| 00:00.46 | *** join/#gpe flip (~flip@users.eiwaz.com) |
| 00:00.47 | flip | hi |
| 00:00.49 | flip | anyone here? |
| 00:00.53 | __moray | yes |
| 00:00.59 | flip | hi __moray :) |
| 00:01.10 | flip | hey listen ive had a ipaq 3670 sitting around like forever |
| 00:01.20 | flip | and i want to put linux on it.. specificaly gpe |
| 00:01.25 | flip | i thought all the cables were the same |
| 00:01.30 | flip | i know i needed the serial cable not the usb |
| 00:01.47 | flip | i just went to best buy and got a serial cable and i see now its for the 3800 series |
| 00:01.54 | flip | and wont plug into my ipaq |
| 00:02.04 | PaxAnima | hm... wonder if I should compress the dictionary as default in dictionary_save() |
| 00:02.08 | flip | anyonw know if i can just snap the little tabs off the end? |
| 00:02.10 | flip | heh |
| 00:02.18 | flip | it LOOKS like itll fit and all the pins wil line up |
| 00:02.23 | flip | if i snap the little tabs off |
| 00:02.26 | __moray | erm, pb_ might know |
| 00:02.29 | flip | no one around here has the serial cradle :( |
| 00:02.35 | pb_ | flip: yah, the 3800 has a different connector. I thought it had more pins than the 3600 version though. |
| 00:02.42 | flip | does it? |
| 00:02.46 | flip | it doesnt look like it |
| 00:02.48 | __moray | didn't they move the power onto the same cable for the 3800 or something? |
| 00:03.05 | flip | it was in a belkin bubble pack so help me god best buy better take it back |
| 00:03.05 | flip | lol |
| 00:03.06 | pb_ | flip: let me check |
| 00:03.14 | flip | crap i cant find it anywhere :( :( |
| 00:03.19 | flip | the 3600 serial wire |
| 00:03.31 | flip | thanks pb |
| 00:04.13 | pb_ | flip: yeah, the 3800 one has 22 pins, the 3600 only 12. |
| 00:04.28 | flip | ugh |
| 00:04.49 | flip | thanks |
| 00:05.05 | flip | it cant be done with the usb cradle right? |
| 00:05.23 | pb_ | flip: not unless you have a CF card as well. |
| 00:05.29 | flip | yea |
| 00:05.30 | flip | i do!! |
| 00:05.42 | flip | here what i got *got it for free* |
| 00:05.57 | flip | a ipaq 3670, usb cradle, CF sleeve and a 16mb flash card |
| 00:06.09 | pb_ | ok, in that case you might just be in luck. |
| 00:06.44 | pb_ | do you have any way to write to the CF card, other than the ipaq? |
| 00:06.52 | flip | no :( |
| 00:07.20 | flip | do i need that ? |
| 00:07.26 | pb_ | well, you can do it without. the problem is that if you install the bootldr and it then turns out you made a mistake with the CF card, you will be stuffed until you get hold of either a CF writer for a PC, or a serial cable for the ipaq. |
| 00:08.06 | flip | whata ya mean |
| 00:08.06 | pb_ | however, if you aren't using the ipaq anyway and you can live with that risk, you may as well go ahead. |
| 00:08.16 | flip | the bootloader will be on the cf card? |
| 00:08.31 | flip | im looking at all the projects |
| 00:08.42 | flip | see im an interface coder.. and i need things to look good hehe |
| 00:08.54 | flip | eQip by the qnx guys looks awesome but its pretty unuseable |
| 00:08.58 | flip | opie looks blah |
| 00:08.59 | flip | heh |
| 00:09.06 | flip | gpe uses gtk so thats cool |
| 00:09.09 | flip | and it looks like gtk2 |
| 00:09.10 | flip | heh |
| 00:09.42 | pb_ | no, you program the bootldr into the internal flash. The problem is that, once you try to program the actual Linux root image, WinCE will (obviously) be erased. If it turns out that for some reason the Linux image you flashed was bad, you will have no way to put a new one onto the card in order to try again. |
| 00:09.58 | pb_ | But the bootldr makes some effort to stop that happening, so you ought to be OK. |
| 00:10.15 | flip | ok |
| 00:10.22 | flip | wanna walk me through it so i dont blow up the ipaq? |
| 00:10.25 | pb_ | anyway, you need to assemble the following files: |
| 00:10.45 | pb_ | latest familiar 0.6.1 bootgpe image, and the corresponding md5sum file |
| 00:10.58 | pb_ | reflash.ctl from handhelds.org/~pb/ |
| 00:11.16 | pb_ | bootldr-2.19.55.bin from handhelds.org/~pb/3600 |
| 00:11.43 | pb_ | latest bootblaster |
| 00:12.06 | pb_ | I guess that's your lot. |
| 00:12.11 | flip | im lookin for familiar now |
| 00:12.23 | flip | theres n osimple login space |
| 00:12.24 | flip | heh |
| 00:12.30 | flip | err download area i mean |
| 00:12.49 | pb_ | http://handhelds.org/familiar/releases/v0.6.1/install/bootgpe-v061-01.jffs2 |
| 00:12.58 | pb_ | http://handhelds.org/familiar/releases/v0.6.1/install/bootgpe-v061-01.jffs2.md5sum |
| 00:13.32 | flip | thanks |
| 00:13.37 | flip | i didnt know which one to get haha |
| 00:13.52 | *** join/#gpe TheMasterMind1 (foobar@h-69-3-0-125.MCLNVA23.covad.net) |
| 00:14.08 | pb_ | and bootblaster is at http://handhelds.org/familiar/releases/v0.6.1/install/BootBlaster_1.18.exe |
| 00:15.45 | flip | do i need familiar's md5sum for anything except to check the jiff? |
| 00:15.48 | flip | or is it required |
| 00:15.56 | florian | gn |
| 00:16.16 | flip | ok sooo |
| 00:16.17 | flip | pb_: |
| 00:16.18 | pb_ | florian: bootldr will need it |
| 00:16.21 | flip | now what hehe |
| 00:16.26 | flip | i got all the stuff |
| 00:16.27 | pb_ | florian: uh, night, ignore what I just said :-) |
| 00:16.38 | __moray | flip: it's a good idea to check the md5sums, given that corruption will be problematic |
| 00:16.53 | __moray | pb_: hm, irssi also refused to tab-complete flip for me at that point |
| 00:16.53 | pb_ | yeah, bootldr will check the md5sum before it allows you to flash an image. |
| 00:16.56 | flip | im doing it now __moray heh |
| 00:17.16 | flip | brb need a drink |
| 00:17.24 | pb_ | flip: copy bootblaster and bootldr onto the ipaq, then follow the instructions for installing the bootldr |
| 00:17.29 | pb_ | http://handhelds.org/familiar/releases/v0.6.1/install/install-bootldr.html |
| 00:17.56 | pb_ | flip: also, copy both bootgpe-* files and reflash.ctl onto the CF card. |
| 00:18.00 | florian | pb_: :-) |
| 00:18.18 | pb_ | __moray: ah, you're using irssi as well? |
| 00:18.43 | flip | ok hold on |
| 00:18.48 | __moray | pb_: yup, apart from anything else because this machine's connection keeps on dying due to wireless trouble |
| 00:18.59 | pb_ | heh. |
| 00:19.38 | __moray | I might swap an Orinoco into it, but I need to make sure I have appropriate drivers for the prism2 on another machine first |
| 00:20.38 | flip | crap i need activesync |
| 00:20.39 | flip | haha |
| 00:20.42 | flip | lemmie go d/l that |
| 00:20.49 | pb_ | heh. |
| 00:22.23 | __moray | (hm, when did this 'prismstumbler' dependent on GPE appear?) |
| 00:22.37 | pb_ | __moray: no idea, never even heard of it. |
| 00:23.14 | pb_ | __moray: ah, it would appear to be a piece of florianware. |
| 00:23.39 | __moray | indeed |
| 00:23.56 | pb_ | but, regrettably, looks like it's gtk1.2 only. |
| 00:24.23 | flip | gpe's gtk2 |
| 00:24.24 | flip | yes? |
| 00:24.42 | pb_ | flip: the one you are installing is actually gtk1, but you can upgrade to the gtk2 version afterwards. |
| 00:24.48 | flip | good |
| 00:24.51 | pb_ | flip: there are no gtk2-based install images yet. |
| 00:26.17 | flip | k moving over to the CF |
| 00:26.23 | pb_ | ok. |
| 00:27.08 | __moray | pb_: it would probably be good if we had GTK2 install images soon actually, since it's not completely straightforward to be sure you've done the upgrade cleanly (and removed all the old stuff) |
| 00:27.15 | pb_ | __moray: indeed |
| 00:27.36 | pb_ | __moray: unfortunately ipkg is completely fux0red on my ARM box right now so I can't build any images. I guess I should try to fix that. |
| 00:27.45 | flip | arm box? |
| 00:27.49 | flip | you mean your ipaq haha |
| 00:27.56 | flip | or do you have a netwinder or something |
| 00:28.03 | pb_ | flip: it's a CATS actually, but yeah. |
| 00:28.10 | flip | cats? |
| 00:28.12 | flip | whats that |
| 00:28.20 | pb_ | flip: ATX motherboard with an SA-110 processor |
| 00:28.29 | flip | oh |
| 00:28.36 | flip | ok! so |
| 00:28.46 | flip | moved is done |
| 00:29.01 | pb_ | flip: ok. so, proceed with installing the bootldr. |
| 00:29.41 | pb_ | __moray: or you could try to prevail on nelson to build some new images, either from unstable or by starting the 0.7 release process. |
| 00:30.19 | flip | whats the "default folder" |
| 00:30.31 | pb_ | flip: "my documents" or something like that, I think. |
| 00:30.40 | flip | my device? |
| 00:30.51 | pb_ | could be. |
| 00:32.16 | pb_ | ask in #handhelds.org if you can't work it out. it's been a while since I did this, I don't remember the details. |
| 00:32.21 | __moray | pb_: hm, so with 0.7 do we get to stealthily switch to GTK2? |
| 00:32.27 | pb_ | __moray: that's the plan. |
| 00:32.33 | __moray | ah good ;) |
| 00:32.44 | pb_ | __moray: though less with the stealth, in fact. |
| 00:32.54 | pb_ | __moray: Nils seems to be more accepting of gtk2 these days. |
| 00:32.59 | __moray | pb_: right. |
| 00:33.16 | pb_ | __moray: and none of the GTK1 applications are maintained any more as far as I can tell. |
| 00:33.23 | __moray | pb_: you may have noticed there's only a gpe-julia2 package... |
| 00:33.35 | __moray | though the '2' seems a bit silly in that context of course |
| 00:34.00 | pb_ | __moray: yah, I was just about to say that. I've been tending to only use the "2" when there is an existing gtk1 package that it would otherwise collide with. |
| 00:34.41 | __moray | well, I suppose I was leaving the space in case anyone felt there should be a GTK1 release still |
| 00:34.48 | pb_ | fair enough. |
| 00:35.03 | flip | backing up windows now. |
| 00:35.13 | pb_ | flip: right |
| 00:35.33 | __moray | pb_: though perhaps we could move everything 'back' to non-2 versions, to make sure the GTK1.2 users upgrade to the new versions |
| 00:35.46 | pb_ | __moray: mightn't be a bad idea, yeah. |
| 00:36.34 | __moray | I suppose it might eventually cause confusion when a genuine version 2 of something comes out, but with the separate version numbers it shouldn't actually cause anything to go wrong |
| 00:36.59 | pb_ | __moray: yeah, we already had that confusion in reverse with gdk-pixbuf, which was at version 2 with gtk 1. |
| 00:37.16 | __moray | moving to non-2 versions would also give a good excuse for clearing GTK1 stuff out of CVS |
| 00:37.53 | __moray | (otherwise it's likely to hang around for ever...) |
| 00:37.59 | pb_ | yes. |
| 00:40.02 | flip | ok |
| 00:40.05 | flip | reboot the ipaq |
| 00:40.11 | flip | that means the little switch on the bottom? |
| 00:40.21 | pb_ | flip: use the tip of the stylus to push the reset button. |
| 00:40.35 | pb_ | flip: it's in the round hole on the bottom, on the opposite side to the main power switch. |
| 00:40.44 | flip | ok |
| 00:40.57 | flip | im at the ipaq screen it wants me to tap on the screen |
| 00:41.01 | flip | so we did good right? |
| 00:41.01 | flip | heh |
| 00:41.04 | flip | its not nuked |
| 00:41.10 | pb_ | so far so good :-) |
| 00:41.16 | pb_ | are all your files on the CF card? |
| 00:41.18 | flip | ok what next |
| 00:41.23 | flip | lemmie get into wince and check |
| 00:42.12 | flip | ok |
| 00:42.18 | flip | there all in there |
| 00:42.24 | pb_ | ok |
| 00:42.29 | flip | its on the cradle |
| 00:42.44 | pb_ | reset the ipaq again, but this time hold down the middle of the joypad/speaker while you do so. |
| 00:43.11 | pb_ | you should get a white screen with a penguin. |
| 00:43.44 | flip | yep |
| 00:43.46 | flip | arm bootloader |
| 00:43.49 | pb_ | cool |
| 00:43.57 | pb_ | now press the record button. |
| 00:44.10 | flip | scanning |
| 00:44.23 | flip | cant read control file |
| 00:44.29 | pb_ | ah. |
| 00:44.36 | flip | its over? |
| 00:44.38 | flip | heh |
| 00:44.40 | pb_ | you will have to go back into wince and find out what's gone wrong. |
| 00:44.59 | pb_ | check that reflash.ctl has the right name, in particular. |
| 00:45.01 | flip | ok no biggie hold up |
| 00:45.34 | flip | md5 is gone! |
| 00:45.36 | flip | i gotta move that over |
| 00:45.38 | flip | maybe thats the deal |
| 00:46.06 | pb_ | you'll certainly need that later, but I don't think that will be the cause of the error you had. |
| 00:46.29 | flip | well reflash.ctl is in there indeed |
| 00:46.36 | flip | in the CD card |
| 00:46.39 | pb_ | hrm |
| 00:46.44 | flip | cf i mean |
| 00:46.51 | flip | let me delete it and move it over again |
| 00:46.54 | pb_ | ok |
| 00:46.56 | flip | maybe it was a transfer problem |
| 00:47.00 | pb_ | could be. |
| 00:47.01 | flip | i dont trust activesync |
| 00:47.03 | flip | ehehe |
| 00:47.51 | pb_ | oh, can you find out what format the CF card is in? |
| 00:48.00 | pb_ | I think bootldr is maybe a bit fussy about what type of fat partitions it will read. |
| 00:48.07 | pb_ | I have no idea how you tell from inside CE though. |
| 00:50.42 | PaxAnima | pb_: gpe-nmf complains about not finding libpopt.so? |
| 00:50.43 | flip | hrmm |
| 00:50.45 | flip | neither do i |
| 00:50.53 | flip | what if i move the ctl to the root partition if you will |
| 00:50.54 | flip | heh |
| 00:50.56 | flip | will that work? |
| 00:51.09 | pb_ | flip: no, it has to be on the cf card. |
| 00:51.46 | pb_ | flip: it's remotely possible it may work better if named in capitals, REFLASH.CTL |
| 00:52.07 | pb_ | flip: oh, and obviously it needs to be in the top level of the card, not in any subdirectory. |
| 00:52.19 | flip | well duh |
| 00:52.23 | flip | lemmie try caps |
| 00:52.24 | pb_ | PaxAnima: yeah, dependency snafu, gstreamer needs it. |
| 00:52.33 | pb_ | PaxAnima: install it by hand for now. |
| 00:53.06 | PaxAnima | yup |
| 00:53.38 | flip | didnt work pb |
| 00:53.45 | pb_ | flip: oh dear. |
| 00:53.47 | pb_ | flip: same problem? |
| 00:54.02 | flip | mmhmm |
| 00:54.09 | pb_ | flip: do you have anything else on the CF card? |
| 00:54.16 | flip | no but NOW |
| 00:54.23 | flip | ce isnt seeing the CF card |
| 00:54.26 | flip | how do i format it? |
| 00:54.38 | pb_ | dunno |
| 00:54.58 | __moray | flip: we're probably not the best people to ask about CE ;) |
| 00:55.04 | flip | neither am i |
| 00:55.07 | flip | hah |
| 00:58.14 | flip | eck |
| 00:58.15 | flip | wtf |
| 00:58.15 | flip | heh |
| 00:58.19 | flip | this has gotta work |
| 01:01.14 | flip | know of a pocketpc channel anywhere? |
| 01:01.15 | flip | heh |
| 01:01.23 | flip | like ppl that actually give a flying shit about this thing |
| 01:01.37 | flip | i just wanna format my card :( |
| 01:02.11 | pb_ | try asking in #handhelds.org, maybe one of the dudes in that channel will know. |
| 01:02.42 | PaxAnima | yay |
| 01:02.54 | PaxAnima | I can now save my dictionaries |
| 01:03.12 | PaxAnima | (that was the easy task on in the TODO) |
| 01:04.08 | pb_ | cool |
| 01:04.19 | PaxAnima | it saves to a compressed format |
| 01:09.20 | flip | shit |
| 01:09.22 | flip | i wanna fix this |
| 01:11.08 | flip | any ideas guys? |
| 01:13.16 | flip | pb_: ? |
| 01:13.58 | pb_ | flip: sorry, not really |
| 01:14.12 | flip | ah |
| 01:19.51 | *** join/#gpe __moray (moray@217.155.67.241) |
| 01:22.25 | spung | cmarqu: ping |
| 01:23.26 | pb_ | __moray: btw, you might like to add a gpe_set_window_icon call to gpe-julia so you get a pretty icon in the titlebar. |
| 01:23.56 | pb_ | __moray: (compare the gpe-julia shot to the others on http://gpe.handhelds.org/projects/GPE-games.shtml) |
| 01:33.00 | *** join/#gpe moray__ (moray@dsl-217-155-67-241.zen.co.uk) |
| 01:33.07 | moray__ | hm |
| 01:33.12 | PaxAnima | there... added mccarthys idea of adding words to the common dictionary as the user types for the irc input ;) |
| 01:33.36 | __moray | network trouble :( |
| 01:34.29 | *** join/#gpe __moray_ (moray@dsl-217-155-67-241.zen.co.uk) |
| 01:34.50 | pb_ | :-/ |
| 01:35.43 | pb_ | __moray: did you see my remarks about gpe-julia and window icons a moment ago? |
| 01:35.48 | __moray | nope |
| 01:36.00 | pb_ | < pb_> __moray: btw, you might like to add a gpe_set_window_icon call to gpe-julia so you get a pretty icon in the titlebar. |
| 01:36.02 | pb_ | < pb_> __moray: (compare the gpe-julia shot to the others on http://gpe.handhelds.org/projects/GPE-games.shtml) |
| 01:36.04 | pb_ | - |
| 01:36.06 | pb_ | there they are again :-) |
| 01:36.10 | __moray | ah yes, thanks |
| 01:36.20 | __moray | I did think of that at one point, but obviously forgot again :) |
| 01:37.14 | pb_ | ha. :-) |
| 01:38.33 | pb_ | __moray: it might also be neat if it zoomed in continuously in some kind of smooth fashion while you held the stylus down. |
| 01:40.09 | __moray | pb_: yes, though I think that's more of a wishlist bug... - it would also be nice if motion events didn't get queued up (try a long drag in Julia mode), and ideally if more precision was used |
| 01:40.45 | pb_ | __moray: you can stop the event queueing thing by using PointerMotionHint, take a look at gpe-gallery's pan-drag code for example. |
| 01:41.11 | PaxAnima | gpe-gallery crashed when I clicked the rotate button :/ |
| 01:41.15 | pb_ | __moray: but agreed, definitely a wishlist item rather than a real bug. |
| 01:42.35 | PaxAnima | pb_: how did the rotation get implemented in gpe-gallery? ie. the user interface? |
| 01:42.49 | pb_ | PaxAnima: I don't remember exactly what we ended up with. |
| 01:43.24 | __moray | pb_: aha, I'll have a look at gpe-gallery then for PointerMotionHint - during my searching the only things I found were people on GTK/GNOME lists suggesting to each other to speed up redrawing, which is an unhelpful (and naive) response |
| 01:43.57 | pb_ | PaxAnima: the original idea was that you would just invoke rotate mode and then use the stylus to spin the image around the centre of the screen, but the rotation code turned out to be too slow for that to be feasible for anything other than small images. |
| 01:44.43 | PaxAnima | pb_: ok |
| 01:45.57 | *** part/#gpe __moray_ (moray@dsl-217-155-67-241.zen.co.uk) |
| 01:46.10 | pigeon | Hmm so, it boots an image via usb? |
| 01:46.28 | pb_ | pigeon: bootldr? that's what NonToxic is working on, but it doesn't work yet. |
| 01:46.38 | pigeon | :) |
| 01:46.49 | pigeon | That was a wrong channel. But anyway :) |
| 01:46.56 | pigeon | I see. |
| 01:47.56 | *** join/#gpe hunterdie (~kb3hag@pcp02465537pcs.chrchv01.md.comcast.net) |
| 02:16.11 | PaxAnima | damn, my ipaq is acting up |
| 02:16.21 | __moray | PaxAnima: how's that? |
| 02:17.48 | PaxAnima | it's gone all crazy... logged in, and suddenly mb died, and the kernel went mad with "Out of memory", and killed all processes.. |
| 02:18.38 | PaxAnima | got into a loop of starting/killingX, and suddenly the ipaq died, and needed a reboot to get started :/ |
| 02:18.45 | __moray | ugh |
| 02:18.54 | __moray | is it working after the reboot? |
| 02:19.19 | PaxAnima | well, I get gpe-login, but I don't dare log in |
| 02:20.12 | PaxAnima | this happened after an upgrade |
| 02:20.14 | __moray | well, it's perhaps worth seeing if the fault is repeatable (after copying off any important data ;) |
| 02:21.09 | PaxAnima | seems to be acting normal now |
| 02:21.27 | PaxAnima | but now I can't suspend the ipaq |
| 02:22.13 | PaxAnima | what was the story of the suspend button? it was controlled by a userland app? |
| 02:24.39 | PaxAnima | I'd like to suspend the ipaq before I go to sleep... |
| 02:24.57 | __moray | PaxAnima: yes - apparently it's better for it to be done in userland, since that gives more flexibility (and compatiblity with Opie, etc.) - the default keylaunch settings, e.g., should work |
| 02:25.26 | PaxAnima | hmm.. it executes /sbin/pm_helper, but seems to fail, somehow |
| 02:25.32 | PaxAnima | pm_suggest_suspend:270 got pid=832 |
| 02:25.33 | PaxAnima | run_sbin_pm_helper:74 pm_helper_path=/sbin/pm_helper |
| 02:25.33 | PaxAnima | : waitpid returned -10 (exit_code=0) |
| 02:25.51 | __moray | hm, someone was saying there was a pm_helper problem just now :/ |
| 02:26.17 | *** join/#gpe prpplague (~joebob@21-190.lctv-b4.cablelynx.com) |
| 02:26.37 | __moray | PaxAnima: doing 'su', 'apm -s' would presumably work for tonight ;) |
| 02:26.49 | PaxAnima | yes, would seem so |
| 02:26.56 | PaxAnima | but I'd like to get it fixed too... ;) |
| 02:31.20 | PaxAnima | hm... putting apm -s in a suspend-script was a baaaaaad idea... |
| 02:31.37 | __moray | erk |
| 02:31.46 | prpplague | lol |
| 02:32.26 | PaxAnima | it appears that S99Suspend doesn't work... it tries to write to some /proc/sys/pm/suspend, but that doesnt exist |
| 02:32.46 | __moray | could again be related to the current broken kernel? :/ |
| 02:32.59 | PaxAnima | hmm... is the kernel broken? |
| 02:33.13 | __moray | well, it's been randomly(?) crashing for me |
| 02:33.59 | PaxAnima | but why would apm -s work and not the suspend button? |
| 02:36.22 | PaxAnima | how silly... it worked when _removing_ S99Suspend from the suspend scripts... |
| 02:37.01 | __moray | PaxAnima: on my ipaq, "apm -s" as non-root *doesn't* usefully work just now, only as root |
| 02:37.11 | __moray | as non-root it redisplays the login screen, but nothing else |
| 02:37.38 | PaxAnima | I guess that the script returned a bad exit code, so the suspend was interupted |
| 02:37.55 | PaxAnima | think jamey or someone else mentioned it a long time ago on #hh.org |
| 02:38.06 | PaxAnima | but now I know... |
| 02:38.41 | PaxAnima | ... or not |
| 02:38.58 | PaxAnima | it doesn't work after I've logged in... |
| 02:41.51 | PaxAnima | oh well... I can suspend it in the login screen... I'll ask pb tomorrow... |
| 04:13.43 | *** join/#gpe walters (walters@verbum.org) |
| 07:48.37 | walters | is there a gpe metapackage? |
| 07:49.19 | walters | ah, task-gpe |
| 09:42.49 | *** join/#gpe mallum (~mallum@pc-80-193-218-21-hw.blueyonder.co.uk) |
| 09:43.29 | mallum | hey |
| 09:45.14 | pigeon | Hello mallum. |
| 10:02.06 | mallum | hey pigeon |
| 10:39.49 | *** join/#gpe pb_ (~pb_@dsl-62-3-66-201.zen.co.uk) |
| 10:42.30 | pigeon | Yo pb_ |
| 10:42.40 | mallum | hey pb_ |
| 10:42.48 | pb_ | hi pigeon |
| 10:42.49 | pb_ | hi mallum |
| 10:54.47 | mallum | pb_: are there still plans for fosdem ? |
| 11:03.18 | pb_ | mallum: I think so, but I haven't made any really concrete plans yet. |
| 11:11.29 | mallum | pb_: I remember seeing an offer in the sun for cheap ferry tickets |
| 11:11.44 | mallum | pb_: usually though, they are only for a day return though :( |
| 11:16.33 | pb_ | Yeah. A return ticket through the tunnel is something like 100 quid, which isn't too bad if you split it between three people. That's usually faster than the ferry, of course. |
| 11:17.43 | mallum | pb_: yeah that sounds good. |
| 13:07.40 | PaxAnima | hi |
| 13:12.19 | mallum | hi Pa |
| 13:14.53 | PaxAnima | todays mission: expand my dictionary system with statistics and compression (online, that is)... |
| 13:16.53 | PaxAnima | maybe do higher level predictive dictionary system, using a statistical dictionary and a normal one, like I do with the irc input |
| 13:21.02 | PaxAnima | pb_: suspend doesn't worked when I log in... in the login window it works |
| 13:21.11 | PaxAnima | -ed |
| 13:22.41 | PaxAnima | pb_: some error on the serial console about "waitpid returned -10", or something like that |
| 13:23.35 | *** join/#gpe prpplague (~JoeBob1@12.148.134.9) |
| 13:32.07 | PaxAnima | morning, prpplague |
| 13:32.14 | prpplague | morning |
| 13:32.22 | prpplague | PaxAnima: any good beer lately? |
| 13:33.26 | PaxAnima | prpplague: no :/ |
| 13:33.48 | prpplague | me either :( |
| 13:37.38 | pb_ | PaxAnima: what kernel are you using? |
| 13:39.24 | PaxAnima | pb_: hmm... not sure |
| 13:39.35 | pb_ | you probably want to use the latest from 2.4.19 feed. |
| 13:40.08 | PaxAnima | damn ipaq-sleep... the default values are not good |
| 13:41.13 | PaxAnima | pb_: 2.4.18-rmk3 |
| 13:42.54 | pb_ | ok, suggest you upgrade. |
| 13:43.00 | PaxAnima | just add the 2.4.19 feed to ipkg.conf? |
| 13:43.17 | *** join/#gpe mccarthy (~mccarthy@granular.che.pitt.edu) |
| 13:43.19 | pb_ | you probably need to install the packages explicitly, the naming has changed. |
| 13:43.25 | mccarthy | pb_: did you get to look at that code? |
| 13:43.25 | pb_ | they're now kernel-sa, etc. |
| 13:43.28 | pb_ | mccarthy: not yet |
| 13:43.36 | mccarthy | :( |
| 13:43.50 | pb_ | mccarthy: sorry. I'll hopefully get a chance later this afternoon. |
| 13:47.22 | PaxAnima | pb_: kernel-sa is the correct one for 3850? |
| 13:47.36 | pb_ | yeah |
| 13:47.41 | pb_ | kernel-pxa is only for 39xx. |
| 13:47.57 | PaxAnima | wht does the kerel have pxa2 in its name? |
| 13:48.32 | PaxAnima | do I have to install all the kernel modules? |
| 13:48.47 | mccarthy | pb_: what version of gtk2 does lingpewidget now depend on? |
| 13:49.18 | pb_ | mccarthy: 2.2.0 right now. |
| 13:49.30 | pb_ | 2.1.5 might work as well, I'm not sure. |
| 13:49.48 | mccarthy | pb_: I *did* need that -lXsettings to compile gpe-calendar.... should I update the CVS Makefile? |
| 13:50.11 | pb_ | mccarthy: no, it shouldn't be required. |
| 13:50.35 | pb_ | mccarthy: libXsettings-client has (should have) libXsettings as DT_NEEDED, so you ought not to need to mention it explicitly. |
| 13:50.59 | pb_ | mccarthy: I'll have a look at it and see if I can figure out what's up. |
| 13:52.10 | mccarthy | pb_: shouldnt we change the dependency of gpe-calendar to libsqlite (instead of sqlite)> |
| 13:52.14 | mccarthy | >=? |
| 13:52.44 | PaxAnima | shouldn't ipkg -recursive remove kernel-modules give me some output? |
| 13:53.35 | pb_ | mccarthy: yes. |
| 13:53.40 | pb_ | PaxAnima: -recursive doesn't work. |
| 13:53.56 | pb_ | mccarthy: presumably libsqlite0 in fact. |
| 13:54.18 | mccarthy | gpe-calendar2 does it correctly.... gpe-calendar does not |
| 13:54.22 | PaxAnima | pb_: oh. does -force-removal-of-dependent-packages work? |
| 13:55.08 | *** join/#gpe mibus (~mibus@164.a.001.ade.iprimus.net.au) |
| 13:55.22 | mibus | hi ppls |
| 13:55.29 | __moray | hi mibus |
| 13:55.46 | mibus | hows things? |
| 13:56.05 | __moray | fine here I think, though as ever too little time to do the things I want |
| 13:56.17 | mibus | i know the feeling :D |
| 13:56.28 | pb_ | mibus: g'day |
| 13:56.28 | mibus | i have the 'excuse' of watching far too much TV tho :( |
| 13:56.29 | mibus | ;) |
| 13:56.43 | mibus | pb_: well night technically. but a good one it is! :P |
| 13:56.48 | pb_ | PaxAnima: no. |
| 13:56.49 | mccarthy | pb_: oops....Luis messed up the gpe-calendar control file (big time)...he uploaded a messsed up version of the "wallpaper" one to gpe-calendar! |
| 13:57.00 | PaxAnima | pb_: hmm.. how do I upgrade all my kernel modules then? |
| 13:57.13 | pb_ | mccarthy: oh, I see. I don't really care too much about the gtk1 packages any more to be honest, but feel free to update the dependencies if you want. |
| 13:57.19 | pb_ | PaxAnima: you will have to remove them individually. |
| 13:57.36 | pb_ | PaxAnima: or complain to TheMasterMind1 and get him to fix ipkg. |
| 13:58.06 | mibus | anyone noticed probs with unstable being uberunstable? dillo wont work (pops up a window then disappears), weird behavior when i strace too |
| 13:58.21 | mccarthy | pb_: why do I always get errors with the po files?! (I dont even really know their purpose) |
| 13:58.31 | mibus | mccarthy: they are for translations |
| 13:58.33 | pb_ | mccarthy: .po files are for translations. |
| 13:58.34 | __moray | mccarthy: I think the build process edits them |
| 13:58.46 | __moray | mccarthy: so therefore a new cvs update causes conflicts |
| 13:58.58 | mccarthy | how do I work around it? |
| 13:59.03 | __moray | (or that happens with some of the translation files anyway) |
| 13:59.10 | mibus | mccarthy: 'make clean' first? |
| 13:59.17 | PaxAnima | there... some xargs magic |
| 13:59.20 | __moray | mccarthy: delete the files causing errors and cvs update, if that doesn't help |
| 13:59.22 | mibus | mccarthy: 'rm -f `find -name \*.o`? |
| 14:00.27 | mibus | pb_: i havent had a chance to check out the new nmf yet... are there UI options for loop & shffle? |
| 14:03.15 | PaxAnima | pb_: 2.4.19 don't have mmc support! |
| 14:08.22 | pb_ | PaxAnima: it should do, there is MMC support in the source tree. |
| 14:08.31 | pb_ | mccarthy: what is the actual problem you are having with the .po files? |
| 14:08.55 | PaxAnima | pb_: oh, it seems to be compiled into the kernel |
| 14:09.25 | mccarthy | pb_: hmmm. I didnt actually look before, I just assumed it was the same one I used to have.... |
| 14:09.35 | mccarthy | pb_: but it seems like it is different.... |
| 14:09.42 | mccarthy | /bin/sh: line 1: recode: command not found |
| 14:09.51 | pb_ | mccarthy: it would appear you need to install recode. |
| 14:10.16 | mccarthy | pb_: I guess so....so I dont know what problem I *used* to have :) |
| 14:10.22 | mccarthy | (because it wasnt that one) |
| 14:11.02 | pb_ | __moray: yeah, the problem with conflicts in .po files is also a nuisance. I think "make all" probably ought not to imply update-po, and you should have to do the latter manually if you want it. |
| 14:13.55 | PaxAnima | pb_: suspend still doesn't work when logged in |
| 14:14.34 | mccarthy | PaxAnima: I am talking with pb about that right now on #hh |
| 14:14.53 | pb_ | PaxAnima: ok, I will look at it later on. |
| 14:15.31 | __moray | pb_: yes, that might be reasonable (I suppose the alternative might be to take them out of CVS, and force the relevant files to be built each time) |
| 14:15.35 | mccarthy | PaxAnima: what kernel are *you* running? |
| 14:15.51 | pb_ | __moray: the .po files aren't generated, they are actually the source for the translations. |
| 14:15.54 | PaxAnima | mccarthy: well, tried 2.4.19 |
| 14:16.23 | mccarthy | PaxAnima: oh |
| 14:16.32 | __moray | pb_: hm, I get confused about which way round that stuff works - which is their reason for causing problems then? |
| 14:16.45 | PaxAnima | mccarthy: but it's the same with 2.4.18 |
| 14:18.06 | pb_ | __moray: make update-po extracts the current strings from the source file and merges them into the .po files. Those changes tend to get checked in along with everything else, which is fine on its face but leads to conflicts when someone else has already updated the files locally and tries to check the changed ones out of cvs. |
| 14:18.31 | mccarthy | PaxAnima: I think it was a relatively recent update that caused the problem though...maybe hh20-ish should work... |
| 14:18.53 | __moray | pb_: ah yes, thanks |
| 14:19.02 | PaxAnima | hmmm.. seems that apm -s won't work as non-root user |
| 14:19.47 | pb_ | yes, that seems to be the situation. |
| 14:20.01 | PaxAnima | gives me the same errors on the serial console: waitpid returned -10 (exit_code=0) |
| 14:22.09 | __moray | in full, I get: |
| 14:22.11 | __moray | run_sbin_pm_helper:76 pm_helper_path=/sbin/pm_helper pm_suggest_suspend:291 got pid=628 <2>pm_suggest_suspend:291 got pid=628 : waitpid returned -10 (exit_code=0) |
| 14:22.38 | PaxAnima | __moray: yes, same as here |
| 14:23.39 | PaxAnima | as opposed to: (flood warning) |
| 14:23.40 | PaxAnima | run_sbin_pm_helper:76 pm_helper_path=/sbin/pm_helper |
| 14:23.40 | PaxAnima | pm_suggest_suspend:291 got pid=705 |
| 14:23.40 | PaxAnima | sa1100_pcmcia_suspend(): suspending socket 0 |
| 14:23.40 | PaxAnima | : waitpid returned 705 (exit_code=0) |
| 14:23.43 | PaxAnima | pm_suggest_suspend: REALLY SUSPENDING NOW |
| 14:23.44 | mccarthy | pb_: do you have a kernel tree that you can "grep -ri waitpid *" for me so that I can see which file may be the problem? |
| 14:23.45 | PaxAnima | rtc_pm_callback: pm callback 0 RTSR=00000000 ICPR=00000000 |
| 14:26.15 | PaxAnima | hmm... waitpid() can't even return -10, according to the man file.. |
| 14:28.10 | mccarthy | PaxAnima: aha. I am getting closer.... |
| 14:29.36 | mccarthy | Pax moray: I am sure that hh20 would work fine.... |
| 14:30.55 | mccarthy | pb_: it is in the pm.c file in the linux/kernel/arch/arm/mach-sa1100/ directory |
| 14:31.26 | PaxAnima | mccarthy: what's the bug? |
| 14:31.28 | mccarthy | from hh20 to hh22+ the handling of usermode suspends changed (incorrectly I gather) |
| 14:31.38 | mccarthy | PaxAnima: I didnt get that far yet... |
| 14:31.44 | PaxAnima | oh |
| 14:33.50 | pb_ | mccarthy: indeed. |
| 14:34.20 | pb_ | mccarthy: the really strange thing is that it works when run as root. Given that apm is setuid root, it shouldn't really make any difference. |
| 14:35.27 | mccarthy | pb_: but the problem is *not* in apm, it is that pm_helper is *not* run as root.... |
| 14:36.51 | mccarthy | hmmm |
| 14:38.30 | mccarthy | pb_: should apm be sending "pm_suggest_suspend_hook" instead of "pm_use_sbin_pm_helper" |
| 14:40.07 | mccarthy | pb_: damn, I know exactly what snippet of code is causing the problems, but I have no idea how it *really* works, so I cannot tell *why* it causes problems... |
| 14:40.13 | PaxAnima | seems to be introduced by a patch by sandman, according to webcvs |
| 14:42.53 | PaxAnima | http://cvs.handhelds.org/cgi-bin/viewcvs.cgi/linux/kernel/arch/arm/mach-sa1100/pm.c.diff?r1=1.15&r2=1.15.2.1 |
| 14:43.39 | mccarthy | PaxAnima: I did 1.15.2.1, but it is around line 240 either way |
| 14:43.49 | mccarthy | I mean 1.15.2.2 |
| 14:44.23 | PaxAnima | yes |
| 14:45.09 | pb_ | mccarthy: right, but the behaviour is reportedly different depending on whether you run apm as root or as a normal user. |
| 14:46.08 | PaxAnima | hmm.. seems like the setuid bit doesn't work, somehow |
| 14:47.24 | PaxAnima | added a dummy script to /etc/suspend that wrote do a file on /, but it didn't show up when running apm -s as a normal user |
| 14:48.11 | PaxAnima | but did when running as root |
| 15:51.52 | *** join/#gpe jsmanrique (jsmanrique@cmr-213-141-51-252.telecable.es) |
| 15:51.55 | jsmanrique | hi |
| 15:57.10 | PaxAnima | hi |
| 16:10.22 | jsmanrique | does any of you know about xfce 4? |
| 16:12.38 | jsmanrique | http://members.home.nl/jbhuijsmans/xfce4-cvs.html |
| 16:12.59 | jsmanrique | I think they have nice ideas using gtk 2.0 for a Window Manager |
| 16:13.10 | jsmanrique | mallum: yours rock |
| 16:15.11 | mallum | jsmanrique: how dare you suggest use of a wm other than matchbox ! |
| 16:15.15 | mallum | jsmanrique: ;-) |
| 16:15.17 | mallum | muahaha |
| 16:15.31 | PaxAnima | the evil mallum awakes! ;) |
| 16:15.32 | jsmanrique | X-D |
| 16:15.42 | jsmanrique | I never suggested that |
| 16:15.57 | jsmanrique | I ask you to see some nice features for your great WM |
| 16:16.42 | jsmanrique | the only one WM |
| 16:17.30 | mallum | jsmanrique: im only joking around ;-) |
| 16:17.37 | PaxAnima | hmm... time to code something usefull, ie. expand the dictionary system |
| 16:17.56 | jsmanrique | it is a 'small' wm |
| 16:18.09 | jsmanrique | well, I told them to look into mb :-) |
| 16:20.45 | mccarthy | pb_: what would be the best way to check for a memory leak? |
| 16:21.01 | pb_ | mccarthy: see the glibc manual |
| 16:21.36 | mccarthy | pb_: couldnt I just check the RSS on top? |
| 16:22.42 | pb_ | mccarthy: yeah, but that may not be completely reliable. |
| 16:22.49 | pb_ | mccarthy: slow leaks will be hard to spot that way. |
| 16:23.15 | pb_ | obviously if you are leaking megabytes and megabytes it should be fairly conspicuous though, |
| 16:26.31 | mccarthy | pb_: i am trying to free the cloned events every time I remake them, but it is seg faulting... :( |
| 16:26.40 | pb_ | mccarthy: ah. |
| 16:27.31 | mccarthy | pb_: is there a better way than to make a new glist of "cloned_events" and then before making new ones iter through the old one and event_destroy (g_free) the old evs? |
| 16:27.56 | pb_ | mccarthy: nope, probably not. |
| 16:28.17 | pb_ | mccarthy: obviously you need to make sure nothing is holding a reference to the cloned events you are about to destroy. |
| 16:29.23 | mccarthy | pb_: does destroying events *within* the iter loop cause problems? |
| 16:29.55 | pb_ | mccarthy: shouldn't do. |
| 16:30.35 | mccarthy | pb_: do I need to check that cloned_events is non-null or will the iter loop simply not go if it is null? |
| 16:31.36 | mccarthy | pb_: and is it better to erase item-by-item in the cloned_events list of simply zap the whole thing after the fact? |
| 16:31.39 | pb_ | mccarthy: depends exactly how you've structured the loop, but it should do nothing when null. |
| 16:31.53 | mccarthy | for (iter = cloned_events; iter; iter = g_slist_next (iter)) { |
| 16:31.57 | mccarthy | event_t ev = iter->data; |
| 16:32.00 | mccarthy | event_db_destroy_clone(ev); |
| 16:32.02 | mccarthy | } |
| 16:32.06 | mccarthy | then |
| 16:32.14 | mccarthy | if (cloned_events) event_db_list_destroy(cloned_events); |
| 16:32.15 | pb_ | mccarthy: looks ok to me. |
| 16:32.24 | pb_ | oh, not that last line. |
| 16:32.32 | pb_ | I think you just want g_slist_free (cloned_events) there. |
| 16:32.51 | pb_ | event_db_list_destroy probably does too much. I think it will try to free the events themselves. |
| 16:33.07 | mccarthy | pb_: thats all event_db_list_destroy foes |
| 16:33.11 | mccarthy | foes=does |
| 16:33.15 | pb_ | oh, ok. |
| 16:33.43 | pb_ | well, in that case I don't see anything wrong with that code. where does it actually crash? |
| 16:34.08 | mccarthy | one sec, make DEBUG=yes.... |
| 16:35.15 | mccarthy | pb_: in g_free(ev) |
| 16:35.31 | mccarthy | (shown as event_db_destroy_clone(ev); above) |
| 16:36.00 | pb_ | what's ev at that point? |
| 16:36.33 | mccarthy | pb_: I dont know....how should i check printf(%x\n, ev);? |
| 16:36.57 | pb_ | are you in gdb? |
| 16:37.10 | mccarthy | pb_: not anymore |
| 16:37.14 | mccarthy | but I can be |
| 16:37.20 | pb_ | well, get back into it, run the program til it crashes, then do "print ev". |
| 16:37.42 | pb_ | you might have to do "up" a few times to get to a point where it understands what ev is. |
| 16:37.47 | mccarthy | pb_: well it goes into g_free then free then _int_free |
| 16:37.49 | pb_ | or just "bt" will probably show it. |
| 16:38.13 | pb_ | mccarthy: right, you will have to step up until you get back to event_db_destroy_clone. |
| 16:38.29 | mccarthy | pb_: ahhh, never "upped" before |
| 16:38.38 | mccarthy | (gdb) print ev |
| 16:38.38 | mccarthy | $1 = (struct event_s *) 0x4212a2d0 |
| 16:39.04 | pb_ | well, that looks like it should be ok. |
| 16:39.16 | mccarthy | one would think.... |
| 16:39.16 | pb_ | at least, it's not NULL or any other obviously-bad pointer. |
| 16:39.41 | pb_ | mccarthy: try linking with electric fence, see if it has anything to say. |
| 16:40.23 | mccarthy | pb_: does it "say" it in gdb? (never used -lefence) |
| 16:40.48 | pb_ | mccarthy: it will print snide messages like "address 0x4212a2d0 not from malloc" to the terminal. |
| 16:40.51 | mccarthy | ElectricFence Aborting: free(43c93d40): address not from malloc(). |
| 16:41.37 | pb_ | ok. well, that means you are either trying to free the same thing twice, or are trying to free some pointer that wasn't originally mallocd. |
| 16:42.57 | mccarthy | pb_: it doesnt appear in my printf(%x) list more than once.... |
| 16:43.35 | pb_ | mccarthy: and there's no danger that it will have been freed anywhere else in the meantime? |
| 16:44.13 | mccarthy | { |
| 16:44.13 | mccarthy | <PROTECTED> |
| 16:44.13 | mccarthy | <PROTECTED> |
| 16:44.13 | mccarthy | <PROTECTED> |
| 16:44.13 | mccarthy | <PROTECTED> |
| 16:44.18 | mccarthy | I dont see how |
| 16:45.12 | pb_ | not necessarily while that loop is running. do you pass the cloned events to any other routine that might decide to free them? |
| 16:45.52 | pb_ | you could attempt to find out by putting a conditional breakpoint on g_free, I guess. |
| 16:45.59 | mccarthy | pb_: hmmm, I dont think so.... |
| 16:46.14 | mccarthy | pb_: does this look right |
| 16:46.17 | mccarthy | new_ev = memcpy (new_ev, ev, sizeof(event_t)); |
| 16:46.48 | pb_ | yeah |
| 16:47.06 | pb_ | though the "new_ev =" is unnecessary; memcpy will just return the same pointer you gave it initially. |
| 17:07.13 | PaxAnima | oooh |
| 17:09.43 | prpplague | mallum: i'm sending some stuff to you via ade next week sometime |
| 17:11.42 | mallum | prpplague: cool |
| 17:24.23 | *** join/#gpe walters (walters@verbum.org) |
| 17:26.32 | walters | so...i will probably need to use stuff like xlibs from the familiar feed, since openzaurus doesn't appear to have them...right? |
| 17:26.59 | mallum | okey matchbox_0.4pre.20030103-0_arm.ipk uploaded |
| 17:27.10 | mallum | I havn't tested myself yet - so use at your own risk ;-) |
| 17:27.42 | pb_ | mallum: cool, what's the news in this version? |
| 17:27.51 | pb_ | walters: I thought oz did have a build of xlibs, though I don't know how new it is. |
| 17:27.56 | pb_ | walters: the one from familiar ought to work OK regardless. |
| 17:28.13 | mallum | pb_: mainly startup-notification fixes, re your suggestions. and a load of minor wierd bits |
| 17:28.27 | pb_ | mallum: righto. well, I shall install it as soon as I get home. |
| 17:28.31 | mallum | pb_: if this ones good I'll dump it into unstable |
| 17:28.36 | pb_ | mallum: great. |
| 17:28.37 | walters | well... |
| 17:28.38 | walters | i get this: |
| 17:28.40 | walters | root@zaurus:~# ipkg install task-gpe |
| 17:28.40 | walters | ERROR: Cannot satisfy the following dependencies for task-gpe: |
| 17:28.40 | walters | <PROTECTED> |
| 17:29.00 | pb_ | walters: ah, I detect a recurring xlibs theme there. |
| 17:29.10 | walters | sorry if i am asking silly questions, i am very new to ipkg and stuff :) |
| 17:29.13 | pb_ | walters: maybe the OZ xlibs package is called something slightly different, or maybe I was just wrong. |
| 17:29.37 | walters | right now i am using: |
| 17:29.38 | walters | src unstable http://openzaurus.org/official/testing/feed |
| 17:29.38 | walters | src unstable http://handhelds.org/download/gpe/unstable/base |
| 17:29.47 | walters | should i add the familiar src feed? |
| 17:30.04 | pb_ | I think adding familiar to an oz ipkg.conf is a dangerous thing to do. |
| 17:30.11 | walters | right, that's what i guessed |
| 17:30.16 | pb_ | maybe check with the dudes in #openzaurus, see what they suggest. |
| 17:30.54 | mallum | pb_: hmm, I just rebooted ( after an upgrade ) and it appears gpe-login is playing the xserver like a yoyo ;-) |
| 17:31.00 | walters | pb_: ok |
| 17:31.04 | pb_ | mallum: check your gpe-login and gpe-session-scripts versions. |
| 17:31.16 | mallum | pb_: I just updated , and theres a new one |
| 17:31.30 | pb_ | mallum: make sure gpe-login got upgraded. |
| 17:31.40 | pb_ | mallum: should be at 0.34 I think. |
| 17:31.42 | mallum | pb_: ok its trying to upgrade it now .. |
| 17:31.48 | mallum | pb_: yeah to 0.34 |
| 17:32.22 | pb_ | walters: or maybe hassle prpplague, I think he was maybe into X-on-Z. |
| 17:40.42 | prpplague | pb_: ya, just been busy |
| 17:41.13 | mallum | hmm, matchbox package bad :( |
| 17:41.21 | pb_ | mallum: oh dear. |
| 17:41.35 | prpplague | walters: gpe lots of little issues on getting a gpe package on the z |
| 17:41.45 | prpplague | walters: nothing major just time oriented |
| 17:42.25 | walters | prpplague: ok...do you have a snapshot of your work? |
| 17:42.27 | pb_ | mallum: did you order a zaurus in the end, by the way? |
| 17:42.47 | mallum | pb_: no, not yet |
| 17:42.52 | prpplague | walters: everything i have is included in the openzaurus buildroot |
| 17:48.22 | walters | ok...i am switching to unstable. |
| 17:59.11 | mallum | new matchbox uploaded ( fixed ) |
| 18:00.16 | *** 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 | that's GPE with a G | *New* matchbox 0.4pre - ipkg install http://handhelds.org/~mallum/downloadables/experimental/matchbox_0.4pre.20030103-0_arm.ipk | happy 2004 |
| 18:11.15 | *** join/#gpe jsmanrique (jsmanrique@cmr-213-141-51-252.telecable.es) |
| 18:11.19 | jsmanrique | hi again! |
| 18:15.57 | mallum | hi jsmanrique |
| 18:16.38 | jsmanrique | finnaly I've bought something for my girlfriend.. of course, not an ipaq :-) |
| 18:16.47 | jsmanrique | Three Kings are coming :-) |
| 18:21.01 | mallum | wtf ? |
| 18:24.44 | jsmanrique | ¿? |
| 18:45.13 | PaxAnima | wow, years go by fast... "happy 2004" in topic ;) |
| 19:01.30 | *** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com) |
| 19:03.02 | mccarthy | pb_: look at my code yet? :) |
| 19:04.28 | pb_ | mccarthy: no, just got home, I'll look at it in a bit. |
| 19:08.26 | *** join/#gpe pierre (~pierre@lyon-1-a7-62-147-19-12.dial.proxad.net) |
| 19:14.42 | mccarthy | pb_: what is the status of gtk 2.2.0 relative to Debian.... (is it in testing?) |
| 19:14.48 | pb_ | mccarthy: not yet |
| 19:14.55 | pb_ | mccarthy: it is in unstable for all architectures other than arm. |
| 19:15.43 | mccarthy | pb_: ok. I asked Nick Duffek to install it on the ipaq's, but he said "no go, until its in debian (preferably testing)" |
| 19:16.07 | mallum | pb_: ah I uploaded a fix mb ipkg |
| 19:16.10 | pb_ | mccarthy: ah ok. I think it should go into testing almost immediately when the arm debs are ready, which will probably be tomorrow. |
| 19:16.11 | mallum | pb_: see topic |
| 19:16.31 | pb_ | mallum: ah yes, you need to add more stars around New so that it looks even newer than the last one. |
| 19:16.43 | mallum | pb_: yeah ;-) |
| 19:17.41 | mallum | pb_: you shouldn't now be able to start 2 instances from single-instance app via mbdesktop |
| 19:17.49 | pb_ | mallum: great. |
| 19:18.34 | mallum | pb_: also it should handle single instace stuff better in general |
| 19:18.47 | pb_ | mallum: btw, would you consider "unbundling" some stuff from the matchbox ipk, so there were separate packages for say matchbox proper, libmbdock, and the panel apps? |
| 19:19.14 | mallum | pb_: though it currently stupidly messes up the mapping over a theme change |
| 19:19.34 | pb_ | mallum: righto |
| 19:19.56 | mccarthy | mallum: should xmonobut be showing up in the mbdock 'add" menu? |
| 19:20.16 | mallum | mccarthy: it should but it dont currently, needs updating |
| 19:20.36 | mallum | pb_: I'll probably unbundle libmb seperatly |
| 19:20.41 | mccarthy | mallum: what does that mean....updating? |
| 19:20.47 | mccarthy | update-menus? |
| 19:20.50 | mallum | mccarthy: fixing it ;-) |
| 19:20.58 | mccarthy | mallum: ok. thanks. |
| 19:21.12 | mallum | mccarthy: matchbox does not use any of the update menu's crap ;-) anymore ... |
| 19:21.53 | pb_ | mallum: ok, that would be cool. it would be nice to have the mini* apps and maybe mbdesktop packaged separately as well if possible, but I guess that isn't so important. |
| 19:24.37 | mallum | pb_: yeah sorting out and versioning libmb is the first step in that direction |
| 19:25.21 | pb_ | mallum: right. |
| 19:35.23 | pb_ | oh, and I checked my hacked miniapm into gpe cvs as base/minibat, in case you or anybody else wanted to look at it. |
| 19:37.17 | mccarthy | pb_: how many times should list_for_period be called at the startup of gpe-calendar? |
| 19:37.40 | pb_ | mccarthy: dunno, I guess each view might want to call it once during startup. |
| 19:38.38 | mccarthy | pb_: that might cause a problem since a view will want to "hang on" to the cloned list while I try to throw it away and make a new one! |
| 19:39.24 | pb_ | mccarthy: ah. |
| 19:39.35 | mccarthy | pb_: I am not sure how to avoid that... |
| 19:39.37 | pb_ | mccarthy: yes, I was a bit worried that kind of thing might be happening. |
| 19:40.04 | mccarthy | pb_: unless we tell list_for which view we are calling it from... |
| 19:40.16 | mccarthy | (and each gets its own clone list) |
| 19:40.32 | pb_ | mccarthy: hrm, that doesn't sound like a good idea. maybe add explicit show and hide methods for views, so they can release the lists when they no longer need them. |
| 19:40.52 | mccarthy | BTW, I got the "max" thing working again (truncating the list) |
| 19:41.02 | mccarthy | (oops, but that will likely cause trouble too!) |
| 19:41.09 | pb_ | heh. |
| 19:42.29 | mccarthy | pb_: would you care if I chcanged the datefont in main to 7 pt instead of 8 pt? |
| 19:42.45 | pb_ | mccarthy: no, but I was planning to remove that stuff altogether so you might not want to bother. |
| 19:42.55 | pb_ | mccarthy: it's only used for the year view nowadays. |
| 19:43.31 | mccarthy | pb_: I know but the year view gets screwy on my desktop because it choses some weird font when you say *-8-* |
| 19:44.07 | pb_ | mccarthy: I think I'm going to just turn the year view off altogether until someone has the enthusiasm to rewrite it in a modern way. Right now it's ugly and fairly useless. |
| 19:44.39 | mccarthy | haha "modern" (as in not three months old) |
| 19:45.00 | pb_ | mccarthy: heh. in this case, "modern" as in using outline fonts not bitmap X core fonts. |
| 19:45.21 | mccarthy | pb_: oh, that part....it was funnier the other way |
| 19:45.26 | pb_ | :-) |
| 19:46.00 | pb_ | mccarthy: if it's easy to rewrite with pango/xft I might have a go at that instead, but I think sufficiently few people actually care about the year view that I'm not inclined to make any heroic efforts to keep it working. |
| 19:47.42 | mccarthy | pb_: I dont get the untimed versus untimed_significant stuff? |
| 19:48.49 | pb_ | mccarthy: untimed is a boolean saying whether to return untimed or timed events. But a boolean doesn't have a "don't care" state, so if untimed_significant is FALSE the code will ignore untimed and just return everything. |
| 19:50.59 | mccarthy | pb_: crap |
| 19:51.33 | pb_ | mccarthy: ! |
| 19:51.35 | mccarthy | pb_: day view alone calls it 24 times |
| 19:51.42 | pb_ | mccarthy: once for each hour, I suppose. |
| 19:51.59 | pb_ | mccarthy: that would kind of make sense, since day view maintains a separate event list for each hour. |
| 19:52.29 | mccarthy | pb_: yes, but the cloners are confusing me now.... |
| 19:52.50 | mccarthy | (I watched Star Wars II: AoTC last night BTW :) |
| 19:53.14 | pb_ | heh, was it any good? |
| 19:53.33 | mccarthy | pb_: terrible acting, but cool effects and a half-decent story |
| 19:53.50 | mccarthy | pretty much only Anakin was a terrible actor |
| 19:54.02 | pb_ | mccarthy: ah :-) |
| 19:54.04 | mccarthy | so overall it was probably slightly better than 1 |
| 19:55.11 | mccarthy | pb_: so do the cloners live forever now instead? |
| 19:55.26 | pb_ | mccarthy: that doesn't seem terribly desirable. |
| 19:55.30 | mccarthy | (so I make them only once?) |
| 19:55.41 | mccarthy | damn and it seemed close.... |
| 19:56.13 | pb_ | mccarthy: can you arrange for event_db_list_destroy to free any clones it encounters? |
| 19:56.55 | pb_ | mccarthy: I think all the views will call that when they are through with a list of events. Or if they don't, they could probably be made to. |
| 20:00.51 | pb_ | mallum: single instance does indeed seem to be working much better now |
| 20:01.33 | mccarthy | pb mallum: what is single instance? do some apps set a single-instance flag or something? and matchbox now listens to it? |
| 20:02.05 | pb_ | mccarthy: the launchers, rather than matchbox itself, but yeah. |
| 20:02.21 | mccarthy | pb_: when do we even *call* event_db_list_destroy? |
| 20:02.29 | pb_ | mccarthy: so if the calendar is already running, and you click on it again in mbdesktop, it won't launch another copy. |
| 20:02.38 | pb_ | mccarthy: dunno, I rather hoped the views would do that at some good time. |
| 20:02.55 | pb_ | mccarthy: they must have _some_ way of disposing of the list of events they get from event_db_list_for_period. |
| 20:10.59 | mccarthy | pb_: there *is* a reason I keep you on this gpe-claendar team ;) |
| 20:11.19 | mccarthy | destroy_list now erases any clones |
| 20:11.25 | pb_ | mccarthy: cool |
| 20:11.37 | mccarthy | pb_: I think it is fine to be uploaded now... |
| 20:11.43 | pb_ | mccarthy: righto. |
| 20:12.27 | mccarthy | pb_: what about the Makefile? should I leave off the -lXsettings? |
| 20:12.41 | pb_ | mccarthy: yeah, leave it out for now, I'll try to work out what's wrong there |
| 20:12.51 | pb_ | mccarthy: I can build a new package here. |
| 20:13.30 | mccarthy | pb_: ok. I have also made (very) minor changes to get rid of most (but not all) of the warnings |
| 20:13.58 | pb_ | mccarthy: great. |
| 20:14.40 | mccarthy | pb_: oops.... it seems that I took out the -lXsettings already and now it works here....strange |
| 20:15.31 | mccarthy | pb_: BTW, I was not at all clever about *editting* clones. Right now it works *exactly* the way that it used to -- edit *any* instance of a recur and it edits all of them.... |
| 20:16.35 | mccarthy | pb_: oops, found another bug....hold off on that cvs commit :) |
| 20:16.48 | pb_ | mccarthy: ok :-) |
| 20:20.18 | *** join/#gpe jg (~jg@dialup-67.30.181.81.Dial.Boston1.Level3.net) |
| 20:20.44 | pb_ | mccarthy: suggest you check the code into cvs when you are happy with it, and I will build a new package. |
| 20:20.48 | pb_ | jg: hello |
| 20:21.24 | jg | pb_: hi |
| 20:21.57 | pb_ | grr damn lilo and his server notices. |
| 20:33.15 | mccarthy | pb_: if I want to stick the uid rather than a pointer to the ev itself in th erow_data, would this work... |
| 20:33.27 | mccarthy | gtk_clist_set_row_data (GTK_CLIST (day_list), row, &(ev->uid)); |
| 20:34.23 | pb_ | mccarthy: probably want just (gpointer)(ev->uid), no & |
| 20:34.53 | mccarthy | pb_: then how can I grab it in the gtk_clist_get_row_data routine? |
| 20:35.30 | pb_ | mccarthy: just cast the return value back to an int. |
| 20:38.43 | mccarthy | pb_: it seems that only the day_view actually destroyed any of its lists (until now) |
| 20:38.58 | mccarthy | pb; but I amhaving trouble figuring out where/when to have month_view do it... |
| 20:39.13 | mccarthy | (and the pop-ups from month view) |
| 20:52.44 | mccarthy | pb_: nevermind. you were erasing them already in month_view, so you just skipped future_view (erasing them TWICE causes a problem) |
| 20:58.13 | mallum | pb_: oh is the gpe-bluetooth thing missing a dep on libsdp ? |
| 20:58.44 | *** join/#gpe jg (~jg@dialup-67.30.181.81.Dial.Boston1.Level3.net) |
| 20:59.51 | pb_ | mallum: Andy just rearranged the bluez packages again, that might be causing a problem. |
| 21:01.01 | mallum | pb_: ah, okie - I'll figure it out |
| 21:06.26 | *** join/#gpe hunterdie (~kb3hag@pcp02465537pcs.chrchv01.md.comcast.net) |
| 21:10.02 | pb_ | mallum: oh, it looks like it might need a rebuild this time. |
| 21:10.22 | pb_ | mallum: looks like the new bluez-sdp includes libsdp.so.2 not libsdp.so.1. |
| 21:12.09 | hunterdie | eh |
| 21:31.33 | mccarthy | pb_: all works at least as well as it did before now :) |
| 21:31.39 | mccarthy | pb_: I have committed changes |
| 21:32.24 | mccarthy | (am making an ipkg for myself now, but I think the *real* package should come from you since it will be linked with the correct library versions) |
| 21:32.38 | pb_ | mccarthy: ok cool, I'll build one in a few minutes |
| 21:32.56 | pb_ | just filing some bugs against bluez packages at the moment :-} |
| 21:33.05 | mccarthy | pb_: ok. |
| 21:33.15 | mccarthy | pb_: BTW, classes start Monday :( |
| 21:33.30 | mccarthy | pb_: probably going to be a lull in joe-coding again |
| 21:34.00 | pb_ | mccarthy: ah. :-/ |
| 21:34.14 | pb_ | mccarthy: well, I guess these things can't be helped. |
| 21:34.17 | mccarthy | pb_: yep, the ipaqs give that weird xsettings error still |
| 21:34.32 | mccarthy | I wonder why? |
| 21:34.43 | pb_ | no idea, it seems very strange. |
| 21:35.08 | mccarthy | thats a pain that I cannot build ipkgs very easily |
| 21:35.54 | pb_ | yes. |
| 21:41.38 | mccarthy | pb_: after you build an ipkg, you should email it to me or something before uploading to unstable.... |
| 21:41.46 | mccarthy | pb_: it seems quite slow with the changes :( |
| 21:43.20 | pb_ | mccarthy: ok |
| 21:43.41 | mccarthy | pb_: i am not sure if that is related to my weird build or simply poor coding :) |
| 21:44.02 | pb_ | :-/ |
| 21:47.19 | mccarthy | pb_: i didnt know that bug was assigned to me :) I *think* sqlite and libsqlite work fine now... |
| 21:47.41 | mallum | pb_: did you install latest matchbox ? |
| 21:48.27 | mccarthy | mallum: I did, why? |
| 21:48.30 | pb_ | mallum: yeah |
| 21:48.41 | mallum | working ok for the both of you ? |
| 21:48.57 | pb_ | < pb_> mallum: single instance does indeed seem to be working much better now |
| 21:49.11 | pb_ | haven't done a lot of testing yet other than that. |
| 21:49.12 | mallum | pb_: cool |
| 21:49.21 | pb_ | but certainly no sign of any problems yet. |
| 21:49.23 | mccarthy | mallum: I believe so, although I cannot get *back* to an app that I "scrolled" off of (from the topbar thingie) |
| 21:49.57 | mallum | mccarthy: yeah, I yet to clean up the panel |
| 21:50.07 | mccarthy | pb_: how did you test the single-instance thing...what is supposed to have only a single-instance? |
| 21:50.09 | pb_ | mccarthy: well, that bug was very old, I'm fairly sure it's been fixed for ages. if not, I guess the submitter can file another :-) |
| 21:50.14 | pb_ | mccarthy: gpe-calendar, gpe-todo, gpe-contacts I think |
| 21:50.31 | mccarthy | pb_: I can start a second gpe-calendar! |
| 21:50.34 | pb_ | mccarthy: check their .desktopf files to make sure. |
| 21:50.42 | pb_ | mccarthy: ah, then you lose. |
| 21:50.58 | pb_ | mccarthy: how are you starting it? |
| 21:51.21 | mccarthy | pb_: oops, I could *before* :) now it seems to be working...maybe I updated but didnt log out. |
| 21:51.37 | pb_ | ah. |
| 21:52.02 | mccarthy | pb_: nope. just got a second todo |
| 21:52.21 | pb_ | mccarthy: check that it really mentions singleinstance in its .desktop. I may not have uploaded a version with that in yet. |
| 21:53.41 | mccarthy | pb_: you are right again, no SingleInstance=True for todo |
| 21:55.30 | pb_ | mccarthy: ah ok, I don't always make a new package straight away for trivial changes like that. Have a go with contacts, maybe that one has it. |
| 21:55.45 | mccarthy | pb; well, calendar seems to work fine |
| 21:55.57 | pb_ | ok,. cool |
| 22:01.36 | *** join/#gpe flypiper (~ISP-1@12-222-141-210.client.insightBB.com) |
| 22:05.04 | mallum | pb_: heh, I finally got the toolbar raised, main win changed bug |
| 22:05.18 | pb_ | mallum: aha! |
| 22:05.29 | pb_ | mallum: did you do anything in particular to provoke it? |
| 22:06.28 | flypiper | pb_: what MLB are you using for your arm box ?? Im thinking about building one:) |
| 22:06.49 | pb_ | flypiper: MLB? |
| 22:07.02 | flypiper | Main Logic Board? |
| 22:07.36 | pb_ | flypiper: ah. one of the Simtec CATS boards. It's an ATX form factor motherboard with an SA-110, couple of SDRAM and PCI slots. |
| 22:08.47 | mallum | pb_: no, it happened after a rotation to landscape, its ok now |
| 22:09.01 | pb_ | mallum: ah. |
| 22:09.17 | flypiper | pb_: THANKS:-) I contacted Netwinder ppl.. and they don't act interested in selling me a box:( |
| 22:10.12 | pb_ | flypiper: yeah, I dunno what exactly is going on with netwinder at the moment. If you offered to take their whole inventory of strongarm boxes off their hands they might go for it :-) |
| 22:10.23 | pb_ | flypiper: or I expect you can find used netwinders on ebay from time to time. |
| 22:10.55 | pb_ | flypiper: however, the cats is more expandable, and they have a 275MHz version now so it should be just as fast as the netwinder. |
| 22:11.32 | flypiper | pb_: LOL.. dunno what I would do with their inventory. And been looking on e-bay.. with no luck. Im gonna google for some hardware and prolly just build one.. I just need the MLB.. got the other HW |
| 22:12.47 | pb_ | flypiper: presumably put them in your garage and sell them one at a time to other people like you (same as tim riker did with the tuxscreens last year). From what I gather, netwinder are just not interested in selling those old boxes in ones and twos because it's a lot of hassle for them and makes them very little money. |
| 22:14.17 | mallum | an xscale cats would be uber cool |
| 22:15.00 | flypiper | pb_: are they located in the UK?? Cause they asked me where I was located.. then no response\cut short the conversation:( |
| 22:15.14 | mallum | flypiper: yes |
| 22:15.18 | flypiper | ahhhh |
| 22:15.26 | mccarthy | pb_: send me email to let me know if your calendar is as slow as mine |
| 22:15.29 | mccarthy | I am outta here |
| 22:15.43 | pb_ | flypiper: simtec? yeah, they are in Lancashire in England. |
| 22:15.51 | flypiper | thanks |
| 22:16.01 | pb_ | flypiper: netwinder are in Canada. |
| 22:16.34 | pb_ | mccarthy: will do |
| 22:16.40 | *** part/#gpe mccarthy (~mccarthy@granular.che.pitt.edu) |
| 22:17.01 | pb_ | mallum: I was talking to vince earlier on, apparently simtec do have some xscale machines but they don't feel the performance is good enough to warrant selling them at the moment. |
| 22:17.35 | pb_ | mallum: they also have some arm9 hardware, but that is no faster than strongarm (maybe a bit slower even) |
| 22:20.32 | pb_ | hmm, is it new that you can't run xcalibrate as anybody other than root? |
| 22:32.04 | flypiper | pb_: Nice looking Board:) |
| 22:33.12 | pb_ | flypiper: yeah, it's pretty cool. |
| 22:45.36 | mallum | pb_: oooh gpe-games2 is pretty nice ;-) |
| 22:45.44 | pb_ | mallum: yeah. |
| 22:45.56 | pb_ | mallum: mccarthy is responsible for most of it. |
| 22:47.25 | PaxAnima | pb_: any tip on using ElectricFence? just run ef ./app ? |
| 22:47.35 | pb_ | PaxAnima: link with -lefence, run as normal |
| 22:47.42 | pb_ | (ideally under gdb of course) |
| 22:47.43 | PaxAnima | ah. |
| 22:48.51 | PaxAnima | thanks |
| 22:50.14 | pb_ | you might have to turn off trapping of malloc(0), because I think some part of xft does that. There's an environment variable you can set for that purpose. |
| 22:50.57 | pb_ | it'll be obvious if you have this problem, because electric fence will abort immediately somewhere in the bowels of libxft. |
| 22:51.05 | PaxAnima | seems to work ok |
| 22:51.11 | pb_ | ok, cool. |
| 22:54.28 | PaxAnima | things are sooooo slow when using electric fence... |
| 22:54.54 | PaxAnima | but I think I found the fault.. |
| 23:09.41 | *** join/#gpe Dr-Proton (~Dr-Proton@ppp-209-232-54-243.dialup.snfc21.pacbell.net) |
| 23:10.00 | *** part/#gpe Dr-Proton (~Dr-Proton@ppp-209-232-54-243.dialup.snfc21.pacbell.net) |
| 23:20.09 | *** join/#gpe dc_ (~dc@modem-91-52-60-62.vip.uk.com) |
| 23:22.02 | dc_ | hey |
| 23:22.53 | PaxAnima | hey, dc_ |
| 23:25.03 | dc_ | weeee :D |
| 23:25.13 | mallum | dc_: hey |
| 23:25.35 | dc_ | PaxAnima: I've been playing Halo with a friend for the last 36 hours =) |
| 23:25.40 | dc_ | mallum: :D |
| 23:25.43 | dc_ | mallum: great game :)) |
| 23:25.59 | mallum | dc_: hey me too ;-) |
| 23:26.06 | PaxAnima | dc_: heh ;) |
| 23:26.06 | mallum | dc_: it f**king rules :_) |
| 23:26.09 | dc_ | mallum: yep. |
| 23:26.25 | dc_ | mallum: quite :D |
| 23:26.37 | dc_ | mallum: co-op is great |
| 23:27.01 | PaxAnima | dc_: next irc input may get very leaky: had to change the structure of the dictionary system, so strdup() and free() are littered all over the source :/ |
| 23:27.09 | mallum | dc_: yeah I've been prepping sid for co-op |
| 23:27.32 | mallum | dc_: is it your xbox or your mates ? |
| 23:27.56 | dc_ | mallum: my mates ;/ |
| 23:28.09 | mallum | dc_: any other good games ? |
| 23:28.25 | dc_ | mallum: but I might consider getting one once I see breed |
| 23:28.42 | dc_ | mallum: well, spliter cell's _not_ bad... |
| 23:29.15 | dc_ | mallum: and sega gt is pretty fun aswel.. |
| 23:29.22 | dc_ | mallum: but, halo :)) |
| 23:29.37 | mallum | dc_: I found splinter cell a bit arkward / complex to control |
| 23:29.57 | dc_ | yes. it would seem. |
| 23:30.02 | mallum | dc_: I got stuck and pissed off with it trying to get past some stupid camera on the training level |
| 23:30.26 | mallum | dc_: I want to SHOOT things, not lark around in shadows |
| 23:30.47 | dc_ | hah |
| 23:30.57 | dc_ | mallum: I can imagine that with you. |
| 23:31.08 | mallum | dc_: is it me or does the ememy AI in halo really good |
| 23:31.32 | dc_ | mallum: I'll assume you've completed halo :)) |
| 23:31.48 | mallum | dc_: no not yet |
| 23:32.03 | mallum | dc_: still loads togo I think, I just play it for 20 mins every few hours |
| 23:32.30 | dc_ | mallum: quite, my mate seems to be in shock. |
| 23:32.39 | mallum | dc_: I just got a rocket launcher :-) |
| 23:33.43 | dc_ | mallum: hehe, well the melee atack is certianly well worth it. |
| 23:34.28 | dc_ | mallum: defently try and get sid to do co-op with you, having two people in the warthog is great fun! |