irclog2html for #gpe on 20030630

00:12.06*** join/#gpe jg (~jg@h005018015b26.ne.client2.attbi.com)
00:12.15dc_morning jg
00:12.35jgdc_: evening...
00:30.16*** join/#gpe Thelonius (~soopaman@h24-66-54-32.wp.shawcable.net)
01:04.44*** join/#gpe dc__ (~dctanner@modem-147-46-60-62.vip.uk.com)
01:32.14*** join/#gpe glm2k (~greylensm@c-66-229-176-251.we.client2.attbi.com)
02:01.26*** join/#gpe flypiper (~deb-user@12-222-141-210.client.insightBB.com)
05:08.03*** join/#gpe glm2k (~greylensm@c-66-229-176-251.we.client2.attbi.com)
06:21.47*** join/#gpe mibus (~mibus@233.c.002.ade.iprimus.net.au)
06:23.32*** join/#gpe Thelonius (~soopaman@h24-66-54-32.wp.shawcable.net)
06:25.12mibus'lo all
07:34.55*** join/#gpe florian (~fuchs@kc.celox-online.de)
07:35.02florianhi all!
07:48.07*** join/#gpe herzi (~herzi@kiwi.mediascape.de)
07:51.58*** join/#gpe HET2 (~diman@chello080110092229.507.15.vie.surfer.at)
08:07.05*** part/#gpe glm2k (~greylensm@c-66-229-176-251.we.client2.attbi.com)
08:17.31*** join/#gpe florian (~fuchs@kc.celox-online.de) [NETSPLIT VICTIM]
08:18.38*** join/#gpe PaxAnima (~petter@b097c.studby.ntnu.no)
08:40.53*** join/#gpe pb_ (~pb@dsl-62-3-66-201.zen.co.uk)
08:41.34pb_morning
08:49.08*** join/#gpe herzi (~herzi@kiwi.mediascape.de)
09:09.01florianhi pb_
09:11.02pb_hi florian
09:12.12floriananything new about a familiar release?
09:12.47pb_No, we still need to fix the hda: lost interrupt bug.
09:13.11pb_Also, I think Russ is on vacation this week, so we probably won't see much release activity.
09:13.35florianoh, this is bad...
09:13.55florianbut this is a reason to see what we can do for aur release in the meantime
09:14.08florians/aur/our
09:14.56pb_Yep.
09:16.33pb_Or, if by any chance you'd like to take a look at the hda bug, that'd be very cool.
09:20.49florianhmm... my knowledge about the kernel internals is not really good, i'm not shure is this would make sense. But i could ask Nils for some time :-)
09:25.37*** join/#gpe mallum (~mallum@82-35-112-58.cable.ubr07.dals.blueyonder.co.uk)
09:25.38florianWhat do we know about this problem?
09:25.45florianmorning mallum
09:26.13mallumhey florian
09:28.44pb_florian: it's bug 856, therer's some commentary there.
09:31.02mallumstargirl: ping
09:31.46florianpb_: ok, thanks.... let me see...
09:32.19pb_I'm going to try out Jamey's suggested workaround.
09:32.31pb_If that works, I think we should probably just go with it for 0.7.
09:44.35florianpb_: The idea to use a new ide driver sounds good too - do we have some clue about the kernel version this appeard first?
09:46.47pb_I think there were some changes in 2.4.20, and more in 2.4.21.  I did try importing the 2.4.21 driver into the handheld.org tree, but wasn't able to make it build.
09:52.02florianhmm... nasty - and since which hh version does this bug persist?
09:52.33pb_Nobody knows for sure, but we speculate it appeared in the change from 2.4.18 to 2.4.19, when all the interrupt code was redone (ie at the same time as that 3800 touchscreen bug).
09:52.48pb_Jamey's "cardctl eject" workaround does work for me, btw.  I added a note to the bug to that effect.
09:53.48florianThats good...
09:54.01florianWher is the error message generated?
09:54.15pb_In the ide driver, I think.
09:54.28florianAh, ok...
10:20.07florianOK, gpe-conf is now able to remove a package :-)
10:21.00pb_Cool :-)
10:22.06florianHmm... creating a real package manager frontend would be much more fun...
10:23.03pb_Yeah.  Should hassle TheMasterMind1 to find out what the current status of libipkg is.
10:23.57florianTheMasterMind1: hi! What's the status of libipkg?
10:24.16mallumI've written some tests to measure full screen blitting on the ipaq
10:24.16florianIn this way? ;-)
10:24.34pb_florian: Exactly!  Might be a little early for him, though.  He's in the USA someplace.
10:25.08florianpb_: ehh, ok... i think i hould ping him later :-)
10:27.13pb_Yep.
10:27.41pb_mallum: what results did you come up with?
10:31.36mallumpb_: these so far;
10:31.54mallum~ # time -p ./test-x
10:31.55mallumreal 7.28
10:31.55mallumuser 3.37
10:31.55mallumsys 0.02
10:31.55mallum~ # time -p ./test-fb
10:31.55mallumreal 3.84
10:31.57mallumuser 3.68
10:31.59mallumsys 0.03
10:32.01mallum~ # xrandr -o left
10:32.03mallum~ # time -p ./test-x
10:32.07mallumreal 5.56
10:32.09mallumuser 3.64
10:32.11mallumsys 0.02
10:32.13mallum~ # time -p ./test-x --no-shm
10:32.15mallumreal 6.19
10:32.17mallumuser 3.75
10:32.19mallumsys 0.31
10:32.21mallumpb_: each does 100 full screen blits of the same data
10:32.28mallumpb_: I'll upload the code in a mo
10:33.00mallumpb_: also there currently compiled with no optimisation
10:34.08mallumpb_: http://www.handhelds.org/~mallum/tests/
10:34.15mallumpb_: and thats on my 3870
10:36.25florianmallum: what's the status of the mbcrystal theme?
10:37.01mallumflorian: just needs some tlc to clear out uneeded icons + maybe updated for mb0.6
10:37.28florianmallum: is the current status packaged?
10:37.39mallumflorian: I think so
10:37.54mallumflorian: you can d/l from mb site. it has a make ipkg target
10:38.23florianmallum: great, thanks
10:39.36pb_mallum: was it intentional that you have the XPutPixel() loop inside the main loop?  If you want to test blit performance, I would have thought that should be outside.
10:42.36mallumpb_: yeah. Although its writing the same thing everytime, I wanted to assume that data will be different on each write - thus the putpixel call
10:44.51*** join/#gpe cmarqu_werk (www@vws1.kens.com)
10:45.07pb_Ah, I see.  Is that how real applications create images for blitting?
10:45.14florianhi cmarqu_werk
10:45.20cmarqu_werkHi all
10:45.46pb_hi cmarqu_werk
10:47.42pb_fwiw, on my h3600 with an optimised build of test-fb, I get:
10:47.46pb_/mnt/ramfs/tmp # time -p ./test-fb
10:47.46pb_real 1.11
10:47.46pb_user 1.08
10:47.46pb_sys 0.00
10:48.06mallumpb_: I think the xputpixel is just a macro which safely sets data in the ximage structure
10:48.18mallumpb_: what flags did you pass to the build ?
10:48.39mallumpb_: with -O2 -mtune=strongarm -fomit-frame-pointer, I get the following ;
10:48.49mallum~ # time -p ./test-x
10:48.50mallumreal 4.36
10:48.50mallumuser 2.41
10:48.50mallumsys 0.07
10:49.12pb_mallum: hm, XPutPixel looks like a function pointer call to me.
10:49.26pb_I used: gcc -O2 -fomit-frame-pointer -march=armv4 test-fb.c
10:50.38mallumpb_: let me move the xputpixel out of the loop and see what we get
10:51.30pb_Looking at the generated code for test-fb, I think it's spending most of its time calculating pixel values rather than actually blitting.
10:51.51*** join/#gpe thr33 (olly@oghm2.kings.cam.ac.uk)
10:53.34mallumpb_: you think I should pregenerate that ? There both doing the same calculation
10:54.05pb_Yes, I think that'd be a good idea.
10:55.06mallumpb_: Im just thinking real world usage. where the image is unlikely to be the same on every copy
10:56.10pb_True, but the image generation process is going to be highly application specific.  If you are just trying to compare X vs framebuffer blit performance, I think you should probably leave that out.
10:56.23pb_You could pregenerate several images and rotate between them, if you're worried about cache effects.
10:56.47*** join/#gpe dc_ (~dctanner@62.60.119.104)
10:57.00*** join/#gpe tmh198 (~tmh198@202-6-129-31.ip.adam.com.au)
10:58.25mallumpb_: right, maybe I'll add some switches
10:58.35dc_morning mallum
10:58.57dc_mallum: my cards are on a city link truck today :-)
10:59.11dc_mallum: so they should be here in the next hour or so
10:59.23mallumpb_: also I think theres maybe more startup time in the X case - connecting to display, opening the window, negotiating the wm etc - Im not sure how much of an impact that makes
10:59.56mallumpb_: you think pregenerate the data, then memcpy it onto the framebuffer ?
11:00.43pb_mallum: yep
11:00.53mallumdc_: hey, cool :-)
11:01.04mallumpb_: okey
11:01.09mallumpb_: I wanna test DGA too
11:02.27pb_right
11:05.02*** join/#gpe cmarqu_werk (www@vws1.kens.com)
11:07.13dc_doh, the school's ringing up asking where I am ;)
11:07.19pb_doh :-)
11:07.34pb_how come you're meant to be at school?  I thought your exams were over.
11:07.51dc_indeed, every other school is over.
11:08.05dc_for some reason our school seems to think we're all still ment to be going there.
11:08.11dc_but hell no, I've finished my exams.
11:08.24mibusyo sucker! :-P
11:08.30mibusnah-nah
11:08.32mibus:-P
11:08.35mibusthats why i like you
11:08.38mibus*uni*
11:08.42mibusi'm frelling nuts :-S
11:08.42mallumpb_: these are pregenerated times ( both with flags )
11:08.42dc_hehe
11:08.43mibusanyway
11:08.43mallum~ # time -p ./test-x
11:08.44mallumreal 1.90
11:08.44mallumuser 0.06
11:08.44mallumsys 0.04
11:08.45mallum~ # time -p ./test-fb
11:08.47mallumreal 1.87
11:08.49mallumuser 1.71
11:08.51mallumsys 0.02
11:08.53dc_mibus: drinking ;)
11:08.55mibusyou don't (usually) get marked on attendance
11:08.58mibusLOL
11:09.00mibusyeah that too!
11:09.09mibusits good to get drunk, uh, i mean, relaxed, before lectures :-)
11:09.18dc_exactly
11:09.22mibusbut the tav only opens at 12! i mean, seriously, where are you supposed to have breakfast? :-D
11:09.36mibus(actually you can normally get grog upstairs too but it's so much less cool :D
11:09.49mibuswhats the drinking age in the UK?'
11:09.53dc_18
11:09.53mallummibus: 18
11:09.59mibussame here
11:10.19dc_I have a cool half-sister though.
11:10.20mallummibus: you'd probably die if you couldn't
11:10.21dc_;-p
11:10.35dc_mallum: oh goh, yeah he woul
11:10.49dc_tying problem
11:10.57dc_*tpying
11:11.00dc_typing
11:11.10mibusmallum: actually i don't drink alcohol all that much
11:11.20mibusthough there was this incident recently with a snorkel and a bottle of beer
11:11.20pb_mallum: mm, strange, I wonder why yours are slower than mine.
11:11.27pb_mibus: hah
11:11.30cmarqu_werkdc_: half-sister... what was her name again...
11:11.33mibusdc_: you're typing is about as good as mine! :-)
11:11.44cmarqu_werkHi mibus
11:11.45mallumpb_: mine are on the 3870
11:11.56dc_cmarqu_werk: shhh!
11:11.59pb_3870 shouldn't be any slower than 3600, afaik.
11:12.06cmarqu_werkdc_: Hrhr
11:12.18mibuscmarqu_werk: hey, long time no see!
11:12.19mibus:-)
11:13.24mallumpb_: dont they have a slightly different setup - 3600 has 'natural' portrait mode, 3870 landscape ?
11:13.54mibusmallum: i thought the 3600's natural mode was landscape?
11:14.07mibusmallum: iirc writing directly to the fb is landscape
11:14.09dc_if only the c700s was.
11:14.15dc_and on that note.
11:14.20pb_mallum: no, both are naturally landscape.
11:14.22mibusmallum: or i'm just being stupid and not understanding you :-D
11:14.22dc_mallum: what's the current state?
11:15.30mallumpb_: hmm, I've just noticed my membpy one isn't actually painting anything :/
11:15.36pb_hah
11:15.41mallumdc_: state ?
11:16.05mibusdc_: so how are your exams going anyway?
11:16.14mibusdc_: i have six this semester for uni, half-way done
11:16.16mibus:-S
11:17.36mallumpb_: I uploaded the newer test-fb.c , do you see anything obvious why its no painting ?
11:18.53pb_I think your length is too big, but that shouldn't stop it from painting.
11:19.12pb_iirc, fix.smem_len is the actual length; you shouldn't be multiplying it by 4.
11:19.50mallumpb_: yes, I looked at fb.h and its defined as a u32, which is 4 bytes right ?
11:20.12pb_yeah.
11:20.28mibusok well i'm off
11:20.37mibuspb_: please remember to do a nmf package for matt :-)
11:20.40pb_oh, I guess if you are writing off the end of the framebuffer you would destroy the palette, which might have deleterious effects.
11:20.43pb_mibus: wilco
11:20.50mibuspb_: if you dont want to upload, email it to me and i'll give it to him :-)
11:21.02mibusoh and i have GPE sort-of building on OSX
11:21.09mibusits very messy :-S
11:21.11pb_ah, cool
11:21.13mibusbut i have gpe-gallery working
11:21.15mibus:)
11:21.21pb_how did you fix the library issues?
11:21.26mibusdamn frigging darwin
11:21.27mibusuh
11:21.29mibus*cough*
11:21.31mibus:)
11:21.34pb_quite.
11:21.34mibusi have a build.sh script
11:21.41mallummibus: osux ?
11:21.48mibusalso, gtk 2.0.6 (which is what fink installs) doesn't have the tray stuff
11:22.00mibusmallum: not quite, more like darwiSUCKS
11:22.02mibus:)
11:22.11mallummibus: install linux on it
11:22.15mibusso not all of libgpewidget can be built
11:22.23mibusyeah im going to eventually (maybe ;)
11:22.24mallummibus: remove that crazed propretry shit
11:22.31mibusi'm looking at just using gentoo live cds
11:22.39mibusmallum: and do my (paid) work how exactly? :-P
11:22.52mallummibus: mol ?
11:23.10mallummibus: you can even run wine under powerpc linux
11:23.13mibusim just going to try and get the ipod to be picked up as a firewire HDD and use a ext2 image on it or something
11:23.36mallummibus: newest benh kernels can read/write hfs+
11:23.39mallummibus: iirc
11:23.49mibusmallum: mol is useful, but then i still use "propretry" "shit" :)
11:23.55mibusmallum: there is a patch for it on sf.net
11:24.22mibusATM im happy with OSX, the whole shared lib thing is the only real complaint i have
11:24.48mibusid rather someone ported MOL to darwin
11:25.16mibus:-)
11:25.25mallummibus: that'll never happen
11:26.12mibushttp://216.170.67.34/demo/bitch.mpg
11:26.19mibusactually someone mentioned starting it
11:26.24mibusbut i havent heard anything since :(
11:26.49mibusmy biggest issue with installing linux on my laptop is I have to repartition it :-(
11:27.04mallumpb_: it was too big, now I get
11:27.06mallum~ # time -p ./test-fb
11:27.07mallumreal 0.61
11:27.07mallumuser 0.46
11:27.07mallumsys 0.00
11:27.18mibusmaybe i can come up with a weird-ass way to boot from a root disk image on a HFS+ partition :-)
11:27.23mibusanyhow
11:27.28mibusmatt wants his PC back so i gotta go :)
11:27.35mibussee you all around :-)
11:27.40mallummibus: laters dudeo
11:27.44mallummibus: good to see you
11:27.55mibusditto!
11:27.56mibus:-)
11:28.08*** part/#gpe mibus (~tmh198@202-6-129-31.ip.adam.com.au)
11:28.53dc_wahoooooo
11:28.58dc_got my cards
11:29.00mallumhehe
11:29.09dc_and I just found out the pci is dual head aswell
11:29.15dc_so I have quad screen
11:29.17dc_rock
11:30.08pb_dc_: rock indeed!
11:30.17dc_now, lets see if mandrakes hardware detection lives upto its name
11:30.18dc_brb
11:31.24pb_mallum: aha, that sounds more reasonable.
11:33.35mallumpb_: I uploaded the new sources
11:33.48mallumpb_: aha, I should set your X packages installing too !
11:34.55mallumpb_: hmm, maybe I should profile the code too, so I can figure out how much time is spent in initial startup before the blit
11:35.05pb_good thinking.
11:36.17mallumpb_: I like you patch comments :-)
11:37.10mallumpb_: can you make that xsf dir become a feed ?
11:43.00pb_an apt source?  yeah, I imagine so.
11:44.38pb_okay, I made a Packages file, try that.
11:50.45mallumcool thx, one sec just eating lunch
11:58.13mallumpb_: hmmm is apt expecting a Packages.gz ?
11:58.39mallumpb_: Im getting Failed to fetch http://people.debian.org/~pb/xsf/Packages  404 Not Found
11:58.55mallumpb_: yet that url exists
12:18.46*** join/#gpe jblinux (~chatzilla@adsl-2-127.QLD.dft.com.au)
12:19.56jblinuxHi florian
12:19.59jblinuxHi pb_
12:24.07florianhi jblinux
12:26.03*** join/#gpe dc_ (~dctanner@modem-98-59-60-62.vip.uk.com)
12:26.16dc_hum
12:26.24dc_so I've got one screen so far.
12:26.24dc_which is a start.
12:26.30dc_mallum: you have the 9000pro and 7000 right.
12:26.45mallumdc_: 7500 agp and 7000 pci
12:27.00dc_mallum: right, got a copy of your X config file you can send me?
12:27.07mallumdc_: wtf are you doing with mandrake anyway ? you a win32 user or summin
12:27.09dc_I can't get the 7000 working
12:27.24dc_mallum: quiet, it has a nicely configured gnome.
12:27.36mallumdc_: what the xfree log give you luser ?
12:27.58mallumdc_: ;-)
12:30.14dc_..not many errors.
12:30.32dc_ah.
12:30.56mallumwhat X version you using too ?
12:31.45dc_(WW) RADEON: No matching Device section for instance (BusID PCI:0:12:0) found
12:31.50dc_4.3.0
12:32.14mallumdc_: lspci shows both cards right ?
12:33.17dc_yeo
12:33.20dc_yep
12:33.31dc_Card:ATI Radeon : ATI|Radeon 7000 QY
12:33.37dc_Card:ATI Radeon : ATI|Radeon RV250 If [Radeon 9000]
12:33.37dc_Card:ATI Radeon : ATI Technologies Inc|Radeon R250 [Radeon 9000] (Secondary)
12:34.45mallumand your specifying the pci id's the xconfig right ?
12:35.01dc_yes
12:35.48dc_1:0:0 and 1:0:1 are the pci ones
12:36.02dc_oh and 0:12:0
12:38.50mallumdc_: try switching the radoen modules to the ones here http://www.xfree86.org/~alanh/
12:41.08dc_right, lets see.
12:43.27*** join/#gpe dc_ (~dctanner@modem-98-59-60-62.vip.uk.com)
12:43.41dc_hrmf
12:43.53dc_I get a driver missmatch
12:43.53dc_and it won't load it
12:44.48mallumdc_: dont you have to update your xfree86 binary as well ?
12:44.55dc_ah
12:45.01dc_sorry, what was the url again?
12:45.30mallum<PROTECTED>
12:45.30mallum<dc_> right, lets see.
12:46.27dc_okay.
12:47.41mallumdc_: make sure you backup your old binary
12:48.14dc_right, I seem to have two on my system
12:48.29dc_on usr/X11R6/bin/XFree86 and the other /usr/bin/X11/XFree86
12:49.43dc_oh right
12:49.45dc_n/m
13:15.38pb_mallum: hm, dunno.  I suppose I can try creating a Packages.gz and see if that helps.
13:17.38mallumpb_: okie cool
13:19.24pb_right, try now.
13:19.34mallumpb_: yep working
13:19.43mallumpb_: many thanks
13:19.58pb_ok, great.
13:21.37mallumpb_: hmmm something still not right
13:21.39mallumI have;
13:21.43mallumdeb http://people.debian.org/~pb/ xsf/
13:21.49mallumin my sources.list
13:21.55mallumapt update work oks
13:22.12mallumbut then when I upgrade I get a load of
13:22.15mallumErr http://people.debian.org xsf/ libxaw7-dbg 4.3.0-0ds5
13:22.15mallum<PROTECTED>
13:22.15mallumErr http://people.debian.org xsf/ libxaw7 4.3.0-0ds5
13:22.15mallum<PROTECTED>
13:22.18mallum......
13:25.50pb_hm.  you could try changing your apt line to "deb http://people.debian.org/~pb/xsf/ ./" and see if that helps.
13:29.43mallumokey
13:31.02mallumworking :-)
13:43.39*** join/#gpe herzi (~herzi@kiwi.mediascape.de)
13:50.02*** join/#gpe pattieja (~pattieja@sc2-24.217.177.39.charter-stl.com)
14:42.02*** part/#gpe cmarqu_werk (www@vws1.kens.com)
16:23.53florianHow do i create translated .desktp file from the .desktop.in?
16:26.30cmarqu_florian: They are msgstrs in the po file.
16:28.23florianoh - i missed to generate the po files... intltool-extract: Command not found
16:33.30floriancmarqu_: hmm i have a .desktop.in - i have a .pot and nothing more .-/
16:34.05cmarqu_florian: Copy the pot file to de.po and change it there.
16:34.52floriancmarqu_: yep, ok - then this is ok - is there a default file generated for the non-supported languages and c?
16:35.23cmarqu_No, it jsut uses what is in the program itself.
16:36.23florianAh, ok
16:48.17florianbbl
17:29.50*** join/#gpe BBrox (~ulmer@bbrox.org)
17:52.45PaxAnimacan someone kick me in my behind so I can fix bugs in psCalc2?
17:59.23cmarqu_PaxAnima: I called it an advanced calculator in the screenshots page, so go and make it one!
18:01.03PaxAnimacmarqu_: ;)
18:01.10PaxAnimawill do
18:01.19cmarqu_Hehe
18:01.29*** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com)
18:05.51*** join/#gpe jg (~jg@dhcp-r13.crl.dec.com)
18:11.20PaxAnimacmarqu_: I atleast changed the window title to "Advaned calculator" ;)
18:11.32cmarqu_PaxAnima: Perfect :)
18:13.35PaxAnimanow to fix a cut'n'paste bug.... if you copy a text, the tags are copied as well, making it paste the text as read-only :/
18:21.23cmarqu_I need a hh14 kernel for gpe-aerial to work with hostap_cs, right?
18:22.07*** join/#gpe florian (~fuchs@p509292D7.dip.t-dialin.net)
18:22.12florianre
18:22.35cmarqu_Ha.
18:22.41cmarqu_florian: 20:21:24     cmarqu_ · I need a hh14 kernel for gpe-aerial to work with hostap_cs, right?
18:26.01floriancmarqu_: i'm not shure - if you are lucky it works with an older version. if you are not so lucky is diesn't work with >= hh14 too.
18:26.44cmarqu_Ah, hmm. It doesn't really work I think...
18:26.45florians/diesn't/doesn't
18:27.18cmarqu_florian: http://handhelds.org/scap/port.17113.png doesn't really look well, right?
18:28.19floriancmarqu_: hmm... looks like a broadcast from an unassociated device
18:28.49cmarqu_Ah. Could be my desktop, maybe I screwed it up.
18:29.01florianbut it tells you something was captured - you'll see if channel switching works whan you look at your xsession-errors
18:30.21cmarqu_Error for wireless request "Set Frequency" (8B04) :
18:30.21cmarqu_<PROTECTED>
18:30.24cmarqu_Ah.
18:30.38cmarqu_(gpe-aerial:705): Gtk-CRITICAL **: file /home/pb/gtk+-2.2.1/gtk/gtktreestore.c: line 635 (gtk_tree_store_get_value): assertion `iter->stamp == GTK_TREE_STORE (tree_model)->stamp' failed
18:30.42cmarqu_(gpe-aerial:705): GLib-GObject-WARNING **: gtype.c:2967: type id `0' is invalid
18:30.44cmarqu_(gpe-aerial:705): GLib-GObject-WARNING **: can't peek value table for type `<invalid>' which is not currently referenced
18:30.54pb_Sounds bad.
18:31.02cmarqu_It might have done the latter when it crashed on me.
18:32.25floriancmarqu_: is the first error message repeated?
18:32.36cmarqu_florian: Yes, several times.
18:33.12cmarqu_Should I try the hh14 kernel?
18:33.46floriancmarqu_: this is the problem with the hostap driver - sometimes with some cards this works - a second later it doesn't work anymore
18:33.47cmarqu_(gpe-aerial:705): Gdk-WARNING **: locale not supported by C library
18:33.47cmarqu_killall: dhcpcd: no process killed
18:33.47cmarqu_Can't connect to localhost.2947: Connection refused
18:33.47cmarqu_SIOCGIFFLAGS: No such device
18:33.49cmarqu_Can't open pcap device: No such deviceError for wireless request "Set Mode" (8B06) :
18:33.52cmarqu_<PROTECTED>
18:33.55cmarqu_Error for wireless request "Set Frequency" (8B04) :
18:33.58cmarqu_<PROTECTED>
18:34.00cmarqu_Ah.
18:34.19cmarqu_(these messages are from before all the 8B04 messages.
18:34.36florianthe first error messages are ok - this happens on startup, which is not that clen...
18:34.47cmarqu_Okay.
18:37.43florianbtw. there is a new version in unstable :-)
18:38.48cmarqu_I updated an hour ago, still newer?
18:39.42floriannope, don't think so - did you run a complete upgrade?
18:40.45cmarqu_Yep.
18:45.14mallumcmarqu_: did you get a chance to add them files to xkbd ?
18:45.44cmarqu_mallum: No. What if I tried it right now?
18:46.53florianhmm... i was just asking myself if the dependencies are correct....
18:47.39cmarqu_florian: It pulled prismstumbler_0.7.0-pre5_arm.ipk
18:47.54floriancmarqu_: and gpe-conf?
18:48.07mallumcmarqu_: cool if you want
18:48.24cmarqu_florian: I installed that by hand before,  so I don't know.
18:48.26mallumcmarqu_: Im gonna dump the new xkbd and mbinputmgr into the feeds I think
18:48.50cmarqu_mallum: Cool. Do you also have xmerlin working okay?
18:50.44mallumcmarqu_: it works , just s l o w startup
18:51.00cmarqu_Ah, yes.
18:52.03mallumsomebody should hack xstroke so it dont have a dock app and can be externally activated
18:55.31pb_cmarqu_: boo hiss
18:55.57cmarqu_Debian has versions 1.4 - 1.7, crazy
18:56.08mallumpb_, cmarqu_: just think of the end user
18:56.34mallumxkbd, mbmerlin and mbinputmgr uploaded
18:57.29cmarqu_make[3]: *** No rule to make target `en_GB.qwerty.tabletpc.xkbd.in', needed by `en_GB.qwerty.tabletpc.xkbd'.  Stop.
18:57.44cmarqu_Ah, I had that before.
18:57.56mallumcmarqu_: hmm, did you update ? or maybe I didn't commit
18:57.59mallumcmarqu_: one sec
18:58.44mallumcmarqu_: yeah I never commited, update now
18:59.05cmarqu_Ok, thanks
19:04.16cmarqu_Argh. This is really convoluted.
19:05.01mallumwhat is ?
19:06.06cmarqu_All those Makefile.{am,in} files. They look so generated as if they are not meant to be touched by humans, yet I probably should?
19:06.34mallumcmarqu_: all you need to worry about is the Makefile.am's ignore everything else
19:06.44mallumcmarqu_: they are really simple
19:06.57cmarqu_Ok.
19:07.18cmarqu_Why doesn't "make distclean" clean stuff like autom4te.cache/output.0
19:07.19cmarqu_?
19:07.43mallumdunno
19:08.12mallumwhy do you want to make distclean ?
19:08.34mallumas your not actually in a 'dist'
19:08.54cmarqu_To see what is in CVS, so that I don't get confused by files that are genereated.
19:09.16cmarqu_Well, to me, make distclean is usually the cleanest form of clean.
19:11.51mallumcvs status ?
19:12.32cmarqu_Yeah, but I just don't want to see them at all. Quickest way would be to rm the whole directory and do a new checkout.
19:14.48cmarqu_Hmm, couldn't I have the whole directory structure for the layouts under xkbd/data, including the control file etc., and then call ipkg-build on it? I guess it could be considered unclean.
19:15.20mallumcmarqu_: have a look at packaging/ipkg/Makefile.am
19:16.58cmarqu_I had...
19:17.23cmarqu_These layouts have their own version numbering and control files.
19:19.38mallumcmarqu_: cant you just make seperate control files for each one ?
19:20.08cmarqu_Yeah, I guess that's the easiest way.
19:20.55cmarqu_I was thinking of generalizing the control files, but it's maybe not worth it.
19:21.59mallumthe x ipkg stuff just has a load of directorys for each package with a list of files and a control file in each
19:22.33mallumtheres a script that goes through and builds packages from each copying them out of an 'installed' temp directory
19:23.01*** join/#gpe dc_ (~dctanner@modem-65-55-60-62.vip.uk.com)
19:23.02cmarqu_Should I have a file of files too, or have it in some Makefile.am?
19:23.11dc_hrmf
19:23.34dc_so tri screen works in windows, and gta3 plays great :D
19:24.11pb_cool
19:25.31mallumdc_: did you get it under X ?
19:25.51mallumcmarqu_: file of files is probably best
19:26.00mallumdc_: are you old enough to play gta3 ?
19:26.13dc_mallum: ;-p
19:26.19dc_mallum: nah, still nothing with X.
19:26.32dc_mallum: it seems it must be some lame driver problem
19:26.44dc_mallum: care to have a look at my config?
19:26.46mallumdc_: what happened when you added the alt X binary
19:27.06dc_nothing new
19:27.20mallumnothing new in X log ?
19:29.07mallumdc_: did gta3 go across 3 heads or just 1 ?
19:29.18dc_just 1 :-(
19:29.29dc_although it should do atleast two
19:29.40dc_as the 2nd card is the full 7000
19:30.01cmarqu_Argh, no. I'm not in the mood for that today.
19:30.50dc_mallum: you're using the radeon driver for your 7000 pci right?
19:32.17dc_one sec
19:35.56cmarqu_Thunderstorm. Going off for a while.
19:37.19florianhuh?!
19:38.55pb_I guess Colin doesn't want to be sitting at the console when his machine is struck by lightning.
19:42.23mallumespecially when hacking auto* stuff
19:42.29mallumdouble whammy
19:42.30mallumhehehe
19:42.44pb_I dunno, the lightning might come as a relief under those circumstances.
19:43.15mallumhaha
20:42.11pb_florian: I decided to have a go at writing a new sql server.
20:48.35mallumpb_: whats wrong with sqlite ?
20:48.48pb_mallum: it's a library, not a server.
20:49.29mallumpb_: ah right, to help with syncing ?
20:49.43pb_yes
20:50.10morayand presumably let several programmes use the same db, once the apps want to do clever cross-db things
20:51.06pb_Yeah.  That bit might be more tricky though.
20:51.09mallumcan only one app at a time access a sqlite db then ?
20:51.47pb_Actually, I thought sqlite did provide database sharing itself, though with no notification of changes.
20:52.22pb_Otherwise, gpe-today would make it impossible to run gpe-calendar, which would be a serious crimp on usability.
20:53.11moraypb_: ah true
20:53.24mallumis there any plans to document the sql scema's used ( if not alread ? )
20:53.33morayI don't know what happens if two things try to write to the db though (the docs may say of course ;)
20:53.49pb_I thought the schema already was documented.
20:55.11mallumit may be a good thing to point people at when they wander in offering to hack a multisync plugin or whatever
20:55.19mallumit should'nt be that hard
20:57.28pb_yeah, indeed
20:57.51mallumwhere is the schema ?
20:59.36pb_in gpe/docs someplace, I thought.
21:03.12pb_oh, hm, maybe I imagined that.
21:04.23*** join/#gpe herzi (~herzi@kiwi.mediascape.de) [NETSPLIT VICTIM]
21:04.23*** join/#gpe mallum (~mallum@82-35-112-58.cable.ubr07.dals.blueyonder.co.uk) [NETSPLIT VICTIM]
21:29.49*** join/#gpe ibot (ibot@208.186.182.172)
21:29.49*** topic/#gpe is GPE Palmtop Environment | http://gpe.handhelds.org | http://www.handhelds.org/z/wiki/CurrentThingsToDo | http://handhelds.org/bugzilla | Familiar 0.7 issues: http://www.handhelds.org/bugzilla/showdependencytree.cgi?id=99 | Visit comp.os.linux.embedded newsgroup | Familiar 0.7-rc3 now at http://familiar.handhelds.org/releases/v0.7/install/
21:52.21*** join/#gpe pigeon (~pigeon@203.206.217.82)
21:52.47florianpb_: this sounds good - keep it simple :-)
22:00.01florian(i was in bg some time....)
22:01.20floriannight all!
22:31.38pb_yeah, looks like there isn't any current schema documentation after all.
22:32.30mallum:\
23:11.23*** join/#gpe xpato (~pato@48-32-89.adsl.terra.cl)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.