**** BEGIN LOGGING AT Fri Oct 14 02:59:56 2005 Oct 14 03:18:21 hrw|work: thks Oct 14 03:29:30 hm, ive bumbped right back into the LONG_MAX undefined issue whilst building bluez on my amd64. I have since seen this with something else and found the fix was a modifcation to limits.h but cant rmeebr what it was. Does anybody know? Oct 14 03:36:53 03koen 07org.oe.dev * rc381987d... 10/: vte: add 0.11.15 Oct 14 03:37:01 03koen 07org.oe.dev * r6e9bd43a... 10/: gtkhtml: add 2.6.3, needed for scilab Oct 14 03:57:07 a-chew Oct 14 05:03:02 morning Oct 14 05:44:22 does the gpe environment use the standard gtk apis? ie can i develop a progon desktop and cross compile itfor gpe on zaurus reasonably easily? Oct 14 05:59:58 yes Oct 14 06:27:40 exelent ! Oct 14 06:43:17 can anybidy recomend a lightweight db to suitable for embedded use? should i use gnomedb ? Oct 14 06:43:39 squlite Oct 14 06:43:43 sqlite Oct 14 06:44:07 ok, im kinda familiar with that so ill go with it;) Oct 14 06:44:27 plus i know it can do what i require Oct 14 07:08:36 hm. anybody here who has hints about benchmarking for comparing PDA vs. workstation performance? Oct 14 07:15:20 "time cat /dev/urandom>/dev/null"? :) Oct 14 07:15:40 fdv: dont you think that comparing them is senseless? Oct 14 07:18:49 hrw|work: of course, to some extent, but I'm working on a system where the work load is shared between a server running on a stationary computer and a client on the PDA, so it would make sense wrt how much work must be done by the server Oct 14 07:19:43 the comparison would be for the report, sort of to state the obvious :) Oct 14 07:21:32 morning Oct 14 08:01:14 join #elinux Oct 14 08:29:20 RP: ping Oct 14 08:30:54 hi Oct 14 08:34:34 hi AvengerMoJo Oct 14 08:35:40 03florian 07org.oe.dev * r8e8ea79d... 10/: gpe-bluetooth: Add new release. Oct 14 08:41:54 hi hrw|work how is going Oct 14 08:42:19 quite good Oct 14 08:42:25 :) Oct 14 08:42:35 I'm still trying to put bootsplash into OE Oct 14 08:42:54 just install Suse10 and test drive a little Oct 14 08:42:57 AvengerMoJo: kernel patch or aplication? Oct 14 08:43:06 kernel patch Oct 14 08:43:12 application is done Oct 14 08:44:37 oz 2.6 kernel has elpp for it Oct 14 08:45:39 re Oct 14 08:50:38 what is elpp Oct 14 08:58:37 bootsplash is better, especially for X86 Oct 14 09:00:22 hum Oct 14 09:00:39 CosmicPenguin: so elpp is working in the currect OE kernel? Oct 14 09:00:49 there is no current OE kernel Oct 14 09:04:46 CosmicPenguin: I mean the 2.6.13 one Oct 14 09:04:57 I'm not sure how I would say it :P Oct 14 09:05:57 Well, its clear that the kernels are organized by architecture Oct 14 09:06:06 And furthermore, each kernel in OE has a specific name Oct 14 09:06:07 and version Oct 14 09:07:13 CosmicPenguin: right .. I would say ...2.6.13 for Spitz :) Oct 14 09:07:59 That is indeed a far better question Oct 14 09:08:04 AvengerMoJo: you sure, that doesn't exist any more as a .bb Oct 14 09:08:05 Unfortunately for you, I cannot answer that Oct 14 09:09:43 you mean the .... the kernel module? or kernel patch for bootsplash? Oct 14 09:10:00 elpp exists in OE for 2.6 spitz. The current stable version is 2.6.14-rc1 Oct 14 09:10:02 I mean there's no .bb file to build a 2.6.13-* kernel for any Zaurus Oct 14 09:10:05 I'm not search I think I have look for it but with no result Oct 14 09:10:18 any howto on elpp RP? Oct 14 09:10:27 I will try it out later tonight Oct 14 09:10:54 Set OZ_KERNEL_ENABLE_ELPP="yes" iirc Oct 14 09:11:05 echo ENABLE_ELPP = "yes" > local.conf Oct 14 09:11:25 sorry, lardman|work is right Oct 14 09:11:40 bit of luck there ;) Oct 14 09:11:40 read linux-openzaurus.inc Oct 14 09:11:53 thank I will take a look Oct 14 09:12:06 the OZ_KERNEL_ENABLE_ELPP is in the bb? Oct 14 09:12:26 you set it in local.conf Oct 14 09:12:38 to see what it does, look in the .inc file Oct 14 09:13:25 ok Oct 14 09:13:26 thanks Oct 14 09:33:46 A basic question: at what point does OE build the cross-compilers? Oct 14 09:34:04 pretty much the first thing it does Oct 14 09:34:15 I'm on a project where I'm on Fedora Core 3 (GCC3.x) and my counterpart has Fedora Core 4) Oct 14 09:34:18 He has GCC4. Oct 14 09:34:46 We have determined (after much sweat) that opie doens't like to build cleanly (with bitbake) using GCC4 Oct 14 09:34:50 It WORKS under GCC3. Oct 14 09:35:16 I was directed to a patch for QT-2.3.10 for GCC4, but it doesn't appear to address the compilation errors we're getting under FC4. Oct 14 09:35:35 So we have to "hand compile" the source that doesn't work. Oct 14 09:35:57 With gcc3.x version of the compiler (arm-linux- etc. I don't have the exact name in front of me). Oct 14 09:35:59 Any ideas? Oct 14 09:36:11 Other patches to make GCC4 work with the current source? Oct 14 09:36:46 Opie shouldn't be built using the native compiler, or is this building some tool which is required to run natively? Oct 14 09:37:48 I'm using OpenEmbedded to build for Familiar LINUX on IPAQ. Oct 14 09:38:15 OE "out of the box" works GREAT / cleanly for a cross compilation on X86 running Fedora Core 3. Oct 14 09:38:21 Does NOT work with Fedora Core 4. Oct 14 09:38:33 (does not compile cleanly). Oct 14 09:38:46 Can you provide any MORE details of how it FAILS to work? Oct 14 09:39:11 Yes, I can provide details. Let me paste the compiler error messages in ... give me a sec. Oct 14 09:39:31 use pastebin Oct 14 09:39:36 ~pastebin Oct 14 09:39:37 methinks pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca/ Oct 14 09:39:59 don't use pastebin.ca Oct 14 09:40:05 pastebin.com would be a better bet Oct 14 09:41:07 pb___: why not .ca? Oct 14 09:41:37 because pastebin.ca is effectively unreachable for many users, including me Oct 14 09:41:46 OK. Oct 14 09:41:56 oh, someone should tell ibot then I suppose Oct 14 09:42:03 I have the compiler errors (lvalue) and have gone to pastebin.com Oct 14 09:42:14 Now how can everyone here see the paste? Oct 14 09:42:21 ( i haven't published yet ) Oct 14 09:42:35 tell us the url Oct 14 09:43:11 ok. URL is http://pastebin.com/393568 Oct 14 09:43:23 This is the compiler output from a "bitbake opie-image" Oct 14 09:43:42 In this particular case my colleague trace it to the pcmcia package. Oct 14 09:43:55 He took it out of the mix and the build process continued ... Oct 14 09:44:02 but broke on "openswan" with the same error. Oct 14 09:44:14 This time removing the package caused the build to fail (dependencies) Oct 14 09:49:44 now see post at http://pastebin.com/393579 talking about the multiple cross-compiler versions. Oct 14 09:50:12 hi to all Oct 14 09:50:51 has anyone included zope in his own distro? Oct 14 09:54:04 cyberchuck: I thought you were talking about native gcc? Oct 14 09:54:39 I'm PRIMARILY concerned with the OE build process. I want to end up with Oct 14 09:54:47 a JFFS2 image that works. Oct 14 09:55:15 When I say "OE build process" I mean a cross-compilation on my X86 box. Oct 14 09:55:34 It would be NICE to be able to use Qt-embedded to develop applications completely on the LINUX box, then Oct 14 09:55:43 use OE to cross-compile and include in a disk image. Oct 14 09:55:58 The QT-embedded part works GREAT on my Fedora Core 3 system. Oct 14 09:56:18 Breaks on Fedora Core 4 due to the compiler incompatibility problem (apparently). Oct 14 09:56:51 Okay, but we are talking cross-compiling then (as Opie can also be built for x86)? Oct 14 09:56:55 Anyway, to repeat: I want to generate a JFFS2 image, cross-compiled from Fedora Core 4 environment. Oct 14 09:57:16 target is iPAQ/Familiar 0.8.2. Oct 14 09:57:23 In which case I'd rm -rf tmp and allow bitbake to regenerate the toolchains Oct 14 09:57:49 how will regenerating the toolchains help? Oct 14 09:57:59 On the Fedora Core 4 system? Oct 14 09:58:04 Because then it won't use 4.0.0 Oct 14 09:58:10 the FC4 system generates GCC4. Oct 14 09:58:27 That's the problem ... FC4 uses GCC4 and apparently generates the toolchain for GCC4. Oct 14 09:58:51 the timestamp on the 4.0.0 toolchain is earlier than that on the 3.4.4 which is supposed to be used Oct 14 09:59:30 which tells me that 4.0.0 existed, then 3.4.4 was built, and the later version (as in version number) is being used Oct 14 10:00:04 I think the fact that is has a pre-installed native 4.0.0 toolchain is by the by Oct 14 10:02:27 my rootfs doesn't include the shutil python module needed for running zope Oct 14 10:10:08 night Oct 14 10:23:04 OK. So removing the existing toolchain will cause bitbake to regenerate version 3.4.x? Oct 14 10:23:07 That doesn't make sense. Oct 14 10:23:39 What I'd like to do is to regenerate the 3.4.x toolchain. Oct 14 10:26:57 can anyone help me? Oct 14 10:27:23 I need to have python-shell ipk installed in my rootfs Oct 14 11:04:36 hi zecke Oct 14 11:05:28 hey Oct 14 11:05:58 how are you? Oct 14 11:06:15 good Oct 14 11:06:22 I just enrolled in the Microsoft Partner Program Oct 14 11:06:41 fear my partnership! Oct 14 11:11:49 pb_: What will you marry? a M$ mice or some kind of software? Oct 14 11:13:42 zecke: I was hoping to marry Bill Gates himself, but apparently I am too late for that Oct 14 11:14:15 however, it seems I am eligible for a Microsoft "Action Pack" as a kind of consolation prize. Oct 14 11:17:28 pb_: wow I bet you get a compiler for half of the price ;) Oct 14 11:18:01 yeah, something like that Oct 14 11:18:39 it costs 199 GBP, but you get ten licences each for Windows XP, Office Professional, Visual Studio, and a pile of other things, plus one licence each for Exchange and Windows Server. Oct 14 11:19:13 obviously the latter are only useful to point at and laugh, but the licences for xp and office are kind of handy to have. Oct 14 11:19:24 pb_: I'm part of this first shot free dealer program by M$ Oct 14 11:19:31 it is called MS Academic Program... Oct 14 11:19:37 ah right, yeah Oct 14 11:19:47 MS do seem to have a program for every eventuality. Oct 14 11:22:33 hi Oct 14 11:23:19 I have created a package inside packages/meta/ in order to define a set of packages Oct 14 11:49:49 do you get a license to kill? (like 007?) Oct 14 11:51:19 only at the higher membership levels Oct 14 11:53:45 hm Oct 14 11:53:47 is it just me, Oct 14 11:53:53 or do these things look silly? Oct 14 11:53:55 http://www.geekzone.co.nz/content.asp?contentid=5325 Oct 14 11:55:40 (they're the new keyboarded phones from nokia. concept is nice, execution, ehh) Oct 14 11:55:43 pb_: Did you ever create a MSI? Oct 14 11:56:27 zecke: not knowingly Oct 14 11:56:33 What's an MSI? Oct 14 11:57:00 micro soft installer? Oct 14 11:57:03 pb_: Miscrofot ??? Installer Oct 14 11:57:18 pb: ipk for the windows world Oct 14 11:57:20 ah, right Oct 14 11:57:47 I haven't created any windows installers for several years. I guess I used to use InstallShield, back in the day. Oct 14 11:58:05 is installshield an MS product, or third party? Oct 14 11:58:23 third party, though I think it (now) uses the windows installer at runtime Oct 14 11:58:55 older versions had their own special proprietary runtime stuff. I think you might still get that if you try to build an installer for windows 95 or some such. Oct 14 12:01:41 ever since winrar has been reasonably stable in handling tar (even tar.bz2) I've been using tar to package up many of my windows bits lately :/ Oct 14 12:01:53 * zecke searches for deployment in Visual Studio .NET Oct 14 12:02:11 (packages created under unix though, you might imagine) Oct 14 12:34:48 * CosmicPenguin read employment Oct 14 12:35:27 hi Oct 14 12:35:40 how often is the ewi546.ewi.utwente.nl snapshot updated? Oct 14 12:35:51 10 minutes? Oct 14 12:37:37 thanks Oct 14 13:15:50 does nybody else here build oe with amd64? Oct 14 13:20:01 few people Oct 14 13:20:27 I do Oct 14 13:20:27 heh Oct 14 13:20:31 obviously Oct 14 13:21:15 CosmicPenguin: when last time you used normal x86? (for build atleast) Oct 14 13:21:24 ;) Oct 14 13:21:57 Its been a while Oct 14 13:24:15 CosmicPenguin, ahh, i thought it was you, but then i thught it may be JustinP or both i spose.. anyhow.. do you get stuck at bluez_utils with the undefined LONG_MAX error ? Oct 14 13:24:44 hmmmm..... Oct 14 13:24:50 I'm pretty sure I've built bluez_utils before Oct 14 13:25:22 But its not in my local tree, so if I did, it didn't break Oct 14 13:26:41 sorry Oct 14 13:29:56 dan2003_: I also build on amd64 Oct 14 13:30:06 k, just im trying to build the gpe-image and it gets stuck there Oct 14 13:30:23 i mreber thif hapening a few months back when i was trying to build the simapd image Oct 14 13:31:36 dan2003_: where is LONG_MAX defined? Oct 14 13:31:44 in limts.h Oct 14 13:31:52 sorry, limits.h Oct 14 13:31:55 check that file on your system Oct 14 13:32:10 and check the build line in the log that's failing Oct 14 13:32:18 I bet it has a -I which points to your system headers Oct 14 13:32:28 should i be looking in my system files or the build tree ? Oct 14 13:32:41 and amd64 linux (on gentoo at least) has "hack" headers which include 64 or 32 bit headers conditionally Oct 14 13:32:49 the system one Oct 14 13:32:59 it's a stub probably...and that's probably the one that's included Oct 14 13:33:17 k Oct 14 13:34:23 lots of builds in OE need to be fixed not to include system headers Oct 14 13:34:29 but I have no idea hwo to fix them for real.... Oct 14 13:34:38 I've been band-aiding what I find Oct 14 13:35:56 k, ive deleted the tmp tree earlier.. so im building again.. but ill check when it gets back the the problem point Oct 14 13:50:48 hello all Oct 14 13:56:46 multiple profiles in the network app in gpe would be really handy ;) Oct 14 13:57:28 or just prfiles! abilty to choose between one of many stored network configs Oct 14 13:58:05 same with opie Oct 14 13:58:18 it has it or needs it? Oct 14 13:59:37 for the machine setting do i put h3600 or h3600.conf Oct 14 14:00:27 dan2003_: if you are thinking of wifi, gpe-wlancfg can do that. Oct 14 14:00:30 installing pcmciautils, I get "update-rc.d: /etc/init.d/coldplug: file does not exist". What's the problem? Oct 14 14:00:50 pb_, and i was just thinging about implementing it ! Oct 14 14:01:10 well, it'd be nice to implement it in gpe-conf Oct 14 14:01:11 heh Oct 14 14:01:30 the problem with gpe-wlancfg was that the profiles were a bit too "in your face", and simple-minded users found them confusing. Oct 14 14:01:50 k Oct 14 14:02:00 dan2003_: need Oct 14 14:02:12 ideally, you need some combination of the two: something that is as easy to use as gpe-conf for simple configs, but also allows you to do the gpe-wlancfg thing in more advanced cases. Oct 14 14:02:21 i need t learn gtk! so ill make that a little learning curve task Oct 14 14:02:28 mreimer: iirc it was removed Oct 14 14:02:36 dan2003_: very good Oct 14 14:02:38 a panel app which can select by essid would be handy Oct 14 14:02:51 yes, that would also be cool Oct 14 14:02:55 hrw: coldplug was removed? so pcmciautils needs a little love? Oct 14 14:03:10 you might check with florian in #gpe, since he is the wirelessmeister. Oct 14 14:03:44 mreimer: coldplug from pcmciautils Oct 14 14:04:16 hrw: do you know of a fix to get pcmciautils installed? Oct 14 14:07:03 mreimer: remove update-rc.d from its postinst Oct 14 14:07:10 thanks hrw Oct 14 14:11:56 dan2003_: any bugs like that that you see which say things are undefined which are normally defined (LONG_MAX, pid_t, etc) file a bug and make sure to mention amd64 in the title Oct 14 14:13:23 k Oct 14 14:13:50 ill try workk out howto fix it first, and fileit anyway if i get too stuck! Oct 14 14:20:34 dan2003_: if you figure it out file a bug as well (unless you have commit access) Oct 14 14:20:54 dan2003_: if you can figure out how to fix it for *real* (the system path isn't included) that would be best. A patch of some kind. Oct 14 14:21:07 dan2003_: if you can't figure that out, try a find/sed hack. Like: Oct 14 14:21:34 find ${S} -name Makefile | xargs sed -i 's:/usr/include:${STAGING_INCDIR}:' Oct 14 14:21:43 in do_compile_prepend() Oct 14 14:21:45 ::sigh:: Oct 14 14:21:56 * JustinP starts upgrading EFL again...I'm 3 "releases" behind! Oct 14 14:26:41 I'm trying to use 2.6.13 + pcmciautils to mount a CF card. Do I just need to plug it in and it will automatically mount? Right now I'm only seeing "add@/devices/platform/pxa2xx-pcmcia.0/0.0" but no /dev/hda etc. Oct 14 14:29:19 * CosmicPenguin fears the lm_sensors build system Oct 14 14:29:31 I can just feel in my bones that its going to be *very* OE unfriendly Oct 14 14:30:03 mreimer: on my c7x0 it happen automatically Oct 14 14:30:23 hrw: so udev 070, pcmciautils 010 is all I need? Oct 14 14:30:47 yep Oct 14 14:30:57 ok, I'll check the kernel side of things. thanks Oct 14 14:31:29 I have udev 070-r4 and pcmciautils 010-r0 Oct 14 14:46:37 ~lart appweb Oct 14 14:46:37 * ibot raises middle finger to appweb Oct 14 14:47:10 ~botsnack Oct 14 14:47:11 reenoo_: :) Oct 14 14:48:30 ~bothack Oct 14 14:48:31 brb, bricking your ipaq Oct 14 14:49:06 * raduga wonders what model of pda ibot is Oct 14 14:49:28 its a little too uppity for an ipaq Oct 14 14:49:40 little too bright and cheerful for a zaurus Oct 14 14:50:14 hm. maybe a Newton? Oct 14 14:50:18 I vote itsy Oct 14 14:50:42 could be. Oct 14 14:50:47 ~itsy Oct 14 14:50:48 well, itsy is the spritual predecessor of the iPAQ - a small research prototype. More information is at http://handhelds.org/Compaq/#Itsy_Summary Oct 14 14:51:58 the itsy actually rather resembles an ipod nano Oct 14 14:54:06 heh - the skiff is on there too Oct 14 14:54:13 I used the skiff toolchain for the longest damn time Oct 14 14:56:36 replacing PSU with other one made my PC more quiet.. Oct 14 14:56:38 hrw: would you mind showing me what you have in /etc/pcmcia? Oct 14 14:57:43 mreimer: bluetooth (755), bluetooth.conf, config.opts, hostap_cs.conf (644 all three) Oct 14 14:57:57 hrw: thanks, that's all I have too Oct 14 14:58:38 np Oct 14 15:06:05 hrw: one more question: what's in /etc/udev/rules.d? Oct 14 15:06:52 devfs-udev.rules local.rules permissions.rules udev.rules Oct 14 15:06:59 thanks again Oct 14 15:09:34 hrw: found it: just needed to load ide-disk Oct 14 15:12:39 mreimer: you ipaq people with lot of modules... Oct 14 15:12:39 ;) Oct 14 15:12:51 :-) Oct 14 15:13:51 I don't like Viva too much but atleast they dont censor videos Oct 14 15:16:14 RP: 2.6.14-rc1 & 2.6.14-rc4 -> "cs: pcmcia_socket0: time out after reset." after BT card insert == no card visible Oct 14 15:16:48 RP: pccardctl eject/insert helps Oct 14 15:17:28 RP: anyway dtl1_cs module is not loaded Oct 14 15:19:47 RP: hostap doesnot load too Oct 14 15:20:04 hrw: hostap should work :-/ Oct 14 15:20:19 I have noticed card ejection/insertion issues with 2.6.14-rc4 Oct 14 15:20:43 I'm hoping 2.6.14-rc1 isn't affected. If it is, we have a problem :-/ Oct 14 15:20:54 I have card ejection/insertion issues with my linksys WCF12 under 2.4.20 in general Oct 14 15:21:03 I'm not sure if its the card, or the kernel :/ Oct 14 15:21:25 RP: 2.6.14-rc4 hostap from kernel does not load Oct 14 15:21:28 card was fine on 2.4.18, but its not impossible the hardware has gone bad since then Oct 14 15:21:44 hrw: Ah, you need external hostap Oct 14 15:21:51 I've never got internal hostap working Oct 14 15:22:06 RP: internal one also does not allow to set WEP on Oct 14 15:22:41 hrw: As it doesn't work at all, that doesn't surprise me Oct 14 15:23:24 will install 0.4.4 external Oct 14 15:23:47 I recommend 0.3.9 Oct 14 15:23:57 i know ;) Oct 14 15:24:12 You just like to break things? :) Oct 14 15:24:25 like to test 0.4.4 Oct 14 15:24:32 Try and see if he firmware is loaded correctly with the internal module Oct 14 15:24:39 I suspect that's half the problem Oct 14 15:25:30 RP: I have card with firmware on card Oct 14 15:25:39 RP: did your ide-removable patch get accepted upstream? Oct 14 15:25:49 There goes that theory :) Oct 14 15:26:02 mreimer: It needs revising and I've not had time to do it Oct 14 15:26:07 Its 2.6.15 material anyway Oct 14 15:26:11 ok Oct 14 15:27:51 RP: we need to disable in-kernel hostap for release if it does not work Oct 14 15:28:04 hrw: It gets built but not installed Oct 14 15:28:06 all is fine Oct 14 15:28:40 think of those users which will ask: should I update hostap-modules-cs to kernel-module-hostap-cs or not? Oct 14 15:29:42 hm.. inserting card does not run anything. Oct 14 15:30:00 hrw: Feel free to disable it in the branch then. I'm indifferent... Oct 14 15:30:07 I'll update udev to r5 Oct 14 15:33:22 RP: you know that -rc4 does not reboot? Oct 14 15:33:35 "Rebooting... Restarting system." Oct 14 15:33:37 "." Oct 14 15:33:42 and stay on it Oct 14 15:33:46 hrw: Do any of the kernels? Oct 14 15:34:01 I usually use SysRq... Oct 14 15:34:06 RP: -rc1 atleast shutdown iirc Oct 14 15:34:49 can anybody tell me anythig about fc-cache/fc-cache within fontconfig? I cant really ask anythign more specific cos i dont have a clue wot it is, just that an issue is associated with it, anythign might be halpful Oct 14 15:35:14 hrw: I'm not aware of anything that's changed. We should probably try and track these bugs... Oct 14 15:35:17 well actually its fontconfig-native Oct 14 15:35:21 fc-cache as name say create cache for fontconfig Oct 14 15:35:35 I upgraded the -dev kernel to -rc4 so we could try and catch them... Oct 14 15:36:14 i'm using -dev one now Oct 14 15:36:43 k, its seems to be trying to write to /usr/X11R6/lib/X11/fonts .. which of coyrse causing problems! Oct 14 15:37:36 but i know a way round it and am detemined to sort out the bluez-util issue first, so ill come back to it later Oct 14 15:46:54 ARGH.. Oct 14 15:47:15 I loaded hostap_cs 0.3.9 by hand, 'ifup wlan0' by hand too Oct 14 15:47:34 and wep is not on (hostap_crypt_wep autoloaded after ifup) Oct 14 15:48:52 and iwconfig report encryption as off Oct 14 15:48:55 fsck Oct 14 15:50:52 screw it. I go sleep Oct 14 15:53:00 cu Oct 14 15:59:13 when bb says NOTE: make install, is that a genric term to says whats happening, or is that the actual command being executed? Oct 14 15:59:23 mickeyl: monotone.vanille.de seems to be down again.... Oct 14 16:05:31 how can i stop bitbake from building in interactive modde without quiting out of i altogtehr (ie ctrl+C) Oct 14 16:06:41 ctrl+c Oct 14 16:06:52 it shouldn't quit interactive mode Oct 14 16:06:56 doesn't for me Oct 14 16:07:21 oh ok, ill try it next time i need to. seconds after i asked it sttoppped anyway at the bluez-util probelem! Oct 14 16:15:55 I'm confused.. prefix=/usr is used all other the place, well it appears in the run , and lof files, where ithought it would have to be giving the path to the inclused withing the build system. for example Oct 14 16:15:57 NOTE: Running /home/mango/oe/build/tmp/work/bluez-utils-dbus-2.21-r2/bluez-utils-2.21/configure Oct 14 16:16:03 is followed by prefix=/usr Oct 14 16:16:26 as well as includedir and many others Oct 14 16:17:05 bie! Oct 14 16:17:32 dan2003_: sorry, not following Oct 14 16:17:44 good night Oct 14 16:19:09 Uhm.. im kinda thiking out loud, but basically i was expecting to see --includedir set to soething llike /home/mango/oe/build/tmp/staging/arm-linux/include where as its set to /usr/include Oct 14 16:19:38 and wondered if im lokkin gin the wrong place Oct 14 16:19:58 run.do_configure and the log for it Oct 14 16:20:42 but then if this is just wrong it would explain why it snot working Oct 14 16:20:46 dan2003_: you're saying that OE is setting /use/include? It *shouldn't* be doing that, no Oct 14 16:21:31 thast wot i thought.. im gonna check th elogs of somthign that has built a second.. Oct 14 16:23:21 hmm its the same for dbus --includedir=/usr/include Oct 14 16:24:03 maybe that's right then Oct 14 16:24:06 it very well may be Oct 14 16:24:26 the includedir while on the Z *is* supposed to be that...it's just the compiling include dir needs to be different Oct 14 16:24:46 im lost at how it can find the right includes.. ahh! Oct 14 16:24:50 yeah Oct 14 16:25:15 see i can see in tthe comile a -I/usr/include like you sugested ealier.. im tryign to track down hwere it comes from Oct 14 16:26:03 03justinp 07org.oe.dev * r214806aa... 10/: appweb: Make v2.0.4 compile with php 5.0.5 (this is hard-coded for now) Oct 14 16:26:40 hehe Oct 14 16:26:41 good luck Oct 14 16:30:25 the oe path to limits.h is included in the compile prior to the /usr/include path and i can see it defined with the file, so i dont get why its reported as udnefined unless its a compiler bug Oct 14 16:33:41 dan2003_: does the one in OE exist in the right place? Oct 14 16:34:01 dan2003_: are you *sure* that it's including the right one? Try altering the system one with some bad code and see if it fails there Oct 14 16:35:48 ill do that, but the irgh one is there.. ive just noticed someting that im sure is about to make me recall something! Oct 14 16:36:42 if (count > SSIZE_MAX) Oct 14 16:36:43 count = SSIZE_MAX; Oct 14 16:37:04 and SSIZE_MAX is defined as LONG_MAX somehwere Oct 14 16:37:18 and i seem to rmeebr having to change this before Oct 14 16:37:46 /usr/include/limits.h:1: error: parse error before "hlyyyyi34" Oct 14 16:37:46 hah Oct 14 16:38:01 thought so Oct 14 16:39:39 yet -I/home/mango/oe/build/tmp/staging/arm-linux/include exists before the -I/isr/include and does contain the file limits.h ! Oct 14 16:45:36 i just introdeced an error into the correct limits.h and that throws up an parse erro prior to LONG_MAX, so its usign one copy in one place and the other lateron Oct 14 16:57:35 the makefile for hcid ithin bluez-dbus has AM_FLAGS hardcoded, wheres in the common Makefile it isnt , i dleted -I/usr/include from this and then form the Make in sdpd also cos it then borked there, and now its continuing. AM_FLAG is from automake right? Oct 14 17:04:27 ive come to the conclusipon that or_runconf should not set --includedir=/usr/include ... but hey i dont know jack ;) Oct 14 17:05:05 off to bed zzZZZzz Oct 14 17:06:21 dan2003_: put it in a bug report Oct 14 17:06:28 dan2003_: let someone else sort it out ;-) Oct 14 17:08:55 hehe Oct 14 17:09:15 * JustinP wrote himself a little script to auto-update efl bb file versions Oct 14 17:09:32 I should have used python....oh well Oct 14 17:23:00 * JustinP marvels Oct 14 17:23:12 these efl packages are compiling without help so far.... Oct 14 17:23:21 no updated to patches even Oct 14 17:23:26 updates Oct 14 17:23:48 aw damn Oct 14 17:23:54 oh well, it was nice while it lasted Oct 14 22:33:54 Hello all .. Oct 14 22:34:18 I have one (hopefully quick) queston: how does one set the default cross compiler version within OE? Oct 14 22:34:40 We've got OE built on a fedora core 4 system which defaults to version 4.0. Oct 14 22:35:03 the tmp/bin tree shows that the cross compiler versions are present for 3.4.4 and 4.0 Oct 14 22:35:26 and if we try to "hand compile" at least one of the packages we're trying to compile, pcmcia-cs, does NOT Oct 14 22:35:33 compile under 4.0 but DOES compile under 3.4.4. Oct 14 22:35:40 anyone out there listening? Oct 14 22:39:07 .... Oct 14 22:39:41 check the local.conf Oct 14 22:55:14 OK. and look for what? Oct 14 22:55:25 I've got DISTRO="familiar" Oct 14 22:55:44 and PREFERRED_VERSION_gcc-cross = "3.4.4" in the "familiar.conf" Oct 14 22:55:51 doesn't seem to have an effect. Oct 14 22:56:13 cyberchuck, what do you mean by "hand compile" Oct 14 22:56:31 and what are you compiling for? Oct 14 22:56:47 hand compile means "invoke arm-linux-gcc-3.4.4" on the pcmcia-cs package Oct 14 22:57:08 outside the OE enviroment? Oct 14 22:57:11 I'm on Fedora Core 4 on X86 trying to compile for Familiar 0.8.2 on and iPAQ 3800 series. Oct 14 22:57:26 I'm trying to USE the OE environment for the cross compilation. Oct 14 22:57:31 you cant compile for familiar 0-.8.2 Oct 14 22:57:44 unless you have a very old repo Oct 14 22:57:47 the problem is that I get compiler errors on (at least) pcmcia-cs as well as konqueror. Oct 14 22:57:55 ok. Oct 14 22:58:27 when I specify familiar-0.8.3 in the DISTRO bitbake says "not supported". Oct 14 22:58:43 you can only use valid distros listed in org.openembedded.oe/conf/distro Oct 14 22:59:01 Ok. "familiar" is listed there. "familiar-0.8.3" is listed there. Oct 14 22:59:10 "familiar-0.8.3" kicks out immediately. Oct 14 22:59:27 "familar" works until pcmcia-cs, then gives an "lvalue" error and terminates the compilation process. Oct 14 22:59:28 what machine did you tell it? Oct 14 22:59:40 MACHINE="3600" Oct 14 22:59:49 thats wrong Oct 14 22:59:53 unless its changed Oct 14 22:59:58 its h3600 Oct 14 23:00:08 oops. "h3600" you're right (that's what i/we used) Oct 14 23:00:19 TARGET_ARCH is unset Oct 14 23:00:22 (default) Oct 14 23:00:32 TARGET_OS="linux" Oct 14 23:00:34 interesting because thats what i have set Oct 14 23:00:44 sec i'll update my db and retry Oct 14 23:01:08 and you're using ?what? for your DISTRO ? Oct 14 23:01:24 and, most importantly what distro and version of X86 LINUX are you using? Oct 14 23:01:39 We're using Fedora Core 4, which defaults to GCC 4. Oct 14 23:01:52 Which bombs qt-mebedded consistently. Oct 14 23:01:56 familiar-0.8.3 and h3600 Oct 14 23:01:57 qt-embedded. Oct 14 23:02:04 interesting. Oct 14 23:02:40 i have gcc 3.4 and gcc4 avail Oct 14 23:02:42 er Oct 14 23:02:43 3.3 Oct 14 23:03:05 OK. Here's what I'm getting right now. Oct 14 23:03:14 RROR: Remove this line or set a dummy DISTRO_VERSION if you really want to bui\ Oct 14 23:03:14 ld an unversioned distro Oct 14 23:03:14 NOTE: exceptions.SystemExit:1 while evaluating: Oct 14 23:03:14 ${@bb.data.getVar("DISTRO_VERSION",d,1) or bb.fatal('Remove this line or set a \ Oct 14 23:03:14 dummy DISTRO_VERSION if you really want to build an unversioned distro')} Oct 14 23:03:18 ugh i should update more often ... Oct 14 23:03:43 which is what you should get from using "familiar" Oct 14 23:04:22 I'm set to "familiar-0.8.3" in "/stuff/build/conf/local.conf" Oct 14 23:04:37 BBPATH is /stuff/build:/stuff/org.openembedded.dev Oct 14 23:04:38 and you removed tmp or atleast tmp/cache Oct 14 23:04:47 NO. Not yet. Oct 14 23:04:55 ./tmp, right? Oct 14 23:04:58 you need to after ANY change to local.conf Oct 14 23:05:06 umm Oct 14 23:05:09 Ahhhhhh. That's a big part of the problem then. Oct 14 23:05:12 build/tmp Oct 14 23:05:21 is where it normally is Oct 14 23:05:37 there IS no "build/tmp". Oct 14 23:05:46 I put everything under "/stuff" as suggested. Oct 14 23:05:50 echo $BBTMP Oct 14 23:06:02 empty. Oct 14 23:06:07 Should that be set to something? Oct 14 23:06:10 yeah that doesnt look right Oct 14 23:06:16 no Oct 14 23:06:18 sec Oct 14 23:07:06 The effect we're getting is that TMP is put in `pwd`/tmp Oct 14 23:07:28 should I go ahead & remove /stuff/tmp or at least /stuff/tmp/cahce ? Oct 14 23:07:34 well what you have set yo build in, yes Oct 14 23:07:52 i would probably toss all of tmp Oct 14 23:07:57 OK. Will do that now. Oct 14 23:08:06 you dont really know what state it is in Oct 14 23:08:12 true. Oct 14 23:08:40 I didn't see anything in the OE wiki that indicated that I had to remove "$PWD/tmp" whenever I made a change to "local.conf". Oct 14 23:08:44 120 to go ... whee Oct 14 23:09:03 most people only set it once Oct 14 23:09:05 Is that a general rule? If so can I add this step to the wiki? Oct 14 23:09:12 set ?what? once? Oct 14 23:09:20 and most cant follow the wiki anyway Oct 14 23:09:29 local.conf Oct 14 23:09:58 ok. I removed "tmp" and still got tERROR: Remove this line or set a dummy DISTRO_VERSION if you really want to bui\ Oct 14 23:09:58 ld an unversioned distro Oct 14 23:09:58 NOTE: exceptions.SystemExit:1 while evaluating Oct 14 23:10:06 seems lots of people have windows-itis Oct 14 23:10:38 you have it set "familiar-0.8.3" Oct 14 23:10:48 yes. In /stuff/build/conf/local.conf Oct 14 23:11:10 your enviroment sounds wrong ... Oct 14 23:11:17 echo $BBPATH Oct 14 23:11:20 OK. Oct 14 23:11:38 BBPATH = /stuff/build:/stuff/org.openembedded.dev Oct 14 23:12:21 hmm Oct 14 23:12:29 what's your BBPATH? Oct 14 23:12:41 and are there more ENV variables I should be setting ? Oct 14 23:12:54 export BBPATH=/backupfiles/biohazard/openembedded/build:/backupfiles/biohazard/openembedded/org.openembedded.dev Oct 14 23:13:01 None that I can see in the "Getting Started" OE Wiki ... Oct 14 23:13:19 most have been eliminated Oct 14 23:13:21 ok. Structure looks similar to mine. Oct 14 23:13:24 yeah Oct 14 23:13:52 You 'have "/backupfiles/biohazard/openembedded/build" and I have "/stuff/build" (first part) Oct 14 23:14:14 pattern looks the same tho. Oct 14 23:14:15 i am curious Oct 14 23:14:22 and your "local.conf" is where? Oct 14 23:14:26 in the "build" subdirectory? Oct 14 23:14:29 did you comment/remove the last line in your local.conf? Oct 14 23:14:34 Yeah, I'm curious too .. and going mildly crazy. Oct 14 23:14:46 No I didn't. Should I have ? Oct 14 23:14:57 whats the last two lines say? Oct 14 23:15:11 take it back. I did comment out the last line Oct 14 23:15:24 #REMOVE_THIS_LINE:="${@oe.fatal('Read the comments in your conf/local.conf')}" Oct 14 23:15:49 build/conf/local.conf Oct 14 23:15:59 that's where mine iz. Oct 14 23:16:19 your permissions are all the same ? Oct 14 23:16:24 and I nuked the whole "/stuff/tmp" directory, which of course includes the cache directory as well. Oct 14 23:16:31 yeah Oct 14 23:17:10 user:group wise Oct 14 23:17:49 cmon monotone ... blech 1/2 way Oct 14 23:18:35 um, I just checked, and I must have set the wrong "local.conf" Oct 14 23:18:43 ? Oct 14 23:18:44 when I just changed to "familiar-0.8.3" I got : Oct 14 23:18:50 ERROR: Unable to parse conf/bitbake.conf (conf/distro/familiar-0.8.3.conf:7: un\ Oct 14 23:18:51 parsed line: '"${@bb.fatal('This file is unmaintained in the org.oe.dev branch,\ Oct 14 23:18:51 use the org.openembedded.oz354fam083 branch instead. You can also remove this \ Oct 14 23:18:51 line, but don't complain if something breaks')}"') Oct 14 23:19:30 oz354fam083 branch ? Oct 14 23:19:48 that's what it says. Oct 14 23:19:57 That branch DOES exist in the monotone library. Oct 14 23:20:04 monotone "repository" ... Oct 14 23:20:06 yeah i expect it does Oct 14 23:20:23 But did you have to do this? Oct 14 23:20:25 you need to pull it Oct 14 23:20:45 comment out the line or pull the oz354fam083 branch? Oct 14 23:21:10 pull the branch by the looks Oct 14 23:21:15 OK. I just uncommented the line. Oct 14 23:21:20 i'll find out for sure here in a min Oct 14 23:21:21 but HOW did you build yours? Oct 14 23:21:28 Or did you attempt a build of an image? Oct 14 23:21:36 We're trying to do a "bitbake opie-image" Oct 14 23:21:50 it DOES do a LOT of the directory tree, but croaks on "pcmcia-cs". Oct 14 23:21:55 i build gpe images generally Oct 14 23:21:57 and , apparently, konqueror. Oct 14 23:22:00 Ah. OK. Oct 14 23:22:18 Well, I uncommented out the "bb" line. Will see what happens. Oct 14 23:22:30 If this doesn't work I'll pull the other repository. Oct 14 23:22:46 Thanks for the sympathetic shoulder / keyboard. Oct 14 23:23:01 i should read the mailing list for monotone changes again i suppose Oct 14 23:23:35 I've been banging on this for about a week now (night stuff, I have a "day job" 9-6PM). Oct 14 23:23:54 It takes SO long to do a from-scratch build ... Oct 14 23:24:30 even on a reasonably fast (2Ghz) machine. Oct 14 23:26:22 anyhooo ... need to go hit the sack. Many thanks for the chat/suggestions; blowing away the "tmp" definitely helped things move along. At least now I know to do that when I change the "local.conf". Oct 14 23:28:08 hmm Oct 14 23:28:17 not seeing anything about taht branch ... Oct 14 23:29:51 When I checked the Monotone repository (the main one from the Getting Started" secton of the Wiki) I DID see that branch. Oct 14 23:29:55 But I did NOT check it out (yet). Oct 14 23:30:17 hmm Oct 14 23:31:19 i dont see that branch mentioned Oct 14 23:32:38 http://oe.handhelds.org/cgi-bin/moin.cgi/Oz354Fam083CherryPicking?action=highlight&value=oz354fam083 Oct 14 23:32:44 see http://monotone.vanille.de/viewmtn/ Oct 14 23:32:45 looks liek an inprogress issue Oct 14 23:34:20 meaning? Oct 14 23:34:49 I went to that bug report, don't see how it helps me ... looks like some developers are working on the other branch. Oct 14 23:34:49 they are branching it for some reason i guess Oct 14 23:34:57 gotcha. Oct 14 23:35:06 But back to my initial point: Oct 14 23:35:09 was posed on the 9th Oct 14 23:35:29 pcmcia-cs fails with an 'Lvalue" compiler syntax error. Oct 14 23:35:34 so yeah you need to pull that branch Oct 14 23:35:36 The arm-linux-gcc version Is $.0 Oct 14 23:35:39 is 4.0 Oct 14 23:35:59 If I use "arm-linux-gcc-3.4.4" on the SAME source file it compiles without an error. Oct 14 23:36:22 So how can I force the compiler to be "3.4.4" ? Which is in the "cross/bin" directory. Oct 14 23:36:31 PREFERRED_VERSION_gcc ?= "3.4.4" Oct 14 23:36:31 PREFERRED_VERSION_gcc-cross ?= "3.4.4" Oct 14 23:36:31 PREFERRED_VERSION_gcc-cross-initial ?= "3.4.4" Oct 14 23:36:40 from the 0.8.3.bb Oct 14 23:36:50 That is in the "familiar.conf" that I modified. Oct 14 23:37:08 And I"m using "familiar-0.8.3.conf" right now. Oct 14 23:37:36 ERROR: Unable to parse conf/bitbake.con Oct 14 23:37:38 how so? Oct 14 23:37:39 You got me further along (thanks again for that) by suggesting the removal of the "tmp" directory tree. Oct 14 23:37:53 Huh? Oct 14 23:37:58 "Unable to parse" ? Oct 14 23:38:01 On YOUR end or MINE? Oct 14 23:38:05 cyberchuck ERROR: Unable to parse conf/bitbake.conf (conf/distro/familiar-0.8.3.conf:7: un\ Oct 14 23:38:06 cyberchuck parsed line: '"${@bb.fatal('This file is unmaintained cyberchuck ERROR: Unable to parse conf/bitbake.conf (conf/distro/familiar-0.8.3.conf:7: un\ Oct 14 23:38:06 cyberchuck parsed line: '"${@bb.fatal('This file is unmaintained in the org.oe.dev branch,\in the org.oe.dev branch,\ Oct 14 23:38:19 yours Oct 14 23:38:31 30 to go ... Oct 14 23:38:31 *I* have gotten past that. Oct 14 23:38:35 ah Oct 14 23:38:51 I had to comment out the "bb" line in "familiar-0.8.3.conf". Oct 14 23:38:53 then you have overriden gcc in your local.conf Oct 14 23:39:09 The one that said "comment out but don't blame us if it breaks". Oct 14 23:39:14 yes. Oct 14 23:39:32 But I wasn't seeing the effect 'cuz I wasn't wiping out the "tmp" directory tree. Oct 14 23:39:41 So once again I thank you ... I got past that now. Oct 14 23:39:49 The compilation is working as we speak. Oct 14 23:39:49 np Oct 14 23:40:10 It'll take at least an hour or so to get to the same point it wuz (pcmcia-cs). Oct 14 23:40:13 and all guessing before monotone finished Oct 14 23:40:19 crossin' my fingers & toes. Oct 14 23:40:24 good. Oct 14 23:40:38 I haven't checked real closely .. do I have to apply to be a Wiki author for OE? Oct 14 23:40:39 well a proper setup should compiel fine Oct 14 23:40:43 no Oct 14 23:41:10 good. Then I'll modify via the Wiki with what I've learned as soon as I get a clean compilation (*assuming* Oct 14 23:41:15 I get a clean compliation :-0) Oct 14 23:41:24 compilation :-0 Oct 14 23:41:33 whats your machine? Oct 14 23:41:52 the iPAQ target is a 368x (I forget what "x" is). Oct 14 23:41:55 host Oct 14 23:42:10 My development machine (host) is a generic X86 running Fedora Core 4. Oct 14 23:42:46 I have qt-embedded for X86 running on it, which means I can develop and Oct 14 23:42:54 ment cpu+ram ,,, takes ~2 hours i think to build a task-bootstrap on mine Oct 14 23:42:54 do UI testing all on my LINUX desktop. Oct 14 23:43:14 3g 3gb Oct 14 23:43:16 we're doing a "opie-image" build. Quite a list of packages. Oct 14 23:43:47 er 2gb Oct 14 23:43:49 Let's see ... 256M Oct 14 23:43:56 ouch Oct 14 23:44:05 (but doeesn't seem to be RAM-bound) Oct 14 23:44:09 yeah you need to buy some ram Oct 14 23:44:23 Speed = 1.8Ghz Oct 14 23:44:34 "top" shows 4M free out of the 256M. Oct 14 23:44:45 and prob a full swap Oct 14 23:44:50 I have another machine (mine at home) which is 1.5Ghz and has 1G of RAM. Oct 14 23:45:17 no, swap is 100M out of 1G allocated. Oct 14 23:45:30 so it doesn't feel RAM-bound. Oct 14 23:45:34 thats not a good allocation ... Oct 14 23:45:48 CPU bound I can believe, maybe ... Oct 14 23:45:51 which means? Oct 14 23:45:57 (good allocation? ) Oct 14 23:46:05 top says Oct 14 23:46:05 you could over buffer Oct 14 23:46:16 which would slow your system down more Oct 14 23:46:24 255700k total, 251272 used, 4932K free Oct 14 23:46:34 but what does free say? Oct 14 23:46:40 swap: 1048568k total, 106624k used 941944k free Oct 14 23:46:45 newer mem managment probably prevents it ... Oct 14 23:46:55 prevents it ? Oct 14 23:47:08 We can get more RAM if it would help the compilation ... Oct 14 23:47:09 prevents allocating more swap than physical ram Oct 14 23:47:18 ah. Ok. Oct 14 23:48:01 cpu is 65% us (user), 25% "wa" (waiting?) Oct 14 23:48:25 between .3% and 10% sy (system) Oct 14 23:49:10 ah. Good. Looks like the bitbake is paying attention to my preference to 3.4.4 compiler. Oct 14 23:49:17 It isn't building the 4.0 compiler at all. Oct 14 23:50:03 this is a Good Thing. Oct 14 23:50:32 whee 52 more revs Oct 14 23:50:44 well, it looks like it is churnin away now. See you on the channel again soon, will modify the Wiki if this works. Oct 14 23:50:56 52 more revs to retrieve (in Monotone) you mean? Oct 14 23:51:04 for oz354fam083 yeah Oct 14 23:51:11 Ok. Oct 14 23:51:20 just added it Oct 14 23:51:55 Could you EMAIL me when you get a chance? "oe-irc@cyberchuck.org" (my throwaway EMAIL for this session) will getcha to me, and then we can exchange "real" EMAILs beyond this. Oct 14 23:52:30 I'd like to do some more brainstormin' with you, partiuclarly after I get this image built . Oct 14 23:52:47 Among other things I'm curious as to your cut on GPE "vs" QT. Oct 14 23:52:52 not sure how useful i'd be, but emte@labotomy.net Oct 14 23:53:32 well originally gpe was a smaller image :) Oct 14 23:53:47 but it does have less goodies Oct 14 23:54:00 mostly due to fewer developers i'd think Oct 14 23:55:01 I hearya. I really like the QT-embedded environment so far (have really just started using it). Oct 14 23:55:08 is there a corresponding environment for GPE? Oct 14 23:55:30 ... gpe is Oct 14 23:55:52 http://gpe.handhelds.org/ Oct 14 23:58:51 Ok. But after a quick look there, i don't see an equivalent to "qt-embedded". Oct 14 23:59:20 Qt-embedded lets you have a totally X86-based development environment. It simulates the screen and stylus (not sure about network) Oct 14 23:59:36 which means you can do all your development and UI testing on X86. Oct 14 23:59:36 ah Oct 14 23:59:44 i take it you dont liek xnest? Oct 14 23:59:55 THEN use OE to cross-compile. Oct 15 00:00:18 tho i think xem might work now as well Oct 15 00:00:34 xen* Oct 15 00:00:39 "xnest" is OK, but if you mean that you have to develop and debug on your target machine then no, I don't use xnest for that. Oct 15 00:00:43 xen? Oct 15 00:01:03 I don't think XEN will do cross-architecture stuff (yet) Oct 15 00:01:09 Great software, tho. Oct 15 00:01:22 Fedora Core 4 has XEN built-in, although you have to enable it. Oct 15 00:01:31 ANd I haven' been able to take the time to do taht yet. Oct 15 00:01:59 i suppose gpe is a bit frendlier to generic programs than opie/qpe Oct 15 00:02:23 since it is using the x11 server Oct 15 00:02:28 The major advantage i see to GPE is that it is based on X, whereas Opie is based on Qt. Oct 15 00:02:40 I agree that porting is probably easier. Oct 15 00:02:52 to GPE from X-based progarms. Oct 15 00:02:54 programs. Oct 15 00:03:04 back and forth for gtk anyway Oct 15 00:03:16 but if you 'r estarting from scratch, "designer" (ui builder) is quite nice so far (again, "so far"). Oct 15 00:03:37 gotta git for now. Oct 15 00:03:41 get some zzzzzs. Oct 15 00:03:43 night Oct 15 00:03:50 thanks for the chat. EMAILed you already ... Oct 15 00:03:57 ACK if you want to continue the conversaton(s). Oct 15 00:04:02 * emte hit sylpheed up Oct 15 00:04:16 bye ... Oct 15 00:04:19 ttys. Oct 15 02:35:01 Hi, something is strange with my OE sources. I downloaded OE.bz2 a couple of days ago since I had not pulled in quite a while and wanted to spare myself the trouble of waiting for days. Now, what I get is at http://pastebin.ca/25563 **** ENDING LOGGING AT Sat Oct 15 02:59:56 2005