| 04:30:24 | mibus | hi all |
| 04:30:43 | | rumour has it crlcam is http://192.58.206.220:9192/ or available at http://sigkill9.dyndns.org/crl.html scaled to 640x480 or http://devweekend.handhelds.org:9192 or don't forget to check crlaudio |
| 04:30:43 | mibus | ibot: crlcam |
| 04:30:59 | | hmmm... crlaudio is http://devweekend.handhelds.org:8000 or don't forget to check crlcam |
| 04:30:59 | mibus | ibot: crlaudio |
| 04:48:19 | | dc was last seen on #handhelds.org 4 days, 7 hours, 5 minutes and 53 seconds ago, saying: teck: trolls are [Wed Oct 9 15:42:26 2002] |
| 04:48:19 | mibus | ibot: seen dc? |
| 04:48:21 | | dc_ was last seen on #handhelds.org 1 days, 5 hours, 58 minutes and 34 seconds ago, saying: night :D [Sat Oct 12 16:49:47 2002] |
| 04:48:21 | mibus | ibot: seen dc_? |
| 04:48:21 | | pb_ was last seen on #handhelds.org 4 hours, 52 minutes and 1 seconds ago, saying: sounds neat. [Sun Oct 13 17:56:20 2002] |
| 04:48:21 | mibus | ibot: seen pb_? |
| 05:33:02 | colin_werk | Morning. |
| 06:38:15 | mibus | runs away to rescue a friend's flat car battery... |
| 06:38:23 | mibus | colin_werk: morning BTW |
| 08:36:52 | pb_ | morning |
| 08:38:43 | Zindar | morning |
| 08:38:54 | Zindar | I'm still working to get my ipaq to work :( |
| 08:39:53 | mibus | returns |
| 08:40:03 | pb_ | g'day mibus |
| 08:40:14 | pb_ | Zindar: still this kernel problem? You probably need a bootldr expert. |
| 08:40:28 | Zindar | pb_: well.. I even tried to reinstall the damn thing... |
| 08:40:38 | pb_ | as in, reflash the whole root filesystem? |
| 08:40:40 | Zindar | the upload of the root-image works fine.. |
| 08:40:46 | Zindar | but it still can't boot |
| 08:40:53 | pb_ | Oh, that's weird. |
| 08:40:58 | pb_ | Does the md5sum match? |
| 08:40:58 | Zindar | I agree... |
| 08:41:10 | Zindar | feels like something is wrong with my flash |
| 08:41:42 | Zindar | pb_: look here... from the bootldr... "ls" works... |
| 08:41:45 | Zindar | but "ls boot" |
| 08:41:58 | Zindar | boot> ls boot |
| 08:41:58 | Zindar | find_inode failed for name=boot |
| 08:41:58 | Zindar | Failed to scan jffs2 file structure |
| 08:42:37 | pb_ | Hmm, right. |
| 08:43:06 | pb_ | Well, I guess Jamey would be the man you need to talk to. |
| 08:43:19 | Zindar | yeah.. I'm mailing the mailinglist alittle later... |
| 08:43:33 | Zindar | so I can get help without being here to much :) |
| 08:43:46 | pb_ | Oh, it wasn't you who wrote to the mailing list already? |
| 08:46:53 | Zindar | no.. did someone write about something simular? |
| 08:46:57 | pb_ | Some guy called Jaap Hogenberg is reporting what sounds like the same problem. |
| 08:47:02 | Zindar | I just joined 2 sec ago |
| 08:47:04 | Zindar | ahh |
| 08:47:18 | Zindar | did he get help? |
| 08:47:22 | pb_ | Not yet. |
| 08:47:28 | Zindar | hmm... |
| 08:47:39 | Zindar | was it "ipaq@hh" or "hh@hh" or......? |
| 08:47:46 | pb_ | familiar@ I think |
| 08:48:01 | pb_ | the subject line was "Problems loading 0.6 rc 1" |
| 08:48:01 | Zindar | hmm..ok.. I'm mailing ipaq@ |
| 08:48:19 | Zindar | maybe it's a rc1 problem... I'm trying pre1 on |
| 08:48:21 | Zindar | now |
| 08:48:43 | Zindar | but I don't think that's it.. |
| 08:48:46 | Zindar | :/ |
| 08:50:10 | Zindar | seams like the same problem (reading his mail now) |
| 08:50:30 | pb_ | Yes. |
| 08:55:06 | mibus | hi pb_ |
| 09:03:52 | Zindar | wow |
| 09:04:02 | Zindar | pb_: it works with pre1 but not rc1 |
| 09:04:13 | Zindar | pb_: seams like something is wrong with it... |
| 09:04:19 | pb_ | weird |
| 09:04:37 | Zindar | well.. yeah... I'll mail familiar@ |
| 09:04:43 | pb_ | Nelson built the most recent rc1 images. Maybe he was doing something differently to the way jamey created the others. |
| 09:04:49 | pb_ | Good idea. |
| 09:05:21 | Zindar | :) |
| 09:18:27 | pb_ | mibus: what's the status with libredalert these days? |
| 09:20:27 | mibus | pb_: havent touched it |
| 09:20:53 | pb_ | mibus: does that mean you think it's ready to go, or just that you've been busy with other things? |
| 09:20:57 | mibus | pb_: russ said he tried to modify atd to do what i mentioned (time.appname-id IIRC) |
| 09:21:02 | mibus | pb_: busy :( |
| 09:21:08 | mibus | pb_: but he said atd was segfaulting |
| 09:21:11 | pb_ | ah. |
| 09:21:20 | mibus | pb_: once i know im free for a day or two im gonna offer a hand |
| 09:21:30 | mibus | pb_: but ATM uni keeps dumping lame assignments on me ;) |
| 09:21:35 | pb_ | hell! |
| 09:25:15 | mibus | pb_: yes... |
| 09:25:55 | mibus | pb_: ive cut out chunks of code from appmgr looking for this mem leak |
| 09:26:01 | mibus | pb_: then realised something - its not appmgr at all |
| 09:26:04 | pb_ | ah. |
| 09:26:07 | mibus | pb_: check nmf's mem usage |
| 09:26:31 | mibus | glances at the clock - 3 minutes before auto-disconnect |
| 09:26:35 | pb_ | heh |
| 09:26:39 | pb_ | so, you think it's GTK? |
| 09:27:31 | mibus | pb_: stands to reason... nmf is like 10MB if im reading it right |
| 09:27:36 | pb_ | hmm. |
| 09:27:37 | mibus | pb_: and thats not with anything loaded |
| 09:29:06 | mibus | ok ciao everyone |
| 09:29:10 | pb_ | bye |
| 09:32:46 | pb_ | yo mallum |
| 09:34:18 | mallum | hay pb_ |
| 10:17:22 | pb_ | hi colin_werk |
| 10:17:33 | colin_werk | Hi pb_ |
| 10:18:07 | colin_werk | Any news? I didn't even read my email from the weekend yet |
| 10:18:33 | pb_ | Hmm, nothing all that noteworthy I don't think. |
| 10:18:51 | pb_ | The guys at CRL spent the weekend making new 0.6-rc1 releases. It seems to be in fairly good shape now. |
| 10:19:45 | colin_werk | Ah, cool. |
| 10:21:22 | pb_ | I gather jg and keithp pounded out some new XRandR stuff as well. |
| 10:26:43 | colin_werk | More news on the GTK2 transition? |
| 10:26:44 | pb_ | They had a rather cool live streaming audio and video feed from the CRL event. |
| 10:26:44 | pb_ | Well, the GTK2 transition is more or less complete. Lots of people (well, me, dc_ and mccarthy at least) are running GTK2 exclusively on our ipaqs. |
| 10:26:44 | pb_ | I gather the gtk2 packages are going to ship with 0.6, but not be the default. |
| 10:26:44 | colin_werk | Good. |
| 10:26:44 | pb_ | Oh, and I made some locale packages (in unstable) so the translations are actually usable now. |
| 10:26:46 | colin_werk | Were you able to fix the transparency problem? |
| 10:26:46 | colin_werk | Ahhhh... |
| 10:26:46 | pb_ | The ipkg magic to install them automatically still needs some work, but you can do it by hand. |
| 10:27:06 | pb_ | Okay. I can probably do it soon enough. |
| 10:27:13 | colin_werk | Good. |
| 10:28:50 | pb_ | Oh, and we have a better dillo2 now, thanks to mallum. |
| 10:29:22 | pb_ | This one includes an image scaling thing too, per request from jg. It seems to work quite well. |
| 10:30:03 | mallum | pb_: yeah it does |
| 10:30:46 | mallum | pb_: at some point I'm gonna hack an html compressing proxy togeather so its cheaper to use over gprs |
| 10:32:05 | pb_ | mallum: cool idea. |
| 10:32:45 | mallum | pb_: Im going to demo the ipaq at this co I interviewed for in a bit |
| 10:33:02 | pb_ | oh right, excellent. |
| 10:34:17 | pb_ | I guess we could make dillo ask for the gzip transfer-encoding too, if it doesn't do that already. I dunno how many servers will actually honour that though. |
| 10:36:24 | mallum | pb_: I think quite a few do ... I know slashdot and freshmeat both do |
| 10:36:41 | pb_ | cool. |
| 10:43:23 | mallum | pb_: you know you've added pango markup to them calender entries right ? |
| 10:45:54 | pb_ | mallum: right |
| 10:46:18 | pb_ | mallum: well, only fairly minimal markup, it just uses <b> for the day labels |
| 10:50:06 | colin_werk | Ah, image scaling, nice. |
| 10:51:46 | colin_werk | The HIG police notes that the "Strewth!" dialog is not yet optimal. |
| 10:51:59 | pb_ | colin_werk: That is the en_AU locale at work. |
| 10:52:09 | colin_werk | :) |
| 10:53:44 | colin_werk | Maybe "Friday 11 Oct" would be better UI-wise than "Fri 11 October" -- I generally know what month I am in, so maybe all the text is a waste |
| 10:54:02 | pb_ | Mmm, that's a thought. |
| 10:54:28 | colin_werk | But the scap really looks nice. |
| 10:54:39 | pb_ | I think gpe-calendar is ---> <--- this close to being able to configure preferences with XSETTINGS, so the user will get the choice of that kind of thing. |
| 10:55:07 | colin_werk | Ah. |
| 10:55:10 | pb_ | Yeah, that's the new pango-based stuff that mallum mentioned a couple of minutes ago. |
| 10:55:23 | pb_ | Gpe-calendar2 0.24 (in unstable) has the new week view now. |
| 10:55:29 | colin_werk | "London Linux Expo" is two items, one of the untimed I guess? |
| 10:55:38 | colin_werk | one of them I mean |
| 10:55:52 | pb_ | It's one item that flows over two days. I need to figure out a way to indicate that the second one is continuing from the previous day. |
| 10:56:05 | colin_werk | Ah. |
| 10:56:45 | pb_ | Probably with a little arrow or other icon in the place where the time would be. |
| 10:56:58 | colin_werk | Maybe displaying three dots, or an icon, yes |
| 11:24:58 | moray | does the dillo2 in unstable do scaling, and if so how do you activate this? |
| 11:32:06 | pb_ | moray: yeah, but only in a fairly primitive way. |
| 11:32:13 | pb_ | you put "halfsize_images=YES" in your .dillorc. |
| 11:32:37 | pb_ | it can also scale the fonts, via "font_factor=0.5" or whatever (default is 0.8 I think) |
| 11:33:15 | moray | since I can't find the documentation, do you know how I can get it to use a proxy? |
| 11:33:22 | pb_ | The pigeon patch has a rather more sophisticated scaling thing, where you get a control in the toolbar that scales the entire page. But that patch seemed a bit scary so I haven't applied it. |
| 11:33:25 | pb_ | Uh, lemme check. |
| 11:33:49 | pb_ | Put "http_proxy=http://localhost:8080/" in your dillorc |
| 11:34:07 | moray | ah, thanks |
| 12:05:03 | nils | pb_: Are you online and reading? |
| 12:56:31 | pb_ | nils: sorry, was away from the keyboard for a while. |
| 12:56:35 | pb_ | I'm back now. |
| 13:07:11 | pb_ | hi mccarthy |
| 13:07:13 | mccarthy | pb_: my gpe-login looks pretty now.... :) |
| 13:07:20 | pb_ | cool :-) |
| 13:07:25 | mccarthy | I like the new look on the week view too |
| 13:07:46 | pb_ | mccarthy: yeah, I think it looks nicer. |
| 13:07:52 | mccarthy | (I would love to help you put pop-ups on it and the year-view, but I am pretty bad at that) |
| 13:08:02 | pb_ | :-/ |
| 13:08:45 | mccarthy | I am OK at *debugging* popups (I helped with the month-view), but I am bad at the "drawing area" stuff :( |
| 13:09:10 | mccarthy | nils is here?! wow, hi! |
| 13:09:34 | mccarthy | nils: can we officially move to gtk2 now? its prettier ;) |
| 13:10:25 | mccarthy | pb_: are you going to fix ownerinfo too? |
| 13:10:37 | pb_ | mccarthy: yeah, unless you would like to have a go. |
| 13:11:01 | pb_ | mccarthy: btw, zecke was just asking me to build a new ipk of sqlite 2.7.x. Would you prefer to take care of that yourself? |
| 13:11:07 | mccarthy | pb_: I havent even looked at what you did.... |
| 13:11:34 | mccarthy | pb_: were you the one that made the libsqlite package? |
| 13:12:37 | mccarthy | pb_: and would moving to 2.7 be backwards compatible? |
| 13:12:53 | pb_ | mccarthy: yeah, I did the libsqlite, but that was just an ipkg-unbuild; mv; rm; ipkg-build type job. |
| 13:13:20 | mccarthy | pb_: I got errors upgrading :( |
| 13:13:24 | pb_ | :-/ |
| 13:14:13 | mccarthy | pb_: I didnt have it installed previously (of course) and it knew to upgrade to it (which is good), but it didnt seem to be *replacing* sqlite since it yelled about overwriting the libsqlite.whatewver file |
| 13:14:41 | nils | mccarthy: Back from lunch ;) |
| 13:14:53 | pb_ | mccarthy: oh, yeah, that. |
| 13:15:02 | mccarthy | nils: hi |
| 13:15:13 | pb_ | mccarthy: I need to get advice from Jamey on what to do there. If you uninstall the old sqlite first it will work. |
| 13:15:16 | nils | mccarthy: I don't think that moving to GTK2 should be done _now_. It is still too big to fit into 16MB flash (IMHO). |
| 13:15:35 | mccarthy | pb_: but you have to -force-depends :( |
| 13:16:02 | nils | mccarthy: We once decided that we first want a stable GTK1.2 based GPE release and then go for GTK2. |
| 13:16:13 | mccarthy | nils: it fits about the same here as gtk1.2 did. maybe jg made X smaller? |
| 13:16:47 | mccarthy | nils: I just thought since the first release was taking so long.... |
| 13:17:55 | nils | mccarthy: Yes, and to make the time not even longer we should stick with GTK1.2. But if the transition is that easy... I really don't know. What do others say? I have no experience with GTK2! |
| 13:18:49 | mccarthy | nils: I think about 90% is already ported (for the most part you simply need to change the Makefile and recompile).... |
| 13:19:23 | mccarthy | I changed many of them myself and there is only 1 or 2 small things that may/may not crop up, but they are not hard... |
| 13:19:32 | mccarthy | (all of our apps are pretty simple after all) |
| 13:20:13 | nils | mccarthy: Yes, but what about space consumption? I had the impression that all parts that make up GTK2 sum up to be more than GTK1.2 was. You also have to count in pango and friends... |
| 13:20:41 | mccarthy | pb_: should we use a "provides/replaces/suggests" in the libsqlite control file? |
| 13:20:49 | pb_ | mccarthy: yes, probably. |
| 13:20:57 | mccarthy | nils: Hmm.... I dont know what is the best way to check yet.... |
| 13:20:59 | pb_ | mccarthy: maybe just "Replaces". |
| 13:21:33 | mccarthy | pb_: I thought "suggests" too for the command line util in case anyone actually used to use it and were wondering where it went... |
| 13:21:38 | nils | pb_: what do you say about this GTK1.2/GTK2 "dispute"? Which is the way you would prefer? |
| 13:21:40 | pb_ | nils: see the mail I just sent you: the gtk2 packages do seem to fit in a 3600 series machine with a reasonable amount of free space left. |
| 13:21:44 | florian | hello! |
| 13:21:48 | ole | Hello |
| 13:22:07 | pb_ | nils: for familiar 0.6, we should stick with the gtk1 packages because they are better tested and there are still a few things that don't work quite right in gtk2. |
| 13:22:24 | pb_ | nils: however, I would be in favour of switching to gtk2 as soon as possible for any new development. |
| 13:23:40 | mccarthy | pb nils: is fam 0.6 supposed to be the first "release" of gpe? |
| 13:23:51 | pb_ | mccarthy: that's more or less what I thought, yeah. |
| 13:24:28 | nils | Is there any deadline already ahead for Fam0.6? |
| 13:24:55 | pb_ | nils: should be any day now, they were trying to finalise a release at CRL over the weekend. |
| 13:25:19 | pb_ | nils: nelson made an 0.6rc1-17 yesterday. |
| 13:25:30 | mccarthy | pb nils: gtk2 1228700 + pango 200868 + libatk0 30956 + libgcc1 11191 |
| 13:25:48 | mccarthy | nils: care to do a ipkg info trace for gtk1.2 for me? |
| 13:25:55 | nils | Arg ... that's absolutely against what we anticipated :( |
| 13:26:25 | nils | mccarthy: Can do so, just have to hook up my iPaq, mompls ;) |
| 13:26:35 | mccarthy | nils pb: I guess we should also compare: libglib2 202847 |
| 13:29:07 | florian | just to think about: what about bringing gpe to a release (e.g. 1.0) with gtk 1.2 and switching to 2 then? |
| 13:30:49 | nils | OK, here is GTK1.2: libgtk1.2 495703 libglib1.2 70568 |
| 13:31:23 | nils | It is just about one third in size! |
| 13:31:54 | moray | mccarthy: are the actual sizes that material, given that a GTK2 setup with lots of GPE apps does fit fine onto e.g. my H3660 (with several megabytes of free space) |
| 13:32:14 | mccarthy | hmmm, it doesnt *seem* to be so much bigger.... I wonder what I left out of my current install?! ;) |
| 13:32:16 | moray | mccarthy: and this is before our additional X server size reductions e.g. |
| 13:32:28 | pb_ | mccarthy: maybe the new packages compress better. |
| 13:32:54 | mccarthy | I think the changes in X happened simultaneously with my move to gtk2 |
| 13:33:00 | pb_ | mccarthy: I think quite a lot of the libgtk size increase is due to pixmaps, for example, which will probably compress more highly than code. |
| 13:33:19 | mccarthy | pb_: oh, maybe thqt too... |
| 13:34:00 | pb_ | but I agree with moray: if it fits, there doesn't seem much point in obsessing over exactly how large the files are. |
| 13:34:02 | mccarthy | either way, I can fit all my stuff with *some* room to spare (about 90% full, all loaded up with PIM, games, dillo etc) |
| 13:34:32 | mccarthy | (I even use Xboard with gnuchess which are both hogs and require extra libs!) |
| 13:34:46 | pb_ | yeah. |
| 13:35:12 | pb_ | btw, I filed a bug on libcrypto asking for it to be made smaller. |
| 13:35:29 | moray | nils: what are you thinking of as making a GPE version 1.0? removal of bugs, or addition of synchronisation etc., or completion of things that could be useful in the apps? |
| 13:36:34 | nils | moray: A 1.0 release should simply be a _stable_ release of what we have now. Most bugs fixed and made somewhat usable and easy to install. New features are really up to 2.0. |
| 13:36:47 | mccarthy | pb_: gpe-kbd, at least on my machine, is needed for the dock-app to work |
| 13:36:59 | pb_ | mccarthy: yeah |
| 13:37:22 | pb_ | mccarthy: I think gpe-kbd as we know it now is more or less obsolescent, which is why I haven't rushed to make a new package. |
| 13:37:22 | mccarthy | pb_: but you listed it in the useless cruft in your email |
| 13:37:27 | nils | Has anyone already made up ready-to-go /skiff cross-dev-packages for GTK2 and friends? |
| 13:37:46 | mccarthy | nils: yes, see the porting page |
| 13:37:56 | pb_ | mccarthy: I think it should be replaced by something that configures xkbd directly. |
| 13:38:04 | pb_ | mccarthy: ie. no need for the wrapper. |
| 13:39:51 | nils | mccarthy: OK then... I think I'll have a look at it too. Maybe I'll change my mind ;) |
| 13:43:20 | florian | has anyone started strip down gtk2 for gpe usage? |
| 13:43:43 | pb_ | florian: not really, it seems to fit OK as it is so people are concentrating their efforts in other areas. |
| 13:45:21 | moray | ooi, does anyone know whether it would be plausible to e.g. emulate the Palm synchronisation stuff? |
| 13:45:57 | florian | pb_: i don't really know ist is good to use gtk2 now, i'd prefer to use gtk1 for now. what are the opportuneties of gtk2? |
| 13:46:12 | pb_ | florian: antialiased text, better i18n, bug fixes, new widgets. |
| 13:46:18 | pb_ | florian: display migration |
| 13:46:43 | florian | pb_: speed? |
| 13:48:31 | pb_ | Massive speed. |
| 13:48:34 | moray | florian: the speed seems fine for me - have you had some problem? |
| 13:50:13 | nils | pb_: massive speed-up or -down ;) ? |
| 13:50:20 | florian | moray: no, not yet - i just want to find out if it is really usable. |
| 13:50:24 | mccarthy | pb_: you think the speed it *better* with gtk2? |
| 13:50:48 | mccarthy | I thought it was kindof slow... |
| 13:51:24 | pb_ | mccarthy: no, I was just being flippant. I think it's about the same actually. |
| 13:57:20 | florian | hmm okay, sounds fine. i'm not used to gtk2 - how much work is it to port the gtk1 applications to gtk2? |
| 13:58:02 | pb_ | florian: usually extremely straightforward, in many cases just a case of recompiling with different flags. |
| 14:00:46 | florian | pb_: maybe i should try this before asking more questions ;-) |
| 14:01:04 | pb_ | heh |
| 14:16:09 | nils | On which ports page am I supposed to find the /skiff cross-dev-pkgs? |
| 14:18:13 | nils | mccarthy: The WiKi page Porting Projects does not have the GTK2 cross-dev packages... |
| 14:19:54 | | rumour has it cross compiler is see toolchain |
| 14:19:54 | pb_ | ibot, cross compiler? |
| 14:19:59 | | toolchain is at http://handhelds.org/z/wiki/Toolchains |
| 14:19:59 | pb_ | ibot, toolchain? |
| 14:20:02 | pb_ | nils: try there |
| 14:31:24 | nils | Err... nice try, but I was looking just for the GTK2 balls, not the complete toolchain ;) |
| 14:32:18 | pb_ | Ah. |
| 14:38:58 | nils | pb_: How do you compile for the iPaq? Native? If not, where do you get your gtk2-cross-dev files from? |
| 14:39:45 | pb_ | nils: yes, native. |
| 14:40:07 | nils | pb_: Err... on iPaq? Really? :) |
| 14:40:35 | pb_ | nils: not on an actual ipaq, but on another arm machine. |
| 14:40:58 | nils | If I start now, what should I get then? GTK2.0 or GTK2.1? |
| 14:41:25 | pb_ | 2.1 probably. |
| 14:41:45 | nils | Oh well ... wish me luck ;) |
| 14:41:48 | pb_ | Good luck :-) |
| 14:46:38 | nils | pb_: BTW: what kind of ARM is it you are using? CATS board? Netwinder? |
| 14:48:40 | nils | pb_: 2.1 is no good idea IMHO. All the other packages like glib, atk, ... are missong in the 2.1 tree... |
| 14:49:28 | pb_ | nils: CATS |
| 14:49:35 | pb_ | nils: use the 2.0 versions |
| 14:49:58 | pb_ | nils: you just need to update Pango to 1.1.1 for gtk 2.1 |
| 14:54:30 | pb_ | nils: I imagine gtk 2.0 would do just as well for your purposes. 2.1 is what is in Familiar unstable though. |
| 15:02:38 | nils | pb_: Ah cool, we also have a CAT(S) here ;) I'll first compile 2.0 there and mayb update to 2.1 later. |
| 15:02:55 | pb_ | nils: right, seems like a good plan, |
| 15:11:39 | mccarthy | nils: oops. I misread your question adn thought you were going native from the start... :( |
| 15:11:48 | mccarthy | sorry for the goose-chase |
| 15:15:07 | mccarthy | nils: no need to rip on ipaq-sleep, jeez. just submit a bug if something doesnt work for you, man! |
| 15:15:16 | mccarthy | (BTW, it works fine here) |
| 15:15:46 | nils | mccarthy: It's OK. As a result I might come up with what I wanted first: Cross-dev .tar.gz packages for /skiff/ environment ;) |
| 15:15:59 | nils | mccarthy: Was not meant offensively, sorry! |
| 15:16:27 | nils | mccarthy: As matter of fact I created ipaq-screensaver long before ipaq-sleep ;) |
| 15:17:49 | nils | mccarthy: If I get time I'll see what I can patch. One thing for example does not work for me: The dimming does not dim the display but switches the backlight off completely. |
| 15:18:54 | mccarthy | nils: the light daemon would "dim" the display... but nobody seemed to care if there was alight daemon or not... so I just turn off the backlight.... |
| 15:20:19 | pb_ | nils: there never seemed much point in trying to control the backlight to any finer resolution than on and off. |
| 15:20:44 | nils | mccarthy: Hmm... I like that function very much. And there is point: Less light, less power consumption! |
| 15:21:40 | pb_ | Right, but you can set the backlight level manually at a brightness you are comfortable with. I don't really see where ipaq-sleep needs to get involved. |
| 15:23:35 | moray | presumably it also saves the backlight's life to turn it off completely? |
| 15:23:37 | nils | pb mccarthy: The lowest light level is very low. In case the user wants more (typically about 50%) the light level would be dimmed to minimum at idle and switched back to user-level when not-idle. This is alomost no cost for the daemon but a significant gain in power consumption. |
| 15:23:40 | mccarthy | pb_: how does one "blank" the screen? should ipaq-sleep do this as well? |
| 15:24:14 | mccarthy | nils: what is the advantage of dimming while idle over shutting off when idle? |
| 15:24:18 | mccarthy | (the backlight that is) |
| 15:24:34 | nils | mccarthy: I had this in ipaq-screensaver too. You will have to notify the user somehow that the iPaq is not sleeping but rather "blind". |
| 15:24:52 | pb_ | mccarthy: mmm, dunno. I used to find it confusing that the ipaq would blank the screen but still actually be running. On several occasions I thought it was suspended when it wasn't, and the battery ran flat. |
| 15:24:52 | nils | mccarthy: You still at least see a little of what is going on? |
| 15:25:05 | mccarthy | nils: yes |
| 15:25:16 | moray | nils: hm, surely this is most likely to happen when people aren't using it though? |
| 15:25:21 | mccarthy | nils: it is just dark, and it jumps back to life when you touch anything... |
| 15:26:22 | Zindar | well.. got to go.. bye! :) |
| 15:26:28 | moray | pb_: I think if you're not on AC you certainly want to suspend fully, yes |
| 15:26:51 | mccarthy | pb_: is there any advantage to a "screen saver" then? (draw silly things, but not keep it all the way black)? |
| 15:26:52 | nils | mccarthy: I see this function as power-saver. When you idle, for example you hold the iPaq in hand but talk to someone, why not drop power soncumption in different levels depending on idle time? First a little, then more and then off alltogether? When you at least still see a little for example you can still see where you should NOT tap on the screen to not start any action... |
| 15:27:34 | nils | mccarthy: Screen-saver was just a name... it does none of those eye-candy things! |
| 15:28:02 | moray | does anyone else see a bug where xkbd on the unsuspending locked screen appears as only a horizontal line (and therefore isn't useful)? |
| 15:28:04 | mccarthy | nils pb: but turning the screen "off" gets confused with suspending... |
| 15:28:43 | nils | mccarthy: In ipaq-screensaver I let the LED blink to show that. |
| 15:28:59 | moray | nils: I suppose that might be helpful if the first stage (dim screen) happened very quickly, then suspending took a few minutes |
| 15:29:06 | mccarthy | nils: hmm, thats not a bad idea... |
| 15:29:18 | nils | mccarthy: After a next while you could also auto-suspend. All levels could be configurable so users could choose what they want or not. |
| 15:29:51 | mccarthy | nils: but I still dont see much use to a dim screen versus an off backlight (I dont see much *difference* in the screen either, they are both quite dim that way) |
| 15:30:01 | pb_ | mccarthy: if ipaq-sleep blanked the screen and did a full suspend a minute or so later, that might be cool. |
| 15:30:20 | pb_ | mccarthy: turning off the LCD will save some power, though in relative terms it doesn't make nearly as much difference as turning off the backlight. |
| 15:30:32 | mccarthy | pb_: I like nils' blinking LED to warn you that it is "thinking" about suspending, but not quite suspended |
| 15:30:35 | nils | mccarthy: If it is quite dark around, a dimmed screen is still readable, a dark one not. That's the difference. |
| 15:30:39 | pb_ | mccarthy: yes. |
| 15:30:44 | moray | nils: though it's better to have good default behaviour than rely on people configuring it / than confuse everyone by a million configuration options... ;) |
| 15:30:58 | pb_ | mccarthy: btw, another (minor) advantage to blanking the screen is that there is more cpu bandwidth available when the LCD is not running. |
| 15:31:25 | bal00 | hi all |
| 15:31:38 | bal00 | can someone give me a hint about this error?? |
| 15:31:40 | mccarthy | pb_: I guess, but I mostly meant "do you get 'burn-in' with LCDs" (so you should stop displaying the same thing...) |
| 15:31:40 | bal00 | arm-linux-ld: libqpe.so.1.5.0: Not enough room for program headers (allocated 6, need 7) |
| 15:31:40 | bal00 | arm-linux-ld: final link failed: Bad value |
| 15:31:43 | moray | pb_: hm, can we sort things out so that if someone's running the media player, the screen blanks but it *doesn't* suspend (until low battery)? |
| 15:31:55 | mccarthy | ba100: nope try opie people |
| 15:31:59 | mccarthy | this is Gpe |
| 15:32:04 | mccarthy | not Qpe |
| 15:32:26 | mccarthy | moray: if there is sound playing it shouldnt suspend |
| 15:32:30 | nils | mccarthy pb: I would like a multi-step sleep: 1. Full display, user set light level -> 2. dimmed backlight to level 1 -> 3. backlight off -> 4. screen off with user notification -> 5. suspend. |
| 15:32:30 | bal00 | it is a problem on the crosscompiler not on the code |
| 15:32:40 | pb_ | mccarthy: I'm not sure. I've never heard of burn-in on an LCD. |
| 15:33:22 | mccarthy | nils: that is fine... you wanna hack on ipaq-sleep? |
| 15:33:23 | nils | mccarthy: AFAIK LCDs do not wear out. |
| 15:33:44 | nils | mccarthy: As I said, when I have some time ;) |
| 15:33:53 | mccarthy | nils: bah, who has time |
| 15:33:57 | moray | mccarthy: (that's what I meant, yes) |
| 15:34:14 | nils | mccarthy: BTW: The display off is a fb ioctl()... |
| 15:34:17 | mccarthy | moray: no, I mean it *does* that already! |
| 15:34:39 | moray | mccarthy: ah, ok - due to a lack of storage on my ipaq, I've never tried ;) |
| 15:34:50 | pb_ | nils: yeah, your multi-step thing seems reasonable. |
| 15:34:54 | moray | mccarthy: (I've only played audio via NFS etc.) |
| 15:35:14 | mccarthy | moray: and it wont suspend while you type on the console or a whole host of other thingies by default... |
| 15:35:24 | moray | sounds good |
| 15:35:25 | mccarthy | you can, of course, set whatever IRQs you want to poll by hand too |
| 15:35:36 | pb_ | bal00: are you using some wacky custom linker script? |
| 15:35:42 | mccarthy | moray: what machine do you use? |
| 15:35:59 | moray | mccarthy: my ipaq's an H3660 |
| 15:36:20 | mccarthy | moray: ok. I just wanted to see if it worked on others... |
| 15:36:29 | bal00 | pb_: no, the gcc command ended with a seg fault |
| 15:36:31 | mccarthy | mine is a 3670 so yours should work exactly the same |
| 15:36:47 | bal00 | so I tryed doing the ld by hand with this result |
| 15:37:26 | pb_ | sounds like your toolchain is broken. |
| 15:37:41 | bal00 | pb_: where can I get a working toolschain from? |
| 15:37:56 | pb_ | bal00: try the monmotha one on handhelds.org |
| 15:38:13 | bal00 | ok, thanks |
| 15:39:38 | nils | is away: ...coming back soon... |
| 15:41:42 | pb_ | yo mallum |
| 15:41:57 | mallum | hey pb_ |
| 15:42:10 | mallum | pb_: Just got back from my travels |
| 15:42:29 | mallum | pb_: they were impressed with the ipaq and are borrowing it to demo to some client tommorow |
| 15:44:01 | pb_ | mallum: cool! |
| 15:46:49 | mallum | pb_: whats even better is I noticed a couple of old macs in there office and I think they'll give me one :) |
| 15:49:02 | pb_ | heh :-) |
| 15:49:14 | pb_ | mallum: what kind of old macs? |
| 15:49:29 | mallum | pb_: they had a 9500/150 and a 8600/200 |
| 15:49:43 | mallum | pb_: both powerpc based and debian supported :) |
| 15:49:51 | pb_ | nice. |
| 15:51:09 | mccarthy | pb_: why is libsqlite0 an ar package rather than a tar? |
| 15:51:27 | pb_ | mccarthy: I guess I must have upgraded my tools while I wasn't looking. |
| 15:51:47 | pb_ | mccarthy: tar is so 2001, ar is the future. |
| 15:56:56 | mallum | pb_: [OT] do you find maildir faster than mbox ? |
| 15:57:08 | pb_ | mallum: yes. |
| 15:57:21 | pb_ | mallum: but I suspect it depends on your usage patterns. |
| 15:57:51 | pb_ | (not that I've actually used mbox for several years: before I switched to maildir I was on MH for a long time) |
| 15:59:04 | mallum | pb_: well I archive my mails in monthly 'boxs' , when it gets near the end of the month the various folders get quite large and mutt can take a little while to open them ... this new machine is only 200Mhz ( though scsi ), so Im just wondering how much of a dog it will be ... |
| 16:00:55 | pb_ | mallum: I think our company mail server is only 200MHz too (and ide!) and it seems to run at a bearable speed. |
| 16:01:46 | pb_ | mallum: it can take a while to open large mailboxes (>20000 messages) but for most stuff it seems to be fine. It's hard to say how much of the overhead is due to imap and how much to maildir. |
| 16:05:40 | mccarthy | pb_: what was the purposed of moving to sqlite 2.7.2? |
| 16:05:53 | mccarthy | (I haqve built ipks and am testing them...) |
| 16:06:01 | pb_ | mccarthy: I'm not sure exactly. Ask zecke. |
| 16:06:16 | mccarthy | I dont know who zecke is... |
| 16:06:20 | pb_ | mccarthy: It sounded like part of it was that he wanted better performance. He was asking about different CFLAGS as well. |
| 16:06:38 | pb_ | mccarthy: /join #handhelds.org and address yourself to zecke. |
| 16:06:59 | pb_ | mccarthy: he is one of the opie.de crew. |
| 16:07:38 | mallum | pb_: Im sort of assuming this machine will have a lot of ram too, so I maybe be able to figure out some sort of evil running the maildir on a ram disk kind of thing |
| 16:08:04 | pb_ | mallum: if you have loads of ram, most of your maildir will end up sitting in the buffer cache anyway. |
| 16:11:38 | mallum | pb_: okey |
| 16:26:11 | pb_ | mccarthy: so, what was the conclusion about ipaq-sleep? Are you going to make it work like Nils suggested? |
| 16:29:54 | pb_ | mccarthy: btw, I asked mibus about libredalert. Apparently he is waiting for nelson to fix some bug in atd. |
| 16:47:58 | nils | is back (gone 01:08:20) |
| 16:48:57 | nils | pb: Err... I think I was supposed to fix it and did so last monday. Patches for atd and the kernel rtc driver are on their ways. And I even made a new patch for atd, as it had a nasty security 'feature' ;) |
| 17:03:13 | mallum | hey dc_ |
| 17:03:36 | mallum | dc_: I was gonna ask you something .... trying to rememeber ... |
| 17:04:44 | dc_ | :) |
| 17:04:51 | dc_ | mallum: hehe |
| 17:05:04 | mallum | ah yes, how did you get that cute little gui control for mplayer ? |
| 17:05:22 | mallum | dc_: when I run with '-gui' I get this ugly green thing which dont even work :( |
| 17:05:33 | moray | themeable, isn't it? |
| 17:08:44 | dctanner_ | heh |
| 17:09:16 | dctanner_ | mallum: you errr ./configure --enable-gui then start with mplayer -gui |
| 17:09:32 | dctanner_ | mallum: but you also need to download a theme from the website |
| 17:09:59 | dctanner_ | downloads winex |
| 17:10:23 | mallum | dctanner_: okey what mplayer version you using BTW ? |
| 17:10:41 | dctanner_ | MPlayer 0.90pre6-2.95.4 |
| 17:10:50 | dctanner_ | heh |
| 17:10:50 | dctanner_ | from source I think |
| 17:10:50 | dctanner_ | ys |
| 17:10:54 | dctanner_ | yes |
| 17:12:28 | mallum | dctanner_: yah same as me ... pre8 seems to be available .. |
| 17:12:40 | dctanner_ | ooh. |
| 17:12:50 | dctanner_ | probly try it out someitme tomorrow then |
| 17:15:31 | mallum | pb_: hey, what do you think 'colors per channel' could mean ( not meaning to sound stupid ) ? |
| 17:15:37 | dctanner_ | ooooh |
| 17:15:40 | dctanner_ | shit... |
| 17:15:43 | pb_ | mallum: I have no idea, what's the context? |
| 17:15:47 | pb_ | dctanner_: yo |
| 17:15:49 | dctanner_ | rm's all his old website backups |
| 17:15:57 | dctanner_ | kicks rm |
| 17:15:58 | dctanner_ | bah |
| 17:16:01 | dctanner_ | pb_: hey |
| 17:16:31 | dctanner_ | oh well, lets hope I don't need them anytime soon! |
| 17:16:38 | pb_ | ha. |
| 17:19:21 | pb_ | ah, moray has made some new screenshots, cool. |
| 17:19:28 | moray | though I'm now confused |
| 17:19:34 | moray | try loading http://gpe.handhelds.org/images/gpe-contacts2.png |
| 17:19:53 | moray | I get a bizarre error - "cannot be displayed, because it contains errors" |
| 17:19:58 | dctanner_ | brb food |
| 17:20:00 | pb_ | moray: you should include the 1337 new week view in gpe-calendar too |
| 17:20:09 | mallum | pb_: this blackbox color code, its very difficult for me to work out the context myself :( |
| 17:20:17 | mallum | pb_: all evil C++ isms |
| 17:20:18 | pb_ | mallum: weird |
| 17:20:29 | pb_ | mallum: if you give me a pointer to the code I'll have a look at it |
| 17:20:33 | moray | pb_: probably - though I've been trying to keep the screenshots page relatively small ;) |
| 17:20:50 | pb_ | moray: also, the gpe-conf picture with the ipaq in seems to be chopped off about 60% of the way down for me. |
| 17:21:10 | moray | pb_: yes, that's not mine - and gpe-conf won't currently run on my ipaq so I can't easily replace it :* |
| 17:21:10 | pb_ | moray: gpe-contacts2 loads fine for me with galeon. |
| 17:21:13 | mallum | pb_: its image.cc in the blackbox source, Im slowly porting it to c ... I have 16 & 24 bit working ok, probably 8bit but greyscale is going a little wierd |
| 17:21:19 | moray | pb_: it's started working for me now |
| 17:21:30 | pb_ | weird |
| 17:21:51 | pb_ | moray: maybe have just one shot of appmgr? |
| 17:22:27 | moray | pb_: yes, as much as it's great, people can probably get the idea from one image... |
| 17:24:13 | pb_ | Indeed. |
| 17:37:43 | moray | if anyone has gpe-mail installed/easily runnable, updated screenshots would be good for that too |
| 17:38:17 | pb_ | oh, you managed to get a shot of gpe-filemanager? cool. |
| 17:38:30 | pb_ | it always consumed all memory and crashed when I tried. :-} |
| 17:38:41 | moray | oh dear... |
| 17:38:53 | moray | yes, it's (apparently) working here |
| 17:39:07 | pb_ | well, to be fair I was running it on a rather large directory (over nfs) |
| 17:39:20 | pb_ | I wanted one with some interesting filetypes in. |
| 17:39:31 | moray | yes, that would be more interesting |
| 17:39:56 | moray | I thought the current one at least would look interesting to Unix geeks etc. |
| 17:40:02 | pb_ | Yeah. |
| 17:40:28 | moray | of course, once things are more usable/stable, we'll want to hide screenshots like that so as not to scare off users ;) |
| 17:40:47 | pb_ | :-) |
| 17:41:29 | pb_ | Ah, your calendar-week is the old week-view. |
| 17:41:42 | pb_ | The new one (on scap I think) is a bit prettier. |
| 17:42:39 | moray | pb_: feel free to update any of them if you have better versions |
| 17:42:54 | pb_ | ok. |
| 17:43:38 | moray | I'm assuming we may as well use the screenshots to show off new features - CVS versions are fine, though presumably it would be stretching things to have too many screenshots around of nils' non-CVS code ;) |
| 17:43:56 | pb_ | Right, yeah. |
| 17:44:15 | pb_ | I think most of the things in the screenshots are at least in unstable, if not 0.6. |
| 17:45:32 | moray | yes |
| 17:45:36 | nils | What non-CVS of mine? |
| 17:46:49 | moray | nils: I was just seeing if you'd notice that really... - I'm sure at various points you've had versions of code that was further advanced beyond what was present in CVS? |
| 17:47:27 | moray | should go and get some food soon |
| 17:49:27 | dc_ | food :) |
| 17:49:33 | dc_ | yum |
| 17:50:26 | moray | dc_: that's what I thought, yes - I just need to summon enough energy to actually get round to eating first |
| 17:51:28 | nils | moray: Sometimes, yes, but currently I think everything should be latest... |
| 17:59:52 | nils | Hmm... while Debian upgrade: "E: Internal Error, Couldn't configure a pre-depend" What is this? |
| 18:00:54 | dc_ | hum |
| 18:01:12 | dc_ | man I really don't want to have to do maths A level :{ |
| 18:02:54 | pb_ | dc_: ha |
| 18:03:08 | pb_ | dc_: maths A-level is okay, I hear it's further maths you need to avoid |
| 18:03:20 | mallum | dc_: its all differentiation |
| 18:03:59 | pb_ | dc_: but steer clear of chemistry at all costs. |
| 18:04:33 | mallum | dc_: yeah pb is right, all that is, is remebering names for stuff |
| 18:04:52 | mallum | dc_: I got a 'U' for A-level chemistry :) |
| 18:05:09 | dc_ | heh |
| 18:05:10 | dc_ | mallum: :) |
| 18:05:13 | dc_ | mallum: well done |
| 18:05:27 | pb_ | mallum: oh cool :-) |
| 18:05:31 | mallum | dc_: thanks, I didn't revise - only need two A-levels for uni |
| 18:06:01 | dc_ | yeah, well I'm not good at chemistry in the slightest way either |
| 18:06:18 | pb_ | mallum: when I went to my university interview, the first question they asked was about organic chemistry. I thought all was lost. |
| 18:06:19 | mallum | dc_: we made moisteriser once |
| 18:06:36 | dc_ | a wha? |
| 18:06:57 | mallum | dc_: I just used to like heating up was, till it got real hot, and then knocked it over the desk like napalm :) |
| 18:07:36 | dc_ | :D |
| 18:07:36 | mallum | wax I mean |
| 18:07:39 | dc_ | cool. |
| 18:08:03 | dc_ | I can do biolidy of cells as my natural science... |
| 18:08:16 | mallum | pooring meths in the bin and waiting for someone to drop a match in was fun too |
| 18:08:21 | dc_ | it seems you need to do one asweel as computer science. |
| 18:08:57 | mallum | dc_: sid says home economics or textiles :) |
| 18:09:08 | dc_ | heh |
| 18:09:18 | dc_ | ;p |
| 18:09:36 | mallum | dc_: general studies is always a guarenteed pass |
| 18:09:49 | pb_ | indeed. |
| 18:10:32 | mallum | dc_: seriously though a good teacher makes a difference to any subject, especially at A-level |
| 18:10:36 | pb_ | dc_: do you know what you want to do at university, or have you not planned that far ahead? |
| 18:10:53 | dc_ | pb_: computer science, most likely |
| 18:10:57 | dc_ | mallum: yeah |
| 18:11:36 | pb_ | dc_: ah cool. |
| 18:12:10 | dc_ | I'm just wondering if I need mathmatics A level to take that |
| 18:12:15 | mallum | pb_: did you get a chnance to poke around the blackbox code ? |
| 18:12:23 | pb_ | dc_: yeah, maths is pretty much essential |
| 18:12:26 | mallum | dc_: I dont think A-level maths is that bad |
| 18:12:43 | dc_ | you think.... |
| 18:12:50 | pb_ | dc_: definitely more important to have maths than computer science A-level |
| 18:13:00 | dc_ | oh well |
| 18:13:04 | dc_ | not at A level. |
| 18:13:12 | dc_ | I want to take compuer science at uni |
| 18:13:14 | dc_ | at a level |
| 18:13:41 | dc_ | well |
| 18:13:43 | dc_ | I'm not sure. |
| 18:13:51 | mallum | dc_: I cant believe your bad at at maths ... |
| 18:13:56 | pb_ | dc_: right, I meant that if you want to do computer science at uni, they are much more likely to want you to have done maths a-level than computer science. |
| 18:14:04 | dc_ | right. |
| 18:14:22 | mallum | dc_: if I can get a maths A-level, Im sure you can :) |
| 18:14:30 | dc_ | :) |
| 18:15:02 | dc_ | maby I'll take biology, maths and film production at a level then :} |
| 18:15:08 | dc_ | or media studies |
| 18:15:11 | pb_ | sounds cool. |
| 18:15:25 | dc_ | that'll atleast be a tad interesting. |
| 18:15:29 | mallum | yeah. taking something less scientific is what I should have done ... |
| 18:15:35 | pb_ | indeed. |
| 18:15:37 | dc_ | hehe. |
| 18:15:58 | dc_ | well, I'll probly take ict and ict comm A level aswell |
| 18:16:01 | pb_ | I wish I'd done a language or arts A-level instead of chemistry. |
| 18:16:06 | dc_ | but they're a push-over. |
| 18:16:10 | pb_ | dc_: what on earth is ict? |
| 18:16:28 | dc_ | the computer courses. |
| 18:16:40 | dc_ | infomation communication technology |
| 18:16:42 | pb_ | ah. |
| 18:16:54 | pb_ | obviously I'm not up with all the new tlas. |
| 18:17:06 | dc_ | they'll all VB and silly hardware... |
| 18:17:12 | dc_ | and lots of lame network diagrams |
| 18:17:19 | moray | they have to rename them frequently once everyone realises they're useless |
| 18:17:25 | dc_ | yeah |
| 18:17:51 | dc_ | but I'm not sure, since they're still pretty useless for me. |
| 18:18:44 | pb_ | hmm |
| 18:18:48 | pb_ | trying to understand ipkg code |
| 18:18:50 | pb_ | :-/ |
| 18:21:26 | moray | 63630 packets transmitted, 11 packets received, 99% packet loss |
| 18:21:31 | pb_ | ha. |
| 18:21:32 | moray | that's not good, is it? |
| 18:21:35 | pb_ | not best. |
| 18:22:31 | moray | and strangely those 11 were between icmp_seq=22569 and icmp_seq=22855 (not at the start or whatever) |
| 18:22:41 | moray | this is to a machine of mine on the wireless network in London |
| 18:22:51 | moray | which appears not to be communicating optimally |
| 18:23:23 | moray | although another machine just next to it is reaching the AP just fine |
| 18:23:52 | moray | (well, or mostly fine - still some packet loss) |
| 18:32:43 | cmarqu_ | moray: did you cvs add http://gpe.handhelds.org/images/gpe-contacts2.png with the -kb option? |
| 18:33:47 | moray | I suspect I forgot, since it might explain the problems - can I persuade cvs to change its opinion on a file? |
| 18:34:05 | cmarqu_ | moray: pb_ says you can do that with cvs admin |
| 18:34:38 | pb_ | I do. |
| 18:34:48 | pb_ | cvs admin -kb xx.png |
| 18:34:55 | moray | pb_: yup, just did that |
| 18:34:59 | cmarqu_ | http://gpe.handhelds.org/images/gpe-conf-buttons.png is fixed now. |
| 18:35:08 | cmarqu_ | Dunno why it didn't get in fully. |
| 18:37:03 | cmarqu_ | mallum: BTW, Roman from art.gnome.org is usually in #gnome, but as "star" |
| 18:57:14 | mallum | cmarqu_: cool, he dont seem to be there at the mo .. |
| 19:05:30 | pb_ | cmarqu_: I think I have a patch now to make ipkg install locales automatically. |
| 19:06:13 | cmarqu_ | Ahhh. Nice. |
| 19:07:12 | pb_ | cmarqu_: The way it works is that you install locale-base-CC packages by hand, to select the set of locales you are interested in. Then, when you install something like gtk, ipkg should automatically select the corresponding set of gtk-locale-CC packages to go with them. |
| 19:07:46 | pb_ | So, we need to start actually shipping some translations with the GPE applications. |
| 19:08:45 | cmarqu_ | Indeed. |
| 19:09:26 | cmarqu_ | Does this package building stuff work? |
| 19:09:35 | pb_ | cmarqu_: I don't know. |
| 19:10:39 | cmarqu_ | BTW, do you have pix from the Linux Expo somewhere? |
| 19:11:08 | pb_ | No. I forgot to take my camera, and I didn't notice dc_ or mallum toting one either. |
| 19:11:21 | cmarqu_ | Hell. |
| 19:11:32 | pb_ | I didn't notice any chicks who could have been our mascot, though we did see a giraffe. |
| 19:11:39 | cmarqu_ | :) |
| 19:13:03 | nils | BTW: Anyone want to meet at Linux-Expo in Frankfurt end of October? |
| 19:14:31 | cmarqu_ | Not me... my car broke, so I guess I'll have to save as much money as I can |
| 19:14:58 | pb_ | Hell. |
| 19:15:27 | cmarqu_ | If I'm unlucky, I'll have to get a new engine. Or dump it :/ |
| 19:15:36 | pb_ | cmarqu_: What's wrong with it? |
| 19:16:16 | cmarqu_ | They are still evaluating whether the valves hit the, uhm, pistons? |
| 19:16:33 | pb_ | Ah, nasty. |
| 19:16:38 | pb_ | Timing belt broke, or something like that? |
| 19:16:46 | nils | cmarqu_: It's sad but sometimes you have to let go :( Had the same with my old car some weeks ago... sniff... |
| 19:16:59 | cmarqu_ | This one belt thing apparently came loose when the water pump broke or something like that |
| 19:17:36 | pb_ | cmarqu_: Yeah, sounds likely. :-( |
| 19:17:55 | cmarqu_ | I was not going too fast, but with 16 valves, I guess there just has to be some damage... |
| 19:18:23 | cmarqu_ | Actually, I need to call this friend of mine to aks him what my options are in the bad case |
| 19:18:30 | cmarqu_ | BBBIAB |
| 19:30:02 | jsmanrique | hi! |
| 19:31:01 | cmarqu_ | Friend said (part of) my engine might fall under goodwill |
| 19:31:03 | cmarqu_ | Hi jsmanrique |
| 19:31:23 | jsmanrique | mallum: has HP answered you? |
| 19:31:27 | pb_ | cmarqu_: oh, that would be good. How old is the car? |
| 19:32:59 | cmarqu_ | pb_: a bit over 4 years. The timing belt is due at 120,000km or 5 years, while I only have 105,000km |
| 19:34:48 | mallum | jsmanrique: no |
| 19:35:08 | jsmanrique | mallum: :( |
| 19:35:11 | mallum | cmarqu_: I had a camrea , but never took no pictures :( |
| 19:36:13 | dc_ | pb_: there was a chick with that giraffe! |
| 19:36:17 | dc_ | :} |
| 19:36:22 | pb_ | dc_: oh yeah! |
| 19:36:38 | pb_ | dc_: damn, opportunity lost there. |
| 19:36:47 | dc_ | hehe |
| 19:36:57 | dc_ | M$ didn't have anything like that at their talk |
| 19:38:54 | mallum | dc_: I think you get the most dollybirds at computer games expo's - lara crofts etc |
| 19:39:16 | dc_ | yeah :D |
| 19:40:58 | dc_ | bbl |
| 20:11:14 | dc_ | heh |
| 20:11:20 | dc_ | damn winex sucks. |
| 20:11:24 | pb_ | :-/ |
| 20:11:28 | dc_ | it won't even run simtower :( |
| 20:17:42 | mallum | dc_: you have that snes binary BTW ? |
| 20:18:13 | dc_ | mallum: he emulator? |
| 20:18:15 | dc_ | the |
| 20:18:21 | mallum | dc_: yeah |
| 20:19:36 | dc_ | luckey you mentioned it :) |
| 20:20:14 | dc_ | email cool? |
| 20:20:28 | dc_ | I have snes9x for arm |
| 20:20:37 | dc_ | and the source it seems |
| 20:20:57 | dc_ | and a fw roms :) |
| 20:21:11 | mallum | dc_: yeah email fine |
| 20:26:06 | pb_ | dc_: cool, what roms do you have? :-) |
| 20:27:55 | mallum | pb_: I just mailed you a simple draft of how migration 'signaling' could work |
| 20:28:03 | pb_ | mallum: excellent |
| 20:34:36 | dc_ | arg |
| 20:34:40 | dc_ | vlc |
| 20:37:14 | dc_ | mallum: just comming |
| 20:42:16 | dc_ | I need sleep sooooon. |
| 20:44:11 | dc_ | hey mallum |
| 20:44:16 | dc_ | jg: :) |
| 20:44:25 | dc_ | damn xchat autocomplete |
| 20:44:41 | PaxAnima | dc_: why use xchat? |
| 20:44:45 | jg_ | dc_: hi... |
| 20:45:51 | dc_ | PaxAnima: it's fun |
| 20:46:08 | dc_ | jg_: how'd the dev weekend go then? |
| 20:46:48 | mallum | dc_: whats up ? |
| 20:47:04 | jg_ | dc_: went well, me thinks. |
| 20:47:10 | bdale | hello. having a problem with gpe apps on familiar 0.6-rc1I could use a hint with. "Unable to load icon "ipaq"" ... any idea what's missing or out of date that should be providing that? |
| 20:47:12 | dc_ | mallum: uploading via dialup, it's coming |
| 20:47:23 | dc_ | jg_: hehe :) |
| 20:48:14 | pb_ | dc_: we have a picture of jg at the developers weekend: http://handhelds.org/~pb/x-men.png |
| 20:48:36 | mallum | dc_: okey np |
| 20:48:54 | jg_ | can someone help out bdale? for those of you who don't know him, he's the debian project leader.... |
| 20:49:12 | pb_ | I was hoping dc_ could help him. |
| 20:49:34 | dc_ | hides... |
| 20:49:48 | dc_ | hehe |
| 20:49:54 | pb_ | dc_: ah, not one of your icons? |
| 20:50:02 | dc_ | pb_: is that image not in the newer packages. |
| 20:50:02 | dc_ | finds cmarqu_ |
| 20:50:10 | pb_ | dc_: apparently not :-/ |
| 20:50:19 | pb_ | dc_: let me invoke the power of cvs |
| 20:50:22 | dc_ | no...it's gpe-conf |
| 20:50:26 | cmarqu_ | Hmm. I find ./gpe-conf/pixmaps/ipaq.png |
| 20:50:28 | pb_ | ah. |
| 20:50:38 | dc_ | maby i needs a new ipk.......... |
| 20:50:40 | pb_ | so gpe-conf is just busted in 0.6 at the moment? |
| 20:50:46 | dc_ | yeah |
| 20:50:49 | pb_ | that was one of the things nelson updated yesterday :-} |
| 20:50:53 | cmarqu_ | What happened to gpe-conf recently? |
| 20:50:54 | bdale | I've been more or less routinely upgrading the familiar installation in-place on this ipaq... but I went about 3 weeks between updates and today's update left things mostly better but not quite working... :-) |
| 20:51:11 | pb_ | darn. |
| 20:51:14 | dc_ | bdale: so you should do an ipgrade. |
| 20:51:24 | jg_ | bdale: sounds like a missing dependency somewhere, somewhen.... |
| 20:51:26 | dc_ | oh... |
| 20:51:28 | dc_ | hum |
| 20:51:31 | pb_ | cmarqu_ / pierre, can you figure out what went on there? I can make a new package if need be. |
| 20:51:40 | cmarqu_ | Not really a dependency... |
| 20:51:42 | bdale | thinks he's up to the current 0.6-rc1 on everything right now |
| 20:51:46 | dc_ | jg_: I'll assume that's you on the left :) |
| 20:52:02 | bdale | looks like keith on the right? |
| 20:52:12 | dc_ | bdale: I'm assuming your using unstable? |
| 20:52:14 | dc_ | bdale: yeah |
| 20:52:20 | dc_ | hehe. |
| 20:52:27 | jg_ | I have a bit more hair than keithp; I'm the big guy. |
| 20:52:35 | bdale | dc_: well, I've had unstable in and out of my ipkg.conf ... right now, I think I only have the two 0.6-rc1 lines. |
| 20:52:43 | dc_ | jg_: how many people turned up then? |
| 20:52:56 | cmarqu_ | bdale: What does ipkg status gpe-conf say? |
| 20:52:57 | jg_ | dc_: at least 15; I don't have a final count. |
| 20:53:13 | bdale | cmarqu_: 0.0.5 |
| 20:53:16 | dc_ | jg_: cool |
| 20:53:23 | cmarqu_ | Okay. |
| 20:53:31 | dc_ | bdale: well running on unstable is always fun ;] |
| 20:53:37 | cmarqu_ | bdale: Do you have a file /usr/share/gpe/default/ipaq.png ? |
| 20:53:46 | bdale | dc_: right, I'm fearless about it... |
| 20:53:46 | cmarqu_ | http://ipkgfind.handhelds.org/details.phtml?package=gpe-conf says that it should be there. |
| 20:53:53 | bdale | cmarqu_: yes |
| 20:54:04 | cmarqu_ | Why is it in that location though... |
| 20:54:05 | bdale | -rwxr-xr-x 1 root root 41507 Oct 14 21:46 ipaq.png |
| 20:54:06 | pb_ | <jamey> bdale pb_: but it seems to be ok if I copy ipaq.png from /usr/share/gpe/default to /usr/share/pixmaps |
| 20:54:09 | pb_ | hmm |
| 20:54:28 | bdale | hrm. |
| 20:54:30 | pb_ | <jamey> oops, that doesn't help after all |
| 20:54:34 | pb_ | :-/ |
| 20:54:38 | dc_ | cmarqu_: it's in the wrong place. |
| 20:54:51 | bdale | is watching both channels (among others), fwiw |
| 20:55:04 | cmarqu_ | Who upladed that package? nelson? mccarthy? pierre? |
| 20:55:16 | pb_ | Lemme look. |
| 20:55:21 | dc_ | bdale: copy it to /usr/share/gpe/pixmaps/default/ |
| 20:55:35 | dc_ | I'm guessing that should fix it. |
| 20:56:10 | dc_ | as the error would be coming from the libgpewidget icon loader that is looking in that directory pb_, no? |
| 20:56:11 | bdale | yep, that fixed it |
| 20:56:17 | pb_ | dc_: right |
| 20:56:22 | cmarqu_ | Hmm. Many of the single applets somehow pull in that pic, even if they don't use it. Kinda bad dependency there too |
| 20:56:32 | dc_ | actualy has the same error to |
| 20:56:42 | dc_ | I havn't ran gpe-conf on my new install yet |
| 20:56:50 | bdale | is more or less happy now. |
| 20:56:51 | bdale | thanks! |
| 20:56:53 | cmarqu_ | Yeah, I get it too. |
| 20:56:58 | mallum | dc_: you can just see the arm of eric 'nelson' clapton in that shot |
| 20:57:07 | dc_ | there we do |
| 20:57:24 | dc_ | mallum: :) |
| 20:58:04 | jg_ | folks, we plan to release .6 as soon as installation directions get updated; so make sure jamey knows of any last second packages that should go into rc1 |
| 20:58:24 | cmarqu_ | Uh oh. |
| 20:58:35 | dc_ | isn't rc1 already out...jg? |
| 20:58:41 | pb_ | FWIW, it seems that mccarthy uploaded the gpe-conf in question. |
| 20:58:49 | cmarqu_ | Aha. |
| 20:58:49 | pb_ | dc_: there have been multiple rc1s. |
| 20:58:58 | mallum | sharpens the knives |
| 20:58:58 | pb_ | dc_: we are up to rc1-17 now :-} |
| 20:58:58 | dc_ | fun |
| 20:59:03 | dc_ | mallum: email sent :) |
| 20:59:10 | jg_ | yes, and if there are fixes that should go into rc1 before it becomes just .6, let jamey know.... |
| 20:59:11 | mallum | dc_: coolio dudeo :) |
| 20:59:30 | bdale | is glad that Debian isn't the only project with release logistics issues... well, not really "glad", but... |
| 20:59:36 | mallum | pb_: did yoy get my mail ? |
| 20:59:39 | cmarqu_ | pb_: Can you make a new package with the icon fixed? |
| 20:59:41 | pb_ | mallum: yeah |
| 20:59:44 | cmarqu_ | I hope the rest just works... |
| 20:59:47 | mallum | pb_: make any sense ? |
| 20:59:50 | pb_ | mallum: haven't really thought about it yet though. |
| 20:59:59 | jg_ | bdale: but we do alot more releases alot more often. there will always be .6.1... |
| 21:00:04 | pb_ | cmarqu_: okay, what do I need to do exactly? |
| 21:00:14 | pb_ | cmarqu_: just move the file to "default"? |
| 21:00:18 | bdale | jg_: we release unstable every day... so there... :-) |
| 21:00:40 | jg_ | and we release unstable every 10 minutes, so there .... :-) :-) |
| 21:00:43 | cmarqu_ | Makefile:$(INSTALL) pixmaps/ipaq.png $(DESTDIR)$(PREFIX)/share/gpe/default/ipaq.png |
| 21:01:15 | dc_ | change that |
| 21:01:22 | pb_ | change it to what? |
| 21:01:30 | cmarqu_ | There is a pixmaps/ part missing |
| 21:01:40 | pb_ | Duh, right. |
| 21:01:40 | cmarqu_ | /usr/share/gpe/pixmaps/default/ |
| 21:01:44 | dc_ | :) |
| 21:01:47 | dc_ | hehe |
| 21:01:57 | pb_ | This is a gtk1 package, right? |
| 21:02:02 | cmarqu_ | Yup. |
| 21:02:15 | bdale | hrm. so, still no contact manager in 0.6-rc1? |
| 21:02:21 | pb_ | Okay. It'll take me a minute to gird my libgpewidget0. |
| 21:02:41 | jg_ | bdale: have you done an ipkg install gpe? |
| 21:02:59 | bdale | not recently... |
| 21:03:08 | bdale | ~ # ipkg install gpe |
| 21:03:08 | bdale | Package gpe (0.01) installed in root is up to date |
| 21:03:15 | jg_ | the old gpe integrated releases just had a subset of the apps by default. |
| 21:03:21 | pb_ | try ipkg install gpe-apps |
| 21:03:32 | bdale | ~ # ipkg install gpe-apps |
| 21:03:32 | bdale | Package gpe-apps (0.11) installed in root is up to date |
| 21:03:44 | pb_ | Hmph. |
| 21:04:00 | bdale | I see calendar, dillo, sketchbook, text editor, and to-do list on the relevant screen... |
| 21:04:35 | bdale | I do *not* have unstable lines in my ipkg.conf right now |
| 21:04:41 | pb_ | Maybe I didn't think contacts worked well enough to put it in the gpe-apps metapackage. |
| 21:04:48 | pb_ | You can probably ipkg install gpe-contacts manually. |
| 21:05:08 | dc_ | ngiht |
| 21:05:10 | bdale | ~ # ipkg install gpe-contacts |
| 21:05:11 | dc_ | night |
| 21:05:11 | bdale | ERROR: The following packages conflict with libxml2: |
| 21:05:11 | bdale | libxml1 |
| 21:05:13 | pb_ | dc_: night |
| 21:05:19 | dc_ | bdale: have fun |
| 21:05:24 | bdale | dc_: thanks! |
| 21:11:07 | pb_ | bdale: ah, that might have been another reason why it wasn't included. |
| 21:11:11 | pb_ | bdale: what is using libxml1? |
| 21:11:48 | bdale | pb_: how do I figure that out? |
| 21:11:59 | bdale | aha |
| 21:12:06 | pb_ | bdale: easiest way is to try removing it and see what ipkg bitches about |
| 21:12:07 | bdale | libglade0 |
| 21:12:27 | pb_ | ok, see if you can do without libglade0. I don't think anything in gpe uses it. |
| 21:12:43 | bdale | python-pygtk wants it... |
| 21:12:47 | pb_ | ah. |
| 21:13:01 | bdale | which leads to utils-py and sysset |
| 21:13:02 | pb_ | in fact, the libxml1/libxml2 conflict is entirely spurious. |
| 21:13:15 | pb_ | The packages can be installed simultaneously with no problems. So you could just force them on. |
| 21:13:42 | bdale | k, forced libxml2 on |
| 21:13:48 | pb_ | cool. |
| 21:13:59 | bdale | and gpe-contacts installed cleanly afterwards |
| 21:14:20 | pb_ | even cooler |
| 21:14:55 | bdale | complains about an icon, though ... /usr/share/gpe-contacts/pixmaps/frame.xpm |
| 21:15:12 | pb_ | oh, hm. |
| 21:15:17 | bdale | seems to be there... restarting required, or something? |
| 21:15:30 | pb_ | oh, you may need gdk-pixbuf-loader-xpm. |
| 21:15:49 | pb_ | or gdk-pixbuf2-loader-xpm if you are on gtk2 |
| 21:15:57 | bdale | installed gdk-pixbuf-loader-xpm |
| 21:16:03 | bdale | nah, still 1 I think |
| 21:16:32 | bdale | yep, it comes up now |
| 21:16:39 | pb_ | ok |
| 21:28:28 | pb_ | cmarqu_: I've uploaded a new gpe-conf. I have no way to test it, however. |
| 21:28:39 | cmarqu_ | Okay. let me try it. |
| 21:29:02 | pb_ | BTW, the build process was a bit broken. It left VERSION unsubstituted in the control file, I had to do that by hand. |
| 21:30:47 | cmarqu_ | Yeah, I never built a package with this. |
| 21:32:30 | cmarqu_ | pb_: Okay, it works. |
| 21:32:40 | pb_ | Cool. |
| 21:32:47 | cmarqu_ | And argh, it include the Login applet which has known bugs. |
| 21:32:48 | pb_ | Will you alert the relevant authorities? |
| 21:32:50 | mccarthy | anyone using *both* gpe and opie? |
| 21:33:02 | pb_ | mccarthy: try antiproxy, he does that kind of wacky stuff |
| 21:33:11 | cmarqu_ | pb_: I'll send a mail to jamey |
| 21:33:17 | pb_ | Cool. |
| 21:33:24 | mccarthy | pb_: ok. I am thinking about interoperability with the opie calendar.... |
| 21:33:32 | pb_ | ah. |
| 21:33:36 | mccarthy | pb_: do they use sqlite yet? |
| 21:33:45 | pb_ | No idea. You would have to take it up with Max. |
| 21:36:24 | cmarqu_ | Anybody tried memtrace? |
| 21:38:00 | cmarqu_ | http://www.uv-ac.de/ipkgs/ - seems nice |
| 21:40:29 | mccarthy | pb_: is ownerinfo pretty yet? |
| 21:40:33 | bdale | so, is there no way to change the timezone from within gpe, or am I missing something? |
| 21:40:33 | mccarthy | ownerinfo2 |
| 21:40:38 | pb_ | mccarthy: no |
| 21:40:45 | pb_ | bdale: not at present |
| 21:41:04 | bdale | ok. so I have to manually replace /etc/localtime? |
| 21:41:07 | pb_ | yup |
| 21:41:20 | bdale | other files on the ipaq already, or do I need to fetch one from elsewhere? |
| 21:41:30 | pb_ | bdale: I think there are some packaged timezones, perhaps in a special ipkg feed. |
| 21:41:35 | bdale | hrm |
| 21:41:58 | cmarqu_ | Actually, jamey closed the my bug so that the timezones feed should be in rc1 already |
| 21:42:04 | pb_ | Ah. |
| 21:42:08 | bdale | cmarqu_: pkg? |
| 21:42:34 | bdale | timezones? |
| 21:42:54 | cmarqu_ | bdale: check what /etc/ipkg.conf references and look at the packages there |
| 21:43:12 | pb_ | or ipkg list | grep timezone might yield some inspiration, possibly. |
| 21:43:23 | cmarqu_ | Hmm, that would be something for the release notes... |
| 21:43:25 | bdale | pb_: timezones, but it's not available |
| 21:43:30 | pb_ | oh. |
| 21:43:45 | cmarqu_ | It was called different then |
| 21:44:00 | cmarqu_ | They were I mean :) |
| 21:44:10 | cmarqu_ | It was single packages |
| 21:44:20 | bdale | I bounce around enough (too much travel!) that being able to wiggle timezones on the fly is important. one thing I really want someday is the ability to enter appointments in the current timezone, and have them be at the right UTC time even when I move to another timezone... |
| 21:44:33 | bdale | PalmOS gets that wrong in my experience |
| 21:44:40 | pb_ | Yeah, there is a bug open on gpe-calendar for that I think. |
| 21:45:05 | bdale | I mean, my boss'es staff meeting happens at the same UTC time no matter where I'm dialing in from... :-) |
| 21:46:16 | pb_ | Right. |
| 21:47:04 | bdale | hrm. I don't see anything obvious, not that I really know what I'm looking for. |
| 21:49:01 | ade | is away: phone |
| 21:49:08 | cmarqu_ | bdale: What are the source lines in your /etc/ipkg.conf? |
| 21:50:34 | bdale | cmarqu_: the only active ones are |
| 21:50:36 | bdale | src v0.6-rc1-base http://familiar.handhelds.org/familiar/releases/v0.6-rc1/base/armv4l |
| 21:50:37 | bdale | src v0.6-rc1-x http://familiar.handhelds.org/familiar/releases/v0.6-rc1/x/armv4l |
| 21:51:01 | cmarqu_ | Hmm. |
| 21:52:23 | cmarqu_ | You need http://www.handhelds.org/feeds/timezones/ |
| 21:54:07 | cmarqu_ | Then do the grep thing again with timezone-* |
| 21:54:22 | bdale | done, and installed timezone-mst ... hrm. |
| 21:54:32 | bdale | ok, it puts things in /etc/timezones |
| 21:55:02 | bdale | yep, symlinking /etc/localtime to the /etc/timezones/* of interest works |
| 21:55:27 | cmarqu_ | We really should have that in the release notes... |
| 21:55:31 | bdale | yes. |
| 21:55:34 | pb_ | indeed. |
| 21:55:45 | cmarqu_ | Let me copy and paste this log into a mail to jamey |
| 21:55:53 | bdale | and frankly, it'd be nice to have a task-all-timezones or something that installs all of the files. they're tiny, after all. |
| 21:56:13 | bdale | I realize that's heretical here, but... :-) |
| 21:59:13 | cmarqu_ | We really need a configurator. Something for a later release. |
| 21:59:18 | pb_ | Yes. |
| 22:00:23 | bdale | must be a usage model thing. the configuration screen seems cluttered with config widgets I'll use rarely if at all, yet things like this that I consider fundamental aren't there yet. |
| 22:01:12 | cmarqu_ | Yeah... |
| 22:01:16 | cmarqu_ | It's hard. |
| 22:01:42 | cmarqu_ | Like with Gnome 2... many people complain about that it is so dumbed down. |
| 22:02:05 | bdale | MS gets this wrong, too, they have a gazillion knobs you rarely ever touch mixed in with the few that you use often. it's be nice to put some of the usability research folks I know to work on this someday, but we might not like their answers... :-) |
| 22:02:09 | cmarqu_ | Okay, jamey pulled the new gpe-conf package |
| 22:02:20 | cmarqu_ | Right :/ |
| 22:02:34 | bdale | the real problem with gnome2, for me, is that vertical panels are poorly supported by most applets. |
| 22:02:36 | bdale | :-) |
| 22:02:42 | pb_ | cmarqu_: "pulled" in what sense? |
| 22:02:58 | nils | Good night ;) |
| 22:03:05 | pb_ | cmarqu_: you mean he put it into 0.6? |
| 22:03:07 | cmarqu_ | pb_: His answer was "done." :) |
| 22:03:12 | pb_ | ah, cool. |
| 22:03:22 | cmarqu_ | I asked: |
| 22:03:23 | cmarqu_ | > we found a serious bug with the gpe-conf package in rc1. |
| 22:03:23 | cmarqu_ | > |
| 22:03:23 | cmarqu_ | > Please update it with |
| 22:03:25 | cmarqu_ | > http://familiar.handhelds.org/familiar/feeds/unstable/packages |
| 22:03:26 | pb_ | clearly not a man to waste words. |
| 22:03:28 | cmarqu_ | > /armv4l/gpe-conf_0.0.6_arm.ipk |
| 22:03:33 | cmarqu_ | So it is the answer we want to hear. |
| 22:03:41 | pb_ | cmarqu_: Phew. |
| 22:04:45 | bdale | yep, 0.0.6 bitched about the copy of ipaq.png I had made... :-) |
| 22:04:51 | pb_ | ah :-) |
| 22:05:07 | cmarqu_ | bdale: I mentioned GPE on #usability already, and e.g. Calum Benson (one of the Sun Gnome guys) is now at least aware of its existence |
| 22:05:29 | cmarqu_ | They probably have lots of other stuff to worry now though |
| 22:05:35 | bdale | no doubt |
| 22:06:15 | bdale | hrm. ok, possibly my last question for today. is there a reasonable way to play ogg files from a nearby http server on the ipaq right now? |
| 22:06:32 | pb_ | depends what you class as reasonable, I guess. |
| 22:06:37 | pb_ | wget | ogg123 would probably do the job. |
| 22:07:04 | bdale | hrm. ok. maybe I've found the first thing I need to write for my ipaq? ;-P |
| 22:07:16 | pb_ | ha. |
| 22:07:26 | pb_ | it would be nice to have that as a module for NMF, but I have never had the time to write it. |
| 22:08:07 | bdale | gpe-nmf-ogg - GPE music player: Ogg Vorbis decoder |
| 22:08:07 | bdale | libvorbis-fixp0 - Fixed point Ogg Vorbis decoder |
| 22:08:07 | bdale | libvorbisfile-fixp0 - Fixed point Ogg Vorbis decoder |
| 22:09:39 | cmarqu_ | > There is a timezones package that contains 137 timezone files. |
| 22:09:39 | cmarqu_ | > Should we just install this? Hmm. du reports it is 1.6M. It's |
| 22:09:39 | cmarqu_ | > in the opie feed in 0.6. zoneinfo by continent is more reasonable |
| 22:09:39 | cmarqu_ | > in size. |
| 22:09:43 | cmarqu_ | (jamey asking) |
| 22:10:11 | bdale | 1.6M ? that doesn't correlate with the size of the individual timezone files in the timezone feed at all... |
| 22:10:59 | mallum | pb_: where in cvs is the xsetting stuff you spoke of - has it all been split up now ? |
| 22:11:02 | bdale | it looks like /etc/MST7MDT is only 877 bytes, for example |
| 22:11:12 | pb_ | mallum: yeah: libxsettings, libxsettings-client, xst, gpe-confd |
| 22:11:12 | bdale | /etc/timezones/MST7MDT, I mean |
| 22:13:01 | mallum | pb_: cool. its still stores to disk using berkley db right ? |
| 22:13:18 | cmarqu_ | bdale: should I Cc: you on my reply to jamey? I need to go soon... |
| 22:13:19 | pb_ | mallum: you can select sqlite as a compile option now, but db is still supported. |
| 22:13:31 | pb_ | mallum: the package in familiar uses sqlite, because gpe machines don't usually have db installed. |
| 22:13:37 | bdale | cmarqu_: sure, if you want. bdale@gag.com is canonical. |
| 22:13:38 | mallum | pb_: yep |
| 22:14:43 | mallum | pb_: I guess I should hack some xsettings functionaility into xkbd, matchbox |
| 22:14:46 | cmarqu_ | bdale: Okay, done. |
| 22:15:19 | pb_ | mallum: sounds cool |
| 22:15:45 | mallum | pb_: is anything else currenlt reading xsettings ? |
| 22:16:20 | pb_ | mallum: not right now |
| 22:17:46 | bdale | Installing gpe-nmf (0.06) |
| 22:17:46 | bdale | Configuring gpe-nmf.../usr/bin/python: can't open file '/usr/sbin/install-menu' |
| 22:17:46 | bdale | pkg_run_script: `/usr/lib/ipkg/info/gpe-nmf.postinst configure' failed: Illegal seek. |
| 22:17:46 | bdale | ERROR: gpe-nmf.postinst returned 2 |
| 22:18:16 | pb_ | eugh. |
| 22:18:21 | pb_ | mallum: http://handhelds.org/scap/port.3491.png :-) |
| 22:18:46 | pb_ | bdale: that kind of thing is usually harmless. |
| 22:18:54 | bdale | oh no. if my kids see frozen-bubble on the ipaq, I'll never see the ipaq again... |
| 22:19:03 | mallum | pb_: ha :) |
| 22:19:09 | mallum | pb_: that your ipaq ? |
| 22:19:18 | bdale | pb_: the icon shows up in the audio screen, tapping it gives nothing. |
| 22:19:29 | pb_ | mallum: well, that's a slight cheat, it is actually running on the CATS over the network, just displaying on the ipaq. |
| 22:19:43 | mallum | pb_: this pure lovely xlib :) -> https://savannah.nongnu.org/projects/xbubble/ |
| 22:19:44 | pb_ | mallum: to run it locally I will need to muster a CF card, and I seem to have left them all at the office. |
| 22:20:13 | pb_ | mallum: cool :-) |
| 22:20:24 | pb_ | bdale: try running gpe-nmf from the terminal? |
| 22:20:35 | mallum | pb_: looks like its jumped a version since I last tried it ... |
| 22:20:36 | bdale | gpe-nmf: error while loading shared libraries: libexpat.so.1: cannot open shared object file: No such file or directory |
| 22:20:44 | pb_ | weird |
| 22:20:54 | bdale | mumbles something about dependency management |
| 22:21:07 | pb_ | I don't think NMF links against expat directly. This is probably some further fallout from the new GTK2 being shoved into 0.6. |
| 22:21:20 | bdale | hrm. could be, I wouldn't know. |
| 22:21:24 | pb_ | GTK2.1 is linked against the latest xlibs, which are still languishing in unstable. |
| 22:22:45 | pb_ | mallum: I'll have to give it a go. |
| 22:25:15 | bdale | pb_: installing libexpat gets past that error, but now it warningsa bout some UTF-8 conversions and then says "gpe-nmf: relocation error: /usr/X11R6/lib/libXft.so.2: undefined symbol: XRenderCompositeText8" |
| 22:25:37 | pb_ | bdale: yeah, I think the gtk2.1 libs will just be completely screwed with 4.1 xlibs. |
| 22:25:48 | pb_ | bdale: Render and all the font stuff has changed quite a lot. |
| 22:26:08 | bdale | so, does that mean it's a lost cause for now, or can I pick up an older or newer copy of something to make everything happy? |
| 22:26:26 | pb_ | bdale: if you fish an older version of libgtk2.0-0 out of the archive, it should start working again. |
| 22:26:34 | bdale | hrm. ok, will go look. |
| 22:26:53 | pb_ | bdale: alternatively, you can install all the xlibs from unstable. |
| 22:27:10 | bdale | pb_: hrm. any downside to doing that? |
| 22:27:23 | mallum | pb_: hmmm, want a big display :*( |
| 22:27:25 | pb_ | bdale: not other than the usual caveats about unstable being less well tested |
| 22:27:29 | bdale | I'm always a fan of moving forward to fix things insted of moving back... |
| 22:27:58 | pb_ | ok, in that case the new xlibs would be your best bet. |
| 22:28:02 | bdale | ok. so if I add an unstable feed to my ipkg.conf, do an update, then what should I install? |
| 22:28:10 | pb_ | let me check |
| 22:28:29 | bdale | unstable feed added, ipkg update complete |
| 22:28:41 | pb_ | xlibs, libxrender, libxft, libxft2 I reckon. |
| 22:28:51 | pb_ | also libfontconfig probably |
| 22:28:58 | bdale | doing it |
| 22:29:11 | bdale | aha. 4.1-5 to 4.2-2 |
| 22:29:24 | pb_ | yup. |
| 22:30:50 | bdale | only xlibs and libxrender updated, the rest it says are already up to date |
| 22:30:57 | pb_ | okay. |
| 22:31:08 | pb_ | does NMF run now? |
| 22:31:09 | bdale | fear. that works better. |
| 22:31:14 | pb_ | cool. |
| 22:31:41 | bdale | ok, so it only groks locally mounted filesystems? |
| 22:32:17 | pb_ | yeah, at the moment. |
| 22:32:35 | bdale | heh. |
| 22:32:39 | bdale | nfs mountes his audio tree |
| 22:33:11 | pb_ | :-) |
| 22:34:08 | cmarqu_ | The intimate guys are planning a meeting in London |
| 22:34:17 | bdale | ah, the awesome fidelity of the internal speaker is... overwhelming. |
| 22:34:18 | mallum | pb_: did you write the add_events function in xsettings-client ? |
| 22:34:28 | pb_ | mallum: not that I remember |
| 22:34:39 | bdale | finds his earbuds ... :-) |
| 22:34:49 | pb_ | mallum: does it look especially inept? |
| 22:35:26 | mallum | pb_: no its very cool :) |
| 22:35:31 | pb_ | ah :-) |
| 22:36:04 | mallum | pb_: I never new there was a 'your_event_mask' window attribute |
| 22:36:17 | mallum | pb_: its named so un-xlibby |
| 22:36:22 | pb_ | heh |
| 22:36:28 | mallum | pb_: actually seems a little 'friendly' named |
| 22:36:46 | bdale | pb_: hrm. is there anyway to hand the playlist thingy all the ogg files in a given directory at once, or does it only handle one at a time? |
| 22:37:12 | pb_ | bdale: hmm, I don't remember offhand. I know you can construct an XML file containing a whole playlist and put that in. |
| 22:37:26 | pb_ | bdale: have you tried just giving it the filename of the directory? That might work. |
| 22:37:33 | bdale | pb_: it's also more than moderately annoying that it always resets to your home directory when you go into the playlist editor thingy |
| 22:37:47 | pb_ | Yes. :-/ |
| 22:38:19 | bdale | pb_: it doesn't highlight anything in the selection box until I get into the directory and select a .ogg |
| 22:38:20 | pb_ | (I don't use the editor much, I just have one monster playlist that includes everything.) |
| 22:38:28 | pb_ | bdale: ah. |
| 22:39:21 | pb_ | bdale: there is definitely stuff in the source to handle adding a whole directory at once, but I'm not sure exactly how you trigger it. |
| 22:39:40 | bdale | I suspect what I'm going to want to do is write the distributed jukebox client I've always meant to write that groks the structure of my server's http export of all my audio collection. where's the source to this? seeing how it works would be a good start. |
| 22:39:54 | pb_ | bdale: in gpe cvs. |
| 22:40:03 | bdale | k, will look later |
| 22:40:27 | pb_ | righto |
| 22:42:30 | bdale | this really is pretty amusing, though... an 802.11 widget that's almost pocketable streaming me audio from the other end of the house... :-) |
| 22:43:06 | pb_ | heh. |
| 22:43:39 | pb_ | if you have another X display, you should be able to migrate the NMF frontend over the wireless link too. |
| 22:43:54 | pb_ | oh, actually, no. |
| 22:43:58 | pb_ | I remember, it needs support in GTK. |
| 22:44:37 | mallum | pb_: but thats coming :) |
| 22:44:42 | pb_ | :-) |
| 22:44:59 | mallum | pb_: read the little spec :-) ! |
| 22:45:04 | pb_ | reading now! |
| 22:45:09 | mallum | ha |
| 22:46:30 | pb_ | looks good. |
| 22:46:58 | pb_ | isn't the X traditional notation <hostname>:<display nr>.<screen nr>? ie dot rather than colon at the end. |
| 22:47:08 | pb_ | not that it really matters either way |
| 22:47:18 | mallum | pb_: yeah, my mistake |
| 22:48:14 | mallum | pb_: hostnames cant currently be UTF8 can they ... ? |
| 22:48:43 | pb_ | Handling errors is kind of difficult though. In the gtk-demo, you open the display ahead of time so I guess the chance that something will go wrong during the actual migration is kind of low. In this case, you might have all kinds of problems like authentication failure, so there does need to be some way for the application to (a) not end up in limbo, and (b) report an error to the user. |
| 22:49:13 | pb_ | mallum: I'm not sure. |
| 22:51:39 | mallum | pb_: yeah, As the app will do the actual migration, it will replace the property text with some sort of error message on failiure. If it succeeds the apps window will get destroyed ( on current display )). The 'migrator' should select for both these events to figure out whats going on |
| 22:51:57 | pb_ | Right. |
| 22:52:08 | ade | is back (gone 01:03:07) |
| 22:52:26 | pb_ | As you say, the app could crash and the migrator would mistake that for a successful migration. but I guess this is no worse than the app just randomly crashing immediately _after_ a successful migration. |
| 22:52:46 | mallum | pb_: it initially deletes the property to a) notify the migrator its attempting to migrate and b) stop any race conditions of mulitple migrators trying to migrate the window |
| 22:53:29 | pb_ | mallum: surely deleting the property would _cause_ a race, not fix it? |
| 22:53:49 | mallum | pb_: I think gtk2 apps set a NET_PID property, so this could be tracked to get a crash ( assuming the window originally started out on the same display ) |
| 22:54:02 | pb_ | ok. |
| 22:54:32 | mallum | pb_: you'd grab the server during the delete I guess ( grabbing it during the whole process would be too long as it can take a few seconds to connect to another display ) |
| 22:55:25 | pb_ | yeah. |
| 22:55:39 | pb_ | mmm, what if the migrator crashes before reading out the error message? |
| 22:56:19 | mallum | pb_: I guess the user would see the migrator crash and be informed that way that something has gone wrong |
| 22:56:44 | pb_ | mallum: right, but the migrated app would be stuck with this error message in its property, which would presumably inhibit further migrations. |
| 22:57:38 | mallum | pb_: no, the property would still exist , so it would not effect any further attempts to migrate. The migrator would just check for the existance of the property rather its actual value |
| 22:58:47 | mallum | s/its/ than its/ |
| 22:58:58 | pb_ | mallum: ok, so what stops another migrator coming along and stomping on the error message before the first one had a chance to read it? |
| 22:59:52 | mallum | pb_: hmm, nothing ..... |
| 23:00:03 | pb_ | also, if the migrated app takes a long time to respond, you could potentially have several migrators all trying to move it at the same time if they get in prior to it grabbing the server and deleting the property. |
| 23:01:10 | mallum | pb_: maybe a migrator should refuse to migrate if the property contains some text and the original error recieveing migrator should reset the property to "" |
| 23:01:40 | pb_ | yeah, I think that would work. Migrators would need to grab the server while checking, to make sure the transaction was atomic, but that should be no big deal. |
| 23:02:03 | mallum | pb_: yep |
| 23:02:26 | pb_ | except that we are then stuck with the case where the original error-receiving migrator crashes, and the property is left hanging around. I suppose the migratee could delete it after a timeout, or if it saw a destroynotify for the original migrator. |
| 23:02:35 | pb_ | s/delete/set to ""/ |
| 23:05:46 | mallum | pb_: hmm, yeah timeout seems probably best |
| 23:06:04 | pb_ | ok cool |
| 23:06:53 | mallum | pb_: They maybe a cleaner way to catch these possible errors - I need to think a little more about it - but I think at its core it should do the job |
| 23:07:13 | pb_ | Yeah. |
| 23:07:19 | pb_ | So, you wanna have a go at implementing it? |
| 23:07:22 | mallum | pb_: I think some cleverly place XSetErrorHandlers may help |
| 23:07:31 | mallum | pb_: sure, tommorow morning probably |
| 23:07:41 | pb_ | excellent |
| 23:07:50 | mallum | pb_: I gotta cut and paste this conversation ... :) |
| 23:08:02 | pb_ | heh |
| 23:08:25 | pb_ | did you actually try out the XSETTINGS stuff, by the way? |
| 23:08:36 | pb_ | I haven't really used it in anger on the ipaq yet. |
| 23:09:52 | mallum | pb_: not yet I think its next on the matchbox TODO list though |
| 23:09:58 | pb_ | cool. |
| 23:10:07 | mallum | pb_: shouldn't take long at all to add |
| 23:10:28 | pb_ | yeah. |
| 23:10:52 | mallum | pb_: is libxsettings a seperate package ? |
| 23:11:01 | pb_ | I think I'm gonna make gpe-calendar use it for preferences, at least initially. The XSETTINGS spec does explicitly forbid random apps from storing their preferences there, but I laugh in its face. |
| 23:11:04 | pb_ | mallum: yes |
| 23:11:30 | pb_ | mallum: libxsettings contains xsettings-common.c, essentially. libxsettings-client is (shock) xsettings-client.c. |
| 23:11:51 | pb_ | xsettings-manager.c is bundled with gpe-confd, on the assumption you probably don't want too many managers. |
| 23:11:53 | mallum | pb_: yeah it makes to much sense for us I think to take the spec too seriouly |
| 23:12:01 | mallum | pb_: yep |
| 23:13:26 | mallum | pb_: I trust being a non-x86 biggot & byte master you added that big endian fix to your xsettings stuff ? |
| 23:13:38 | pb_ | mallum: yeah, I think so. |
| 23:14:16 | pb_ | mallum: however, the protocol between xst and gpe-confd will break if they run on machines of different endianness at the moment. I think I can fix that quite easily, just haven't done it yet. |
| 23:14:24 | mallum | pb_: BTW I meant non 'x86 buggot' rather than 'non x86' biggot |
| 23:14:36 | pb_ | phew :-) |
| 23:14:38 | mallum | s/buggot/biggot/ oops ;) |
| 23:17:43 | pb_ | right |
| 23:17:50 | pb_ | must be my bedtime by now. |
| 23:17:55 | mallum | pb_: I'll add xsettings for mb theme and cursor thingy ( which can change on the fly ), and the xkbd layout |
| 23:17:56 | pb_ | night guys |
| 23:17:59 | mallum | pb_: gnight |
| 23:18:01 | pb_ | mallum: sounds excellent |
| 23:40:58 | ade | is away: with the fairies |
