irclog2html for #gpe on 20030211

00:11.57*** join/#gpe mallum (~mallum@pc-80-193-218-21-hw.blueyonder.co.uk)
00:11.57*** join/#gpe walters (walters@verbum.org)
00:11.57*** join/#gpe paul ([Snm8WiMCa@hibernia.jakma.org)
00:11.57*** join/#gpe Bullet-Man-2 (~vadim@bgp01031502bgs.sothfd01.mi.comcast.net)
00:11.57*** join/#gpe moray (~moray@moray-ascendit.tsps1.freenet6.net)
00:11.58*** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com)
00:11.58*** join/#gpe flypiper (~netwinder@12-222-141-210.client.insightBB.com)
00:11.58*** join/#gpe cmarqu (colin@buug.mind.de)
00:11.58*** join/#gpe drkness (~sharath@nwkea-http-1.sun.com)
00:11.58*** join/#gpe eah (~seeah@pro.proact.se)
00:11.58*** join/#gpe pigeon (~pigeon@CPE-203-45-126-186.nsw.bigpond.net.au)
00:11.58*** join/#gpe spung (~rewt@a213-22-98-118.netcabo.pt)
00:13.11*** join/#gpe mallum (~mallum@pc-80-193-218-21-hw.blueyonder.co.uk) [NETSPLIT VICTIM]
00:13.11*** join/#gpe walters (walters@verbum.org) [NETSPLIT VICTIM]
00:13.11*** join/#gpe paul ([Snm8WiMCa@hibernia.jakma.org) [NETSPLIT VICTIM]
00:13.11*** join/#gpe Bullet-Man-2 (~vadim@bgp01031502bgs.sothfd01.mi.comcast.net)
00:13.11*** join/#gpe moray (~moray@moray-ascendit.tsps1.freenet6.net) [NETSPLIT VICTIM]
00:13.11*** join/#gpe flypiper (~netwinder@12-222-141-210.client.insightBB.com)
00:13.11*** join/#gpe cmarqu (colin@buug.mind.de) [NETSPLIT VICTIM]
00:13.11*** join/#gpe drkness (~sharath@nwkea-http-1.sun.com)
00:13.11*** join/#gpe eah (~seeah@pro.proact.se) [NETSPLIT VICTIM]
00:13.11*** join/#gpe pigeon (~pigeon@CPE-203-45-126-186.nsw.bigpond.net.au) [NETSPLIT VICTIM]
00:13.11*** join/#gpe spung (~rewt@a213-22-98-118.netcabo.pt) [NETSPLIT VICTIM]
00:35.50*** part/#gpe jimSnore (~jimhill@216.184.13.24)
00:47.31*** join/#gpe stargirl (hmw26@student.cusu.cam.ac.uk)
01:02.22*** join/#gpe jimhill (~jimhill@216.184.13.24)
01:14.14*** part/#gpe TheMasterMind1 (foobar@h-68-164-97-230.MCLNVA23.covad.net)
01:44.32*** join/#gpe jg (~jg@dialup-63.214.102.11.Dial1.Boston1.Level3.net)
02:29.14*** join/#gpe jimhill (~jimhill@216.184.13.24)
03:11.23*** join/#gpe pigeon (~pigeon@CPE-203-45-126-186.nsw.bigpond.net.au)
05:24.22*** join/#gpe __moray (moray@dsl-217-155-67-241.zen.co.uk)
05:53.20*** join/#gpe __moray (moray@dsl-217-155-67-241.zen.co.uk)
06:33.09*** join/#gpe __moray_ (moray@dsl-217-155-67-241.zen.co.uk)
08:45.50*** join/#gpe colin_werk (www@vws1.kens.com)
10:00.55*** join/#gpe drkness (~sharath@nwkea-http-1.sun.com)
10:45.49*** join/#gpe drkness (~sharath@nwkea-http-2.sun.com)
11:26.35*** join/#gpe pb_ (~pb_@dsl-62-3-66-201.zen.co.uk)
11:27.16pb_morning
12:07.01*** join/#gpe paul (~paulj@s0.alphyra.bbnplanet.net)
12:18.55*** join/#gpe colin_werk (www@vws1.kens.com)
12:55.49*** join/#gpe PaxAnima (~petter@v058b.studby.ntnu.no)
12:56.08PaxAnimamorning
13:27.38*** join/#gpe __moray (moray@dsl-217-155-67-241.zen.co.uk)
13:48.28*** join/#gpe __moray (moray@dsl-217-155-67-241.zen.co.uk)
14:57.35*** join/#gpe florian (~fuchs@dialin-145-254-100-150.arcor-ip.net)
14:57.40florianhi all!
14:59.17PaxAnimahi, florian
15:01.33pb_hi florian
15:01.34florianwhere does mbdesktop look for it's default background?
15:05.01floriani can't believe how many things are broken in gpe-conf :-(
15:05.09pb_:-(
15:12.39pb_florian: looks like it reads /usr/share/pixmaps/gpe-default-bg.png or something like that, judging from the source.
15:12.47pb_or the filename you give it on the command line.
15:15.34florianpb_: ok, thanks. i'll wait to add config options until xsettings work with this
15:15.46pb_good idea.
15:16.06pb_there's been some discussion of background images on the gpe@ mailing list, as you probably saw alreay.
15:16.07florianhope so...
15:21.03florianyep - gpe-conf could handle this in addition to the setting if the default application manager...
15:21.12*** join/#gpe prpplague (~JoeBob1@12.148.134.9)
15:23.24florianboth should look at xsettings like mallum suggested. an alternate solution (to handle it by gpe-conf) would be to add a xsetting that points to the application that should draw the background
15:24.34colin_werkflorian: the idea for gpe-login/gpe-ownerinfo was that it reads /etc/gpe/gpe-default-bg.png which would be a symlink to /usr/share/pixmaps/gpe-default-bg.png that root could point to a different file if wanted
15:24.44colin_werkThis was before xsettings
15:25.40colin_werkmbdesktop is in the user domain though,so a file in /etc/gpe/ would not cut it anyway, and now we *have* xsettings...
15:29.44floriancolin_werk: yep, i know - this should be the solution with no setting in xsettings, it think
15:30.40florianis mbdesktop able to use xsettings now?
15:36.25florianbbl
15:59.54*** join/#gpe pattieja (~pattieja@www.pcxperience.com)
16:01.24*** part/#gpe colin_werk (www@vws1.kens.com)
17:02.15*** join/#gpe dc_ (~dc@62.60.118.18)
17:05.59pb_dc_: afternoon
17:06.03dc_pb_: hey
17:06.12dc_pb_: I had a wonderfull idea today/
17:06.40PaxAnimahi, dc_
17:06.41pb_dc_: oh yes?
17:06.49dc_pb_: a way to intergrate simple voice reconisation into gpe-login
17:06.53dc_pb_: using gzip :}
17:07.00pb_dc_: heh, cool
17:07.01dc_pb_: and it seems to work very well
17:07.11pb_great!
17:08.13PaxAnimadc_: gzip?
17:09.13dc_gzip
17:09.35PaxAnimahow?
17:10.34dc_PaxAnima: well
17:11.03dc_PaxAnima: because gzip compresses by replacing repeated data with a pointer to it's first occurence.
17:11.37dc_PaxAnima: if you first take the gsm file with the voice password you want to check against.
17:11.56dc_PaxAnima: then append it to the one you just recorded.
17:12.16dc_PaxAnima: and if the compression ratio above a certian value it matches.
17:12.52PaxAnimaah, I see.. rather clever ;)
17:14.43PaxAnimanot the most scientific or fault tolerant method, but could work
17:14.54dc_quite
17:15.23dc_it seemed to yeild quite good results when I was testing it on my iPAQ today
17:16.20PaxAnimanot sure how the GSM files are, but I guess the volume of the speech and background noise would affect the recognition
17:25.30cmarqudc_: Looks like gpe-irc should replace olirc.
17:25.34cmarquERROR: Package gpe-irc wants to install file /usr/share/pixmaps/gpe-irc.png
17:25.41cmarquBut that file is already provided by package olirc
17:26.45pb_hm, yes.
17:27.00pb_it's a bit naughty of olirc to be shipping that file anyway, but gpe-irc should indeed replace it.
17:27.10pb_I've also put gpe-irc into the gpe-apps2 metapackage in place of olirc now.
17:27.19cmarquGood.
17:28.19pb_and, yesterday I made a crude gpe-taskmanager thing to kill errant applications, which hopefully will be enough to convince jg that xkill should be removed from the launcher menu.
17:28.54cmarquOh, yeah.
17:31.30pb_I'll run some new bootstrap images this evening.
17:33.17*** join/#gpe mallum (~mallum@pc-80-193-218-21-hw.blueyonder.co.uk)
17:33.55mallumhey
17:34.01pb_hi mallum
17:34.18mallumhi pb
17:34.32dc_hi mallum :)_
17:35.13*** topic/#gpe by mallum -> GPE Palmtop Environment | http://gpe.handhelds.org | list of tasks on the wiki: http://www.handhelds.org/z/wiki/CurrentThingsToDo | http://europa.debian.org/bugzilla | bootgpe2 images now in http://handhelds.org/~pb/unstable/ | Familiar 0.7 issues: http://www.handhelds.org/bugzilla/showdependencytree.cgi?id=99 | Please every body hammer matchbox, reporting bugs to bugzilla
17:35.22mallumhey dc_
17:40.11dc_pb_, mallum: I have decided to get a 5400 in the end.
17:40.16pb_dc_: cool
17:40.25mallumdc_: ah nice
17:40.32dc_but I guess it'll come with alot of work.
17:40.41dc_i.e. drivers.
17:40.47pb_yeah
17:40.47mallumdc_: alot of pestering jamey you mean ;-)
17:40.59dc_mallum: well yes :))
17:41.03mallumhehe
17:42.03dc_and I also have the airport card installed in the iBook...
17:42.05mallumId like one of them clam shells
17:42.10mallumdc_: ah cool
17:42.12dc_but I currently lack any sort of ap.
17:42.14dc_mallum: :)
17:42.23pb_dc_: heh.  well, I guess that's a step in the right direction.
17:42.39pb_dc_: APs aren't very expensive anymore.
17:43.09dc_pb_: quite so, but I'm currently waiting on the reply from a friend who may be able to give us an airport.
17:43.18pb_ah, indeed.
18:03.07*** join/#gpe grib (~grib@dsl093-098-046.wdc1.dsl.speakeasy.net)
18:20.53*** join/#gpe paul (~paulj@itg-gw.cr008.cwt.esat.net)
18:36.15*** join/#gpe jg (~jg@dhcp-r17.crl.dec.com)
18:37.36jgmallum: you around?
18:37.51mallumjg: 10mins i will
18:38.01*** join/#gpe walters (walters@verbum.org)
19:15.28*** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com)
19:19.13mallumjg: hey, back now
19:22.43jgmallum: so how went the aliases?
19:23.28mallumjg: havn' had a chance to try yet :(
19:23.58mallumjg: crazy matchbox hacking last week, then fosdem over weekend, non-hacking day today
19:24.05jgah, so.
19:24.09jgHow was fosdem?
19:24.14mallumjg: hopefully get a chance to try this week though :-)
19:24.25mallumjg: good
19:24.40mallumjg: discovered rmk is really a crazed pixie man
19:24.45mallumrms rather
19:24.57jgah, yes, RMS.  Pain in my rear end...
19:25.11jgdefinately crazed.
19:25.18jgThough zealot might be a better term.
19:25.34mallumId never heard him speak beofre, he sounds like a young william s borroughs minus the massive amounts of heroin abuse
19:26.12mallumjg: no movement with him and the font front then ?
19:28.14jghe's been succeeding at alienating Gnome.  He's not amenable to reason: both Eben Moglen and Bradley Kuhn are ok with what we are doing.
19:29.43cmarqumallum: Can you please add the "-sr" switch to the rxvt calls in matchbox/util/mbterm.desktop and matchbox/data/mbdock/finddockapps.sh?
19:29.56cmarqumallum: This make rxvt have the scrollbars on the right.
19:30.20pb_cmarqu: we should rather set rxvt.scrollBar_right in .Xresources
19:30.46jgmallum, pb: so here's the release plan: Russ is working on splitting unstable into familiar base/X related, and opie feeds; then we do a pre-1 through pre-N, and then a release.  Cworth's going to look at fixing xstroke this week.....
19:30.58cmarqupb_: Do we have a central Xresources file?
19:31.09pb_cmarqu: yeah, in /etc someplace.
19:31.21mallumjg: sure, cool
19:31.37jgmallum, pb_: Russ thinks he'll have the feed split by around the end of the week.
19:31.48mallumcmarqu: do you want to just stick it in buugzilla and I'll figure it out
19:31.55pb_cmarqu: it's probably called Xdefaults.  It gets loaded from /etc/X11/Xinit.d somewhere.
19:31.56cmarqumallum: Okay.
19:32.41pb_jg: sounds good to me.
19:34.50pb_now we just need to lean on people to fix the more heinous bugs. :-}
19:35.22jgpb_: yes, and close ou bugs that are now irrelevant.
19:36.41mallumjg: if I were to purchase one of them hp tablet pc's do you think you could obtain info on its ts internals ?
19:36.45paulpb_: can i ask you a question or two about bootldr? (ie are there any PXA builds of 2.19.56 around? if not, why do i get:
19:36.47paulboot-pxa.s: Assembler messages:
19:36.47paulboot-pxa.s:1702: Error: missing ')'
19:36.51paul(etc..)
19:37.00pb_paul: there used to be a 2.19.56 in feeds/bootldr/pxa, but it seems to have been blown away.
19:37.01jgmallum: talk to Jamey: he's got one and has been playing with it some.
19:37.12paulwhen trying to compile BOOTLDR_2_19_56 on ipaqX.handhelds.org
19:37.13mallumjg: oh cool :-)
19:37.29paul(i set CC=gcc-3.2 and removed CROSS_COMPILE)
19:37.41pb_paul: what does line 1702 actually say?
19:38.18paul.long ( ( (MSC_RT_ROMFLASH | MSC_RBW32 | MSC_RDF(14) | MSC_RDN(3) | MSC_
19:38.18paulRRR(2)) <<  0 )  /* MEMBNK0 150ns flash */ \
19:38.26paul(no \n in there)
19:38.44jgpb_: mallum: working toward a release should incent people to close out bugs.  I know it helps my own motivation.
19:38.52pb_jg: indeed.
19:38.58paulHEAD has the same problem. (am i using wrong assembler?)
19:39.13pb_paul: possibly.  Emacs seems to think the brackets match up OK.
19:39.27paulwhats the assembler on arm?
19:39.30pb_paul: you could try running it through cpp and checking the preprocessed source, in case something is going wrong at that stage.
19:39.35pb_pauI: I use gas 2.13.
19:39.50jgpb_: now I just have to find the bits I saved off my old laptop disk....
19:41.08jgpb_: there is some of them, now I just have to find my toolchain....
19:41.16pb_jg: :-)
19:42.01cmarqu/etc/X11/Xinit.d/60xXDefaults loads /etc/X11/Xdefaults if it exists, but it's not in any package. Should gpe-session-scripts ship it?
19:42.39pb_cmarqu: I guess so, yeah.
19:42.55paulpb_: yeah.. the ipaqX.handhelds.org cluster seem to have multiple gccs installed, maybe its using 2.95 cpp. (but i've exported CPP=cpp-3.2 and no joy :( )
19:42.58cmarqupb_: qOkay, I'll file a bug against gpe-login then.
19:43.06pb_cmarqu: ok, thanks
19:43.44pb_cmarqu: will you back out the change you made to gpe-terminal?
19:44.02cmarqupb_: Yup.
19:44.07pb_cool.
19:46.02pb_I'm going to reflash my 3900 with the hh8 image so that I can test the new user experience a bit.
19:47.19*** join/#gpe BBrox (~ulmer@bbrox.org)
19:52.05pb_cmarqu: I guess I could even try to make rxvt support XSETTINGS for the scrollbar position, which would be neat.
19:52.36pb_(though of course I still haven't quite succeeded in making it support display of text yet, so I guess we should take these things one at a time)
20:16.50cmarquhttp://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3400165422&category=31499 -- that's probably how dc started out :)
20:18.00pb_cmarqu: :-)
20:23.38pb_darn, the session is broken in the pb8 bootstrap images.
20:23.39paulhh8? :)
20:23.43paula donde? :)
20:23.55pauloh kernel
20:24.07pb_oh, heh, it looks like it is only broken for h3900.
20:24.42pb_paul: not hh8
20:25.22paulpb_: want anyone to test pb8? :)
20:26.27pauland anyone with a working 2.19.56 bootldr want to upload it and put it on the web somewhere?
20:26.50pb_yeah, I will reinstate it in the feed in a minute
20:27.08pb_or rather in about 30 minutes, when the bootstrap images are done uploading and I have some upstream bandwidth again.
20:27.54paulexcellent :)
20:28.17paulGPE is /very/ cute btw. ta very much for all the work!
20:28.56pb_glad you like it :-)
20:29.35paulnow if i just can get the full 64MB of DRAM recognised.. i might be able to banish the OOM killer that haunted me with gpe2 on my 3830.
20:29.49pb_heh.
20:30.36paulthe linux arm tree in CVS - it doesnt seem to be built with -Os.
20:30.54paulalso, couldnt a few of the hash tables be reduced in size?
20:31.01pb_It was at one time.  I'm not sure exactly what the deal is there.
20:31.20pb_Yes.  In netwinder cvs we have a CONFIG_SMALL_BUFFERS option that winds down some of the hash and buffer sizes, but it never seems to have made it into any other tree.
20:31.34pb_If you wanted to retrieve the patch from netwinder.org I would happily apply it to handhelds.
20:33.02paulwow.. you wouldnt happen to know the date this patch was commited to netwinder CVS?
20:33.17pb_no, it would have been a long time ago.
20:33.31paulwell, CONFIG_SMALL_BUFFERS is a start i guess. :)
20:33.42pb_I think you can get a big diff that includes all the differences between netwinder cvs and the "upstream" rmk tree; it ought to be fairly easy to grep it for CONFIG_SMALL_BUFFERS and extract the relevant stuff.
20:33.59pb_check ralphs's home directory, or ask the dudes in #debian-arm
20:34.08paulk
20:35.07paulor i might just ask on l-k. i've grepped and found /most/ of the hash tables i think. what else can be scaled down?
20:36.03pb_I'm not sure.  One project would be to ditch the whole VT subsystem, which is never really used for much on an ipaq, but that might be a more major undertaking.
20:36.22paulthats an idea.
20:37.45paulwonder how intertwined vt handling is with things like serial console and ptys.. (probably too much)
20:38.23pb_I think the main problem at the moment is that you can't turn CONFIG_VT off but still keep the framebuffer.
20:38.34paulah.
20:38.59pb_Ideally you would want to keep the VC-switching code and the raw fb interface, but chop out the actual VT100 emulation and text console parts.
20:39.17paulideally yes.
20:40.16pb_I have a suspicion it may be a bit tangled up with keyboard input as well, which is something else we would want to keep.
20:40.23paulyes.
20:43.27*** join/#gpe florian (~fuchs@pD9E5FAAE.dip.t-dialin.net)
20:43.30florianhi!
20:45.17paulhi
20:50.36pb_hi florian
20:56.53moraypb_: did the VT stuff ever get cleaned up yet in 2.5? I remember there was discussion about it, but I haven't been following exactly what's gone in
21:15.49pb_moray: no, I'm not sure either.
21:24.23*** join/#gpe TheMasterMind1 (foobar@h-68-164-97-230.MCLNVA23.covad.net)
21:38.51paulpb_: re my bootldr prob -> i had CONFIG_MACH set to IPAQ in config.local.mk, from closer reading of config.mk that should apparently be H3900 instead.
21:41.26paula question, if you compile bootldr - what is the bootldr-c340-<ver> file?
21:43.01paulsame size as bootldr-<ver> but different md5sum.
21:44.47paulah, offset bootldr is loaded into ram.
21:51.39pb_paul: yes, exactly that.
21:52.38paulheh.. but 2.19.47 only sees 0x04000000 of my ram :)
21:52.48paulno.. hang on.
21:52.53paulthat is 64MB.
21:53.04pb_yes.
21:53.10paulhmm.
21:53.49paulso why does 'load 0xc3400000' hang as i send bootldr-c340?
21:54.06pb_ok, there is a h3600 bootgpe2-pb8 image in the usual directory now.  The 3900 one was screwed up so I deleted it.
21:54.24paul(seems like a really really handy way to make sure the bootldr i built is functional - and avoiding bricking the thing if it does have some major problem)
21:54.39pauldrat
21:57.06paulpb_: the 48MB flash thing: if it makes the problem any easier, having 32MB, and a seperate 16MB segment would be fine from a useability POV. (i take problem is trying to hide some kind of bank switching between 32MB bank of flash ROMs and an additional 16MB bank of flash ROMS?) thats what mount is for. :) (if that isnt the problem, nm)
22:01.41paulanyway.. happy hacking!
22:01.44paulbfn
22:13.52PaxAnimaevening
22:19.27*** join/#gpe flypiper (~netwinder@12-222-141-210.client.insightBB.com)
22:55.16*** join/#gpe Bullet-Man-2 (~vadim@bgp01031502bgs.sothfd01.mi.comcast.net)
22:56.25floriangood night!
22:57.28PaxAnimanight, florian
23:46.36PaxAnimawould be cool to do a scorched earth type of game
23:51.36*** join/#gpe pigeon (~pigeon@CPE-203-45-126-186.nsw.bigpond.net.au)
23:52.49PaxAnimahi, pigeon

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.