have ./setup-postgres.pl Feb 19 13:04:50 (and it's postgres 7.4) Feb 19 13:13:20 zecke|breakfast: read ~/SETUP.ewi Feb 19 13:19:26 hi zecke|breakfast Feb 19 13:20:20 koen: so you want my precious tbox3 code? Feb 19 13:22:00 does anyone know a way to turn off the blinking cursor on a VT? Feb 19 13:22:42 CoreDump|home: look at QtE code, it does a special echo Feb 19 13:22:52 CoreDump|home: I will search it now Feb 19 13:23:01 zecke: thanks a ton zecke Feb 19 13:23:20 bootsplash uses \033[?25l\000 which doesn't work for me Feb 19 13:23:26 err wait Feb 19 13:24:02 ::ioctl(KDSETMODE:KD_GRAPHICS) Feb 19 13:24:30 that is probably better, but not applicable from within a shell script? Feb 19 13:24:36 one more reason to have a python shell Feb 19 13:24:46 yea Feb 19 13:25:11 OT: All Java based UML tools suck Feb 19 13:25:24 perhaps even s/Java// Feb 19 13:25:59 that "l" looked like a "1" (one) in my editor.... Feb 19 13:26:19 mickeyl: how far is umbrello by now? (kde's uml tool) Feb 19 13:26:39 mickeyl: most tools are already bad at creating class diagrams :} Feb 19 13:26:59 yeah, class diagrams - that far Feb 19 13:27:13 but i need uml 2.0 components Feb 19 13:27:26 mickeyl: http://uml.sourceforge.net/screen.php Feb 19 13:28:34 mickeyl: which java tools did you try? Feb 19 13:30:07 koen: you toyed with my minimo sources? Feb 19 13:30:26 zecke: yes, I like toying with minimo Feb 19 13:30:46 zecke: poseidon, magicdraw, umodel, microtool objectif, ... Feb 19 13:30:54 more or less all suck Feb 19 13:30:59 guess i'm going back to visio Feb 19 13:30:59 heh Feb 19 13:31:31 mickeyl: there is this one starting with A* and I think of 'Aggro Berlin' :} Feb 19 13:31:46 argouml? Feb 19 13:32:01 mickeyl: also try the builtin tool in eclipse and maybe even try the proprietary plugin for eclipse Feb 19 13:32:02 argo's just the OSS fork of poseidon Feb 19 13:32:07 mickeyl: right Feb 19 13:32:09 ah okay Feb 19 13:33:54 'I already asked for people able to recompile oe using pdaXrom's libc but noone answered ... ' Feb 19 13:34:00 * koen wonders why.... Feb 19 13:34:01 ;) Feb 19 13:38:45 he would have better demanded... Feb 19 13:39:51 koen: okay, you can svn co tinderbox3 from my home directory Feb 19 13:40:25 mickeyl: http://www.omondo.com/ <- did you try this one? Feb 19 13:42:15 koen: svn co file:///home/zecke/tbox_repo/tinderbox3 Feb 19 13:42:33 koen: or svk mirror it first Feb 19 13:46:31 CoreDump|home: I stopped searching but I think this is okay Feb 19 13:47:23 zecke: I guess my response was swallowed by a reconnect, sorry. I already have found the magic echo sequence Feb 19 13:47:40 CoreDump|home: I have seen it l vs 1 :) Feb 19 13:47:53 oh ok =) Feb 19 13:48:40 say hello to instant display of gpe-bootsplash with even the largest of themes =) Feb 19 14:08:29 hello Feb 19 14:08:53 koen: you guessed just right the theme relationship to that abiword bug. Feb 19 14:10:29 koen: Now I am wondering... would that be a abiword bug or a theme bug? Anything I can do to try to debug that? (bug #1520) Feb 19 14:15:38 poli: I suspect it's a bug in the clearlooks engine or gtk+ itself, abiword just highlights it Feb 19 14:18:13 koen: Something I can do to help debug that? Feb 19 14:18:27 attack it with gdb I guess Feb 19 14:18:42 pb_ would have better instructions Feb 19 14:21:24 koen: I have versions of both GTK+ and abiword with symbols and non-stripped but I saw gdb die when trying to run --sync abiword. :( Feb 19 14:25:42 koen: Oh! Great, I go the backtrace! :) Feb 19 14:30:55 koen: do you want to continue with the setup - I'm busy in getting the tinderclient working Feb 19 14:31:06 zecke: sure Feb 19 14:31:12 svn just refused to work Feb 19 14:31:15 (as usual) Feb 19 14:31:21 lol Feb 19 14:31:29 koen: in what way did it refuse? Feb 19 14:31:41 "Go aways your license has been revoked" Feb 19 14:31:43 berkely erros Feb 19 14:32:14 lol... Feb 19 14:32:26 wait a second :} Feb 19 14:32:49 koen: why is your svn defaulting to berkely? Feb 19 14:32:56 no idea Feb 19 14:33:02 svn sucks Feb 19 14:33:04 that's why ;) Feb 19 14:33:28 agreed Feb 19 14:37:53 koen: try again Feb 19 14:39:14 hi all Feb 19 14:43:54 koen: bt results have been posted on #1520 Feb 19 14:46:24 http://www.linuxdevices.com/news/NS6787213741.html Feb 19 14:48:45 koen: and? Feb 19 14:49:07 i saw ... so nice Feb 19 14:49:21 also pegasusII board are nice :) Feb 19 14:50:18 i would see something like this with a Xscale 80219 and compation .. such as Kontron XP8 Feb 19 14:51:33 koen: how is your 'loft' system? Feb 19 14:52:15 zecke: it has power now, it's just waiting for me to buy a switch and a null modem cable Feb 19 14:52:35 koen how much it cost ? Feb 19 14:52:52 gremlin[it]: $350 Feb 19 14:53:03 zecke: http://dominion.kabel.utwente.nl/koen/loft/images/loft-ps_small.jpg Feb 19 14:53:45 koen: nice couche Feb 19 14:54:06 zecke: yeah, the couches arrived yesterday Feb 19 14:54:21 hurray for parent giving away their 'old' couches ;) Feb 19 14:54:29 parents* Feb 19 14:54:52 red, Al loves red Feb 19 14:56:15 zecke: btw, cirrus also uses opie Feb 19 14:56:21 not qtopia Feb 19 14:56:25 * koen chuckles Feb 19 14:57:55 zecke: it seems pg isn't running Feb 19 14:58:09 at least not listening on localhost Feb 19 15:04:46 koen: did you edit the hba conf? Feb 19 15:06:26 zecke: yes Feb 19 15:06:53 koen: how do you test the connection? Feb 19 15:07:06 bitbake:/home/zecke# grep tinder /etc/postgresql/7.4/main/pg_hba.conf Feb 19 15:07:10 host tbox tinderuser 127.0.0.1/32 md5 Feb 19 15:07:17 zecke: nmap and telnet Feb 19 15:20:30 koen: tbox is the database name? Feb 19 15:22:37 poli: it should be Feb 19 15:23:09 zecke: I know. But I thought tbox might be an unusual name for a DB, so that's why I asked. Feb 19 15:26:20 poli: unusual in what sense? Feb 19 15:26:57 zecke: never seen a dB called box. :) Feb 19 15:35:04 koen: what is the error you get? Feb 19 15:35:34 DBI connect('dbname=tbox;host=localhost','tinderuser',...) failed: could not connect to server: Connection refused Feb 19 15:36:28 koen: try using psql Feb 19 15:36:57 zecke: the setup-postgres script bails out Feb 19 15:38:25 koen: OT: linux boots on a imac :) Feb 19 15:38:39 imacs are ugly Feb 19 15:38:46 gentoo as well Feb 19 15:39:09 koen: what other apple desktop system would you buy? Feb 19 15:39:51 zecke: probably a mac mini Feb 19 15:40:05 zecke: but I use my powerbook as desktop anyway Feb 19 15:40:10 koen: I still wait for the intel ones Feb 19 15:40:15 The new mac minis are already intel? Feb 19 15:40:21 koen: I need a intel system for work :} Feb 19 15:40:24 poli: not yet Feb 19 15:40:48 ~lart pdx.freedesktop.org/cvs Feb 19 15:40:48 * ibot installs PocketPC on pdx.freedesktop.org/cvs's PDA Feb 19 15:40:50 zecke: psql does access tbox Feb 19 15:40:59 using -l localhost? Feb 19 15:41:03 and what user are you? Feb 19 15:42:13 koen: By net or local? Feb 19 15:42:50 probably local Feb 19 15:43:02 ostgres@bitbake:/home/tinderbox3/tinderbox3/sql$ psql -h localhost -U tinderuser -W tbox Feb 19 15:43:06 Password: Feb 19 15:43:08 psql: could not connect to server: Connection refused Is the server running on host "localhost" and accepting TCP/IP connections on port 5432? Feb 19 15:44:58 koen: if you specify -h localhost the client will do a network connection, and not use Unix sockets. Feb 19 15:45:12 koen: if you do not specify -h localhost, it will try unix sockets Feb 19 15:45:43 koen: your line on pg_hda.conf only covers network connections, for unix sockets you should replace "host" with "local" Feb 19 15:46:51 koen: So, psqsl -h localhost -U tinderuser -W does a network connect (host entry on pg_hba.conf) and psql -U tinderuser -W will do a sockets connect (local entry on pg_hba.conf) Feb 19 15:48:05 koen: and just to make sure: you have to reload (or restart) the postmaster daemon when you change pg_hba.conf Feb 19 15:49:42 gosh the tinderbox documentation sucks Feb 19 16:02:16 poli: I've done a /etc/init.d/postgresql-7.4 restart Feb 19 16:02:19 a few times Feb 19 16:05:06 koen: Oh... Feb 19 16:05:11 koen: Now I remember... Feb 19 16:05:21 koen: The order of that line IS important... Feb 19 16:06:10 koen: If there is any other line before that one that your user will fit, like the catch-all host all all 127.0.0.1 255.255.255.255 ident sameuser Feb 19 16:06:41 koen: you will never get to your auth line... so, make sure it is before any other line that your user/host combination might fullfill Feb 19 16:08:12 poli: it seems the server isn't even listening to port 5432 Feb 19 16:10:15 koen: is tcpip_socket enabled in postgresql.conf ? Feb 19 16:12:49 koen: telnet localhost 5432 is refused? Feb 19 16:20:01 I am tring to build OE for the Nokia770 and it fails on building the libosso_0.8.9 package with the following error: | Searching for the D-BUS | checking checking for d-bus... configure: error: Can not locate the D-BUS, please use the switch "--with-dbus-root-dir" to specify its location. | FATAL: oe_runconf failed Feb 19 16:21:39 After a surch base on some guidens from zecke I found that problem is that configure is looking for .service in dbus-1.0/service , but I can't seams to find this file and location any were in the OE distrebution or build directory Feb 19 16:21:54 Any one ran into this type of problem? Feb 19 16:27:25 Insid the libosso-0.8.9 package in configure.ac you can see the following lines: # Looking for DBUS Feb 19 16:27:25 echo [Searching for the D-BUS] Feb 19 16:27:25 AC_MSG_CHECKING([checking for d-bus]) Feb 19 16:27:25 if test "x$DBUS_ROOT" != "x" ; then Feb 19 16:27:25 if test -x "$DBUS_ROOT/bin/dbus-daemon-1" ; then Feb 19 16:27:27 DBUS_ETC="$DBUS_ROOT/etc" Feb 19 16:27:29 DBUS_VAR="$DBUS_ROOT/var" Feb 19 16:27:31 AC_MSG_RESULT([found in $DBUS_ROOT]) Feb 19 16:27:33 else Feb 19 16:27:35 AC_MSG_RESULT([NOT found in $DBUS_ROOT]) Feb 19 16:27:37 AC_MSG_ERROR([Can not locate the D-BUS, please use the correct location "--with-dbus-root-dir" to specify its location.]) Feb 19 16:27:42 fi Feb 19 16:27:44 else Feb 19 16:28:46 It looks lke the error is a result of not finding a valid path in $DBUS_ROOT, my question is any one can help me with pointing me to what need to be fixed. Thanks a lot for the help Feb 19 16:33:57 ghellman: it probably needs a newer dbus with nokia patches Feb 19 16:34:30 i'm assuming there is no way to get something back that move just oblitorated Feb 19 16:36:04 poli: setting tcpip to true fixed it, thanks Feb 19 16:40:41 zecke: step 1 - 3 have been done Feb 19 16:44:20 koen: good, (I don't remember which steps these where..) Feb 19 16:45:03 zecke: apache and postgres Feb 19 16:45:15 koen: ah then you can setup the tinderbox alone :) Feb 19 16:45:42 zecke: http://ewi546.ewi.utwente.nl/tinderbox/ is already working :) Feb 19 16:47:49 koen: good - you can hack it a bit (e.g. give your superuser rights) Feb 19 16:48:04 koen: it uses bugzilla authentication Feb 19 16:51:56 koen: now you could start a cron job for the tbox_updater Feb 19 16:52:04 hi Feb 19 16:52:28 zecke: hourly? Feb 19 16:52:30 hey hrw Feb 19 16:54:31 ~seen mickeyl Feb 19 16:54:37 mickeyl is currently on #oe #opie #handhelds.org. Has said a total of 11 messages. Is idling for 3h 22m 52s, last said: 'argouml?'. Feb 19 16:55:19 koen: were can I get a newer dbus with the neccecary nokia patches? Feb 19 16:56:34 koen: yes that is fine Feb 19 16:57:06 ghellman: I wanted to say http://repository.maemo.org/pool/maemo/ossw/source/d/dbus/ , but that only has 0.23.1 :( Feb 19 16:57:34 zecke: can tbox_updater run as root? Feb 19 16:57:51 zecke: or does it has to run as tinderuser/tindbox3? Feb 19 17:01:35 wow, this new 1U server is easily the loudest machine in the whole building Feb 19 17:03:10 never liked to be close to those 1U machines Feb 19 17:03:23 it only has two tiny fans, but they seem to spin at immense speed Feb 19 17:03:34 kind of like having a small jet aircraft parked on the bench Feb 19 17:03:39 at previous work we used Compaq one as DVB machine - loud as hell Feb 19 17:04:23 hm, does anybody remember the magic commandline thing to turn off fbcon in debian-installer? Feb 19 17:05:03 ah, here we go. "linux debian-installer/framebuffer=false" Feb 19 17:05:10 pb__: F3 or F4 Feb 19 17:09:19 oh, right, I never knew about that Feb 19 17:09:23 ah well, it's installing now in any case Feb 19 17:09:51 annoyingly as well, the VIA motherboard didn't fit in the 1U case without removing the audio connector. very tedious. Feb 19 17:11:53 zecke: you should have write access to /home/tinderbox3 as well Feb 19 17:13:28 pb__: epia? Feb 19 17:14:42 yeah Feb 19 17:14:56 there, another two mailinglist subscriptions dropped Feb 19 17:14:57 it's one of the epia-m boards with a vertically mounted 3-way audio socket Feb 19 17:15:12 gosh I'm annoyed Feb 19 17:15:21 this HTTP post is not working Feb 19 17:15:29 the top socket fouls the backplate, and the socket assembly sticks out of the top of the case by about four millimetres. Feb 19 17:15:33 zecke: how unfortunate Feb 19 17:20:53 03koen 07org.oe.dev * r3c8d7192... 10/conf/distro/familiar-unstable.conf: familiar-unstable: remove myself as maintainer as the hh.org folks have made it clear they want to take over again. It was a fun couple of years, good luck with it Feb 19 17:23:04 someone tested sources availability during last months? Feb 19 17:23:22 hrw: yes, see my post from some days ago Feb 19 17:24:34 ah.. yep - I even replied to it. Feb 19 17:27:36 hrw: I'm not working on familiar anymore, and I kind of suspect that the hh.org people dislike OE and won't use it, so .oz354fam083 is strictly OZ now Feb 19 17:27:39 argh.. kmail always hung for minute when read koens mail Feb 19 17:28:05 koen: in other words: Familiar drop OE? Feb 19 17:28:12 hrw: probably Feb 19 17:28:34 I was informed yesterday that they have a complete roadmap ready and stuff Feb 19 17:29:03 So instead of wasting my time I decided to work on something usefull Feb 19 17:29:06 (angstrom) Feb 19 17:31:52 cu Feb 19 17:31:55 cu Feb 19 17:31:55 have to go Feb 19 17:34:26 koen: oh Feb 19 17:38:22 zecke: ? Feb 19 17:38:35 koen: regarding familiar and hh.org crew Feb 19 17:40:04 zecke: it's unfortunate, but all the 'OE sucks' comments the past years were a small clue Feb 19 17:40:48 zecke: but I'm looking ahead, and hope we can cooperate with them on arm stuff with angstrom Feb 19 17:41:50 luckily http post is working again Feb 19 17:41:55 :) Feb 19 17:42:56 zecke: I'm going to make some chili (and eat it), ping me if you need some assistance with setting up tbox on ewi Feb 19 17:44:32 you rang mastah? Feb 19 17:45:02 chili... I would love to eat it... maybe sometime Feb 19 17:46:22 koen, :( Feb 19 17:46:48 its too bad that hh.org thing that way Feb 19 17:46:53 think* Feb 19 17:47:24 but in relation to it OE has had some serious issues lately Feb 19 17:47:59 lately being the last year-ish Feb 19 17:48:20 with bitkeeper, DNS, monotone, broken tools, etc Feb 19 17:55:13 then again hh.org had issues for quite a while as well Feb 19 18:04:00 man bitbake's error reportings sucks as well Feb 19 18:04:22 yeah it does Feb 19 18:04:31 File "", line 28, in tinder_format_http_post Feb 19 18:04:47 lol Feb 19 18:05:15 28 is not even the offset in this method :} Feb 19 18:07:10 hey zecke, while i think of it, is there a list somewhere of what doesnt work fully in 2.6 and the arm chips? Feb 19 18:08:39 i'd liek to do this dbus thing but its going to be a large amount of work for 2.4(not sure if its possible even) and wondering if it would be better to try and help more with 2.6 Feb 19 18:10:13 koen|food: I didn't realise familiar felt like that. Its rather sad :-( Feb 19 18:11:37 if someone need I have hostap-daemon 0.4.8/0.5.1, hostap-modules 0.3.10/0.4.7, wpa-supplicant 0.4.8/0.5.1 recipes - http://ewi546.ewi.utwente.nl/tmp/hrw/OE-recipes/ (will push them tomorrow) Feb 19 18:11:37 yeah :( Feb 19 18:12:08 hrw|gone, i think shadows was playing with wpa-supplicant last night Feb 19 18:12:27 er no it was simcop2387 Feb 19 18:13:01 someone familiar with it should give info does WPA make sense without openssl or not - we currently ship 0.2.6 build without ssl support Feb 19 18:13:46 that was the discussion they were having lastnight i belive Feb 19 18:14:07 i think they decided that you need sll Feb 19 18:14:10 ssl* Feb 19 18:15:02 it gives an ssl error without it or something Feb 19 18:15:07 you want the irc log? Feb 19 18:15:26 give me date & time - I have own log Feb 19 18:15:32 looking now Feb 19 18:16:22 hrw|gone: shipping openssl by default is a bit icky licensewise Feb 19 18:16:29 and it's ~1.5MB Feb 19 18:16:48 Feb 18 19:47:23 shadows hey why the heck does wpa-supplicant-nossl_0.2.6-r3 build? Feb 19 18:16:51 starts there Feb 19 18:16:57 hmm Feb 19 18:17:08 10:23:10 AM Feb 19 18:17:12 is my current time Feb 19 18:17:24 19:19 here Feb 19 18:17:26 my backlog starts on 26.01.2006 Feb 19 18:17:34 19:19 too here Feb 19 18:17:39 lol Feb 19 18:17:47 i ment for time-offset reference Feb 19 18:18:15 **** BEGIN LOGGING AT Wed Feb 2 22:41:42 2005 Feb 19 18:18:27 so we probably end with no wpa-supplicant in images and req users to install it by hand ;( Feb 19 18:18:30 i think its time to rotate that log .. Feb 19 18:18:34 cu for now Feb 19 18:18:45 cant we use a different ssl library? Feb 19 18:19:01 emte: probably, but someone has to look at it Feb 19 18:19:06 hrw|gone: it depends on your target usage, but in Familar (for example) shipping the non-ssl supplicant in the images would be the right answer. Feb 19 18:19:13 I imagine openzaurus is probably the same. Feb 19 18:19:56 * emte vaguly remembers a "small" ssl library that was gpl Feb 19 18:21:02 http://www.peersec.com/matrixssl.html Feb 19 18:21:05 i think that was it Feb 19 18:21:13 emte: http://www.matrixssl.org/ Feb 19 18:21:13 hey people Feb 19 18:21:15 hrw|gone: alternatively, if you could build the supplicant against gnutls then that would be fine, given that (iirc) we ship those libraries already. But virtually no users would benefit from it, and the ones that would can probably install it for themselves, so I wouldn't lose too much sleep over it. Feb 19 18:21:16 heh, too slow Feb 19 18:21:25 50KB total footprint Feb 19 18:22:05 hmm Feb 19 18:22:17 lists arm7 but nothing before ... Feb 19 18:22:31 crikey, what earlier cpus do you want? Feb 19 18:22:41 before arm7 was pretty much the dark ages Feb 19 18:23:17 pb_: you coming to FOSDEM? Feb 19 18:23:21 I'm not sure that linux even still runs on anything earlier than arm7 Feb 19 18:23:22 mithro: yeah Feb 19 18:23:50 I will be staying at the official Team Opie hotel, which I think is the Astrid. Feb 19 18:23:59 i thought the sa-1110 was pre arm7 ... Feb 19 18:24:04 you staying at Astrid? Feb 19 18:24:23 no, the sa-1110 was roughly contemporaneous with the arm8 Feb 19 18:24:25 how much did it cost you? Feb 19 18:24:30 ahh Feb 19 18:24:39 er, dunno, I'll have to look it up Feb 19 18:24:41 hold on Feb 19 18:25:03 154 euros Feb 19 18:25:13 for 2 nights? Feb 19 18:25:14 for 2 night, right? Feb 19 18:25:16 yeah Feb 19 18:25:21 hmm matrixssl is already in OE Feb 19 18:25:34 matrixssl? Feb 19 18:25:39 mithro: http://www.chab.be/ is a bit cheaper, if you don't mind hostels Feb 19 18:25:44 http://www.peersec.com/matrixssl.html Feb 19 18:25:54 mithro, discussing wpa-supplicant Feb 19 18:26:07 and avoiding openssl licensing Feb 19 18:26:46 62 euro per night at the astrid seems good then? Feb 19 18:27:27 koen: how close is it to the Astrid? Feb 19 18:28:01 mithro: no idea Feb 19 18:28:10 koen: the first tinderbox log has been posted :} Feb 19 18:28:16 I don't even know where the astrid is Feb 19 18:28:23 zecke: cool Feb 19 18:28:30 http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=1&logfile=20060219193209.log Feb 19 18:28:32 zecke: is it working now? Feb 19 18:28:50 zecke: rock Feb 19 18:28:58 lol Feb 19 18:29:16 hm, not rock, I just get "No such log found" Feb 19 18:29:25 Software error: Feb 19 18:29:25 No such log found! at /home/tinderbox3/tinderbox3/showlog.pl line 29. Feb 19 18:29:30 yup me too Feb 19 18:29:43 pb_: I removed all logs :) Feb 19 18:29:52 http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=2&logfile=20060219193433.log Feb 19 18:29:52 zecke: very prudent Feb 19 18:30:01 zecke: can I be your deputy ;) Feb 19 18:30:04 ? Feb 19 18:30:05 hey pb_ , i've a question for you Feb 19 18:30:25 emte: what is your question? Feb 19 18:30:26 koen: you can be a superuser as well Feb 19 18:30:50 pb_, since sysfs doesnt exist for kernel 2.4 do you have any ideas how to get kernel events onto the dbus ? Feb 19 18:31:03 pb_: tbox3 has less features than tbox2, but it appears to be more solid Feb 19 18:31:03 what sort of kernel events? Feb 19 18:31:14 emte: the general answer is "you can't", but there is a bridge for hotplug Feb 19 18:31:38 zecke: tinderbox3 is the one that freedesktop uses, right? Feb 19 18:31:44 right Feb 19 18:31:49 * pb_ lost in a maze of twisty tinderboxes, all alike Feb 19 18:31:55 * france is back (gone 22:43:53) Feb 19 18:31:59 yeah the hotplug stuff is tehre as well as terminal login, looking more for device information like cpu, battery, fans, etc Feb 19 18:32:13 pb_: well you only have three versions - with different set of features Feb 19 18:32:24 btw just for peoples information :P Feb 19 18:32:25 emte: I don't think you get any events for those things at all in 2.4 Feb 19 18:32:29 Then on the 21st I catch a ferry at 1900 Feb 19 18:32:29 overnight to Salerno (arriving about 1400 on the 22nd) in Italy. On the Feb 19 18:32:29 23rd I catch a fast train direct to Milan at 0830 (arriving about 1600). Feb 19 18:32:29 Then I catch a plane from Milan (leaving 0840) to Brussels (arriving at Feb 19 18:32:29 1005) on the 24th in the morning! Feb 19 18:32:35 mithro: thanks for keeping us posted Feb 19 18:32:57 zecke: I am easily confused Feb 19 18:33:15 pb_, yeah thats what i kind of thought :( thanks Feb 19 18:33:35 pb_: are you sure you want to commit to GPE not QPE? Feb 19 18:33:42 * zecke tries to exploit pb_ Feb 19 18:33:58 zecke: remind me which one of those uses X? Feb 19 18:34:10 pb_: you mean Y? Feb 19 18:34:15 lol Feb 19 18:34:24 gtkfb ? Feb 19 18:34:32 oh, maybe I do Feb 19 18:34:39 is that QPE? Feb 19 18:35:15 quite possibly extinct Feb 19 18:36:37 does anybody have an idiot's guide to creating data DVDs on debian? I'm not sure which packages I need. Feb 19 18:36:38 anyone know of a list of sa-2.6 problems ? Feb 19 18:36:48 hi does anybody have any problems building font-config native on the release branch? for gpe-image I think the fault is the "/home/norm/.fonts": Invalid argument? is this an error in my setup? Feb 19 18:36:53 for some reason the cdrecord maintainer seems to be making some kind of moral stand against dvd support Feb 19 18:37:08 pb_, i get it occasionally Feb 19 18:37:16 it seems to depend on DVD type ... Feb 19 18:37:33 dvd-rw, I think Feb 19 18:37:44 I forget exactly which kind of drive I bought for this machine Feb 19 18:38:27 like i said i run into it occasionally, sometimes it doesnt like me and other times it works fine ... Feb 19 18:38:52 but i am fairly positive it doesnt liek DVD+R Feb 19 18:39:06 I know k3b can do it, but that's a bit too hardcore for me Feb 19 18:39:40 i've been using nero lately Feb 19 18:40:09 heh, I don't think nero is in debian Feb 19 18:40:14 but thats more to do with lightscribe support (soon hopefully) Feb 19 18:40:22 pb_: rewritables or not? Feb 19 18:40:24 nah you have to buy it Feb 19 18:40:27 mithro: yeah Feb 19 18:40:37 dunno about rewritables Feb 19 18:40:48 well, non rewriteables would be a start Feb 19 18:40:49 heh Feb 19 18:40:51 i have Feb 19 18:40:53 ii dvd+rw-tools 5.21.4.10.8-4 DVD+-RW/R tools Feb 19 18:40:53 ii dvdrtools 0.2.1-1 DVD writing program Feb 19 18:41:00 i dont expect and open version of lightscribe will popup any time soon Feb 19 18:41:00 ah, maybe dvdrtools is the magic bullet Feb 19 18:41:10 and then i just use cd creator in nautilis Feb 19 18:41:12 or... not Feb 19 18:41:19 E: Couldn't find package dvdrtools Feb 19 18:41:29 pb_: Hello, Phil! :) I was wondering if the back trace information I managed to obtain with gdb for bug #1520 is enough, or if I should obtain some other info? Feb 19 18:41:33 hrm... Feb 19 18:41:36 poli: hello! Feb 19 18:41:54 pb_: unstable? Feb 19 18:41:59 poli: you need to figure out what function in abiword the final address in the backtrace (the 0x005something one) corresponds to Feb 19 18:42:01 mithro: no, stable Feb 19 18:42:47 poli: if this is an abiword you built yourself, and you have the unstripped binary on hand (on your build host), that should be just a matter of invoking arm-linux-addr2line on it. Feb 19 18:43:34 if that isn't the case, things are more difficult. in that situation, the easiest option is probably just to rebuild it and retest with a binary that you did build yourself. Feb 19 18:43:54 pb_: I can't get the abiword I compile with symbols and unstripped to run, gdb dies in the process (I assume not enough memory). Feb 19 18:44:10 poli: that's okay, you can strip the one on the ipaq Feb 19 18:44:29 stripping doesn't change the layout of the executable, it just throws away the debug info Feb 19 18:44:41 just keep a copy of the binary before stripping to refer to later Feb 19 18:44:56 pb_: Ok... What can I use to strip it? :) Feb 19 18:45:10 poli: arm-linux-strip Feb 19 18:45:20 strip it on the build host, then copy it to the ipaq Feb 19 18:45:57 you might find that stripping with "--strip-debug" gives you a binary that's small enough to use but still enables gdb to print symbol info Feb 19 18:46:19 pb_: i'm on unstable Feb 19 18:46:22 dunno about stable Feb 19 18:46:26 okay, thanks Feb 19 18:46:30 do you have dvdrecord? Feb 19 18:46:45 no Feb 19 18:46:50 maybe I need to get those bits from unstable Feb 19 18:46:51 Package dvdrecord is not available, but is referred to by another package. Feb 19 18:46:56 However the following packages replace it: Feb 19 18:46:56 dvdrtools Feb 19 18:47:00 odd Feb 19 18:48:14 pb_: I have a copy of arm-linux-strip I downloaded with a toolchain. Is there some other one in OE? Feb 19 18:48:23 poli: check your tmp/cross/bin directory Feb 19 18:48:36 it should be either in there, or in tmp/staging/.../bin Feb 19 18:48:46 pb_, there is a version in stable ... Feb 19 18:48:55 ah, I wonder why I don't have it Feb 19 18:49:06 of dvdplusrw Feb 19 18:49:10 ah Feb 19 18:49:25 lion:/home/pb# apt-cache search dvdplusrw Feb 19 18:49:25 lion:/home/pb# Feb 19 18:49:26 hrm Feb 19 18:49:39 ftp://ftp.us.debian.org/debian/ Feb 19 18:49:45 is the only stable feed i have Feb 19 18:49:58 okay Feb 19 18:50:03 well, I don't need to fix this right now in any case Feb 19 18:50:12 dvd+rw-tools (5.21.4.10.8-2) Feb 19 18:50:23 gotta wait for a replacement for the damn drive, which arrived bent into an amusing shape, first Feb 19 18:50:31 lol Feb 19 18:50:46 need better couriers Feb 19 18:51:31 I think it was the supplier's fault, actually. Feb 19 18:51:47 the packaging wasn't damaged, and the drive was inside about a mile of bubble wrap Feb 19 18:52:02 so, I'm pretty sure it was like that when they sent it out Feb 19 18:52:17 hmm Feb 19 18:52:32 did you insult the salesman? Feb 19 18:54:13 heh. no, not knowingly. Feb 19 18:55:38 now, the next thing I need is a plan for providing encrypted file service to windows clients. Feb 19 18:55:43 not really sure how to go about that. Feb 19 18:55:47 pb_: Got the new position. How do I find it in the non-stripped version? Feb 19 18:56:03 poli: use arm-linux-addr2line Feb 19 18:56:09 (see "man addr2line") Feb 19 18:58:19 isnt that what corba is all about? Feb 19 19:01:56 pb_: addr2line can't find any of the stated addresses. I wonder if that address maybe in another file that is not AbiWord-2.4? Feb 19 19:02:42 possibly. how big is AbiWord-2.4? Feb 19 19:02:59 pb_: a little over 15M Feb 19 19:03:04 you can check /proc//maps on the ipaq while the binary is running to find out what files make up its image Feb 19 19:03:57 pb_: how weird... theere isn't a dir for its proc number... Feb 19 19:04:14 pb_: ops my mistake Feb 19 19:05:38 i've said it before and i'll say it again ..... Feb 19 19:05:46 * ade|desk hates SAP R/3 Feb 19 19:06:34 pb_: 00452000-0070f000 rwxp 00452000 00:00 0 [heap] Feb 19 19:06:46 pb_: the address is 0x005955a8 Feb 19 19:06:59 ah, that is weird Feb 19 19:10:06 pb_: the other libX11 address seem right Feb 19 19:10:20 pb_: how can that breakpoint be on the heap? Feb 19 19:10:43 dunno. if you disassemble the code at that address, does it look sensible? Feb 19 19:10:49 it's possible that gdb is just confused Feb 19 19:10:59 see ya all at FOSDEM! Feb 19 19:11:25 mithro: see you Feb 19 19:12:19 pb_: newbie question: how do I do that? :) Feb 19 19:12:42 disassemble 0x005955a8 0x005955c0 Feb 19 19:12:43 or some such Feb 19 19:13:28 pb_: a8 is a ldreqb r4, [r9], #-224 Feb 19 19:13:54 pb_: rest of them until bc are andeq r0, r0, r1 for the first one and andeq r0, r0, r0 for all the others Feb 19 19:14:19 * poli is ARM-illiterate Feb 19 19:14:59 that looks like garbage Feb 19 19:15:16 try starting a few instructions earlier; change ...a8 to ...90 or something Feb 19 19:15:38 0x005955a0: andeq r0, r0, r0 Feb 19 19:15:38 0x005955a4: andeq r0, r0, r9, asr r0 Feb 19 19:15:38 0x005955a8: ldreqb r4, [r9], #-224 Feb 19 19:16:04 mm, that's no code Feb 19 19:16:14 looks like gdb has gotten the address wrong Feb 19 19:16:34 pb_: any way to fix that? Feb 19 19:17:32 * emte compiles gpe-image for the h3600 to findout what doesnt work Feb 19 19:17:59 poli: not that I can think of right now, other than grovelling through the stack by hand Feb 19 19:18:04 pb_: maybe attaching gdb after the proccess is already running might make a difference? Feb 19 19:18:08 (which wouldn't be all that difficult, I guess, for one level) Feb 19 19:18:19 poli: worth a go, but I doubt it will change anything Feb 19 19:19:39 hmm Feb 19 19:19:52 less ../org Feb 19 19:19:55 bah Feb 19 19:19:56 i think my 1gb sd card has badblocks :( Feb 19 19:23:22 pb_: I am thinking on setting up a temporary swap partition on an CF I have and try to get the debug version to run. Feb 19 19:23:43 good idea Feb 19 19:24:01 that'll be fun Feb 19 19:24:13 swap over loopback device over nfs! Feb 19 19:25:06 over bluetooth! Feb 19 19:26:00 over ssh tunnel! what else Feb 19 19:27:02 pb_: that's weird.... arm-linux-strip apperently corrupted a --keep-debug-only version I made... can't run even outside gdb "Syntax error: word unexpected (expecting ")") " Feb 19 19:27:25 oh dear Feb 19 19:28:01 Well, I will get a swap later when I come back and run the full version. Hope that will give us some information. Feb 19 19:28:19 pb_: thanks for the help! :) Feb 19 19:42:43 ah found the answer to the error for bitbake fontconfig-native was to chmod u-w /home/norm/.fonts Feb 19 19:43:33 simcop2387: nfs over network is not working Feb 19 19:43:44 simcop2387: use network block device instead :) Feb 19 19:44:20 heh Feb 19 19:44:51 ok so, swap on nbd over ssh tunnel over bluetooth, what else can i fit in there :) Feb 19 19:44:53 but there must be a better way, anybody know of one? Feb 19 19:45:28 or is there something wrong in my setup Feb 19 19:47:40 RP: I did add prinks....should just a straight printk("rc int"); work? Feb 19 19:47:45 RP: still saw nothing Feb 19 19:48:13 RP: I suppose I should add some to the attach/detach functions, although I *do* know that they're happening since a new input device is registered Feb 19 19:48:20 RP: and thanks for evtest, that looks very useful Feb 19 19:49:32 RP: I'll try further debug messages to see if I can figure this out Feb 19 19:52:37 hi reenoo Feb 19 19:52:52 hey pb_ Feb 19 19:52:55 evening all Feb 19 19:53:36 hi Feb 19 19:54:02 hey reenoo Feb 19 19:54:48 ok i've got an apparrant opie problem, with the opie-image i built last night something keeps mucking up /etc/resolv.conf on my sd-card when i boot with altboot, according to fsck it contains deleted inodes and causes some i/o errors Feb 19 19:54:57 it happens during init Feb 19 19:55:05 hi shadows Feb 19 19:55:07 yo koen Feb 19 19:55:36 the wpa-supplicant needs to either be removed (not good) or upgraded to 0.4.7 Feb 19 19:55:58 the newer versions of wpa-supplicant have support for updating the config file using the wpa-supplicant socket interface Feb 19 19:56:09 this means you no longer need to hand-edit the wpa-supplicant.conf Feb 19 19:56:16 yay! Feb 19 19:56:30 I tried once but got lost in all the params Feb 19 19:56:45 just have a program do the configuration via the socket control interface, and the changes will be saved (provided you had the correct option in the config file already allowing the config updates) Feb 19 19:56:54 it works great on my debian laptop computer Feb 19 19:59:35 it then makes making an opie or gpe gui for it much easier and nicer Feb 19 19:59:43 oh yeah Feb 19 19:59:57 there's an example gui written for Qt Feb 19 20:00:26 it's called 'wpa_gui' and is included with a recent wpa-supplicant release Feb 19 20:00:29 hmmm Feb 19 20:00:42 qt is a bit too heavy to justify installing it Feb 19 20:01:06 oh, on gpe-image it doesn't make sense Feb 19 20:01:18 I had it installed to test lyx :) Feb 19 20:01:40 better to integrate new code into the network devices configuration Feb 19 20:02:08 why is wpa-supplicant-nossl dragged into the rootfs of a spitz openzaurus gpe-image build? Feb 19 20:02:28 and how to fix that so instead there is the 0.4.7 version of wpa-supplicant Feb 19 20:02:45 wpa-supplicant is probably GPL Feb 19 20:03:14 and the clamshells duplicate handheld-common.conf for the most part iirc Feb 19 20:03:51 i tried modifying it, but i don't know what i'm doing there. i managed to make wpa-supplicant build, but only wpa-supplicant-nossl was on the rootfs Feb 19 20:03:59 would someone mind having a look at it? Feb 19 20:04:29 the PREFERRED_VERSION thing that RP added shouldn't be necessary Feb 19 20:04:45 it's ignored anyways because the thing explicitly depends on wpa-supplicant-nossl Feb 19 20:04:56 shadows: again, if wpa-supplicant is GPL we can't ship a version linked against openssl. you're free to do that for your own images though. Feb 19 20:05:00 ohhh Feb 19 20:05:04 is that the issue huh Feb 19 20:05:17 i didn't understand the licensing clash Feb 19 20:06:42 reenoo: if it works with matrixssl, that would be okay then? Feb 19 20:07:46 or maybe gnutls Feb 19 20:08:04 * france is away: Away Feb 19 20:08:30 gnutls would be fine (and is probably in the images already). I don't know about the licensing situation with matrixssl Feb 19 20:08:45 GPLed looks like Feb 19 20:08:52 i'll try with gnutls then Feb 19 20:10:11 yeah, gnutls would be fine Feb 19 20:10:36 the reason we ship the -nossl supplicant at the moment is that openssl is a big library and most (nearly all) users don't need the tls functionality anyway Feb 19 20:11:03 hm Feb 19 20:11:17 gnutls is indeed already in the images, so if the supplicant can be linked against that it would give the best of all world Feb 19 20:11:28 i will focus on that goal Feb 19 20:11:32 =) Feb 19 20:12:10 okay, excellent Feb 19 20:13:31 reenoo, you read my reply you you re http://www.handhelds.org/~bugzilla/show_bug.cgi?id=1551 ? Feb 19 20:15:21 as i dont think its a duplicate Feb 19 20:16:55 well, you put three issues in a single bug report Feb 19 20:17:32 03freyther 07org.oe.dev * r154c5bc8... 10/classes/tinderclient.bbclass: Feb 19 20:17:32 classes/tinderclient.bbclass: Feb 19 20:17:32 Begin of the rewrite to support the tbox3 protocol Feb 19 20:17:36 03freyther 07org.oe.dev * r44337a3d... 10/classes/tinderclient.bbclass: Feb 19 20:17:36 classes/tinderclient.bbclass: Feb 19 20:17:36 -Save the assigned machine id to a file Feb 19 20:17:36 and restore it when needed. BitBake does Feb 19 20:17:36 not have a 'persistent' cache one could use Feb 19 20:17:37 three? Feb 19 20:17:40 across files... Feb 19 20:17:42 -Post the configuration directly to the server Feb 19 20:17:44 03freyther 07org.oe.dev * re216e540... 10/classes/tinderclient.bbclass: Feb 19 20:17:46 classes/tinderclient.bbclass: Feb 19 20:17:48 -Send the right messages to the TinderBox Feb 19 20:17:50 -Read the log from the logfiles and send it to the tinderbox Feb 19 20:17:52 -Add the wish for a more sane and more featurefill reporting Feb 19 20:17:54 (e.g. less data, less often) Feb 19 20:17:58 03freyther 07org.oe.dev * r55fb04a1... 10/conf/tinder.conf: Feb 19 20:18:00 conf/tinder.conf: Feb 19 20:18:02 Update to the tinderclient.bbclass. We only report Feb 19 20:18:04 using HTTP as this is the only way to do reports Feb 19 20:18:06 for tinderbox3 Feb 19 20:18:14 anyone feels like testing? Feb 19 20:18:16 mmm in my view its one issue causing related problems Feb 19 20:18:27 emte: your own interpretation of them actually. it's a bit difficult to figure out what the actual problems are Feb 19 20:18:58 yeah, i'll agree Feb 19 20:19:19 zecke: testing what exactly Feb 19 20:19:28 our new tinderbox :) Feb 19 20:19:33 hm Feb 19 20:19:36 not me. Feb 19 20:19:39 emte: the time on h3600s is reset to 1970 on shutdown or reset. there's nothing we can do about that. Feb 19 20:19:44 there are no real reports of related behavior, it seems peopel just accept it or there are few h36 users reporting it Feb 19 20:19:49 shadows: do you do regular OE builds? Feb 19 20:20:03 zecke: not very regular. i rebuild a lot while i'm fixing bugs yes Feb 19 20:20:18 actully reenoo it resets to image build date Feb 19 20:20:36 today i will be working on libmb font problem, and also the wpa-supplicant Feb 19 20:20:37 emte: no, that's /etc/timestamp Feb 19 20:20:48 there is no /etc/timestamp Feb 19 20:20:56 reenoo: the time shouldn't be getting reset on h3600s. it's only h3900s that have this problem. Feb 19 20:21:06 h3600s have the other problem, where the clock gains 20 minutes on reboots Feb 19 20:21:16 if its a propershutdown it resets to shutdown time but that is not stored in /etc/timestamp either Feb 19 20:21:38 emte: what you are describing does sound exactly like /etc/timestamp. Feb 19 20:21:55 but it doesnt exist Feb 19 20:22:29 pb_: so h3800s and h3600s differ in that respect? my h3870 certainly does the 1970 thing (and always has). Feb 19 20:22:38 reenoo: hm, that's odd Feb 19 20:22:42 no, h3800s and h3600s should be the same Feb 19 20:22:47 and my 36$something, too Feb 19 20:23:03 it definitely did use to work on h3600s. maybe it's gotten broken since. Feb 19 20:23:11 (well, modulo the 20-minute thing) Feb 19 20:23:28 emte: what does the end of your /etc/init.d/bootmisc.sh look like? Feb 19 20:23:46 paste the last half-dozen lines here Feb 19 20:24:02 k, sec Feb 19 20:24:09 it shut itself down again Feb 19 20:24:13 * reenoo reboots h3870 to check Feb 19 20:25:23 if test -e /etc/timestamp Feb 19 20:25:23 then Feb 19 20:25:23 date -s `cat /etc/timestamp` Feb 19 20:25:23 fi Feb 19 20:25:31 tahts the part your looking for i expect Feb 19 20:26:48 it would work if /etc/timestamp ever got created i'd think Feb 19 20:27:36 that is indeed the bit Feb 19 20:27:36 * emte marks bug as duplicate Feb 19 20:27:48 so what is supposed to generate timestamp? Feb 19 20:28:03 it's created during the image build process, and by the halt and reboot scripts. Feb 19 20:28:16 er, s/and by/and updated by/ Feb 19 20:28:34 just to be sure, try deleting that section from bootmisc.sh and confirm that the problem still occurs. Feb 19 20:28:35 seems all those are failing except build Feb 19 20:28:54 the problem will occure Feb 19 20:29:11 test -e /etc/timestamp will never pass Feb 19 20:30:33 does "grep 2006 /etc/*" find any files? Feb 19 20:30:53 root@h3600:~# grep 2006 /etc/* Feb 19 20:30:53 grep: /etc/resolv.conf: No such file or directory Feb 19 20:30:53 /etc/version:200602091743 Feb 19 20:31:01 (replace 2006 with whatever your year is getting set to, obviously, if that isn't 2006) Feb 19 20:31:31 seems unlikely that it's coming from /etc/version, but I guess anything is possible Feb 19 20:32:03 it gets set to feb 09 2006 Feb 19 20:35:01 pb_: ok, I have to correct myself. it doesn't happen on a shutdown/reboot or a hard reset now. I guess it only happened with a flat battery. sorry for the confusion Feb 19 20:35:23 reenoo: ah right. yeah, in case of flat battery there is obviously nothing that can be done. Feb 19 20:37:11 pb_: yeah, I guess the only time my ipaq is reset is when the battery runs low after two or three weeks. thus the confusion ;) Feb 19 20:37:24 hehe Feb 19 20:37:56 hmm Feb 19 20:38:18 hrm, drat, my boss is awake. seems like I have to write this document I've been putting off all weekend. Feb 19 20:38:22 * pb_ invokes abiword Feb 19 20:38:32 heh Feb 19 20:38:41 http://oe.pastebin.com/563233 Feb 19 20:39:08 if its supposed to do this on reboot etc, why does it only test if file exists ? Feb 19 20:39:34 so that it only updates /etc/timestamp if it exists Feb 19 20:39:41 rather than creating it if it didn't previously exist Feb 19 20:39:49 i am wondering if this means /etc/timestamp was supposed to be created in the initial image? Feb 19 20:39:58 pb_: hmm... I guess I should go back to working on my logic assignments too Feb 19 20:40:17 I guess you should. Feb 19 20:40:35 emte: depends on your DISTRO. I think openzaurus has it; familiar doesn't. Feb 19 20:40:56 (for reasons outlined in bug 1526) Feb 19 20:42:51 yeah ... i cant say i fully agree with it Feb 19 20:43:25 are the osb*.bb files broken? or is this my bitbake? Feb 19 20:44:01 zecke: or both! Feb 19 20:44:25 which bits do you disagree with? Feb 19 20:44:26 so my solution is to ingnore the fact that my ipaq cannot keep decent time? Feb 19 20:44:31 ignore* Feb 19 20:44:50 well, no, your solution is to figure out where your clock is getting set from and stop that from happening. Feb 19 20:46:10 but i dont think thats the problem .... its more that the ipaq shuts down on long suspends and RTC does not keep running Feb 19 20:46:31 is there a mirror for cvs://anoncvs:anoncvs@pdx.freedesktop.org? Feb 19 20:46:51 if the ipaq is in the cradle it can remain suspended indefinatly Feb 19 20:47:11 the behavior only heppens when its on battery Feb 19 20:47:15 happens* Feb 19 20:48:35 * emte thinks on it for a few Feb 19 20:49:13 Need some help: The passphrase for my private key for OE is not accepted by monotone (I used the key half year ago). I tried with my standard passwords, no success. Is there a chance that my key is *disabled*? Feb 19 20:52:19 yuecelm: sure is, there was a change a few month back when all keys had to be re-activate manually Feb 19 20:56:02 To reactivate my key: Is this specified in the mailing list 'Enabling push access again' from koen? Feb 19 20:57:55 yuecelm: right Feb 19 21:01:09 And this issue stops me to change my passphrase (with 'monotone chkeypass ')? Feb 19 21:01:23 yuecelm: no Feb 19 21:01:52 yuecelm: I had an issue where a single byte char is nowadays a multibyte char (modern distributions) Feb 19 21:02:10 yuecelm: I had to write the pw into the monotonerc to unlock they key :) Feb 19 21:04:20 is freedesktop.org up or down? Feb 19 21:05:52 * Philippe is back (gone 46:16:27) Feb 19 21:07:48 03koen 07org.oe.dev * r398c56dd... 10/ (3 files in 2 dirs): Feb 19 21:07:48 Angstrom: Feb 19 21:07:48 * add main distro config file Feb 19 21:07:48 * add 2006.9 release file Feb 19 21:07:48 * add angstrom-version to brand the rootfs Feb 19 21:07:48 NB: eabi is not enabled yet Feb 19 21:12:11 zecke, since the only time the IPAQ stores the (semi)correct time is durring an explicit shutdown -R, which means that it writes(still has) the time in the sa-1110's RCNR register? So that leads me back again to a problem with the suspend code? Feb 19 21:14:08 how it gets intially set to the image compile date i've no clue ... or atleast i cannot find any traces of why Feb 19 21:14:32 hey, attention all, the X11 related stuff needs to be updated Feb 19 21:14:39 xlib has been moved to the git tree Feb 19 21:14:56 for which X11 Feb 19 21:14:58 anoncvs is no longer available from pdax.freedesktop.org; use anoncvs.freedesktop.org instead Feb 19 21:15:02 not sure Feb 19 21:15:11 it's broken at least when building openzaurus for clamshell (spitz) Feb 19 21:16:28 koen@dominion:/data/build/oe/monotone/org.openembedded.dev$ grep pdx -rn . | wc -l Feb 19 21:16:29 65 Feb 19 21:16:30 ouch Feb 19 21:16:48 should be move to $FDO_CVS or something Feb 19 21:16:51 moved* Feb 19 21:21:09 koen: sed ^_^ Feb 19 21:21:34 :) Feb 19 21:25:54 * emte likes short shorts Feb 19 21:26:21 okay how do i turn console logging back on in my ipaq>? Feb 19 21:26:43 not that enableing loging via syslog helps durring a reset Feb 19 21:28:20 pb_: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=4&logfile=20060219223119.log Feb 19 21:28:23 any idea Feb 19 21:29:03 zecke: ah, you hit the bitbake bug Feb 19 21:29:15 zecke: http://handhelds.org/hypermail/oe/65/6559.html Feb 19 21:30:00 haha Feb 19 21:30:08 where is this holger you are talking about? Feb 19 21:30:24 zecke: seems that your autoreconf is way old Feb 19 21:30:32 AUTOV is 1.4-p6 Feb 19 21:30:34 that's no good Feb 19 21:30:57 koen: could you update it? Feb 19 21:31:08 zecke: update what? Feb 19 21:31:16 autoreconf? Feb 19 21:31:26 zecke: I guess your bitbake has forgotten to build autoconf-native or something. Feb 19 21:31:35 ah okay, then this is the bug :} Feb 19 21:31:39 er, automake-native Feb 19 21:32:27 yah, that does sound like a bug Feb 19 21:32:41 drat, dropped my ipod on the floor again. gotta stop doing that. Feb 19 21:33:12 pb_: blame apple - you might end up with a refund this way Feb 19 21:33:16 good plan Feb 19 21:33:27 they shouldn't have made them so slippery Feb 19 21:34:19 luckily they also seem to be quite bouncy Feb 19 21:34:38 and it's kind of cool that they automatically pause when the headphone plug falls out. Feb 19 21:35:53 pb_: which one do you cal yours? Feb 19 21:36:02 that was to make dropping it a "feature" Feb 19 21:36:02 :P Feb 19 21:36:07 and I hope you run linux on it Feb 19 21:36:58 zecke: one of the older ones. I think it was the model before they added video support. Feb 19 21:37:04 heh, no, no linux Feb 19 21:37:28 pb_: you are likey my gal... Feb 19 21:37:45 ! Feb 19 21:38:13 that was a bit out of the blue Feb 19 21:38:25 hehe Feb 19 21:38:43 well I asked her if she wants to run linux on it... Feb 19 21:38:57 and she didn't see the extra value of a liberated device Feb 19 21:40:01 ah Feb 19 21:40:04 yeah, in that case I am like her Feb 19 21:40:13 you should just showed her a pic of a guy holding a baby and the caption that says "let linux liberate your ipod" Feb 19 21:40:23 :P Feb 19 21:40:52 I like running linux on my computers, even little ones like the ipaq, but I have no great desire to make my dvd player or my car or my ipod run linux. Feb 19 21:41:10 lol Feb 19 21:42:09 i'd rather my car ran linux than windows ... Feb 19 21:42:26 I used to feel the same way about my cellphone, but Motorola's UI sucks so much that linux would, in fact, be an improvement there. Feb 19 21:42:32 even if it just had a bash prompt or something Feb 19 21:42:39 lol Feb 19 21:43:28 * emte gives up and installs pixiehunter on my ipaq Feb 19 21:43:55 i am now convinced its the pixies who set the intial time on reset Feb 19 21:57:22 zecke: thinking of the xlib thing that shadows mentioned earlier, you might find http://lists.freedesktop.org/archives/xorg/2006-February/013113.html interesting. Feb 19 21:57:40 obviously Keith's point 3 is of little relevance, but still. Feb 19 21:58:29 1) is questionable from what i've read in HG Feb 19 22:00:46 pb_: what does XCB mean? :} Feb 19 22:01:12 zecke: X C Bindings, iirc Feb 19 22:01:50 basically a new low-level (!) API for driving X. Feb 19 22:02:18 oh keith is at intel :} Feb 19 22:02:27 need to watch Lost on DVD - how wonderful :} Feb 19 22:02:44 zecke: you have it on DVD? Feb 19 22:03:05 koen: well I gave it Al :} Feb 19 22:03:18 can make a decentral safety copy if you want to :} Feb 19 22:03:36 zecke|lost: I'm looking for season 2 Feb 19 22:03:39 zecke: massey has even done some formal methods work with XCB, which might appeal to you. Feb 19 22:04:13 http://www.usenix.org/events/usenix02/tech/freenix/full_papers/massey/massey_html/index.html Feb 19 22:20:45 anyone tried evlog ? Feb 19 22:21:40 tring to figure out if it works with syslog or is a replacement Feb 19 22:26:17 okay any ideas where i might find the clock getting set? Feb 19 22:26:34 i've ran out of spots to look Feb 19 22:32:30 03koen 07org.oe.dev * r4a58674f... 10/packages/linux/ep93xx-kernel_2.6.15.bb: ep93xx kernel: new patch from Lennert Buytenhek Feb 19 22:40:50 * Philippe is away: visual contact - melancholic dreams Feb 19 22:44:23 03koen 07org.oe.dev * r7a3ad19f... 10/packages/linux/ep93xx-kernel_2.6.15.bb: ep93xx kernel: use r1 patch which includes a defconfig Feb 19 22:53:53 is gnumeric suposed to compile? Feb 19 22:54:25 yes Feb 19 22:54:38 everything outside nonworking/ is supposed to compile Feb 19 22:56:24 * france is back (gone 02:48:22) Feb 19 22:56:30 .\ Feb 19 22:58:26 'night all Feb 19 23:04:42 night koen|sleep Feb 19 23:20:05 * emte wonders why we unpack all the kernel26 docs ... Feb 19 23:42:09 koen|sleep: good nite Feb 19 23:42:20 gosh OE is currently in a miserable state Feb 19 23:43:21 zecke|lost: oh dear Feb 19 23:43:26 sorry to hear that you are lost as well Feb 19 23:43:36 hehe Feb 19 23:43:38 good nite Feb 19 23:43:46 ah, zecke is found Feb 19 23:43:47 hurrah Feb 19 23:43:57 http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=4&logfile=20060219230624.log Feb 19 23:44:21 good nite guys Feb 19 23:44:31 at least I have an agenda for the next week :} Feb 19 23:47:18 zecke: heh. it's building the wrong kernel again? Feb 19 23:49:02 looking at the config, it is Feb 19 23:49:18 reenoo: yes, wrong DISTRO was set Feb 19 23:49:49 ah, right. that would do it. Feb 19 23:50:11 mm. should try to get those jlime dudes to set COMPATIBLE_HOST in their kernel .bb files. Feb 19 23:50:54 cc1: error: invalid option ‘l’ Feb 19 23:50:58 wow, what crazy option is that? Feb 19 23:51:08 * pb_ phears the sh3 kernel Feb 19 23:51:08 heh Feb 19 23:51:13 hehe Feb 19 23:51:20 * reenoo watches his terminal freak out Feb 19 23:51:52 good nite Feb 19 23:51:57 zecke: night night Feb 19 23:52:06 at least this tbox looks like it will work more than 42 hours Feb 19 23:52:06 pb_: arguably, we should just move that stuff to nonworking and see if anyone complains. smells like bitrot... Feb 19 23:52:34 ~lart bots Feb 19 23:52:35 * ibot nabs the moon and broadsides bots with the sea of tranquility Feb 19 23:53:13 reenoo: dunno. I imagine it works fine if you're actually targeting superh. Feb 19 23:53:43 pb_: yah, guess I'm just a bit grumpy ;) Feb 19 23:54:06 been fighting this stupid grading bot all day long... Feb 19 23:54:18 doh Feb 19 23:55:19 well, I finished writing that document, so I guess my work for the day is done. Feb 19 23:55:27 first order logic is undecidable and well it's hitting that area Feb 19 23:55:29 which is lucky, because there seem to be only two minutes of the day left Feb 19 23:56:22 so, effectively it just gives me the middle finger for several assignments Feb 19 23:56:32 heh, that's no good Feb 19 23:56:47 * reenoo nods Feb 19 23:57:25 my solutions are quite likely correct, but there's literally no way to prove that Feb 20 00:00:29 mm, that is unfortunate Feb 20 00:03:40 so is "correct" even well-defined in this case? :-) Feb 20 00:04:48 yes Feb 20 00:05:42 a solution is correct if the set of it and the expected answer is first order consistent Feb 20 00:06:38 checking whether it is FO con is unfortunately only semi-decidable Feb 20 00:07:15 and the prover used in the grading bot fails to find a proof (in finite time) Feb 20 00:08:41 but you could demonstrate such a proof, if only the bot would let you? :-) Feb 20 00:09:12 hi has anybody tried a build from scratch without the sources just that I have had some problems with fontconfig and now cairographics cvs access not downloading Feb 20 00:10:41 njs: if I had the expected solution, probably yes. (not manually though, but with a semi-automatic prover) Feb 20 00:11:10 *sigh* I want to take classes where this kind of problem arises :-) Feb 20 00:11:57 heh. take classes on logic and you will run into this Feb 20 00:12:23 or maybe advanced theoretical CS Feb 20 00:12:47 the class on logic I took was all old-school pencil-and-paper-and-chalk Feb 20 00:12:59 and now I don't have time to take math classes anymore :-( Feb 20 00:13:18 from what I have seen on the site the bb files are wrong.. or maby me (ps any body see me) Feb 20 00:13:34 Harvy: see you, but I'm clueless, sorry. Feb 20 00:14:21 Harvy: cairo is moving to git, I believe. and anoncvs is being moved around on fd.o Feb 20 00:14:56 cheers just a bit in the dark bout the problems trying to do a clean build Feb 20 00:15:20 i have seen the change to git Feb 20 00:16:00 njs: heh, yeah, logic at a CS faculty ;) preferably one where people do research on theorem provers and stuff like they do here >:) Feb 20 00:16:24 but the cvs access is pserver:anoncvs@cvs.ca Feb 20 00:18:33 I can dragg ALL the files down by hand to my sources dir but would prefer a working setup Feb 20 00:19:02 any developers on at the moment? Feb 20 00:19:58 reenoo: yeah. graydon keeps tantalizing me with interesting formal methods stuff, but so far I have resisted getting dragged in :-) Feb 20 00:39:46 Harvy: i don't have commit access, or it would have been fixed Feb 20 00:41:24 if you could let somebody who does it would help Feb 20 00:41:47 just trying to do a build from the start Feb 20 00:42:06 why change part of the files if the rest will break tomorrow? Feb 20 00:42:11 cheers dude Feb 20 00:42:29 you could just use the other repo ... if it still exists Feb 20 00:42:46 what other repo Feb 20 00:43:16 treke's Feb 20 00:44:14 surely a rc should build tho Feb 20 00:45:03 you make false assumptions Feb 20 00:45:27 we have no control over other projects sources or what they do with thier source trees Feb 20 00:46:41 well the inital download is from the oe server why is it not there anymore Feb 20 00:46:56 you have the wrong ip Feb 20 00:47:05 dns currently does not work Feb 20 00:47:48 dns for oe... is wrong Feb 20 00:47:56 what happened with openzaurus.org expiring? Feb 20 00:48:05 is that project dead (in favor of Angstrom) ? Feb 20 00:48:12 * emte has no clue why the dns issues exist Feb 20 00:48:31 what does the mailing lists say? Feb 20 00:48:44 what is the ip then Feb 20 00:51:32 * emte is looking throught the list now Feb 20 00:51:39 through* Feb 20 01:02:38 hmm Feb 20 01:02:40 going to have to sign off soon because I have work tommrow UK but will log cheers Feb 20 01:02:51 oesources.org works for me now Feb 20 01:03:57 192.216.230.225 Feb 20 01:05:01 i will give it a try tommrow ...ttfn Feb 20 01:05:08 k Feb 20 01:07:56 * emte grumbles at kernel configure loops Feb 20 01:08:13 and the play by play after Feb 20 01:21:22 i think the 2.6 kernel has some serious bloat issues ... Feb 20 01:21:38 are you on amd64? Feb 20 01:21:53 what do you mean bloat issues Feb 20 01:21:57 i am talking about jsut unpacking it Feb 20 01:22:04 /dev/hdc1 111G 105G 0 100% /backupfiles Feb 20 01:22:14 build$ rm -Rf tmp/work/h3600-linux/handhelds-sa-2.6-2.6.12-hh0+cvs20060219-r0 Feb 20 01:22:25 /dev/hdc1 111G 93G 13G 89% /backupfiles Feb 20 01:22:39 takes up 13Gb of room Feb 20 01:22:48 with the *.o yeah Feb 20 01:23:01 no, this is uncompiled Feb 20 01:23:06 source code itself is not so bad, considering it has support for a bazillion targets Feb 20 01:23:11 erhM? Feb 20 01:23:18 it never made it past ./configure Feb 20 01:24:18 the sources for kernel 2.6.15.4 on my desktop machine unpacked top out at 605M Feb 20 01:25:04 i just pulled the source tar out of /sources and am unpacking by hand Feb 20 01:25:17 i'll do a du -h and see what i get Feb 20 01:25:44 remember if you have any sort of broken filesystem like reiserfs, 'du' will be incorrect Feb 20 01:27:31 reiseerfs is not broken Feb 20 01:27:54 lol Feb 20 01:27:58 i use XFS Feb 20 01:28:10 283M kernel26 Feb 20 01:28:24 xfs is painful Feb 20 01:28:29 looks liek i generated 13Gb of garbage in taht compile loop Feb 20 01:28:39 i trust it far more than reiser Feb 20 01:28:47 XFS is excellent if you have no hardware or software or power problems Feb 20 01:29:00 and the moon is in the correct phase Feb 20 01:29:10 with god herself smiling on you Feb 20 01:29:12 i've never had an issue with XFS in 5 yrs Feb 20 01:29:36 cd kernel26 Feb 20 01:30:10 you have never had an un panned shutdown in 5 years Feb 20 01:30:19 sure i have Feb 20 01:30:35 living on an island we have a couple powerouts every year Feb 20 01:30:41 and xfs was ok Feb 20 01:30:44 yup Feb 20 01:31:21 you got lucky have yu tried fiery os Feb 20 01:31:25 rieser was unrecoverable and ext3 ... lucky it suports ext2 mode Feb 20 01:31:55 fiery os, no Feb 20 01:32:25 emte: I never had any problem with reiserfs ... and so far reiser4 was the best I tested Feb 20 01:32:28 never had aproblem with reisier Feb 20 01:32:47 fier rons on photocopiers Feb 20 01:32:56 i perfer my data being recoverable Feb 20 01:33:08 Harvy, oh wait Feb 20 01:33:19 you mean Fiery ... the RIP server Feb 20 01:33:25 yeah i've used it Feb 20 01:33:33 highend printing applications Feb 20 01:33:59 more from a user perspective, never messed with it Feb 20 01:34:19 and if it dies so does the HW Feb 20 01:34:27 emte: recoverable?? it was I have a huge tape backup system ;) Feb 20 01:34:33 s/was/why Feb 20 01:34:34 lol Feb 20 01:34:57 * emte doenst have a huge tap backup ... or even a small tape backup Feb 20 01:34:59 rip is right Feb 20 01:35:24 * chouimat goes back to this ugly perl crap Feb 20 01:35:25 rest in peaces Feb 20 01:35:44 rip as in raster image processing ... lol Feb 20 01:36:13 firey is great if your trying to print a 300MB+ image Feb 20 01:36:40 emte: ttfn thnks for the help Feb 20 01:38:08 np Feb 20 01:38:10 brb food Feb 20 01:53:22 Is there any way to get ipkg to update critical system packages without crashing the system in the process? I don't want to have to worry about my system getting b0rked every time I run ipkg upgrade and something like libc6 is in the list :/. Feb 20 02:01:43 ipag definatly needs some love on the client side Feb 20 02:01:48 ipkg* Feb 20 02:02:32 Well, I guess another question would be, is ipkg actually intended for updating live/on-line systems? Feb 20 02:03:05 yes Feb 20 02:04:00 http://handhelds.org/moin/moin.cgi/Ipkg Feb 20 02:05:50 should i be using the xlibs.bbclass directly? Feb 20 02:06:11 emte: Thanks. Feb 20 02:07:00 ipkg is a great tool ... the frontend on the client needs some work Feb 20 02:07:42 shadows, what do you mean using? Feb 20 02:07:50 like, inheriting it? Feb 20 02:08:14 * emte isnt sure Feb 20 02:08:18 currently theres a number of packages that define the CVS tree seperately, but they are all from freedesktop anon cvs Feb 20 02:08:25 so it should be fixed Feb 20 02:08:29 i'm going to submit a patch Feb 20 02:08:34 i wanted to know if that's okay Feb 20 02:09:25 i belive koen suggested changing it to ${FDO_CVS} Feb 20 02:09:30 or something like that Feb 20 02:09:37 okay Feb 20 02:09:42 it seems fdo changes every couple months Feb 20 02:09:43 where would FDO_CVS be defined? Feb 20 02:09:46 no Feb 20 02:10:06 it doesn't change every couple of months, this has been the case for 18 weeks and no one bothered to read the ML and update it Feb 20 02:10:11 now i'm updating it Feb 20 02:10:28 its less 6months since fdo last changed their sources structure Feb 20 02:11:07 and i belive in local.conf with the others Feb 20 02:12:45 Feb 19 13:22:51 koen should be move to $FDO_CVS or something Feb 20 02:13:05 oh Feb 20 02:13:19 well these particular ones are grabbing from the xlibs branch of fdo anoncvs Feb 20 02:13:24 so that actually looks okay Feb 20 02:13:37 i'm changing it to an xlibs thing, easy enough to change later Feb 20 02:13:43 xcb apperently is the only change Feb 20 02:13:57 atleast as far as fdo's list has stated Feb 20 02:14:51 Feb 19 14:03:26 pb_ zecke: thinking of the xlib thing that shadows mentioned earlier, you might find http://lists.freedesktop.org/archives/xorg/2006-February/013113.html interesting. Feb 20 02:15:13 * shadows reads Feb 20 02:50:47 hmm Feb 20 02:50:55 i'm going through all the files now and updating Feb 20 02:51:02 has BBDEBUG stopped working ? Feb 20 02:51:04 it's an unbelievable quantity of files Feb 20 02:51:28 i think thats why $FDO_CVS was suggested Feb 20 02:51:52 i split it up into three bbclass'es Feb 20 02:51:59 it works more easily that way i think Feb 20 02:51:59 we had to do the same thing before for all the fdo projects a few months back Feb 20 02:58:14 hrm Feb 20 02:58:44 only 3 more bb's to go Feb 20 02:58:47 then i will make a diff Feb 20 03:05:30 hmm Feb 20 03:05:33 silly silly Feb 20 03:06:01 http://bugs.treke.net/show_bug.cgi?id=706 Feb 20 03:10:49 hmm Feb 20 03:10:55 JustinP, you awake? Feb 20 03:11:04 or pb_ Feb 20 03:11:54 can one of you please push http://bugs.treke.net/show_bug.cgi?id=445 Feb 20 03:12:09 it shoudl have been done months ago Feb 20 03:22:13 emte: yeah, just wasn't online Feb 20 03:22:17 emte: what's up? Feb 20 03:22:33 actually ignore that request Feb 20 03:22:41 that change wont work Feb 20 03:23:37 strange ... why is it failing ... Feb 20 03:24:40 k Feb 20 03:25:29 NOTE: Handling BitBake files: / (1596/3074) [51 %]ERROR: /backupfiles/biohazard/openembedded/org.openembedded.dev/packages/linux/handhelds-sa-2.6_cvs.bb:27: unparsed line: 'do_configure()' while parsing /backupfiles/biohazard/openembedded/org.openembedded.dev/packages/linux/handhelds-sa-2.6_cvs.bb Feb 20 03:25:53 why is it not parsing ... Feb 20 03:26:02 shadows: good...except that you don't need to change PR if you change the SRC_URI that way... Feb 20 03:34:42 okay it worked this time Feb 20 03:34:52 i must of had a bad char Feb 20 03:35:06 JustinP, you wanna push this http://bugs.treke.net/show_bug.cgi?id=445 Feb 20 04:13:38 well i got h3600 sa-2.6 atleast configureing and partially building :) Feb 20 04:13:41 http://oe.pastebin.com/563879 Feb 20 04:14:08 heed to figure out what happened to arch/arm/mach-sa1100/h3600.c:47:36: asm/hardware/gpio_keys.h: No such file or directory Feb 20 04:14:15 need* Feb 20 04:28:31 hrm... Feb 20 04:42:48 emte: I do? Feb 20 04:45:05 * emte shrugges Feb 20 04:45:18 looks like i'll be adding somemore stuff to it soon Feb 20 04:48:23 well, if it's not working..... Feb 20 04:56:50 there Feb 20 04:57:01 cleaned out soem files now i should be able to find it Feb 20 05:02:35 JustinP: i was told that all builds need PR Feb 20 05:03:03 its debatable on your changes i think Feb 20 05:03:25 well someone needed to do it Feb 20 05:03:27 i did it. Feb 20 05:03:40 no, not the changes themselves Feb 20 05:03:52 if PR neede to be changed Feb 20 05:03:56 needed* Feb 20 05:04:25 emte: that should be "do_configure ()" i think Feb 20 05:04:27 notice the space Feb 20 05:04:33 make sure it has the space AFAIK Feb 20 05:04:46 nah it wasnt that Feb 20 05:05:00 it was an invis char i inserted Feb 20 05:05:04 ohhh okay Feb 20 05:05:48 teaches me to copy and paste out of my browser Feb 20 05:05:52 :P Feb 20 05:06:08 via gpm Feb 20 05:06:41 hurry up slocate .... Feb 20 05:06:49 i need smaller hdds Feb 20 05:07:02 then it wouldnt take so long Feb 20 05:08:43 AHA!!!! Feb 20 05:08:51 /asm-arm/arch-sa1100/h3600_gpio.h Feb 20 05:09:09 silly path change Feb 20 05:09:24 hrm ... Feb 20 05:09:43 on second thought ... how many files do i have to patch now .... 6? Feb 20 05:18:18 shadows: 1) I was wrong, 2) no Feb 20 05:18:52 shadows: I told you that all builds need PR and they don't *need* them if they haven't been updated (i.e. default is r0) Feb 20 05:19:29 shadows: and also, SRC_URI changes like the ones in your patch don't need a PR bump. There's no build change and no actual difference in the ipks built so there's no need to make people rebuild Feb 20 05:19:54 shadows: of course, to fix those SRC_URIs we could just use a simple sed script... Feb 20 05:20:05 on the otherside a PR bump is a marker for the URI change Feb 20 05:20:07 Is it some usual thing for gdb to "miss" addresses? Feb 20 05:20:21 I am getting an error from an address on the heap :S Feb 20 05:21:30 emte: marker? Feb 20 05:21:57 indicator when the URI change happened Feb 20 05:22:12 from a historical aspect Feb 20 05:23:33 * emte shrugges Feb 20 05:23:51 like i said it could be debated either way Feb 20 05:24:00 not needed Feb 20 05:24:03 the revision # is enough Feb 20 05:24:26 if the code or build doesn't change there's no reason to bump PR Feb 20 05:24:31 PR is the revision number Feb 20 05:24:50 PR is used to tell the build system to build a new version and the installer to install a new version Feb 20 05:24:58 there is no reason to rebuild or upgrade with this change Feb 20 05:25:10 if you already have it built, using a new source will make no diff Feb 20 05:25:53 RP: looks like the interrupt isn't being called at all Feb 20 05:26:02 correct, but i belive if you already that the source version it will ignore the URI anyway Feb 20 05:26:12 s/that/have Feb 20 05:26:29 RP: and nothing is happening when I insert/remove the remote (although inserting and removing the module shows some output now) Feb 20 05:26:53 * emte gives JustinP a bigger interrupt hammer Feb 20 05:26:55 emte: ? Feb 20 05:27:14 emte: a new URI will cause a new fetch, yes Feb 20 05:27:32 okay, noted Feb 20 05:27:36 really? never noticed it here ... Feb 20 05:27:38 i'm going off what people told me to do before Feb 20 05:27:40 emte: it won't use the old tarball as the tarball name has the URI in it Feb 20 05:27:44 cept with cvs which is normal Feb 20 05:27:49 and also my experience w/ Gentoo Feb 20 05:27:53 shadows: I must not have explained it very well before Feb 20 05:28:02 hm Feb 20 05:28:13 feel free to patch it yourself ;) Feb 20 05:28:33 * emte notes my sources dont have a uri Feb 20 05:28:35 emte: do you know anything about kernel interrupts? Feb 20 05:28:40 i feel it necessary to bump all the revisions and see what breaks Feb 20 05:28:48 JustinP, not enough to help you Feb 20 05:29:06 shadows: clean/rebuild Feb 20 05:29:09 'cause right now, gpe/opie targets are going to be possible to build Feb 20 05:29:32 s/are/are not/ Feb 20 05:29:33 shadows meant: 'cause right now, gpe/opie targets are not going to be possible to build Feb 20 05:30:35 JustinP: the whole thing is going to need an overhaul soon enough Feb 20 05:31:00 this is temporary "emergancy fix" until it gets changed Feb 20 05:34:00 ... Feb 20 05:34:05 what's the "real" fix? Feb 20 05:34:26 BTW, if anyone is willing to check this out, the interrupt isn't being called: http://oe.reversefold.com/sharpsl-rc.patch Feb 20 05:39:39 JustinP: the "real" fix is to mirror snapshots on an OE mirrorsite Feb 20 05:39:42 and reference that Feb 20 05:43:49 shadows, thats not a real fix Feb 20 05:44:04 it has its own inherit issues Feb 20 05:44:24 mainly maintainance and staleness Feb 20 05:45:41 the real fix is to have *releases* which work Feb 20 05:46:05 lol Feb 20 05:47:02 either that or use the releases and then add patches which get it up to the point in CVS that we need Feb 20 05:54:46 right Feb 20 05:54:59 that's what i mean, we shouldn't be depending on some random foo cvs server Feb 20 05:57:14 lolalright Feb 20 05:57:19 alright* Feb 20 05:57:27 lest see what my current mangling has done Feb 20 05:58:39 wanna see my patch so far? Feb 20 05:58:41 http://oe.pastebin.com/563937 Feb 20 05:58:56 516 lines of fun Feb 20 05:59:38 hmm Feb 20 05:59:42 it didnt liek it Feb 20 06:01:06 hmm Feb 20 06:01:09 i wonder Feb 20 06:02:31 emte: jeez..... Feb 20 06:02:39 * JustinP leaves the ever-changing bug alone Feb 20 06:02:54 who me? Feb 20 06:02:56 :P Feb 20 06:03:12 the first 490 lines are changes in the defconfig Feb 20 06:04:05 basically the difference between the defconfig and runing it and saving it again via make menuconfig Feb 20 06:04:34 :-| Feb 20 06:04:56 i think the kernel-crew's defauly is a bit stale Feb 20 06:04:59 default* Feb 20 06:05:13 hence me fixing header paths Feb 20 06:08:35 whoops spelling mistake :( Feb 20 06:20:43 hmm Feb 20 06:23:15 grr Feb 20 06:24:06 http://www.handhelds.org/hypermail/kernel-discuss/13/1335.html Feb 20 06:24:12 * emte must liek to waste my time Feb 20 06:59:27 grrrr Feb 20 06:59:31 interrupt! Feb 20 07:13:40 okay finally ... Feb 20 07:13:57 i got 2.6.15 patched and compiling Feb 20 07:27:26 hey gremlin[it]work i found the threads on your kernel stuff :) Feb 20 07:29:48 yes :) Feb 20 07:30:06 i wish i had found em earlier Feb 20 07:30:32 i'll send a patch this evening for micro_key ... Feb 20 07:30:34 why ? Feb 20 07:30:55 spent a while patching paths before i found yours Feb 20 07:31:50 for h3600 ? Feb 20 07:31:53 yup Feb 20 07:33:34 it's about 6/9 months i start to take care oh h3600 for kernel 2.6 ;) Feb 20 07:34:18 yeah, its just not in the dev repo ... not sure which branch your on Feb 20 07:35:25 no i'm working just on kernel not on oe ... Feb 20 07:35:56 yeah, thats what i am thinking Feb 20 07:36:55 i have done some change on oe, i have a different h3600 machine to build with k26 ... Feb 20 07:44:46 probably i'll put my difference on my site ... Feb 20 07:55:24 morning Feb 20 08:00:01 Good morning all ! Feb 20 08:17:08 mmornign Feb 20 08:51:35 good mornning all Feb 20 09:04:10 woohoo! Feb 20 09:04:16 I'm getting interrupts Feb 20 09:05:53 re Feb 20 09:07:36 and output Feb 20 09:07:42 no events, though Feb 20 09:07:49 at least I'm getting closer Feb 20 09:11:47 morning Feb 20 09:14:15 hey CoreDump|home Feb 20 09:17:22 yay ipkg upgrade fubar'ed my ROM yet again Feb 20 09:17:54 that makes no sense Feb 20 09:18:03 if it's Read Only, how can ipkg mess with it? Feb 20 09:18:45 ~lart koen for pointing out the obvious Feb 20 09:18:45 * ibot hooks into a hydrant and hoses koen down for pointing out the obvious Feb 20 09:18:56 :) Feb 20 09:19:22 mhhh with a cold shower koen can completelly wakeup :) :) :) Feb 20 09:19:43 koe off-topic : what's your ppc board ? Feb 20 09:20:23 gremlin[it]work: the only powerpc board I have is my powerbook Feb 20 09:20:34 the rest is either arm, armeb or mips Feb 20 09:20:37 the loft board ? Feb 20 09:20:43 that's armeb Feb 20 09:21:04 morning all Feb 20 09:21:08 gremlin[it]work: http://www.giantshoulderinc.com/hardware.html Feb 20 09:21:11 hey RP Feb 20 09:21:31 hi RP CoreDump|home Feb 20 09:21:46 * CoreDump|home yawns Feb 20 09:22:15 hi XorA Feb 20 09:22:19 morning Feb 20 09:22:37 http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenEmbeddedBuilds Feb 20 09:22:44 we have a working tinderbox again Feb 20 09:22:51 "every project has a phase when all engineers should be kill and product shipped" is my feeling about OZ 3.5.4 Feb 20 09:22:53 thanks to Zecke Feb 20 09:23:09 hey hrw|work & XorA Feb 20 09:23:39 did anyone try abiword recently? It compiles and installs fine but crashes on start-up with an X error Feb 20 09:23:56 hrw|work: =) Feb 20 09:24:20 CoreDump|home: http://handhelds.org/~bugzilla/show_bug.cgi?id=1520 Feb 20 09:24:43 CoreDump|home: we had plans to ship it before fosdem Feb 20 09:24:55 CoreDump|home: currently I feel that we will ship it in may or later... Feb 20 09:24:56 koen: great thanks! Feb 20 09:25:10 hrw|work: what's holding it up? Feb 20 09:25:44 CoreDump|home: fscking keylauncher/matchbox problem in gpe Feb 20 09:26:03 CoreDump|home: I vote for removing keylaunch at all from c*0 devices Feb 20 09:26:05 then remove keylauncher and configure the keys via matchbox's config file Feb 20 09:26:09 yep Feb 20 09:26:09 hrw|work: don't ship keylaunch Feb 20 09:26:43 interestingly enough, GPE from .dev works fine heh Feb 20 09:27:07 or maybe the newer xserver-common wasn't pushed into .dev only .oz Feb 20 09:27:55 Also, personally, I would disable locale-base-* postinst completely Feb 20 09:28:41 ~seen mickeyl Feb 20 09:28:43 mickeyl is currently on #oe #opie #handhelds.org. Has said a total of 11 messages. Is idling for 19h 57m 3s, last said: 'argouml?'. Feb 20 09:30:51 RP: oh, hey, morning Feb 20 09:31:01 RP: I'm getting interrupts :-) Feb 20 09:31:16 RP: it looks like it's all good, except for the ranges for the buttons..... Feb 20 09:31:47 RP: the values output by get_remocon_raw() are not unique for the different buttons... Feb 20 09:32:21 JustinP: That doesn't sound good :-/ Feb 20 09:32:23 RP: some of the buttons give different values, but most don't Feb 20 09:32:46 JustinP: Could your debugging code be interfering with the measurements somehow? Feb 20 09:32:54 RP: ATM, only vol up and vol down will work...:-| Feb 20 09:33:02 I guess we need to figure out exactly what its measuring... Feb 20 09:33:20 I doubt anything I added could have interfered....but I may be wrong Feb 20 09:33:34 MAX1111_REMCOM could be wrong, though Feb 20 09:33:45 you had it commented as possibly 0u, so I made it that Feb 20 09:34:17 not sure what it refers to exactly, but it's used in the call to sharpsl_pm_pxa_read_max1111 to get the values I'm talking about Feb 20 09:34:28 http://oe.reversefold.com/sharpsl-rc.patch Feb 20 09:34:46 JustinP: Its the ADC channel on the MAX1111 I think the headphones are connected to Feb 20 09:34:56 See the Maxim MAX1111 datasheet Feb 20 09:35:16 ah.....well, I am getting different values for some of the buttons.... Feb 20 09:35:51 that's a start Feb 20 09:37:02 I'm just recompiling with altered constants to that I can hopefully see something come across the input device ^_^ Feb 20 09:37:28 I'd compare how its calling the max1111 functions with that in the sharp driver. Feb 20 09:37:52 I'm wondering if corgi_ssp might be broken somehow... Feb 20 09:38:28 03koen 07org.oe.dev * r244cbc11... 10/conf/distro/angstrom-2006.9.conf: angstrom-2006.9: set distro version to test Feb 20 09:39:53 * JustinP should have looked at that Feb 20 09:40:10 hmmmm Feb 20 09:40:18 1u Feb 20 09:40:31 with a bunch of bit-shifting.... Feb 20 09:40:52 (1u << MAXCTRL_PD0_SH) | (1u << MAXCTRL_PD1_SH) | (1u << MAXCTRL_SGL_SH) | (1u << MAXCTRL_UNI_SH) | (0 << MAXCTRL_SEL_SH) | (1u < (I have no idea what u means) Feb 20 09:42:28 It means unsigned so its just a 1 character Feb 20 09:42:50 ah....wonder why I hadn't learned that previously... Feb 20 09:43:22 is the shitfing to have it work runtime with different devices or something....? Feb 20 09:44:43 Its just so you can see the bitbstatus of all the regsisters like PD0, PD!, SGL, etc Feb 20 09:45:38 could that be the problem then? since we're sending in 0u (or 1u now) Feb 20 09:46:55 morning all Feb 20 09:47:02 do13_: morning Feb 20 09:47:04 JustinP: The commands should be identical Feb 20 09:47:06 hi dirk Feb 20 09:47:37 Hi Richard, JustinP Feb 20 09:47:49 RP: so 1u == all that bit-shifting? Feb 20 09:48:08 * JustinP sees no MAXCTRL_PD0_SH in the 2.6.15 source Feb 20 09:48:14 JustinP: no, See + return corgi_ssp_max1111_get((channel << MAXCTRL_SEL_SH) | MAXCTRL_PD0 | MAXCTRL_PD1 Feb 20 09:48:14 + | MAXCTRL_SGL | MAXCTRL_UNI | MAXCTRL_STR); Feb 20 09:48:41 hmmm, ok Feb 20 09:48:54 oh Feb 20 09:48:55 pff Feb 20 09:49:03 I suspect MAXCTRL_PD0 = (1u << MAXCTRL_PD0_SH) Feb 20 09:49:05 they were defined in the patch Feb 20 09:49:17 hi Dirk Feb 20 09:49:28 Its also defined in the sharpsl_pm code Feb 20 09:49:43 hey Marcin Feb 20 09:49:43 can the http://ipkgfind.nslu2-linux.org/ package search be trusted? Finding it hard to believe that libupnp is not in there yet Feb 20 09:50:02 oh.... Feb 20 09:50:06 ~lart me Feb 20 09:50:06 * ibot pours hot grits down the front of justinp's pants Feb 20 09:50:17 cyphunk: it can be trusted Feb 20 09:50:29 03koen 07org.oe.dev * r0f3ce981... 10/conf/distro/angstrom.conf: Feb 20 09:50:29 angstrom.conf: Feb 20 09:50:29 * set default maintainer Feb 20 09:50:29 * put angstrom and version into image names Feb 20 09:50:31 OE & OZ sucks badly when it comes to PR and informations... Feb 20 09:50:32 cyphunk: if it builds, ask NA|Zzz to include it when he wakes up Feb 20 09:51:06 koen: it does build.... will do Feb 20 09:52:26 RP: is the git patch for lib/bb in svn yet? Feb 20 09:52:36 it seems we might need it for xlibs :( Feb 20 09:52:54 hmmm, no... Feb 20 09:53:35 koen: iirc git fetcher is in trunk Feb 20 09:54:06 looks like 0u would give me the same thing as the Sharp code Feb 20 09:54:23 hrw|work: yes, but it's broken Feb 20 09:54:38 hrw|work: it needs http://www.rpsys.net/openzaurus/temp/bitbake_git-r0.patch Feb 20 09:57:53 http://bugs.openembedded.org/show_bug.cgi?id=610 is useless probably (its gpe) Feb 20 09:57:58 ok, it's 2AM, I'm too tired to keep hacking Feb 20 09:58:31 RP: I would appreciate it if you could take a quick look and see if perhaps the corgi_ssp stuff is broken... Feb 20 09:58:32 koen: not yet. Its not going to happen until this evening. Too many other things to worry about today :-/ Feb 20 09:59:00 RP: it could also be the MAXCTRL constants defined in the patch perhaps... Feb 20 09:59:05 RP: no hurry, I have it applied locally :) Feb 20 09:59:08 JustinP: I suspect the bit bang code from the original sharp driver might ben needed Feb 20 09:59:16 lovely Feb 20 09:59:37 It the bit bang code was for the max1111, that would be a problem Feb 20 09:59:49 * RP can't remember which it belonged to Feb 20 10:00:28 well, it seems like I'm getting about 3 bits in my value...largest I've seen is 5 Feb 20 10:01:08 of course I'm completely speculating Feb 20 10:01:42 anyway, I'm off to bed. Perhaps tomorrow I can look into the ssp code, although I of course no nothing about that either.... Feb 20 10:02:02 at least it's coming along Feb 20 10:02:07 night Feb 20 10:02:28 JustinP: I'll try and have a look at the ssp code to refresh my memory. 'night Feb 20 10:02:59 'night JustinP Feb 20 10:06:32 ~seen mickeyl Feb 20 10:06:35 mickeyl is currently on #oe #opie #handhelds.org. Has said a total of 11 messages. Is idling for 20h 34m 55s, last said: 'argouml?'. Feb 20 10:12:08 someone wanna submit a patch for perl_5.8.7.bb Feb 20 10:12:23 ldc: what sort of patch? Feb 20 10:12:27 For gentoo the following fixes are needed Feb 20 10:12:30 sed -i -e "s,usrinc='/usr/include',usrinc='${STAGING_INCDIR}',g" config.sh-${TARGET_ARCH}-${TARGET_OS} Feb 20 10:12:30 #fix some hardcoded path's in perl. Feb 20 10:12:30 sed -i -e "s,'/usr/include/,'${STAGING_INCDIR}/,g" ${WORKDIR}/perl-5.8.7/ext/Errno/Errno_pm.PL Feb 20 10:15:16 just cant understand why they use hardcoded path's in errno_pm.pl Feb 20 10:16:34 ldc: You should attach to a bug in bugs.treke.net Feb 20 10:18:46 xora: well, not really using a clean OE tree, but this problem exists in the clean OE tree too Feb 20 10:23:08 ldc: well bug report it anyway, its a bug that needs fixed and perl has a maintainer Feb 20 10:38:09 NOTE: multiple providers are available (glibc, glibc-intermediate); Feb 20 10:38:09 NOTE: consider defining PREFERRED_PROVIDER_virtual/arm-linux-libc-for-gcc Feb 20 10:38:36 Am I right to assume that if PREFERRED_PROVIDER is not set, OE uses the first one in the list? Feb 20 10:39:00 no idea, but in your case both options are fine Feb 20 10:39:14 thanks =) Feb 20 10:39:17 although the glibc-i way is a bit longer Feb 20 10:40:11 * CoreDump|home goes with glibc Feb 20 10:40:40 with eabi we'd need glibc-i Feb 20 10:41:09 One would thing something like that would be set in openzaurus*conf Feb 20 10:42:16 ~lart minimo mirror Feb 20 10:42:17 * ibot smacks minimo mirror up side the head with a clue-by-4 Feb 20 10:42:27 03koen 07org.oe.dev * reb1daf07... 10/packages/linux/ (4 files in 4 dirs): handhelds-pxa: update to 2.6.15, include patches and new defconfigs Feb 20 10:53:07 03xora 07org.oe.dev * r1b17d5ea... 10/packages/offlineimap/offlineimap_4.0.11.bb: Feb 20 10:53:07 offlineimap_4.0.11.bb : add a new version with some proper dependecies Feb 20 10:53:07 on the right python modules. **** ENDING LOGGING AT Mon Feb 20 10:59:56 2006