[12:41] hello to all. i have a small question about a missing package. [12:41] i have searched to find the package opie but didnt had any luck till yet. [12:41] does anybody know if opie is available as a package for the arm structure [12:41] here is a nice build how to by the way that works great [12:41] http://opie.handhelds.org/cgi-bin/moin.cgi/BuildOpie [12:41] i need it only now for the arm arch [12:42] xperia: Which package seems to be missing? [12:44] opie itself http://opie.handhelds.org/cgi-bin/moin.cgi/BuildOpie [12:44] open embedded has this package. angstrom too [12:44] only ubuntu mobile is missing it [12:46] Which package? [12:46] As far as I can tell, that "opie" is a collection of stuff. [12:47] persia: http://gitorious.org/opie/opie [12:48] well it is same like gpe only from my side of view sometime better [12:48] it can easy be build in less than 30 min with this instruction [12:49] http://opie.handhelds.org/cgi-bin/moin.cgi/BuildOpie [12:49] i have allready builded it and on open embedded it can be builded it too as the recipes exist [12:49] it is however missing in ubuntu mobile [12:51] Used to be available. Let me see if I can figure out why it's gone. [12:54] persia: ohhh wooow if it could be maked availble especially also for the arm arch this would be awesome. [12:54] Opie is still developed as you can see on gitorious. [12:54] opie has a full desktop with full working mail programms, multimedia programms and so on. [12:54] some screenshots of the the older releases [12:54] http://opie.handhelds.org/gallery/main.php [12:55] We try really hard not to make anything available for only one architecture. [12:55] So that people can use the same mobile environment whether they have i386/armel/powerpc handhelds. [12:57] persia: ohh woow this would be a heavy boost for this quite good gui for mobile devices [12:57] it is allready used on some mobile pocket pc phones [12:57] http://i37.tinypic.com/144b90o.jpg [12:57] http://i37.tinypic.com/24mhqps.jpg [12:57] http://wiki.xda-developers.com/index.php?pagename=Kovsky Linux [12:59] if opie would be availble for ubuntu-mobile this would be the perfect thing as it would increase the possibility running ubuntu mobile on the newest pocket pc successfull [12:59] at the moment most of th linux ditro runns on angstrom or oe [13:00] Unfortunately my debtags search is eating my browser. [13:00] it should be very easy to compile it as described on the step by step build instruction. i have build it more than once successfull. if you have any question just ask. [13:00] My suspicion is that some time in 2004 or 2005 someone stopped maintaining opie in Debian, and it got dropped for lack of further interest. [13:03] the maintainer told me this too as some deb packages should be availble. but opie is still developed and used. last patch is from [13:03] Thursday February 04 2010 [13:03] http://gitorious.com/opie/opie [13:03] i am working on opie too as a hobby coder :-) [13:04] I can't find any good links that indicate when it was or wasn't added or removed from Debian. [13:04] My strongest recommendation would be to work at getting it into Debian. [13:04] This will then also be available in Ubuntu and all other Debian derivatives. [13:05] so i should try at debian in this case [13:06] xperia: That's my recommendation, just because Debian has maintainers, so there's a better chance of keeping it there. [13:06] Adding stuff to Ubuntu that doesn't get added to Debian is usually a temporary matter except for special stuff highly interesting to some specific Ubuntu flavour. [13:09] okay i have allready asked at the debian chanell [13:11] OK. Please feel free to ask more here if you have questions or need help. [13:11] I just think it needs someone to care for it, and just tossing it in the repository is not best. [13:13] well they have told me to package it but i have no experience with packaging it. if somebody would help me the first time i would do it however [13:14] I could certainly do that, but not right now. Would you be available around this time tomorrow? [13:15] Err, no, that won't work :/ [13:15] So a little later than now tomorrow, or another day or another hour, catch me here. [13:16] yeah great ! then we see us tomorow some hours later than yet or another day [13:16] thanks persia :-) [13:17] Happy to help. === doko_ is now known as doko [13:47] persia: or to others for some strange reason opie exist in ubuntu [13:47] https://launchpad.net/ubuntu/+source/opie/2.40~dfsg-0ubuntu1 [13:47] it need just to be updated [13:48] xperia: That's a different "opie", isn't it? [13:48] * persia thought that was a smartcard thing [13:49] well the whole discussion is done on debian-arm [13:49] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=511582 [13:49] Debian bug 511582 in opie "packaged opie 2.4, fixing testsuite failures on arm/armel" [Serious,Open] [13:50] it should be the same package from my point of view [13:50] maybe it just dont has everything enabled like mp3 players email programms and such as this can be set all optional to be builded [13:51] See, the current "opie" package is a one-time passwod manager. [13:51] You7re talking about a *different* opie. [13:52] yeah the people at debian told me that too but this is very strange why does they use the same name for this programm [13:53] That happens a lot. [13:53] Most folk don't seem to check carefully before choosing a name. [13:57] Looks like opie the password management system started in 1995, and opie the environment started. in 2001 [13:58] hmmm ha ha ha very strange [14:00] heh. [14:00] but the very strange thing is that this person develop also the libpam-opie that is used in the palmtop environment too [14:00] for example opie-server that sync with the palmtop environment [14:01] heh. [14:03] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=553416 [14:03] Debian bug 553416 in libpam-opie "libpam-opie: Adding support for pam-auth-update" [Wishlist,Open] [14:04] the whole packages are 100% related to opie [14:05] need maybe to cantact him and ask what is the actuall process [14:10] persia: it looks like really this programms are not related to opie palmtop environment [14:11] I don't think it's related at all, actually. [14:13] yeah. i can only remember that libpam was used in opie while the compiling. thats why i have thinked they are related to opie palmtop environment especially for syncing purposses [14:13] thanks for the clarification [14:17] okay see you all next time. bye ! [14:27] NCommander: Hey! [14:27] persia: sorry about that [14:27] Are you still using Android? [14:27] persia: yes, and still doing some upstream hacking on it [14:28] (low level, but I know how to port android to new platforms) [14:28] Is the hack to use the test demo environment to run inside Ubuntu still viable? [14:28] Also, has the issue that it requires the binary-only Java been resolved? [14:28] persia: using SIMULATOR mode? [14:28] no [14:28] yeah. [14:28] no? [14:28] What happened? [14:28] I believe that got ripped out with eclair [14:28] it was depricated with android 1.0 [14:29] And its the wrong way to do it, we should properly expose /dev/binder for the Android userspace [14:29] Can we? [14:29] persia: its required to build Android, but it might be fixable to use icetea instead (there are some issues with building older Android releases on Java 6, but I think patches landed with 2.1 to finally fix that, I can poke in android-porting if need be) [14:30] persia: I think we have the android bits in the kernel in staging. It should be a matter of flipping them on, and recompiling [14:31] That's all? [14:31] persia: well, Android isn't X based [14:31] so basically, what we need to bring android up is the following [14:31] 1. android ported to x86 (there are two separate projects that do this) [14:32] 2. /dev/binder support [14:32] 3. chroot environment [14:32] 4. Allocate arange of UIDs for android to use (android assigns each app its own UID for security reasons) [14:32] 5. Either trap writes to /dev/graphics/fb0 or use androidvncserver and some voodoo [14:33] Why do we need x86? [14:33] persia: dalvik has platform specific bits. Unless you want to run Android in an emulator [14:33] I'd rather use a VNC server. What sort of voodoo? [14:33] chroot is easy. [14:33] persia: android assumes /dev/graphics/fb0 is going to be there. You'll have to teach it thats its not, and just use VNC instead [14:34] NCommander: android drivers were removed from the mainline kernel staging [14:34] NCommander: Why use an emulator? Why not just run on hardware to which it's ported for now, and encourage bugfixing for the rest. [14:34] persia: because any android app that uses the NDK would then not run [14:34] amitk: ugh, reason? [14:35] unmaintained by google, no fixes to move them to drivers [14:35] NCommander: Hrm? I'm confused. Tell me why I need porting to x86 or an emulator to run on my Netwalker. [14:35] amitk: ****, there goes the easy flip switch for binder. How hard would it be to add to our mainline [14:35] persia: oh ... ugh [14:35] persia: I didn't think about that. I still think netbook == i386 :-/ [14:36] Silly. [14:36] persia: you'd still want to use android-x86, since it has support for mice and not just touchpads [14:36] There's i386, powerpc, and armel hardware out there mostly. Some amd64 stuff is coming out recently as well. [14:36] NCommander: we won't do it (apply a random unmaintained patch with no hope of being mainlined) unless you have a very good reason, as always :) [14:37] persia: android isn't 64-bit clean last time I checked unfortunately, so it won't run on amd64 unless you want the pain of porting it [14:37] amitk: Is this an Ubuntu Kernel position? What if they were pushed mainline? [14:37] persia: google made comments saying that they weren't interested in mainlining [14:37] persia: we had in enabled in staging in Karmic [14:37] *it [14:37] NCommander: That's fine. I don7t really care about porting right now. I can live with just armel for a while. [14:37] amitk: Ah, I understand. [14:38] persia: NCommander: best write to ubuntu-kernel ML if you want the official version of the story [14:38] this is the abridged version [14:38] amitk: First I want to find out how big a project the whole thing ends up being :) [14:38] amitk: roughly speaking, how large is /dev/binder suppotr? [14:38] *support [14:40] NCommander: dunno, but look at sha id b0a0ccf in git for the commit that removed them