00:12.54 | *** join/#gpe ade|indy (~adrian@unimatrix.ex.ac.uk) |
00:14.40 | *** join/#gpe thomas__ (~thomas@202.88.173.5) |
01:35.22 | *** join/#gpe prpplague (~joebob@21-190.lctv-b4.cablelynx.com) |
01:48.41 | *** join/#gpe jg (~jg@dialup-67.30.181.144.Dial.Boston1.Level3.net) |
02:07.08 | *** join/#gpe hookah (~erik@c-558972d5.012-58-73746f32.cust.bredbandsbolaget.se) |
02:17.53 | *** join/#gpe paul ([Cyn+oKAEt@hibernia.jakma.org) |
03:50.35 | *** join/#gpe paul ([CC7IoJhsp@hibernia.jakma.org) |
05:10.58 | *** join/#gpe drkness (~sharath@nwkea-http-2.sun.com) |
05:13.17 | drkness | hi guys.. anyone knows the nomenclature of the kernel version numbers? i.e if I have a kernel 2.4.19-rmk4-pxa2-hh10 what do the rmk numbers, the pxa identifiers and the hh identifiers denote? |
09:39.15 | *** join/#gpe florian (boor@vmax.unix-ag.uni-siegen.de) |
09:39.21 | florian | morning! |
09:41.18 | *** join/#gpe mallum (~mallum@pc-80-193-218-21-hw.blueyonder.co.uk) |
09:41.53 | *** join/#gpe PaxAnima (~petter@v058b.studby.ntnu.no) |
09:42.31 | florian | PaxAnima: hell |
09:42.35 | florian | PaxAnima: hello |
09:42.57 | PaxAnima | hi, florian |
09:45.18 | PaxAnima | indeed ;) |
09:49.30 | pigeon | Or, you can use PaxAnima's minipredict! |
09:51.02 | PaxAnima | pigeon: well, it depends if florian uses "hell" more than "hello" ;) |
09:51.15 | pigeon | lol ;) |
10:00.44 | *** join/#gpe drkness (~sharath@nwkea-http-1.sun.com) |
10:24.30 | *** join/#gpe mallum (~mallum@pc-80-193-218-21-hw.blueyonder.co.uk) |
10:39.26 | PaxAnima | hey, mallum |
10:42.26 | mallum | hey PaxAnima |
10:42.52 | *** join/#gpe pb_ (~pb_@dsl-62-3-66-201.zen.co.uk) |
10:43.34 | pb_ | morning |
10:43.35 | mallum | hey pb_ |
10:43.46 | pb_ | hi mallum |
10:45.47 | *** join/#gpe drkness (~sharath@nwkea-http-2.sun.com) |
11:04.06 | mallum | pb_: I discovered last night you can do a ctrl-alt-backspace on the c700 and it will kill off X :-) |
11:06.23 | pb_ | aha. |
11:12.31 | mallum | pb_: I just noticed a post on the arm kernel list that someone has Linux 'nearly' booting on a loox |
11:15.01 | pb_ | oh right, neat. |
11:15.41 | mallum | pb_: he seems to be having problems with the MMU, the post is mar 09 - titled MMU problems on xscale |
11:27.26 | florian | bbl |
11:30.06 | mallum | pb_: do you how you'd get the current vt from a script ? |
11:32.40 | pb_ | No. You might have to write a C program for that. |
11:34.48 | mallum | pb_: I think 'openvt' will do what I want |
11:35.36 | pb_ | Okay. What is it you're trying to do? |
11:36.16 | mallum | switch to next freevt, run ts_calibrate, exit back to orinal vt |
11:36.40 | pb_ | Shouldn't ts_calibrate take care of the VT switching itself? |
11:37.19 | mallum | not sure, guess I could patch it to do so |
11:40.44 | pb_ | I think it should. It's a framebuffer-based app, right? |
11:41.04 | pb_ | It's never appropriate for those to just scribble over the foreground VT; they should allocate one, switch to it, do their stuff, then switch back. |
11:44.11 | mallum | pb_: yes it definetly seems to be doing something in that respect |
11:44.30 | mallum | pb_: I need to find some docs on these vt ioctls to figure it out |
11:46.37 | mallum | pb_: the console-programming howto seems to have the info I need |
11:47.37 | mallum | pb_: yes it appears it does switch vt :-) |
11:49.40 | pb_ | ah right. |
11:51.01 | mallum | pb_: I wonder why opie dont do it, its only approx 10 lines of code :( |
11:51.30 | pb_ | I think they feel it's SEP (more specifically, Trolltech's). |
11:52.07 | PaxAnima | SEP? |
11:52.11 | pb_ | someone else's problem |
11:52.34 | PaxAnima | ah |
11:59.42 | pb_ | I guess one could file against libqte to that effect and see what happens. |
12:01.13 | *** join/#gpe colin_werk (www@vws1.kens.com) |
12:01.55 | colin_werk | Hi. |
12:02.29 | PaxAnima | hi, cmarqu_werk |
12:02.48 | pb_ | Hi Colin |
12:10.18 | pb_ | hm, I wonder if bluez supports bluetooth mice yet. |
12:24.32 | cmarqu_werk | pb_: http://handhelds.org/scap/port.3841.png -- shouldn't GTK wrap before "to" instead of cutting it off? |
12:25.00 | cmarqu_werk | Oh, and "might" is also missing. Hmm. |
12:26.16 | cmarqu_werk | And the buttons should be a bit closer together... I though I had fixed that. |
12:26.25 | mallum | cmarqu_werk: run matchbox with -use_lowlight yes , and you get a nice effect on that logout dialog |
12:28.58 | cmarqu_werk | mallum: The "shading the background" thing? |
12:29.45 | mallum | cmarqu_werk: yeah, it only happens on modal dialogs now |
12:29.51 | pb_ | cmarqu_werk: Yeah, it should. I don't think GTK word wrapping works terribly well at the moment to be honest. |
12:30.19 | pb_ | cmarqu_werk: I fiddled the wrapping parameters in the current build, though, which may have been a mistake. I will try putting them back and see if matters improve. |
12:32.00 | *** join/#gpe florian (~fuchs@dialin-145-254-094-099.arcor-ip.net) |
12:41.17 | cmarqu_werk | pb_: Cool. |
12:42.11 | cmarqu_werk | mallum: Do you think one could live with using -use_lowlight all the time? This would prevent me from opening xkbd, right? |
12:49.46 | mallum | cmarqu_werk: no not really, not with gpe anyway |
12:51.55 | mallum | cmarqu_werk: well gpe on the ipaq anyway, or other devices without a keyboard |
12:52.28 | cmarqu_werk | I see. |
12:53.17 | mallum | cmarqu_werk: one thing I was thinking of though, is adding an extra propreitry hint that dialogs could set to force a lowlight, even if the option is turned off |
12:53.30 | mallum | cmarqu_werk: that may be useful for the logout dialog ? |
12:58.30 | moray | mallum: it would seems reasonabl |
12:58.43 | moray | mallum: if we had that we could use it for other important messages too perhaps |
12:59.01 | moray | mallum: e.g. a low battery warning etc. (?) |
12:59.17 | mallum | moray: yes, anything that didn't require keyboard input |
12:59.29 | mallum | moray: and was 'urgent' |
12:59.32 | moray | mallum: yes |
12:59.56 | mallum | moray: there may even be an urgency hint for windows ... I think the HIG defines it but I dont know if it exists |
12:59.58 | moray | mallum: (and that doesn't require you to look at the programs behind, obviously, which 'in-application' things might do) |
13:00.01 | moray | ah right |
13:00.50 | mallum | unfortunatly theres no way of getting the hack to work without covering the enitre display |
13:01.52 | moray | mallum: well, I think it might get confusing otherwise at any rate |
13:04.15 | cmarqu_werk | Yep, that would be a cool feature. |
13:05.51 | cmarqu_werk | It would require that gpe-question gets a new "--urgent" switch which enables it, since gpe-logout is a mere shell script wrapper around gpe-question.. |
13:05.52 | mallum | cmarqu_werk: okey I'll add it, when I do some mb hacking later |
13:06.19 | cmarqu_werk | http://mail.gnome.org/archives/usability/2002-January/msg00088.html |
13:07.07 | cmarqu_werk | This even talks about the window title. |
13:09.22 | mallum | cmarqu_werk: ah cool, I will take a look in a mo ... |
13:39.09 | mallum | pb_: X 4.3-3 packages uploaded. fixes for vt / tslib stuff. initial go at xmodmaping. fc-list packaged |
13:47.18 | pb_ | great. |
13:50.38 | mallum | theres also an updated tslib package which basically fixes the old ts.conf |
14:27.08 | pb_ | ok. I will either reflash or repair my h3600 later today and try those out. |
14:31.34 | cmarqu_werk | pb_: What package should set overcommit_memory (if we decide to enable it in familiar)? familiar-base? Or sysvinit? |
14:36.06 | pb_ | cmarqu_werk: I dunno. Either of those would be OK, or it could even go in a separate package. |
14:51.08 | cmarqu_werk | Hmm, an extra package would probably be overkill, considering that it also make the Packages file larger an all that. |
14:51.26 | pb_ | True. |
14:51.36 | pb_ | Well, I guess familiar-base would be my choice. |
14:51.41 | cmarqu_werk | familiar-base seems more natural to me, but sysvinit has many files in /etc/init.d |
14:52.17 | pb_ | Ideally we would like to keep sysvinit as close as possible to the upstream source, without adding too many Familiar specific things to it. |
14:52.24 | cmarqu_werk | Well, jamey controls both of these packages, so he can always choose what he likes. |
14:52.28 | pb_ | Indeed. |
14:52.29 | cmarqu_werk | pb_: ACK |
14:57.22 | *** join/#gpe prpplague (~JoeBob1@12.148.134.9) |
14:57.49 | cmarqu_werk | Argh, a7r is the default maintainer for familiar-base. |
14:58.14 | pb_ | Oh yeah. |
14:58.54 | cmarqu_werk | I'm rudely reassigning this to jamey :) |
15:01.57 | pb_ | Right, good plan. |
15:17.17 | mallum | pb_: hmm, ts_calibrate does not seem switch vt immediatly. you have to click the touchscreen, then it appears |
15:19.36 | mallum | I guess it needs to flush something |
16:19.45 | mallum | pb_: ping |
16:21.40 | pb_ | mallum: hi |
16:22.45 | mallum | pb_: looking through ts_calibrate code, and reading up of vt switching it seems stuff should work :( |
16:23.10 | mallum | pb_: though it appears to open/close framebuffer, then re-open it again strait after |
16:23.31 | mallum | pb_: any idea why they'd want to do that ? |
16:25.13 | pb_ | nope, I can't think of any reason to do that. |
16:25.58 | mallum | pb_: I think I'll remove it and see what happens |
16:26.10 | mallum | pb_: it could of course just be the crappy fb driver on the c700 |
16:26.30 | mallum | pb_: dd if=/dev/zero of=/dev/fb0 seems to cause a kernel crash :( |
16:26.31 | pb_ | what's the actual problem you're having? |
16:26.35 | pb_ | ah. |
16:26.46 | mallum | <mallum> pb_: hmm, ts_calibrate does not seem switch vt immediatly. you have to click the touchscreen, then it appears |
16:27.16 | pb_ | mmm, seems fairly unlikely that the fb driver would be to blame for that, though you never know. |
16:27.44 | pb_ | does strace reveal anything about what is going on? |
16:28.14 | pb_ | I guess you could also try it on the ipaq / cats / x86 desktop and see what happens there. |
16:29.43 | mallum | pb_: yeah |
16:29.48 | *** join/#gpe paul ([UlmZTPupC@212.17.60.230) |
16:31.50 | pb_ | hi paul |
16:49.20 | *** part/#gpe thomas_ (~thomas@202.88.173.5) |
17:03.02 | paul | hi pb |
17:46.23 | pb_ | can anyone reproduce http://europa.armlinux.org/cgi-bin/bugzilla/show_bug.cgi?id=177? |
17:49.42 | paul | pb_: yes |
17:50.31 | paul | pb_: exactly as described (well, i logged out - not rebooted). |
17:50.36 | pb_ | right. |
17:50.45 | paul | second time - wont start. it will start from cli though. |
17:51.05 | pb_ | I wonder what could be wrong. |
17:51.18 | paul | odd |
17:51.43 | pb_ | very. well, I guess it sounds a bit like a matchbox issue. Does the hourglass go away correctly when the program has started? |
17:52.29 | cmarqu | pb_: Me too. |
17:53.12 | pb_ | Okay, so it's pretty widespread then. |
17:53.25 | cmarqu | Funny thing is, it works now with gpe-edit. |
17:53.28 | pb_ | Strange I've never seen it. |
17:53.30 | paul | pb_: no hour glass |
17:53.43 | pb_ | paul: no hourglass at all? |
17:53.58 | cmarqu | Ah, gpe-nmf is harmed too. |
17:54.04 | paul | pb_: if i open something else, gpe-contacts will start again (hourglass appears and goes away once its open as yu'd expect) |
17:54.12 | cmarqu | hourglass the first time, going away. 2nd time, no hourglass. |
17:54.16 | paul | pb_: not when it wont start. |
17:54.21 | pb_ | paul: ah, I see. |
17:54.35 | paul | pb_: you tap the icon, it 'highlights' (ie it gets the tap event) - nothing happens. |
17:54.47 | pb_ | paul: ok. sounds like a bug in the desktop then. |
17:54.48 | cmarqu | Yep. No CPU usage. |
17:54.52 | paul | seems so |
18:02.43 | cmarqu | I have a methodical problem. Suppose I want to make the label in a GtkMessageDialog (created with gtk_message_dialog_new()) selectable. How would I find out what the elements in GtkMessageDialog are? By looking at the GTK 77source? |
18:02.53 | mallum | pb_: sounds like its a problem with startup notification and single instance |
18:03.42 | mallum | pb_: are all them apps listed marked as single instance in there .desktop files ? |
18:04.27 | pb_ | Yeah, I think so. |
18:05.44 | cmarqu | gpe-edit is not, and it doesn't have that problem. |
18:09.03 | pb_ | bbl |
18:10.44 | mallum | paul: does it happen all the time for you ? |
18:11.32 | cmarqu | mallum: It does for me I think. |
18:12.01 | mallum | hmm, I wonder how that crept in ... |
18:12.29 | paul | mallum: yes. it wont start if i close gpe-calendar, etc.. ie if nothing else is running, it wont start. |
18:12.34 | paul | (starts from cli though) |
18:13.25 | mallum | paul: yeah I just checked it, broke for me to :( |
18:13.41 | paul | mallum: tis in your capable hands so :) |
18:21.30 | *** join/#gpe jg (~jg@dhcp-r25.crl.dec.com) |
18:24.38 | paul | hi jim |
18:27.07 | jg | paul: hi. |
18:27.49 | mallum | paul: it actually seems a little more random than that |
18:27.58 | paul | mallum: oh? |
18:28.19 | mallum | paul: if you open 2 apps, like the calender and then the todo, close one it'll reopen ok |
18:28.42 | paul | and if you close them all and then try reopen contacts? |
18:28.44 | mallum | paul: the problem seems to be when just one app has it open |
18:29.02 | mallum | paul: then it will go alittle screwy I think |
18:29.34 | paul | where's the problem? "when just one app has it open" - has what open? (/me is curious) |
18:31.24 | mallum | paul: it set a property on the root window mapping binary names to window id's |
18:31.43 | mallum | paul: when all the apps participating in this, it seems to sometimes not close the list |
18:31.49 | paul | ah. |
18:31.53 | paul | it being? |
18:31.57 | mallum | paul: leaving the last one dangling |
18:31.57 | paul | one of the apps? |
18:32.05 | mallum | paul: yeah lastone closed |
18:32.13 | mallum | paul: so it thinks it still exists |
18:32.47 | mallum | paul: I will make a new package with fix |
18:32.55 | paul | woohoo :) |
18:33.17 | paul | how difficult will it be to add/fix scrolling on menus that are bigger than screen? |
18:33.54 | mallum | paul: yeah I will do that |
18:33.59 | paul | woohoo :) |
18:34.06 | mallum | paul: but Im kind of having a break from any heavy mb development |
18:34.13 | paul | k |
18:34.21 | mallum | paul: I have lots of other stuff to do, documenting etc |
18:34.35 | mallum | paul: its on the todo list though |
18:34.56 | paul | mallum: thats no excuse! what do you think we pay you for! (oh wait, we dont... :) ) |
18:36.34 | paul | actually.. that brings up a question.. how do you guys manage to find all the time to spend on gpe? |
18:36.40 | jg | mallum: there are a few people who use the VGA out card: ergo the need for Xipaq, rather than Xfbdev. |
18:38.02 | jg | mallum: All I get with your new packages is an X cursor on the screen: something busted.... |
18:38.42 | mallum | jg: xipaq has deps on h3600 kernel headers though, that is bad |
18:39.14 | jg | mallum: what's it hurt? |
18:40.11 | jg | mallum: or build and package both, and use provides to do the right dependency thing. |
18:40.50 | mallum | jg: all you get is the ts struct ( not needed anymore as tslib ) and the xtra key defines ( fixed with modmap ) |
18:41.06 | mallum | jg: and for that you have to mess around getting the hh kernel |
18:41.36 | mallum | jg: that as far as I can work is all xipaq is more than xfbdev |
18:41.46 | jg | mallum: problem: xmodmap would have to be run every server reset (e.g. logout). |
18:42.07 | mallum | jg: yep, as will the fontpath setting etc |
18:42.43 | mallum | jg: gpe-dm reruns the session scripts on a logout/login anyway |
18:42.54 | mallum | jg: as well as running the server with -noreset |
18:43.25 | mallum | jg: and nullxdm uses -noreset flag |
18:43.32 | jg | mallum: people with either the HP or Voyager card won't be happy.... |
18:44.24 | mallum | jg: we can make a voyager xserver seperate |
18:44.41 | mallum | jg: there are much more people with zaurus's than voyagers ... |
18:44.52 | jg | mallum: that is my suggestion: build/package both, and use provides to solve the issue. |
18:44.56 | mallum | jg: I didn't think voyagers even worked anymore ? |
18:45.05 | jg | mallum: I know of no reason they shouldn't. |
18:45.13 | mallum | jg: send me a voyager card so I can test and I'll package it |
18:45.18 | jg | mallum: at some point soon, we'd like to get the Margi card going. |
18:45.33 | jg | mallum: don't have one to send you, unfortunately. |
18:45.55 | mallum | jg: do you have one ? |
18:46.04 | jg | mallum: not sure I do. I do have an HP card. |
18:47.05 | *** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com) |
18:48.28 | mallum | jg: so you could test and build a HP enabled xserver ? |
18:49.19 | jg | mallum: I'm building the sever, but now get the following messages.. (have you seen them before)....Xext/libext.a(xres.o): In function `ProcXResQueryClientResources': |
18:49.19 | jg | xres.o(.text+0x4a8): undefined reference to `RegisterResourceName' |
18:49.19 | jg | xres.o(.text+0x548): undefined reference to `ResourceNames' |
18:49.33 | jg | mallum: as soon as I can get the server to link, I can test.... |
18:50.08 | mallum | jg: you need to add that -Dresource extra to the KDRIVE_EXTRA_DEFINES |
18:53.50 | mallum | jg: how are you starting X by the way , abd what revision of my patches are you running ? |
18:54.27 | jg | mallum: thanks for the pointer: I had missed that addition. -DXResExtension fixed it. It does say there is a bug elsewhere in the Imakefile stuff, as that ought to get set by the extension declaration. |
18:54.48 | jg | mallum: I just got it to link this instant, so now I can try running it. |
18:55.07 | jg | mallum: but right now, the packages I have installed on my iPAQ are useless, so I can't do anything. |
18:55.18 | mallum | jg: how did you start X ? |
18:55.19 | jg | mallum: have any idea why your new packages break things? |
18:55.38 | mallum | jg: yes, they are not finished :-) |
18:55.38 | jg | mallum: reboot.... |
18:55.50 | pb_ | jg: xfree86.cf does put the XRes define in XFree86ServerDefines, but kdrive seems to use something of its own rather than looking at that variable. |
18:55.53 | mallum | jg: and is X getting started by gpe-dm or nullxdm ? |
18:56.44 | jg | mallum: gpe-dm |
18:56.58 | jg | mallum: I also note several shells running /etc/X11/Xinit... |
18:57.01 | jg | Not good. |
18:58.13 | jg | mallum: I guess I'll wipe the unit and reinstall... Sigh.... |
18:58.15 | mallum | jg: were trying to improve the startup procedure. gpe-dm may or maynot work with it currently. I suggest you use nullxdm for now |
18:58.47 | mallum | jg: just remove the X packages and install the ones from unstable |
18:58.59 | mallum | dinner time. |
18:59.03 | mallum | paul: that bug is fixed |
18:59.07 | jg | mallum: I've got a phone call to make; I'll just reinstall. |
18:59.17 | jg | pb_: sounds like we need to fix kdrive.cf |
18:59.22 | paul | mallum: oh, excellent - very quick! |
19:00.34 | pb_ | jg: probably. |
19:01.55 | pb_ | jg: by the way, the only change required to make gpe-dm work with mallum's packages should be to remove /etc/X11/Xserver and make it a symlink to /etc/X11/X. |
19:09.35 | mallum | pb_, jg_: what could be happening to jg, is the server is appearing to hang when actually its the switch to ts_calibrate happening |
19:10.33 | mallum | pb_, jg_: taping the display should fix things. If it dont remove ts_callibrate from Xinit, and run it manually |
19:10.48 | mallum | my next set of packages should fix that |
19:15.41 | mallum | jg: ping |
19:16.16 | *** join/#gpe florian (~fuchs@p509291F5.dip.t-dialin.net) |
19:16.22 | florian | hi all! |
19:16.29 | pb_ | hi florian! |
19:16.39 | jg | mallum: I'm just reloading from scratch... Proper patch for kdrive.cf coming your way.... |
19:18.09 | jg | mallum: I'd restarted the reload 5 seconds before pb_ told me how to fix it... Oh well, I have to make a phone call now anyway. |
19:18.42 | mallum | jg: cool |
19:19.28 | mallum | jg: please bear with me, Im attempting to improve/fix alot of stuff and I make no promises them packages will work perfectly, if work at all |
19:20.21 | jg | mallum: fine. NP. |
19:20.38 | mallum | jg: I definetly wouldn't attempt to install them with some sort of reliable external shell connection :-) |
19:20.40 | jg | mallum: I just want to get to the point I can chase the xset fp+ problem.... |
19:21.20 | jg | mallum: that one seems to have you scared, and so it seems like one I should go after.... |
19:21.25 | mallum | jg: we should be there I think |
19:21.35 | mallum | jg: that one is easy ;-) |
19:22.05 | jg | mallum: it should be.... |
19:22.25 | mallum | jg: I think the clues are in the 4.3 release notes regarding fonts |
19:22.56 | mallum | jg: they seem to suggest mkfontdir etc have been updated for handling ttf fonts |
19:23.56 | jg | mallum: probably right... In which case, what I have installed on my laptop should do fine to rebuild them. |
19:25.17 | jg | mallum: I'll go read the release notes now... I normally just build and install, and ignore the docs... |
19:26.39 | mallum | jg: cool, I have you patch btw. I will add it |
19:29.11 | mallum | pb_: OZ dont use ldconfig , just LD_LIBRARY_PATH . is that safe ? ( seems to work ok ) |
19:29.26 | pb_ | mallum: slow but safe, yes |
19:29.46 | mallum | pb_: will prelink still work ? |
19:29.50 | pb_ | no |
19:30.45 | *** join/#gpe dc_ (~dc@modem-126-42-60-62.vip.uk.com) |
19:31.43 | pb_ | dc_: evening |
19:31.48 | mallum | pb_: woah, how come its 450k ? |
19:31.48 | dc_ | hey :) |
19:31.58 | mallum | hey dc_ |
19:32.06 | pb_ | mallum: what binary are you talking about? |
19:32.21 | mallum | ldconfig |
19:32.26 | pb_ | static linked |
19:32.37 | mallum | ah of course |
19:33.44 | PaxAnima | <PROTECTED> |
19:34.01 | jg | pb_: ouch: maybe it should get relinked against a smaller C library.... |
19:34.18 | pb_ | jg: actually I think we should just link it dynamic and be done. |
19:34.19 | mallum | pb_: is the speed difference that big with/without it ? |
19:35.05 | pb_ | dunno, you'd have to make some measurements. |
19:35.10 | jg | pb_: and what happens if the config file gets busted? You are toast... But maybe that's ok in our environment.... |
19:36.00 | pb_ | jg: most people find it easier to reflash than try to repair a busted system anyway. But, if you were determined, 95% of the time you could probably rescue the situation with LD_LIBRARY_PATH. |
19:36.41 | jg | pb_: I guess I buy that... Go for it: I expect it would save much of that space.... |
19:38.45 | mallum | I off ppl, probably be back later |
19:40.25 | jg | pb_: one of the little things done really right about the Palm is that it is clever enough to capitalize fields when appropriate.... Have you given that any thought in the gpe context? |
19:43.29 | *** join/#gpe dc__ (~dc@62.60.121.154) |
19:48.30 | pb_ | jg: Yeah, that's an interesting idea. I must admit it's not something I've ever given any thought to. |
19:50.39 | moray | pb_: hm, if we had some cleverer field-editing system that knew how to talk to the input methods it could be told you were trying to input different classes of things (e.g. a name / a number) |
19:50.57 | pb_ | moray: Yeah, that's more or less what I was thinking too. |
19:51.07 | moray | pb_: inputting a name it could then also know to complete from your contacts database, etc., rather than the app itself having to provide that functionality |
19:51.21 | pb_ | Yes, that'd be cool. |
19:51.22 | moray | pb_: this would also (in principle) let us do work completion etc. properly |
19:51.37 | moray | pb_: you really need bidirectional information flow for that, not just keystrokes -> input method -> app |
19:51.43 | moray | (word completion) |
19:51.56 | pb_ | yeah, I figured work completion was too much to hope for. |
19:52.01 | moray | :) |
19:52.31 | pb_ | but yes, I agree. Maybe you could try to come up with some framework for how this might be implemented? |
19:52.57 | moray | pb_: sure, though I can't do anything in the next few days I expect |
19:53.18 | pb_ | I suppose the obvious application-side paradigm would be subclasses of GtkEntry for inputting different class of thing. |
19:53.41 | pb_ | moray: no rush |
19:54.15 | moray | pb_: well, in the end I'd like to see e.g. the handwriting recogniser appearing in-place rather than at the foot of the screen |
19:54.47 | pb_ | yeah. |
19:55.18 | pb_ | the other (rather simpler) thing that people keep mentioning is that it would be nice for the keyboard, or other selected input method, to automatically pop up when you put focus in an entry widget. |
19:55.27 | moray | ah right |
19:56.02 | moray | but yes, some kind of input mediator could know to bring it up when you need to type (if you have the keyboard set as the method to use) |
19:57.23 | jg | pb_: if you use a palm, you notice its lack *very* quickly.... |
19:58.08 | pb_ | jg: right. |
19:58.30 | PaxAnima | jg: not hard to implement, and very usefull.. I have it in psPIM |
19:58.40 | jg | pb_: Microsoft gets this wrong on the tablet PC's.... |
19:59.20 | moray | pb_: it doesn't seem wrong to integrate this stuff at the GTK level though, presumably? |
19:59.25 | jg | pb_: Alan Nemeth was bitterly complaining to me about this earlier today, as he tried to write "nemeth" to his mail account, and it made it "Nemeth". |
19:59.42 | jg | pb_: needless to say, the UNIX system his mail was on cared..... |
19:59.53 | moray | pb_: as in, I think you want to be able to aim for something that integrates into applications' UIs |
20:00.08 | jg | (Microsoft *always* capitalizes; this is probably worse that what we do). |
20:00.10 | pb_ | moray: agreed |
20:00.12 | moray | jg: ah yes |
20:00.26 | pb_ | jg: yeah, I hate that kind of behaviour in Word as well. |
20:00.40 | jg | pb_: only the app really knows.... |
20:00.51 | pb_ | jg: right. |
20:00.54 | moray | jg: their email stuff also seems to change everything into "Surname, Firstname" <someone@example.com> too |
20:01.26 | jg | moray: he was just logging into a java based mail system that is on a UNIX system from XP.... |
20:01.43 | jg | moray: so Redmond think just screwed him up. |
20:02.09 | moray | jg: I can imagine it would be irritating, yes |
20:02.46 | moray | you want to be sure of what you're doing before you alter user actions (like things they've typed in) |
20:03.42 | moray | as pb_ said, the same thing seems to happen the whole time in MS Word; I dislike the way, for example, that if I have to use it and type in a date, it will change the date's format into the MS-approved one! |
20:04.12 | jg | moray: that is probably controllable in some obscure UI panel in Word, btw.... |
20:04.13 | moray | I assume this stuff is configurable, but I doubt most people know how to do that, and I never use one MS system for long enough for it to be worth looking |
20:04.29 | jg | moray: I once explored to find such things.... |
20:05.13 | moray | jg: yes, but I have an MS OS on any of my own systems; if I change it on a relative's computer (say), they're liable to get upset since it'll turn out they were used to the existing behaviour |
20:17.29 | florian | anyone an idea how the busybox implementation of syslog works? |
20:22.30 | pb_ | at what level? It just logs to a ring buffer in RAM. |
20:25.11 | florian | i looked at the source of logread, that looks quite easy and works well. |
20:25.28 | pb_ | florian: right, it just uses some sysv shared memory to peek at the buffer. |
20:27.25 | florian | now i want to get everything new from this buffer - i detach from shm, attach again an look if the pointer to buffers tail changed... |
20:30.09 | florian | This seems to change never, only if i restart the app i get new log entrys :-( |
20:30.50 | PaxAnima | what's libmimedir for? |
20:31.43 | florian | handling of vcard data... afaik |
20:32.07 | PaxAnima | ah |
20:34.10 | jg | pb_: I think mallum is right about the font path problem. I can nfs mount the font directories from my laptop running XFree86 4.3 and |
20:34.17 | jg | everything seems to work. |
20:34.32 | pb_ | jg: right. |
20:36.46 | jg | I take that back... |
20:37.22 | jg | I can add bitmap fonts, but I get a failure adding TTF fonts..... |
20:37.32 | pb_ | ah. |
20:42.21 | pb_ | florian: wanna live dangerously? ipkg install http://handhelds.org/feeds/unstable/gpe-mimedir-tools_0.0_arm.ipk |
20:44.36 | florian | pb_: i like living dangerously ;-) |
20:44.59 | pb_ | florian: ah good |
20:45.28 | pb_ | well, you should find that package provides "send-vcard" and "import-vcard" commands. |
20:46.14 | pb_ | They probably don't work very well at the moment, but I can at least get it to dump out some of my contacts. |
20:46.41 | pb_ | I'm kind of hoping this will be sufficient to motivate someone into making them better :-) |
20:50.09 | florian | ok, seems to work... now i'd need something to test |
20:51.05 | pb_ | Evolution can import and export vCard, if that helps. |
20:51.30 | pb_ | My cellphone seems to be able to as well, though I've never really tested it. |
20:53.18 | pb_ | I'll build some openobex packages next. Maybe then we can get IrDA and Bluetooth interchange working. |
20:56.48 | florian | This would be cool... what about gpe beam? |
20:57.27 | pb_ | Well, gpe-mimedir-tools is actually built from the gpe-beam source directory. The actual beaming functionality doesn't exist yet, but I doubt it would be hard to write. |
20:58.00 | florian | this sounds like we all could need more time :-) |
20:58.06 | pb_ | Yeah. :-/ |
20:58.37 | pb_ | There's also still the question of actual syncing. I have a plan for that, but it will be a moderate amount of work to implement it. |
20:58.48 | pb_ | Firstly, we would really need a working usqld. |
20:58.51 | PaxAnima | pb_: multisync? |
20:59.27 | florian | maybe we should add these details what's to do to CurrentThingsToDo wiki... |
21:00.03 | pb_ | florian: good plan. |
21:00.24 | pb_ | PaxAnima: I plan to use kitchensync initially, but if someone wanted to have a go with multisync that would be cool as well. |
21:01.51 | PaxAnima | kitchensync is KDEs sync thing? |
21:01.57 | pb_ | Yes. |
21:04.11 | PaxAnima | hmm.. just compiled my PIM and it seg.faults on the ipaq :/ any way to debug it? |
21:04.18 | PaxAnima | cause it works on the desktop |
21:04.20 | pb_ | gdb |
21:04.33 | *** join/#gpe Jacmet (~jacmet@98.adsl0.abc.worldonline.dk) |
21:06.34 | florian | interesting, i have one application that segfaults only when crosscompiled... |
21:11.58 | PaxAnima | pb_: how do I list the fonts installed on the ipaq? |
21:12.14 | pb_ | xlsfonts for core fonts; fc-list for client side. |
21:13.51 | PaxAnima | don't have fc-list |
21:17.31 | PaxAnima | anyways, found the bug |
21:23.46 | *** join/#gpe dc___ (~dc@modem-98-50-60-62.vip.uk.com) |
21:33.55 | PaxAnima | if anyone feels like testing my old PIM, please run ipkg install http://handhelds.org/~paxanima/files/pspim_0.0b_arm.ipk |
21:36.16 | jg | pb_: jamey claimed over lunch he'd been playing with multisync, IIRC.... |
21:36.27 | pb_ | jg: ah right. |
21:52.41 | cmarqu | pb_: import-vcard? Whee! |
21:54.35 | cmarqu | pb_: They are not really meant to be invoked by users directly, right? In that case, I'd rather call them vcard-{import,export}. Common stem and all. |
21:54.50 | pb_ | cmarqu: Yeah, good point. |
21:55.40 | pb_ | cmarqu: This was really just some random development code that I packaged up so people would have something to play with. I didn't put any real effort into it. |
21:56.10 | pb_ | Syncing with Emacs would be cool. How does that work on the host side? |
21:56.21 | cmarqu | ERROR: Cannot satisfy the following dependencies for gpe-mimedir-tools: |
21:56.25 | cmarqu | libmimedir0 |
21:56.32 | pb_ | Did you ipkg update first? |
21:56.38 | pb_ | libmimedir0 should be in unstable by now. |
21:56.45 | cmarqu | Oh, I didn't. |
21:56.59 | pb_ | In fact, gpe-mimedir-tools itself should also be in unstable by now, so no need to use the full URL even. |
21:57.14 | cmarqu | Well, emacs... BBDB is the do-it-all addressbook |
21:57.29 | cmarqu | There is code to dump vcards and also to read it back in |
21:57.33 | pb_ | Yeah, I'm just reading its documentation. |
21:57.35 | cmarqu | Merging is not easy though. |
21:57.41 | pb_ | Ah. |
21:58.34 | cmarqu | I always planned to rearrange the BBDB section on the emacswiki, but didn't get to it. |
22:04.47 | cmarqu | dpkg --remove --force-all brltty |
22:04.56 | cmarqu | Removing brltty ... |
22:04.56 | cmarqu | error in control file: `Section' value not specified at /usr/sbin/install-docs line 644, <IN> line 6. |
22:04.59 | cmarqu | dpkg: error processing brltty (--remove): |
22:05.01 | cmarqu | <PROTECTED> |
22:05.04 | cmarqu | Errors were encountered while processing: |
22:05.06 | cmarqu | <PROTECTED> |
22:05.09 | cmarqu | Argh! |
22:05.16 | cmarqu | What more can I force? |
22:11.28 | cmarqu | /mnt/hda2/vcards # import-vcard < Alexander_Stielau.vcf |
22:11.32 | cmarqu | [...] |
22:11.37 | cmarqu | no such table: contacts_urn |
22:14.28 | moray | cmarqu: I'd suggest editing the preinst / postinst that's going wrong |
22:14.35 | moray | rm even, not inst |
22:15.15 | moray | cmarqu: unless it's a systemwide database that's genuinely corrupted, in which case fixing it might be sensible... |
22:16.11 | cmarqu | moray: Cool, that worked. Thanks. |
22:17.08 | florian | good night |
22:19.13 | cmarqu | florian: Night |
22:38.23 | cmarqu | ERROR: Cannot satisfy the following dependencies for libgpewidget1-ro: |
22:38.26 | cmarqu | virtual-locale-ro |
22:55.03 | cmarqu | Ah, importing works when you already have a contacts DB (I did not, for root at least). |
22:55.26 | cmarqu | The contact is in the DB now, but gpe-contacts doesn't show it. |
22:57.49 | pb_ | cmarqu: Ah, yeah, I guess I should create some more locale-base packages. |
23:00.04 | pb_ | Ah, now I see why Florian was asking about syslog. |
23:03.09 | cmarqu | Is he doing a syslog viewer? |
23:03.27 | pb_ | Seems so. |
23:03.34 | cmarqu | Ah, scap. |
23:03.50 | pb_ | Also gpe-conf-logread in CVS :-) |
23:03.57 | cmarqu | Cool. |
23:04.23 | cmarqu | I once wrote such a thing, my first GTK app ever. :) |
23:04.30 | pb_ | Yes. I'm not totally sure that this kind of thing strictly belongs in gpe-conf, but still. |
23:06.09 | cmarqu | pb_: Seems that none of the imported vcards show up (I edited the source vcards a bit even), but they get dumped with send-vcard, together with the ones I put in directly with gpe-contacts. |
23:06.17 | pb_ | cmarqu: Hm, strange. |
23:06.34 | pb_ | cmarqu: Maybe you could inspect the database by hand and see if there is anything obviously different about the imported ones? |
23:07.06 | pb_ | Perhaps the importer is forgetting some important thing that gpe-contacts cares about, though I can't immediately think what that might be. |
23:07.22 | cmarqu | With a hex editor? |
23:07.29 | cmarqu | more is not well suited... |
23:07.31 | pb_ | Try the sqlite console. |
23:07.37 | pb_ | ipkg install sqlite |
23:07.37 | cmarqu | Ah. |
23:07.40 | pb_ | sqlite ~/.gpe/contacts |
23:07.46 | pb_ | select * from contacts_urn; |
23:07.50 | pb_ | select * from contacts where urn=NNN; |
23:08.26 | pb_ | is approximately the kind of thing you need. There should be one record in contacts_urn per person, and the number you get from that table is what you use as the NNN key to retrieve data from the contacts table. |
23:09.26 | cmarqu | Uhm, case is different. |
23:09.27 | cmarqu | sqlite> select * from contacts where urn=1; |
23:09.27 | cmarqu | 1|NAME|Fischer, Katarina |
23:09.27 | cmarqu | sqlite> select * from contacts where urn=2; |
23:09.27 | cmarqu | 2|name|Alexander Stielau |
23:09.45 | cmarqu | URN 2 has some more info too. |
23:09.52 | pb_ | Ah, hm. I guess case is not meant to matter, but gpe-contacts may not be that bright. |
23:09.57 | pb_ | Let me check the code. |
23:11.01 | pb_ | Yeah, seems that this may be the problem. It's using case-sensitive comparisons in a couple of places. |
23:11.08 | cmarqu | Ah. |
23:11.15 | pb_ | You could try amending that with sqlite and see if they start showing up. |
23:11.26 | pb_ | update contacts set tag='NAME' where tag='name'; for example |
23:11.28 | cmarqu | Sure... how? |
23:11.31 | cmarqu | Ah. |
23:11.44 | pb_ | (I think, anyway. My SQL is a bit rusty.) |
23:12.15 | cmarqu | YES! |
23:12.18 | cmarqu | Works. |
23:12.21 | pb_ | Aha, cool. |
23:12.31 | pb_ | Well, it's easy to fix gpe-contacts to not care about case. I'll do that immediately. |
23:12.51 | cmarqu | Great. |
23:13.07 | *** join/#gpe PaxAnima_ (~petter@v058b.studby.ntnu.no) |
23:13.18 | cmarqu | Two cool new things in one week: a fine gpe-nmf with ALSA and a gpe-contacts with actual data :) |
23:13.44 | pb_ | :-) |
23:14.40 | cmarqu | tiny:~/.gpe$ ls -1 /mnt/hda2/vcards/ |wc -l |
23:14.40 | cmarqu | <PROTECTED> |
23:14.55 | cmarqu | That should stress it a bit :) |
23:15.10 | *** join/#gpe dc_ (~dc@modem-196-52-60-62.vip.uk.com) |
23:15.11 | pb_ | Crumbs, yes. :-) |
23:15.55 | dc_ | right |
23:16.04 | dc_ | time to write my finaly english essay :\\ |
23:16.32 | cmarqu | You will all be swallowed by gpe-contacts, resistance futile is! |
23:20.21 | *** join/#gpe dc__ (~dc@modem-112-51-60-62.vip.uk.com) |
23:20.29 | pb_ | dc_: :-/ |
23:20.59 | dc__ | pb_: quite, this satellite has been delayed so much! I can't wait till it gets installed. |
23:21.16 | pb_ | dc__: do you have a date for that yet? |
23:21.22 | dc__ | arg. |
23:21.26 | dc__ | this is very anoying...hrm |
23:21.44 | dc__ | pb_: well no, we've been waiting for them to contact us with one for quite some time. |
23:21.54 | pb_ | ah. |
23:22.02 | dc__ | pb_: and it also seems their call center is very busy most of the time. |
23:22.06 | dc__ | even more anoying is paypal. |
23:24.17 | pb_ | What's the problem with paypal? |
23:24.23 | pb_ | Are they still rejecting your dad's credit card? |
23:25.20 | dc__ | we have to fax them the creditcard and my dads licence. |
23:25.28 | pb_ | suck |
23:27.08 | *** join/#gpe dc___ (~dc@62.60.119.246) |
23:31.03 | dc___ | bah |
23:34.37 | pb_ | cmarqu: Okay, gpe-contacts2 0.10 is uploaded. It was a bit harder to fix than I expected, but have a go with it and see how you get on. |
23:39.32 | cmarqu | That fix worked, but it only shows the name of each vcard. |
23:39.42 | cmarqu | No phone, no email etc. |
23:41.16 | pb_ | Ah, hm. |
23:41.27 | pb_ | Okay, let me poke at it some more. |
23:41.52 | pigeon | What does it mean when a girl mobile messages me just to tell me she's going to have a haircut today? |
23:42.26 | cmarqu | pigeon: "Don't be shocked when you see me again honey" |
23:42.28 | cmarqu | :) |
23:42.53 | pigeon | (For two reasons) |
23:43.14 | cmarqu | pb_: Also, I get "Invalid byte sequence in conversion input" when I import a vcard with umlauts -- I don't know the encoding though (yet) |
23:43.30 | cmarqu | pigeon: reason two: you thought the same? |
23:44.19 | pb_ | cmarqu: Ah, it probably wants everything in UTF-8. I guess the importer ought to have a munge-charsets option. |
23:44.26 | pigeon | cmarqu: 1) the hair... 2) she's not my honey :\ |
23:45.07 | cmarqu | pb_: Well, that's fine for now I'd say. |
23:45.34 | cmarqu | pigeon: Wait until you see 1), then maybe 2) doesn't matter anymore. |
23:45.50 | pigeon | Hehehe |
23:47.30 | cmarqu | pigeon: a good friend wants to drag *me* to a haircut, a fancy one :/ She says that if she has to spend a lot of time in the morning, I can too. |
23:47.52 | cmarqu | Eeek. |
23:48.02 | pigeon | Heh |
23:48.08 | pigeon | Fancy haircut eh. |
23:48.17 | pb_ | cmarqu: fear |
23:48.19 | pigeon | Does that come with a fancy pants as well? |
23:48.38 | cmarqu | pigeon: How did you know that? |
23:49.10 | cmarqu | pb_: I'll make sure it happens *after* the dev weekend :) |
23:49.14 | cmarqu | If at all... |
23:49.45 | pigeon | ;) |
23:49.51 | pb_ | cmarqu: Hm, I can't reproduce that failure here. Maybe the problem you are having now is caused by something else (like the importer actually using different tags, not just different case) |
23:50.13 | cmarqu | Could be. Hmm. |
23:50.19 | pb_ | I think we did go through a period of changing all the database tags that gpe-contacts uses, and it's possible that the importer dates from another era. |
23:50.42 | pb_ | You can find the ones it understands by looking at the layout.xml file. |
23:51.32 | pb_ | Or it might be that the importer is just broken, and not writing that information out in the database at all. |
23:51.40 | pb_ | None of these possibilities would surprise me too much. :-} |
23:53.37 | pigeon | cmarqu: We have a lot of fancy pants related stuff here in the office ;) |
23:53.40 | cmarqu | The latter could be it, I cannot find more info in the DB |
23:53.51 | cmarqu | pigeon: Oh.... |
23:54.04 | pb_ | Ah, okay. It might be misunderstanding some of the vCard data that comes back from libmimedir then. |
23:54.13 | pb_ | I'll investigate tomorrow. |
23:54.43 | cmarqu | Cool. |
23:55.11 | cmarqu | Oh yes, bedtime. Night. |
23:55.30 | pb_ | Night. |