**** BEGIN LOGGING AT Sun Aug 13 02:59:57 2006 Aug 13 03:36:12 03lenehan 07org.oe.dev * r9afd3807... 10/ (1 packages/man/man_1.5p.bb): man 1.5p: Remove the spurious .debug directory from the main package. Aug 13 03:52:22 03lenehan 07org.oe.dev * r7e448654... 10/ (1 packages/pcre/pcre_4.4.bb): pcre 4.4: Remove the spurious .debug directory from the main package. Aug 13 06:02:51 03xora 07org.oe.dev * r694c7ac9... 10/ (3 files in 3 dirs): Aug 13 06:02:51 xmms_1.2.10.bb : remove extra quotes in xmms-config so the binconfig Aug 13 06:02:51 path munging works. Aug 13 06:02:53 03xora 07org.oe.dev * rd528cddc... 10/ (1 packages/xmms/xmms-mad_0.10.bb): xmms-mad_0.10.bb : add the mad xmms plugin Aug 13 07:24:48 ~seen mr_nice_slacker Aug 13 07:24:54 mr_nice_slacker was last seen on IRC in channel #oe, 10h 33m 13s ago, saying: 'woglinde: but i hope it will work for x29a as well'. Aug 13 08:12:30 morning all Aug 13 08:13:28 http://www.rpsys.net/openzaurus/temp/madness1.log :) Aug 13 08:19:55 RP: whats that? Aug 13 08:20:25 mithro: bitbake using multiple threads Aug 13 08:24:15 cool Aug 13 08:24:19 i guess that, but wasn't sure Aug 13 08:24:22 does it work well? Aug 13 08:24:32 mithro: Not sure yet :) Aug 13 08:25:10 in theory if the dependency graph is correct, I can't see what would cause problems? Aug 13 08:26:24 hey mithro & RP Aug 13 08:27:58 mithro: in theory, it should just work :) Aug 13 08:33:12 It certainly feels faster.... Aug 13 08:33:52 ah, the gentoo feeling ;) Aug 13 08:34:34 Its locked my processor at 100%... Aug 13 08:35:06 cat "RP" > /sys/class/bitbake/scheduler Aug 13 08:37:07 RP: is the number of threads configurable? Aug 13 08:37:18 and should be petition a T2 from sun yet? Aug 13 08:38:29 koen: At the moment its a hardcoded number but it is easily changed to be Aug 13 08:38:59 * koen votes for PARALLEL_BAKE Aug 13 08:39:06 koen: Ideally, we'll be able to control the number of different types of tasks running based on their properties (disk activity, network activity etc) Aug 13 08:39:16 koen: not SUPER_BAKE_OFF Aug 13 08:39:25 :) Aug 13 08:39:40 XorA: nice work on xmms Aug 13 08:40:13 koen: its what happens when a user asks for features, and seeing as none of the other GPE media players seem to work it was obvious that it needed done Aug 13 08:40:22 At the moment its very alpha - the code has about five areas marked "FIXME" :) Aug 13 08:40:42 RP: do you want to put that info in to the recipe, or put a scheduler in bitbake that looks at IO and cpu Aug 13 08:40:43 ? Aug 13 08:41:13 koen: It'll probably be a conf file alongside (or in) local.conf Aug 13 08:41:15 * XorA wonders why lm-sensors seems to think his fan is either spinning at -1 RPM or 12000 RPM Aug 13 08:41:50 XorA: stop rotating your case ;) Aug 13 08:42:16 there used to be a calibration file for lmsensers Aug 13 08:42:33 koen: I think the it87 sensor just doesnt quite work, the fan is actually off Aug 13 08:42:39 koen: atm, I'll settle for adding a single thread version to bitbake head using the new dependency tree code. We can then worry about the multi thread version :) Aug 13 08:42:49 when the fan runs, the speed is more beleiveable Aug 13 08:43:15 12,000rpm isn't that fast. My bike engine redlines at 14,500 :) Aug 13 08:43:16 hi all Aug 13 08:43:35 RP: at 12,000 RPM I think my shuttle would be sliding across the desk :-) Aug 13 08:43:57 hi eFfeM Aug 13 08:44:02 XorA: I used to have hw temperature control board attached to the cpufan and psufan Aug 13 08:44:05 hey eFfeM Aug 13 08:44:07 now at 84000 RPM Aug 13 08:44:51 XorA: on a cyrix 133MHz, it could be passively cooled when using an athlon heatsink :) Aug 13 08:45:27 koen: I have found the shuttle will passive cool when idle, just needs to ramp fan up when Im baking Aug 13 08:47:10 XorA: add a hook in bitbake to enable/disable the fan :) Aug 13 08:47:49 RP: does your code fix 'bitbake -s' in HEAD? Aug 13 08:48:33 koen: now that would be a cool hack :-) INHERIT += "fan" Aug 13 08:50:33 RP: did you try Liams new version yet? Aug 13 09:10:15 * koen likes the 'conservative' scheduler for cpufreq Aug 13 09:11:19 * koen likes the 16hours of (estimated) battery life of his h2200 Aug 13 09:14:24 koen: 16 hours bloody hell, my Z only goes about 4-5 Aug 13 09:14:35 * XorA needs more PDA's Aug 13 09:15:20 XorA: try dimming the backlight more aggressively Aug 13 09:15:38 the h2200 has a transflective screen, so ambient light helps a lot Aug 13 09:23:55 koen: do you actually get 16hours? Aug 13 09:26:00 ~ping mr_nice_slacker Aug 13 09:26:04 * XorA gets 48-72 with his htc wizard Aug 13 09:26:06 pong mr_nice_slacker Aug 13 09:26:25 ping mr_nice_slacker Aug 13 09:26:31 mithro: after 2 hours it said 14 hours left, and it goes up and down when using it Aug 13 09:26:54 mithro: it looks pretty accurate, but I haven't done a real big test Aug 13 09:27:28 leave it running for 16 hours? Aug 13 09:27:59 koen: I saw that bug and will aim to look into it at some point Aug 13 09:28:03 XorA: not yet Aug 13 09:28:32 RP: scenario stuff doesnt work with new asoc, but I just compiled it, I see if I got time to run it on device before I get picked up Aug 13 09:29:06 XorA: I've not tried the scenario code - I'll settle for ASoC working for now :) Aug 13 09:30:17 RP: I havent tried it, but I had it compiled, but the semaphore changes broke it Aug 13 09:31:06 03eFfeM 07org.oe.dev * r73a0fd7a... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages.bb: readded madwifi to the packages list as it is fixed Aug 13 09:36:51 anyone here? Aug 13 09:37:07 dan: no Aug 13 09:37:22 hi Graeme! Aug 13 09:37:32 good to hear about xmms Aug 13 09:38:02 but without the double size screen thing working it is too small to use on a 640x480 display :( Aug 13 09:38:26 no, it's too small on a 200dpi screen Aug 13 09:38:38 it's the right size on my 800x600 100dpi screen Aug 13 09:38:47 dan: maybe you could scour google and see if its a known bug and find me a patch, Im out this afternoon :-) Aug 13 09:39:13 I'll have to get either gpe-vlc or pymp running otherwise Aug 13 09:39:43 I only have 2 probs with OZ now Aug 13 09:41:00 RP: damn, that new asoc patch gives me static Aug 13 09:41:21 XorA: :-( Aug 13 09:42:20 RP: I shall have to see about getting the machine to lrg before static randomely dissapear again Aug 13 09:42:27 1- I can't get a GUI tool for turning my wifi card off to work. CF/SD card monitor crashes my Z if I try ejecting the CF and the PC card tool doesn't let me eject despite asking for root password (which is no good) Aug 13 09:42:51 2- I want to log in to GPE as root Aug 13 09:43:04 then I'll be happy Aug 13 09:44:00 XorA: That might help although if you change anything, you'll not know if it was the code changes or just random :-/ Aug 13 09:45:19 RP: Im altbooted so I can do dev work on my normal flash while static occurs on altboot Aug 13 09:45:44 XorA: :). I should really play with altboot on spitz Aug 13 09:45:55 surely the graphical ejection of cards isn't still on the TODO list? Aug 13 09:47:52 dan: found a fix for doulbe size crash Aug 13 09:48:03 whoop! Aug 13 09:48:04 dan: XLIB_SKIP_ARGB_VISUALS=1 xmms Aug 13 09:48:10 to run program Aug 13 09:48:38 can't that option be compiled in? Aug 13 09:48:42 XorA: could you add that to the .desktop? Aug 13 09:48:55 koen: what .desktop :-) Aug 13 09:50:01 koen: xmms needs an icon and a .desktop, and Im about to go out to lunch Aug 13 09:50:25 thanks to gentoo for that tip Aug 13 09:50:48 03koen 07org.oe.packaged-staging * r45e665c8... 10/ (1 README docs docs/packaged-staging.xml): convert README to docbook and move it to docs/ Aug 13 09:50:52 03koen 07org.oe.packaged-staging * r248a24a5... 10/ (1 docs/Makefile docs/html.css): docs/: reuse makefile and css from org.oe.documentation Aug 13 09:50:52 If docbook doesn't annoy me to much I can have a look at a .desktop Aug 13 09:50:56 03koen 07org.oe.packaged-staging * rbe4c1459... 10/ (1 docs/packaged-staging.xml): docs/packaged-staging.xml: make it validate Aug 13 09:51:00 03koen 07org.oe.packaged-staging * r51d026a1... 10/ (1 docs/packaged-staging.xml): docs/packaged-staging.xml: add markup for code and patches Aug 13 09:51:14 and of course: Aug 13 09:51:17 * koen -> lunch Aug 13 09:51:19 :) Aug 13 09:51:55 I'm going to set up a OZ dev environment but OZ will be useless to me if I can't get these two probs ironed out- I'll both answers to the FAQ as that's where they should be Aug 13 09:52:06 add both Aug 13 09:52:39 dan: setting up an dev environment is always good, means you can port apps accross yourself then submit the .bb changes if needed Aug 13 09:53:04 dan: Im off work tomorrow so Ill look at making you an icon then Aug 13 09:53:10 XorA: Yeah, but being able to write and eject discs under OZ is even more critical Aug 13 09:53:19 XorA- thanks! Aug 13 09:54:09 dan: there is no fancy gui on gpe, but cardctl eject works well Aug 13 09:54:18 dan: or just pull an SD card out Aug 13 09:55:31 XorA: Oh your joking! This has instantly eclipsed any other missing OZ feature! This is the one major feature missing from having a fully graphical envirionment Aug 13 09:56:17 I'm stunned its got so far without this feature Aug 13 09:56:25 dan: you could try matchbox-applet-cards Aug 13 09:56:53 XorA: Doesn't work- I'll file a bug now I've been told this Aug 13 09:57:12 I just presumed GUI card management MUST work, long ago Aug 13 09:57:40 dan: I eject cards like once a week at most :-) Aug 13 09:58:32 anyway time to go downstairs and wait for car Aug 13 09:58:56 I frequently hop on and off the net on my Z, and having to open a console, log in as root and then issue a command is a lot of hassle for something that could be done with one/two clicks/taps Aug 13 09:59:07 bye XorA! Aug 13 09:59:44 what about logging into OZ as root? Aug 13 10:00:11 I mean logging in under gpe-dm as root Aug 13 10:02:13 I don't really understand the connection between hopping on and off the net and this business of ejecting cards. Aug 13 10:02:21 Can you explain that a bit more? Aug 13 10:03:47 I've got the logging in as root bit sorted- so its just the card problem now Aug 13 10:03:54 pb_: yes Aug 13 10:04:27 I want the matchbox card tool to work as it does under pdaXrom Aug 13 10:05:27 which is how? Aug 13 10:05:36 I guess most people here don't know a whole lot about pdaXrom Aug 13 10:05:43 I want OZ/GPE to be fully operational without dropping to the command line and so this tool needs to work Aug 13 10:05:52 hold on Aug 13 10:06:53 At the moment, if I'm logged in as root or a normal user, the CF/SD monitor crashes OZ when I click on either of the cards Aug 13 10:07:42 on top of that it on says 'CF card' regardless of there being a wifi card or CF memory card inserted Aug 13 10:08:20 besides I think it is actually referred to the internal MD and thats why OZ freezes when I click on it Aug 13 10:09:17 under cacko, when I close the web browser, it asks if I want to close my net connection, hence turning off my wifi card (which drains much power of course) Aug 13 10:10:14 under pdaXrom, I just click on the matchbox card applet then click on my wifi card icon and it turns it off, or unmounts it if it is a CF memory card Aug 13 10:10:28 okay, I see Aug 13 10:10:50 so what you want is a way to turn the wifi card off without physically unplugging it. right? Aug 13 10:11:00 yes Aug 13 10:11:11 and without going to the cli Aug 13 10:11:30 2 clicks is all it should take Aug 13 10:12:51 right Aug 13 10:12:54 please do file a bug report for that Aug 13 10:14:10 i most certainly will, and I'm pumping the priority right up :) Aug 13 10:14:16 heh Aug 13 10:14:50 hey Aug 13 10:15:03 pb_: mine as well Aug 13 10:15:04 hey zecke! Aug 13 10:15:14 zecke: rock Aug 13 10:15:31 and it is funny that to build Qtopia, X11 is required... Aug 13 10:15:43 ?? Aug 13 10:16:00 that was totally off topic Aug 13 10:16:09 I'm struggling compiling Qtopia2.2. (once again) Aug 13 10:16:45 03koen 07org.oe.dev * r7af26a86... 10/ (1 packages/mikmod/mikmod_3.2.2-beta1.bb): mikmod: depend on ncurses Aug 13 10:16:50 03koen 07org.oe.dev * rcca0d62f... 10/ (5 files in 4 dirs): chmlib: add 0.38 Aug 13 10:17:28 sigh, where has all my memory gone again? Aug 13 10:17:31 * pb_ stabs firefox Aug 13 10:17:54 ah yeah, firefox has 1500MB and evolution has 700MB Aug 13 10:18:32 pb_: Nobody would ever need more than 640kb of RAM Aug 13 10:18:41 zecke: right Aug 13 10:18:58 opie still runs in 640k, I gather Aug 13 10:19:13 pb_: actually we have decided to stay below 320kb Aug 13 10:19:18 good plan Aug 13 10:19:29 leave some room for future expansion Aug 13 10:19:36 pb_: only this bloody linux counts the mmapped framebuffer as used memory as well... Aug 13 10:19:47 yeah, crazy linux Aug 13 10:20:40 pb_: there is one solution to firefox. Apply mallums patch and view the serverside pixmaps Aug 13 10:20:51 which patch is that? Aug 13 10:21:13 fwiw, the problem I am having is with client side bloat, not server side Aug 13 10:21:22 pb_: eek Aug 13 10:21:30 1.5gb is much Aug 13 10:21:38 root 4457 0.9 6.1 267348 126768 ? S Aug01 152:21 /usr/X11R6/bin/X :0 -dpi 96 -audit 0 -auth /var/lib/gdm/:0.Xauth -nolisten tcp vt7 Aug 13 10:21:44 pb 11560 0.6 34.4 1535940 714756 pts/0 Sl Aug02 104:48 /usr/lib/mozilla-firefox/firefox-bin -a firefox Aug 13 10:21:53 ouch Aug 13 10:21:56 you can see my xserver is quite petite Aug 13 10:22:17 pb 20408 0.5 26.5 718864 552392 pts/13 Sl Aug09 30:11 evolution Aug 13 10:22:18 maybe it decided to mirror the internet? Aug 13 10:22:30 but, firefox and evolution themselves are consuming all my resources Aug 13 10:22:59 hehe Aug 13 10:23:12 what does an office machine need? Mail+Internet Aug 13 10:23:15 still, client side bloat is easy to deal with by just restarting firefox Aug 13 10:23:18 and if that fits within the ram... Aug 13 10:23:19 hi zecke, pb_ Aug 13 10:23:23 zecke: heh, quite Aug 13 10:23:35 zecke: http://www.rpsys.net/openzaurus/temp/madness1.log Aug 13 10:23:46 zecke: still, I dread to think what it is like to use gnome on a machine with less than 2GB of ram Aug 13 10:24:40 pb_: painful is the answer :-/. Every now and again evo will swap the machine to death Aug 13 10:24:49 RP: your new depency code? Aug 13 10:24:53 hi rp Aug 13 10:24:58 zecke: yes Aug 13 10:25:07 zecke: and multithreading :) Aug 13 10:25:20 oh :) Aug 13 10:25:35 Pruned 1610 inactive tasks, 1239 left Aug 13 10:25:44 hmm that is much Aug 13 10:25:55 worse, since my desktop machine is a 32-bit sempron, I can't put much more RAM in it without changing the cpu and motherboard. Aug 13 10:25:56 and what is a self reference? Aug 13 10:26:10 * pb_ stabs ia32 Aug 13 10:26:18 zecke: A package that thinks its depends on itself :) Aug 13 10:26:31 pb_: I thought you could address up to 48gb using one of the extensions... Aug 13 10:26:44 zecke: yeah, but I think those are only in xeons or something Aug 13 10:26:54 iirc, 32-bit AMDs don't have them Aug 13 10:27:13 pb_: you could try to give more space to userspace Aug 13 10:27:27 do you have the 1:3 2:2 or 3:1 split? (whatever that is...) Aug 13 10:27:59 pb_: or try using KDE. Mail/Kontact is broken to a degree you just go back to mutt Aug 13 10:28:15 good plan Aug 13 10:28:21 I never really got on well with mutt though Aug 13 10:28:26 maybe I should go back to MH Aug 13 10:28:34 pb_: xmh?! Aug 13 10:28:54 a friend of mine used xmh until four weeks ago... when he switched to ubuntu... Aug 13 10:28:55 well, exmh Aug 13 10:29:01 ah right Aug 13 10:29:15 the old xmh is pretty ghastly Aug 13 10:29:50 I have never seen the written name of and from the sound 'x' and 'ex' is pretty close Aug 13 10:29:58 I always quite liked the command-line mh though. Aug 13 10:30:04 he was only annoyed by not having a working Euro sign Aug 13 10:30:04 ~seen mickeyl Aug 13 10:30:15 mickeyl was last seen on IRC in channel #oe, 21h 54m 17s ago, saying: 'hey pb_'. Aug 13 10:30:16 but you shouldn't have any problems with this currency Aug 13 10:30:37 RP: you force me to think about tinderclient Aug 13 10:30:39 I think I used exmh until about 2000 or so. Aug 13 10:30:58 I gave it up mostly because I wanted to store my mail on an imap server instead of on my client Aug 13 10:33:02 zecke: Sadly, tinderclient isn't going to like multithreading :-( Aug 13 10:33:40 hi all Aug 13 10:36:09 ... another package I can not build because something with libqte-mt.so.2 seems to be wrong/missing: http://pastebin.ca/129093 Aug 13 10:36:24 do I have somewhere a mistake in my OE configuration/setup? Aug 13 10:37:08 yes, you're trying to use qt/e Aug 13 10:37:42 koen: do I have to modify the local.conf then? Aug 13 10:40:20 I should really have more faith in my code but I'm amazed my first multithreaded build is actually working quite well - hasn't crashed yet :) Aug 13 10:41:36 pb_: i use gnome on my machine with 768mb of ram and it works fine Aug 13 10:42:06 mithro: Try using large imap folders with evo and it doesn't :-( Aug 13 10:43:13 RP: weird, I don't have any problems with an IMAP folder with 22 thousand messages Aug 13 10:43:32 koen: that's the only reference I've found in the conf.: PREFERRED_PROVIDERS = "virtual/qte:qte virtual/libqpe:libqpe-opie" ... is this the one which needs correction? Aug 13 10:44:12 mithro: maybe it's related to the number of messages. Aug 13 10:44:28 mithro: Try 222 thousand and it gets more interesting Aug 13 10:45:08 RP: o well, I've found that evolution is using less ram with most new releases Aug 13 10:45:31 That can only be good :) Aug 13 10:46:15 yah, that does sound good Aug 13 10:46:37 the other problem I have with evo is that if you try to display a single large message, that also causes it to consume an inordinate amount of memory. Aug 13 10:47:00 it's not such a trauma on my desktop machine, but on my laptop I need to be quite careful about what mails I open Aug 13 10:47:14 03koen 07org.oe.dev * r7a6ef70c... 10/ (1 packages/xmms/xmms_1.2.10.bb): xmms: make rebuilds work Aug 13 10:47:19 I've seen that problem too :-/ Aug 13 10:50:19 pb_: the render in evolution is changing to a "display only" render I belive which should fix that problem (kind of like evince works) Aug 13 10:50:36 very good Aug 13 10:50:52 I'm not quite sure I understand what that means, but I guess it's good news Aug 13 10:51:57 pb_: it only renders part of the message currently displayed Aug 13 10:52:07 ah, I see Aug 13 10:52:14 yeah, that would help Aug 13 10:53:33 i know there is a large effort to "trim the fat" of gnome as well Aug 13 10:53:48 by the way, can anybody recommend an socket 939 motherboard where the sata actually works with linux? Aug 13 10:54:10 I have this one server which keeps crashing with disk problems. very tedious. Aug 13 10:54:22 mithro: the gnome-performance list is now dealing with getting gtk/cairo/pango to run faster on ARM :) Aug 13 10:54:38 koen: cool Aug 13 11:00:08 i love my "deskbar" however Aug 13 11:00:15 even if it is a performance hog Aug 13 11:05:07 can someone give me a tip what I need to change to solve this problem: http://pastebin.ca/129093 (koen indicated I should not use qt/e - but how/what do I need to change)? Aug 13 11:05:32 koen: that is ggod news! Aug 13 11:06:02 pb_: are two storb obviously slower than one stor on ARMv5/XScale? Aug 13 11:06:55 when can spitz users expect to see an angstrom beta? Aug 13 11:06:59 zecke: generally yes, each store will take 1 cycle Aug 13 11:07:17 zecke: so, two storebs will take two cycles whereas a single long store will take only one Aug 13 11:08:27 pb_: obviously. I had to fin out that load, or, store on the same pixel for the complete screen is really, really slow Aug 13 11:08:38 what UI/DE does angstrom use? I take it its based upon X11/kdrive? Aug 13 11:09:17 dan: with the lack of a Free Qtopia4 we have X11 and DirectFB as options for Gtk+ Aug 13 11:10:04 Eh? Is that because you need qt4 to be lgpl- its GPL right? Aug 13 11:10:33 dan: no. It is because Qtopia4 is only proprietary Aug 13 11:10:35 zecke: yah, that would be slow Aug 13 11:11:03 dan: anyway. I think angstrom will ship the GUIs that have a dedicated maintainer within the angstrom community Aug 13 11:11:57 pb_: we have used 'fading' to make that slow thing look nice Aug 13 11:12:26 zecke:oh right! Qtopia4 is propietary. What was the last GPL release of qtopia and whats going on with Opie and qt4? Aug 13 11:13:25 dan: Opie and Qt4. Qt4 is awesome, we lack the man power to port it and then we will have community issues/acceptance. "Why did you continue Opie and did not use the award winning Qtopia" Aug 13 11:13:49 dan: this means Opie is pretty much stalled ATM waiting for TT to wake it up Aug 13 11:14:05 TT? Aug 13 11:14:10 Trolltech Aug 13 11:14:16 oh yeah! Aug 13 11:15:28 I wonder what OS Sharp will choose for the next Z? Aug 13 11:15:49 windows ce ? Aug 13 11:16:02 Aug 13 11:16:17 dan: I think they will continue with QPE1.4 and gcc1.4 and linux1.3 Aug 13 11:16:38 yeah, I think they know when they are on to a good thing Aug 13 11:16:56 :) Aug 13 11:17:02 the magical ROI (Return Of Investment) Aug 13 11:17:03 they already shipped w-zero with CE.. Aug 13 11:17:19 nchip: yeah, I was going to say that. sad. Aug 13 11:17:51 and on FOMA cellphones they use symbian Aug 13 11:18:05 if anything, they lack focus Aug 13 11:18:12 oh the w-zero looks nice Aug 13 11:18:49 the w-zero is made by HTC- looks like they'll be making the next Z I'd imagine Aug 13 11:26:25 dan: angstrom is a distribution, not an environment Aug 13 11:26:43 dan: and as zecke said, it will ship with whatever is has maintainers for Aug 13 11:27:01 http://www.angstrom-distribution.org/ is quite informative Aug 13 11:27:22 seen that site, must've missed something Aug 13 11:42:19 it seems that the problem I face is related to RTTI & exceptions; is there a way to disable them (-fno-rtti -fno-exceptions) globaly via local.conf? Aug 13 11:43:01 pgfeller_: take a look at bitbake.conf Aug 13 11:43:05 pgfeller_: CXXFLAGS e.g. Aug 13 11:43:27 zecke: thanks - I'll do Aug 13 11:43:34 also we link using gcc Aug 13 11:43:40 so we shouuld -lsupc++ Aug 13 11:43:49 or -lstdc++ Aug 13 11:45:28 zecke: in my bitbake.conf CXXFLAGS is missing, do you have me an example string which should work? Aug 13 11:46:45 which bitbake.conf did you look at? Aug 13 11:47:06 ./env/bitbake/1.4.2/conf/bitbake.conf Aug 13 11:47:27 pgfeller_: We have $BBPATH for a reason Aug 13 11:47:35 pgfeller_: bitbake wants a 'conf/local.conf' Aug 13 11:47:42 it searches this inside the BBPATH Aug 13 11:48:10 env | grep BBPATH: BBPATH=/stuff/build:/stuff/org.openembedded.dev Aug 13 11:48:47 pgfeller_: What I wanted to say: Open the right FILE Aug 13 11:49:04 which could be either /stuff/build/conf/bitbake.conf Aug 13 11:49:12 or /stuff/org.openembedded.dev/conf/bitbake Aug 13 11:49:14 or /stuff/org.openembedded.dev/conf/bitbake.conf Aug 13 11:50:30 thanks - found it @ /stuff/org.openembedded.dev/conf/bitbake.conf Aug 13 11:51:08 I assume section "build flags and options" is the place to look @ Aug 13 11:51:37 it might not be too wise to generally disable rtti Aug 13 11:52:48 do you have an idea what could cause those problems http://pastebin.ca/129093 ? 2 packages are affected until now (zudoku & libvncserver) ... Aug 13 11:53:36 pgfeller_: see above Aug 13 11:54:21 -lstdc++ ... shall I add it to the reciept, or to bitbake.conf? Aug 13 11:54:59 pgfeller_: that depends on libvncserver Aug 13 11:55:13 pgfeller_: we link with gcc so there are a couple of things to find out Aug 13 11:55:26 pgfeller_: is it using c++ at all? Aug 13 11:55:33 pgfeller_: what c++ apps does it link to Aug 13 11:58:15 zecke: thanks for the tips ... I'll try to figure this out. Aug 13 12:12:51 zecke: do you know of any reason why libpng isn't built wih autotools? Aug 13 12:13:32 pb_: no Aug 13 12:13:52 woglindehad a patch for that Aug 13 12:14:11 JustinP: http://www.angstrom-distribution.org/feeds-updated-with-fancy-frontend Aug 13 12:17:36 zecke: thanks Aug 13 12:50:49 koen: I will be on aim/icq/jabber but be offline from irc this week Aug 13 12:50:59 * zecke doesn't feel too good emotionally Aug 13 13:04:42 hi Liam Aug 13 13:04:51 hey Richard Aug 13 13:05:14 hi lrg Aug 13 13:05:19 lrg: With the latest ASoC, XorA said he heard static on corgi :-/ Aug 13 13:05:21 hey Phil Aug 13 13:06:35 RP: I think there may be too little time between the mute and the link shutdown. I'll try and add a little delay between the two events Aug 13 13:06:53 RP: any news on poodle ? Aug 13 13:07:06 lrg: I've not tested yet Aug 13 13:07:14 RP: np Aug 13 13:19:32 hi all Aug 13 13:20:03 x29a. ping Aug 13 13:29:20 moirning Aug 13 13:29:23 morning, as well Aug 13 13:32:15 morning mickey Aug 13 13:35:01 hail mickey|thesis Aug 13 13:49:06 woglinde: hi - can you please give me the link to zeckes git tree on handhelds.org - i lost my bookmarks and i want to start my kernel build tries with a "fresh" one. with google i can't find it - thx Aug 13 13:49:08 hi mickey|thesis Aug 13 13:50:31 koen: haret is quite cool (only read the site yet - thx. But now i am in the windows ce driver searching hell. I need one for my realtech - hopefully i will find one Aug 13 13:50:40 http://handhelds.org/~zecke/simpad/linux-2.6/ Aug 13 13:50:53 woglinde: thx :) Aug 13 13:52:12 woglinde: i will immediatly put it on opensimpad.org - so i can't lost it a second Aug 13 14:07:11 morning Aug 13 14:07:55 hi chouimat Aug 13 14:13:13 ping mr_nice_slacker Aug 13 14:16:22 mr_nice_slacker: what else could we try? Aug 13 14:16:31 x29a: hi Aug 13 14:16:37 hi Aug 13 14:17:19 x29a: did you reflash it? in your message the flash procedure ended at block 4 mine ended at block _5_ Aug 13 14:18:39 yeah i reflashed it, verified, read it, no diff Aug 13 14:18:55 addr: 0x000A0FFE Aug 13 14:18:55 to be exact Aug 13 14:19:15 x29a: well ok that should it be Aug 13 14:19:24 x29a: wow - shit Aug 13 14:19:34 especially since the readmem-file is just fine Aug 13 14:19:42 i mean, it boots Aug 13 14:20:08 x29a: which is the position it fails? Aug 13 14:20:15 zecke: ping Aug 13 14:20:17 the one from nopaste Aug 13 14:20:41 ~seen zecke Aug 13 14:20:48 zecke was last seen on IRC in channel #oe, 1h 29m 59s ago, saying: 'koen: I will be on aim/icq/jabber but be offline from irc this week'. Aug 13 14:20:55 x29a: please give me the link a second - i have only the one with the flashing procedure Aug 13 14:22:50 How much disk space do I need for the OE database and tree? Aug 13 14:23:18 approx? Aug 13 14:23:21 mr_nice_slacker: http://nopaste.php-q.net/232181 Aug 13 14:23:44 mr_nice_slacker: could that point to a broken ram? Aug 13 14:24:12 i hope not Aug 13 14:24:23 cause then ill be the testing victim for the ram-mod Aug 13 14:24:33 are we talking 100s of megs here? Aug 13 14:25:08 or gigs? I pretty short atm Aug 13 14:25:20 define "pretty" Aug 13 14:25:43 i have 3.5G free Aug 13 14:26:06 should do i think Aug 13 14:26:18 just from my good sense of estimating Aug 13 14:26:19 ;) Aug 13 14:26:56 Its not going to download ALL the code to EVERY app in OE though is it?? Aug 13 14:27:16 I just want to compile 1 GTK app Aug 13 14:27:16 i think thats a matter of the right settings Aug 13 14:27:35 dan: 85MB for the db and 131M for the OE tree Aug 13 14:27:45 thanks koen! Aug 13 14:27:52 thats fine Aug 13 14:28:14 it seems that I've found (with the help of this channel) the solution to Bug1279 (Build of zudoku failed) ... Aug 13 14:28:38 mr_nice_slacker: talk to me, im lost! Aug 13 14:28:39 dan: but it will compile the depencies Aug 13 14:28:44 x29a: sorry i have no clue but next point on the bootup (http://pastebin.ca/129412) is MQ200 - Revision ID=2 ... Aug 13 14:29:05 x29a: :( Aug 13 14:29:14 is there a bootloader which could load memtest86 or so? Aug 13 14:29:27 now I have 2 files: zudoku_1.1.bb & zudoku_1.1.bb.org, can someone give me a tip how to create a proper patch? Aug 13 14:30:00 x29a: need to google about memtest86 - mom Aug 13 14:30:21 mr_nice_slacker: or whatever memtesting program Aug 13 14:31:04 x29a: well memtest86 won't work it is for x86 Aug 13 14:31:27 x29a: unfortunate i don't know a memtest programm Aug 13 14:32:32 I get an error when trying to use monotone to update my database Aug 13 14:32:48 x29a: may there are other jtag utilities which can check the memory? Aug 13 14:32:59 mtn: fatal: std::logic_error: error in roster:1:2:E: wanted symbol 'format_version', got TOK_SYMBOL with value V Aug 13 14:33:01 x29a: or have 0.5.1 a option for it? Aug 13 14:33:11 mr_nice_slacker: with simpadSL.rom i can get into monitoring tool Aug 13 14:33:18 mr_nice_slacker: no, 0.5.1 doesnt Aug 13 14:33:38 dan: what command did you use? Aug 13 14:33:48 dan: and did you follow GettingStarted? Aug 13 14:34:05 mtn --db=/home/d/oe/OE.mtn pull monotone.openembedded.org org.openembedded.dev Aug 13 14:34:16 that path is correct Aug 13 14:34:19 x29a: monitoring tool? Calling Monitor ===================== Monitor: ?? Aug 13 14:34:26 I unbzipped it Aug 13 14:34:34 (the .mnt) Aug 13 14:34:35 mr_nice_slacker: "press esc for monitoring" Aug 13 14:35:14 x29a: so my taken image could be wrong! Aug 13 14:35:29 no Aug 13 14:35:33 since its working for you Aug 13 14:36:34 x29a: well - i didn't wrote the flash full with zeros - so it could be that the missing part is on a other position? Aug 13 14:36:38 dan: and you checked the md5sum after downloading? Aug 13 14:36:58 x29a: i can make a 16 MB image of the hole first flash bank? Aug 13 14:37:00 not done that yet Aug 13 14:37:14 mr_nice_slacker: wonna make a full readmem of 16 megs tonight? Aug 13 14:37:24 x29a: i start it right now Aug 13 14:37:28 mr_nice_slacker: takes forever, but than we know Aug 13 14:37:50 -rw-r--r-- 1 d d 107323392 Aug 13 05:31 OE.mtn Aug 13 14:38:06 mr_nice_slacker: i still have the 16mb file which i read when the pad initially broke Aug 13 14:38:12 we could diff that to your then Aug 13 14:38:53 pgfeller: diff -u zudoku_1.1.bb.orig zudoku_1.1.bb > zudoku_1.1.bb.patch Aug 13 14:39:15 cyrilRomain: thanks :-) Aug 13 14:39:46 cyrilRomain: ehm Aug 13 14:39:56 cyrilRomain: why not use mtn diff? Aug 13 14:40:11 mtn diff zudoku_1.1.bb Aug 13 14:40:21 we use a SCM for a reason, you know Aug 13 14:40:24 x29a: well it could take a while *g Aug 13 14:40:24 koen: ehm.. yes :) Aug 13 14:40:37 mr_nice_slacker: especially with your slow connection Aug 13 14:40:56 koen: thanks Aug 13 14:41:07 cyrilRomain: your command breaks if zudoku_1.1.bb.orig isn't the 'original' Aug 13 14:41:23 x29a: well - i really have to build a better cable Aug 13 14:41:37 koen: you're right. mtn diff if safer Aug 13 14:41:39 x29a: i have read a bit about that HaRET Aug 13 14:41:47 s/if/is Aug 13 14:41:56 x29a: if i get it right we don't have to modify the source Aug 13 14:42:06 mr_nice_slacker: yeah, i think its the later version of bootblaster which brought me to my jornada568 linux Aug 13 14:42:17 x29a: it has a scripting language Aug 13 14:42:24 cyrilRomain: and shorter + less work :) Aug 13 14:42:31 mr_nice_slacker: nice Aug 13 14:42:48 * cyrilRomain update his 'too old' patch script :) Aug 13 14:43:00 mr_nice_slacker: u know what jtagcommand "discovery" does? Aug 13 14:43:08 x29a: and you can check your hardware with it Aug 13 14:43:16 x29a: no Aug 13 14:43:34 x29a: but you can test it out ;) Aug 13 14:43:40 mr_nice_slacker: but not without wince _AND_ bootloader Aug 13 14:43:52 * cyrilRomain remove his script even Aug 13 14:44:17 x29a: well - yes and a working network card driver Aug 13 14:45:44 Is somevar[flag] += a valid expression? Aug 13 14:46:08 (for bitbake class files to append to an existing flag) Aug 13 14:46:09 x29a: arg - i hope we are able to bring your pad back to life Aug 13 14:49:03 x29a: i will upload a image i have taken yesterday with wrong arguments (was 54MB) but the first 16 MB should be the flash. you can shorten your waiting time with it Aug 13 14:49:28 x29a: i will also uploade the other one if it will be complete Aug 13 15:03:21 koen: I'd bunzipped the file before checking the md5. I don't think bzipping it again would work. Thing is monotone said it was probably a monotone error and I hardly ever have problems downloading as I browse under Linux on a good connection Aug 13 15:07:31 mr_nice_slacker: 16777216 2006-08-07 01:30 2.5.3_bootldr_after_broke.rom Aug 13 15:07:40 thats my readout Aug 13 15:12:00 x29a: thx for the boost link - i will put the stuff in the wiki right now Aug 13 15:18:48 03koen 07org.oe.packaged-staging * r5bdd8a03... 10/ (1 docs/packaged-staging.xml): docs/packaged-staging.xml: add more structure Aug 13 15:18:52 03koen 07org.oe.packaged-staging * rd16e3e68... 10/ (1 docs/packaged-staging.xml): docs/packaged-staging.xml: describe how to package native applications Aug 13 15:25:48 where can i find the perl feed for 3.5.4.1? Aug 13 15:36:25 dan: try asking in #openzaurus Aug 13 15:37:34 mr_nice_slacker: sry, got disconnected: http://pastebin.ca/129627 Aug 13 15:38:12 x29a: http://opensimpad.org/images/1/18/Test.img Aug 13 15:39:57 x29a: i am really interested in testing that bost stuff Aug 13 15:40:26 x29a: quite cool options y,Y : FLASH Burst Test Aug 13 15:41:35 x29a: p,P d,D or m,M : PCMCIA RAM Data Test or PCMCIA Memory Test - verry cool stuff Aug 13 15:41:47 x29a: we really need to pach the loader Aug 13 15:42:11 x29a: i will start right after the second image is complete Aug 13 16:07:13 mr_nice_slacker: hehe, ok, im into different stuff for right now, ill join you later maybe Aug 13 16:12:16 RP: packaged-staging probably needs some for of locking combined with a multithreaded bitbake Aug 13 16:12:51 mr_nice_slacker: waaaah, cant detect flash anymore Aug 13 16:13:23 mr_nice_slacker: false alarm, after wiggling and pushing it worked again, somehow the simpad must have been off Aug 13 16:13:29 koen: What kind of locking? Aug 13 16:13:59 koen: I guess it depends how the packaged staging is working... Aug 13 16:15:08 I've now resolved all the dependency issues and the multithreaded build completed :) Aug 13 16:15:19 unbelievable Aug 13 16:15:31 The most annoying "feature" of it is is will build the images twice, once for do_rootfs and once again for do_build... Aug 13 16:15:44 koen: cool Aug 13 16:16:02 RP: either the 2.6.17 kernel has major problems or upgrading doesn't work properly.... Aug 13 16:16:28 JustinP: 2.6.17 works ok for me... Aug 13 16:16:56 Only three problems with poky's metadata were highlighted which was nice :) Aug 13 16:17:23 RP: I got a lock-up yesterday and have been noticing some very odd behavior....such as plugging in headphones causes "6" tp be repeated over and over.... Aug 13 16:17:29 but hold on...perhaps it's my fault Aug 13 16:17:38 x29a: do you flash the pad at the moment? Aug 13 16:17:49 RP: for native packages (e.g quilt-native) it will clean out staging and package the remaining stuff to avoid relocation troubles Aug 13 16:18:04 RP: so you have an empty staging a few moments in time Aug 13 16:18:30 RP: and two ipkg processes writing to the same status file is bad as well Aug 13 16:18:41 koen: Ouch. That will break things :-/ Aug 13 16:18:45 RP: nothing we can't solve however :) Aug 13 16:18:55 koen: indeed - there are ways and means :) Aug 13 16:19:10 koen: Give each thread a different staging area ;-) Aug 13 16:19:18 * RP considered this earlier... Aug 13 16:19:30 (for a different problem) Aug 13 16:19:50 zecke: hi - do you remember with pathches you patched the kernel 2.6? or did you used a vanilla kernel? I want to build a patchset against vanilla kernel - thx Aug 13 16:19:51 RP: relocation problems Aug 13 16:20:06 RP: you can't move quilt once built, etc Aug 13 16:20:07 grrrr...why did it make CONFIG_SHARPSL_RC=y..... Aug 13 16:20:11 * JustinP rebuilds as a module Aug 13 16:20:12 koen: I thought you were avoiding them? :) Aug 13 16:20:27 mr_nice_slacker: yeah Aug 13 16:20:29 RP: yes by only using one staging dir Aug 13 16:20:32 with the 16mb thing Aug 13 16:20:44 koen: hmmm :-/ Aug 13 16:20:56 RP: that only applies to native packages Aug 13 16:21:06 mr_nice_slacker: use the power of git Aug 13 16:21:12 RP: for the other packages we have sed :) Aug 13 16:21:18 actually i was erasing all the blocks, so i could do something in between now Aug 13 16:21:24 mr_nice_slacker: wotcha want? Aug 13 16:21:37 but there is not much in my git tree anyway Aug 13 16:22:35 koen: There is a simple workaround - we can intercept the filesystem io calls and rewrite the paths. Just needs a bit of dynamic linker magic ;-) (this was one of kergoth's ideas) Aug 13 16:22:48 well, "simple" :) Aug 13 16:23:08 RP: iirc gpe-appmgr has the magic Aug 13 16:23:48 RP: I think 2 or 3 packages will be a pain, but we'll attack that when we get to that point Aug 13 16:24:10 x29a: did you started to flash this image http://opensimpad.org/images/1/18/Test.img? Aug 13 16:24:22 mr_nice_slacker: not yet Aug 13 16:24:35 should i? Aug 13 16:24:48 zecke: so i have to get into git - ok. you started with a git from kernel.org? Aug 13 16:24:59 koen: We'll also be able to lock bitbake into single thread mode and still benefit from the new depends code Aug 13 16:25:10 x29a: would be nice to know if it will work - it is the image from yesterda Aug 13 16:25:17 RP: is it in svn yet? Aug 13 16:25:21 k, ill start it, mr_nice_slacker Aug 13 16:25:42 x29a: the actual one is at addr: 0x0043C000 and needs to get to 0x01000000 Aug 13 16:25:54 koen: No, I need to discuss this with zecke. The changes are quite heavy :) Aug 13 16:26:08 mr_nice_slacker: wotcha mean?flashmem 0 Test.bin wont do? Aug 13 16:26:17 mr_nice_slacker: yes Aug 13 16:26:24 mr_nice_slacker: ah, your progress, i c Aug 13 16:26:28 zecke: ok thx :) Aug 13 16:26:30 RP: put it in a branch? Aug 13 16:26:34 mr_nice_slacker: but I think the hh.org cvs is/was more far than my git tree Aug 13 16:26:40 koen: I've been polishing out bugs. There is one remaining thing it doesn't do yet comapred to the current bitbake - try building an alternate provider when one fails... Aug 13 16:26:41 *unfished* crap Aug 13 16:27:08 RP: heavy changes? me tosses a coing to form an opinion Aug 13 16:27:09 The infrastructure is there to handle it though... Aug 13 16:27:29 RP: do bitbake collections work? Aug 13 16:27:38 zecke: I never touched them Aug 13 16:27:50 We even have a test case for this bit of code :) Aug 13 16:27:58 zecke: I'll show you the patch in a few minutes, once I find a final bug... Aug 13 16:28:11 don't hurry Aug 13 16:28:15 zecke: How would I run the tests? Aug 13 16:28:17 I don't feel like reviewing stuff Aug 13 16:28:27 RP: I hope I have written a README Aug 13 16:28:38 zecke: Ah :) Aug 13 16:28:39 you need to set PYTHONPATH and another variable for sure Aug 13 16:29:56 zecke: I'll probably check these changes in over the next week or so in bits if you're agreeable. Some of the patch is just common sense rearranging of code Aug 13 16:30:07 zecke: http://cvs.handhelds.org/cgi-bin/viewcvs.cgi/linux/kernel26/ ? do you have a cvs link for me? Aug 13 16:32:13 RP: I'm sure I will be fine Aug 13 16:32:46 RP: one organisational question is if we should release a BitBake 1.6.0 from the current trunk? Aug 13 16:33:01 RP: it has an enormous parsing speed up by using mithro's COW Aug 13 16:33:40 zecke: That would seem wise before I add a whole load of new code... Aug 13 16:33:48 RP: well...my wifi card no longer comes up automatically....I have to ifup it... Aug 13 16:33:58 zecke: hehe ok it must be the right address -> 15 months zecke MQ200/SIMpad Backlight and LCD :) Aug 13 16:34:06 mr_nice_slacker: yes Aug 13 16:34:28 zecke: Although we do have known dependency problems which I can't fix in the existing code :-/ Aug 13 16:34:40 (the new code removes them) Aug 13 16:34:56 RP: can we 'detect' it in the old code? Aug 13 16:35:16 RP: zaurusd no longer switches headphone on Aug 13 16:35:31 zecke: If we could, I could probably fix the old code :-/ Aug 13 16:35:54 RP: attempting to suspend (power button) makes my Z go to a screen with a blinking cursor and never stop.... Aug 13 16:36:21 RP: even moving the battery slider doesn't turn it off Aug 13 16:36:33 JustinP: I have noticed some wifi issues. The headphone problems is more likly an ASoC bug depending on which version of ASoC you're using Aug 13 16:36:39 ibot: botmail for greg2: new images available without altboot Aug 13 16:36:46 I've not seem suspend/resume fail though Aug 13 16:37:16 well....I'm using newest default kernel in oz354x and newest packages from it Aug 13 16:38:06 JustinP: Anything in dmesg from the suspend issue? Aug 13 16:39:15 RP: ummm....how would I see output form dmesg when I have to kill it? Aug 13 16:41:31 JustinP: I wasn't sure quite how locked up you meant - often when it does that, it resumes a short while later and you'll see some process wouldn't freeze in dmesg Aug 13 16:42:28 hmmm....well, I may not have left it long enough I suppose... Aug 13 16:43:04 If it doesn't do it after a minute or two, its not going to. I'm not exactly sure what times out Aug 13 16:43:49 hmmmm, well perhaps I'll try it again later Aug 13 16:43:52 and try a new image Aug 13 16:49:40 zecke: http://www.rpsys.net/openzaurus/temp/bitbake_multithread-r0.patch Aug 13 16:49:51 zecke: I will document the functions before I commit them ;-) Aug 13 16:50:22 HINT: Document before you write the functions Aug 13 16:50:29 I know you know what you want to do... Aug 13 16:51:17 zecke: That would presume I knew what I was doing in advance ;-) Aug 13 16:52:08 RP: I think you do (structured thinking) Aug 13 16:52:22 RP: that is pretty much to review, I won't do that now :} Aug 13 16:52:37 hehe time.sleep(0) Aug 13 16:53:02 zecke: np, I just thought I'd make is available as a first draft in case anyone objects to the general direction of it :) Aug 13 16:53:28 zecke: and the bits in methodpool.py ;-) Aug 13 16:53:52 zecke: and the dump_data functions writing to /tmp :) Aug 13 16:53:53 "don't try to emulate a SCM in vi" Aug 13 16:55:17 and I guess I need to fix the shell :) Aug 13 16:56:41 bleh Aug 13 16:56:48 I hate writing docs Aug 13 17:02:06 * RP -> food Aug 13 17:03:17 i don't want to bag anyone - i find the project really cool and so. But one of the most important disadvantages (for newbies) of oe is the leak of documentation. Aug 13 17:03:31 oh no Aug 13 17:03:37 we're leaking docs ;) Aug 13 17:03:47 koen: They are marked as confidential Aug 13 17:03:57 koen: we signed three NDAs to get them :) Aug 13 17:03:58 koen: my english is terrible Aug 13 17:04:10 mr_nice_slacker: yes, we know. You can rescue us Aug 13 17:04:20 mr_nice_slacker: we at least have a org.openembedded.documentation branch Aug 13 17:04:26 zecke: ah, I think I signed 4 Aug 13 17:05:08 zecke: i will try to help - but i can't promise it. Aug 13 17:05:31 mr_nice_slacker: no promise needed, just try Aug 13 17:05:44 giel: ping? Aug 13 17:06:05 giel: do you happen to be at the BMCO meeting on thursday? Aug 13 17:12:26 03koen 07org.oe.packaged-staging * ra5ceb493... 10/ (1 docs/packaged-staging.xml): docs/packaged-staging.xml: add another solution for the glibc-intermediate problem Aug 13 17:15:30 zecke: if i wrote a really good documentation for openembedded you would think about writing a driver for simpad smartcard? Aug 13 17:16:13 mr_nice_slacker: yes, but there was a driver already (for 2.4er) Aug 13 17:16:42 zecke: so i will try to give my best to document oe ;) Aug 13 17:17:19 zecke: i need 2.6 due to evdev Aug 13 17:54:15 BTW: anyone knows "Gantz" the anime? Aug 13 17:55:39 mr_nice_slacker: which addr was the last one on your image from yesterday? Aug 13 17:55:54 im at 0x001... at the moment Aug 13 18:05:34 x29a: The 16 MB one? shuld end at 0x01000000 Aug 13 18:06:24 x29a: 0x01000000 = 16777216 = 16 MB Aug 13 18:06:47 x29a: 16777216 Byte Aug 13 18:08:10 x29a: The adresses are all in byte except the ones the bootloader shows they are in bit Aug 13 18:09:10 x29a: the second one is now at addr: 0x008B4000 so it shouldn't take that long. then i will diff the 2 images Aug 13 18:09:56 how to get openembedded.org wiki log in? Aug 13 18:10:33 mr_nice_slacker: you don't need one Aug 13 18:10:53 koen: but i want to have one :( Aug 13 18:11:08 koen: no way to get one? Aug 13 18:12:35 become an OE core member :) Aug 13 18:13:10 koen: regardles of this, we need to change this 'policy' Aug 13 18:13:16 or hassle someone to come up with a policy Aug 13 18:13:25 koen: is drupal that dangerous? Aug 13 18:13:43 zecke: no Aug 13 18:14:34 zecke: we have no website maintainer Aug 13 18:16:01 03eFfeM 07org.oe.dev * r42fa696a... 10/ (3 files in 3 dirs): Aug 13 18:16:01 uclibc: added patch to compile uclibc for thumb; patch is from Aug 13 18:16:01 http://busybox.net/lists/uclibc-cvs/2006-January/008903.html Aug 13 18:16:34 koen: maybe our website should call for one? Aug 13 18:16:57 zecke: if you want it to Aug 13 18:17:13 zecke: it's one of the items for OEDEM, if it isn't solved by then Aug 13 18:18:02 koen: I have some sort of arranged for Rooms... Aug 13 18:20:16 it's only cause if i add a lot of stuff it would be cool to see my name somewhere Aug 13 18:20:45 mr_nice_slacker: understandable Aug 13 18:21:36 send patch for the org.openembedded.documentation branch, so we can do mtn commit --author=slacker Aug 13 18:21:42 patches* Aug 13 18:24:16 zecke: I have the 9pfs module loaded on my ipaq :) Aug 13 18:25:29 koen: ok Aug 13 18:26:19 koen: what is that? Aug 13 18:26:28 koen: I would know Plan9 FS Aug 13 18:26:45 zecke: yes, plan9fs Aug 13 18:27:00 koen: and you have booted plan9? Aug 13 18:27:04 zecke: so adding the plan9 support libs and tools is now on my todo Aug 13 18:27:12 zecke: I haven't booted plan9 yet Aug 13 18:27:45 My ipaq is full of bunnies? Aug 13 18:30:05 zecke: i saw some litte movements of porting plan9 to the simpad ;) Inferno would have the same drivers i think Aug 13 18:32:27 I judge Operating Systems by Mascots Aug 13 18:32:29 1st Plan9 Aug 13 18:32:35 2nd classic BSD Aug 13 18:32:39 3rd Linux Aug 13 18:32:42 4th Minix Aug 13 18:32:49 oh damn I forgot darwin Aug 13 18:32:54 I think 2nd Darwin Aug 13 18:33:03 I agree Aug 13 18:33:13 I'd install plan9 just to get the bunny Aug 13 18:33:40 I hate to be a man... We love bunnies :( Aug 13 18:34:29 openbsds puffy is also quite cool Aug 13 18:34:44 mr_nice_slacker: nice, not too nice though Aug 13 18:34:53 mr_nice_slacker: I love the posters+pictures Aug 13 18:34:57 +thsirts Aug 13 18:35:29 zecke: yes they are cool Aug 13 18:37:05 hmm I still need to pickup older OpenBSD posters Aug 13 18:37:57 zecke: they sold them on their site - and they give them on linuxtag for free or donation Aug 13 18:38:32 mr_nice_slacker: Someone gave me two posters, they are stored at the www.spline.de room Aug 13 18:38:44 koen http://linuxtogo.org/~zecke/Web-Site/2575F43D-5C4F-498E-87DC-3F9CA6197DFE.html Aug 13 18:38:51 koen: fun with iWeb+PhotoBooth Aug 13 18:40:12 * RP returns Aug 13 18:40:51 zecke: you shaved! Aug 13 18:40:58 RP: wb Aug 13 18:41:04 koen: just yesterday Aug 13 18:41:27 koen: I'm not as popular as joco... Aug 13 18:43:13 hmmm Aug 13 18:43:21 yes? Aug 13 18:43:26 Like my Shark? Aug 13 18:43:29 zecke: it seems only one OE core hackers has his beard left Aug 13 18:43:50 who? Aug 13 18:43:56 koen: it will grow again Aug 13 18:43:59 mickeyl Aug 13 18:44:21 he will shave after turning in his Dis Aug 13 18:44:55 koen: I think I won't be able to shave regulary.... and it will grow again Aug 13 18:51:27 has anyone tested mtn 0.28 with oe? Aug 13 18:51:39 oe.org run 0.28 Aug 13 18:51:41 runs* Aug 13 18:51:51 (actually 0.28 + patches) Aug 13 18:52:26 koen: ah - which patches? cause i get this error (http://pastebin.ca/129853) Aug 13 18:53:19 mr_nice_slacker: your monotone is buggy Aug 13 18:53:42 0.26, 0.27 and 0.28 from venge.net work here Aug 13 18:53:47 koen: would this explain my problem earlier? Aug 13 18:53:54 mr_nice_slacker: I'm running mtn 0.28 on MacIntel Aug 13 18:54:39 0.28 on macppc and 0.28 debian/x86 Aug 13 18:55:12 hm.. Aug 13 18:56:28 which boost and gcc versions you are using? Aug 13 18:56:50 none Aug 13 18:57:09 osx/ppc binary from venge.net/monotone and on debian the debian binary Aug 13 18:57:21 apt-get install monotone Aug 13 18:57:43 koen: i don't have debian ;) Aug 13 18:58:01 koen: could you do a mtn --full-version for me? Aug 13 18:58:06 then use a static binary from venge.net/monotone Aug 13 18:58:25 koen: i will try the statc binary Aug 13 18:58:37 monotone 0.28 (base revision: 8c6ce7cb2ccd21290b435e042c2be4554ec6a048) Aug 13 18:58:38 Running on : Darwin 8.7.0 Darwin Kernel Version 8.7.0: Fri May 26 15:20:53 PDT 2006; root:xnu-792.6.76.obj~1/RELEASE_PPC Power Macintosh Aug 13 18:58:38 C++ compiler : GNU C++ version 4.0.1 (Apple Computer, Inc. build 5341) Aug 13 18:58:38 C++ standard library: GNU libstdc++ version 20050421 Aug 13 18:58:38 Boost version : 1_33 Aug 13 18:59:29 koen: thx Aug 13 18:59:38 monotone 0.28 (Basis-Revision: 8c6ce7cb2ccd21290b435e042c2be4554ec6a048) Aug 13 18:59:38 Laufe auf : Darwin 8.7.1 Darwin Kernel Version 8.7.1: Wed Jun 7 16:19:56 PDT 2006; root:xnu-792.9.72.obj~2/RELEASE_I386 i386 Aug 13 18:59:42 C++-Kompiler : GNU C++ version 4.0.1 (Apple Computer, Inc. build 5250) Aug 13 18:59:44 C++-Standard-Bibliothek: GNU libstdc++ version 20050421 Aug 13 18:59:46 Boost-Version : 1_32 Aug 13 19:00:42 * zecke needs to get XCode2.2 Aug 13 19:01:21 * koen needs to get a 64bit apple notebook Aug 13 19:01:22 oh 2.4 is even out Aug 13 19:01:46 koen: when did they release such thing? Aug 13 19:02:24 they didn't Aug 13 19:02:34 even Core Duo2 is 32bit? Aug 13 19:03:26 the one from venge.net did the job :) - i have to recompile mine Aug 13 19:04:59 dan: try the binary from venge.net adn see if that solves your problems Aug 13 19:05:13 zecke: no idea, I don't follow intel stuff that much Aug 13 19:06:48 drat Aug 13 19:06:50 938MB Aug 13 19:06:54 whehe Aug 13 19:07:27 * zecke anticipates the next episode of Gantz Aug 13 19:07:33 a really disturbing Anime Aug 13 19:08:25 ガンツ Aug 13 19:08:35 zecke: anime is always disturbing Aug 13 19:09:00 hehe Aug 13 19:09:48 Can anyone remind me why do_fetch has nostamp set? Aug 13 19:10:26 RP: I want it, specially for SRCDATE=now Aug 13 19:10:43 RP: but pb_ and kergoth judged it as superf? (now I lack the name) Aug 13 19:10:50 RP: as we have the .md5 doing the same thing Aug 13 19:10:53 oh i have spelled it out to fast - it don't work (http://pastebin.ca/129868) with the venge version for me. Aug 13 19:11:36 zecke: superfluous Aug 13 19:11:48 thanks Aug 13 19:12:01 RP: I think we should add a stamp for do_fetch Aug 13 19:12:27 zecke: but kergoth/pb_ think otherwise? Aug 13 19:12:36 no Aug 13 19:12:43 pb_ didn't mind last time I asked Aug 13 19:13:08 I think it could probably use a stamp as well Aug 13 19:14:01 Re-running all the fetch routines for a new image generation seems pointless as we'd do nothing with any new fetched data and the data that was fetched should still be valid... Aug 13 19:14:26 RP: yes, but it shouldn't download stuff Aug 13 19:14:38 but stamping is the right thing to do (specially with SRCDATE=now) Aug 13 19:15:11 zecke: It doesn't which is something I guess :) Aug 13 19:16:17 if the 'stamp' flag could handle a list of filenames for the stamps for that task,itd be easy for the base class to populate that with the result of the downloads (stamp for each file, md5, whatever) Aug 13 19:16:44 kergoth: stamp after having downloaded every file? Aug 13 19:16:56 kergoth: and for the other files .md5 should work aynway Aug 13 19:17:31 kergoth: Why would you need to do that though? Either the fetch completed or it didn't? Or am I not realising how do_fetch works? Aug 13 19:18:32 I found a nice way to express the task dependencies btw: http://www.rpsys.net/openzaurus/temp/loc.conf - simliar to bbdepcmd but renamed as bbdepcmd isn't setup quite right for our new needs Aug 13 19:19:01 I'll add those bits into the appropriate classes Aug 13 19:21:06 RP: the crux of the locking problem: http://linuxtogo.org/~koen/OE/SoC2006/ch03s04.html Aug 13 19:21:27 koen: no, I just tried that version of monotone off whatever that site was and I'm getting the same error Aug 13 19:21:29 I'm also wondering about do_rootfs[nostamp] = 1. That's causing images to get generated twice in the multithreading. A better idea might be to put DATETIME into PR... Aug 13 19:25:25 koen: packaging native packages with ${MACHINE} in ipkgarchs? :-/ Aug 13 19:25:41 RP: yeah, that one sucks Aug 13 19:27:07 RP: I need MACHINE to be in there to use the same status file for staging/x86, cross and staging/$ARCH Aug 13 19:28:57 koen: :-/ Aug 13 19:29:58 RP: I'm more worried about the fact that OE says libgcc's arch is 'arm' even if it has clz and bx isns Aug 13 19:30:26 koen: That is something that needs fixing Aug 13 19:30:48 koen: probably a bad interaction between multimachine and cross packages? Aug 13 19:31:01 RP: no, dumb hardcoding in the .inc Aug 13 19:31:17 ick Aug 13 19:31:29 koen: ouch :-( Aug 13 19:31:44 kergoth: Did you see the new dependency code? Aug 13 19:31:46 PACKAGE_ARCH = TARGET_ARCH Aug 13 19:32:12 koen: Does it need that at all? Aug 13 19:32:14 which will return arm instead of Aug 13 19:32:35 RP: it's a cross package, otherwise you'd get x86 Aug 13 19:32:41 like gcc-cross-locale Aug 13 19:32:50 s/x86/BUILD_ARCH/ Aug 13 19:33:17 koen: ah, right. So it needs to be taught about multimachine's tricks... Aug 13 19:33:54 RP: multimachine doesn't matter I think Aug 13 19:34:04 cross.bbclass should really save the old PACKAGE_ARCH somewhere where cross packages can extract it from Aug 13 19:34:57 mickeyl, hrw and I didn't get caught by it since we build the strongarm toolchains before the xscale ones Aug 13 19:36:41 koen: multimachine has a bug Aug 13 19:36:49 koen: I have arm-darwin in my tmp dir Aug 13 19:36:53 zecke: multiple ones Aug 13 19:37:57 zecke: $HOST_OS instead of TARGET_OS in multimachine.bbclass? Aug 13 19:38:06 zecke: staging only has a arm-foo-eabi over here Aug 13 19:38:22 zecke: instead of armv5-foo-eabi and arm-foo-eabi Aug 13 19:39:05 koen: That's strange as mine does show armv5te-linux... Aug 13 19:40:39 weird Aug 13 19:41:12 shaving sucks Aug 13 19:42:24 how stable is bitbake? Aug 13 19:42:35 ie. does it make sense to upload it to debian ? Aug 13 19:43:04 I think 1.4.2 has been pretty stable Aug 13 19:44:13 lrg: I shall try those patches tomorrow as I already booked the day off work :-) Aug 13 19:44:27 koen: did you get your GSoC T-Shirt yet? Aug 13 19:44:30 XorA|gone: np Aug 13 19:44:37 ~lart HR Aug 13 19:44:38 * ibot runs at HR with an origami Swiss Army knife, and inflicts a nasty paper cut Aug 13 19:53:10 mr_nice_slacker: k, still flashin, im off to bed Aug 13 19:55:46 x29a: good night Aug 13 19:56:05 x29a: hopefully it will be a nice awakening with a running simpad Aug 13 19:56:22 x29a: i am in the hell of finding a wince network card driver Aug 13 19:56:31 x29a: to get haret output Aug 13 19:57:54 which network card? Aug 13 19:58:10 x29a: longshine Aug 13 19:58:21 x29a: i think it is a realtech chipset Aug 13 19:58:50 x29a: fcc id: Mq4fe1kmx Aug 13 19:58:51 id have a xircom card around which i got running already unter wince2.0 Aug 13 19:59:04 x29a: cool Aug 13 19:59:24 x29a: which chipset? Aug 13 20:00:15 hopefully your right with the nice surprise tomorrow morning Aug 13 20:00:22 dont know the chipset, sry Aug 13 20:01:23 x29a: yes i hope too Aug 13 20:01:47 i mean, we are still not at the and, there is your new image still and BOST Aug 13 20:02:19 x29a: yes bost - i will patch it emediatly after i have taken the new image Aug 13 20:02:28 great Aug 13 20:02:32 and i gotta go to bed Aug 13 20:02:33 x29a: can't wait to check it :) Aug 13 20:02:36 hehe Aug 13 20:02:46 a lot of info for the wiki! Aug 13 20:03:03 x29a: i have started a site in the wiki about it Aug 13 20:03:04 zecke: only the gsoc tshirt from 2005 Aug 13 20:03:20 x29a: and a new category knowledge base Aug 13 20:03:25 great Aug 13 20:03:54 x29a: see you tommorow Aug 13 20:04:22 mr_nice_slacker: just for the log, "booting 2.5.3-hh directly wont work"-"BOST unlock"-"image xp with jtag"-whatelse? Aug 13 20:05:02 x29a: ? Aug 13 20:05:22 the knowledge gathered this weekend Aug 13 20:05:37 x29a: ah well yes and don't forget haret Aug 13 20:05:45 x29a: it is really cool Aug 13 20:05:54 yeah right Aug 13 20:15:51 gah. Including DATETIME in PR doen't work as the the cache doesn't realise the PR variable changed :-( Aug 13 21:11:05 zecke, koen, other mac addicts: did you know emacs is available as a cocoa app? http://sourceforge.net/project/showfiles.php?group_id=148174&package_id=163427&release_id=423235 Aug 13 21:13:45 poushag: I didn't know about that one, but I do use Aquamacs: http://aquamacs.org/ Aug 13 21:17:35 justinp do you think you will try the one i posted? Aug 13 21:18:13 poushag: perhaps, if I remember and have time next I go into work Aug 13 21:18:56 i havent yet used emacs in any situation - but its the only really capable thing ive found for directory/file comparison and reconciliation as a cocoa app (filemerge suxxors) Aug 13 21:19:18 didnt know about aquamacs - the two must be very similar Aug 13 21:21:59 nite Aug 13 21:22:01 but if aquamacs is more osx-ey then i will probably prefer it Aug 13 21:22:49 justinp does cmd-h work on aquamacs? (to hide it) Aug 13 21:33:19 ~botmail for x29a: here is the link to the new image http://opensimpad.org/images/7/77/Hole.img Aug 13 21:40:28 justinp did you set up oe on mac osx? Aug 13 22:08:24 ~seen zecke Aug 13 22:08:34 zecke was last seen on IRC in channel #oe, 46m 34s ago, saying: 'nite'. Aug 13 23:24:07 03rpurdie * r560 10bitbake/lib/bb/build.py: Aug 13 23:24:07 trunk/bitbake/lib/bb/build.py: Aug 13 23:24:07 * Add version of stamp_is_current that works with cached data Aug 13 23:24:07 * Tweak mkstamp to work over broken? NFS filesystems (simply opening Aug 13 23:24:07 the file for write access wasn't enough to change the file's timestamp) Aug 13 23:24:08 * Have add_task create _task_deps variable containing various task flags Aug 13 23:24:10 for cache use Aug 13 23:31:47 * kergoth peruses build.py Aug 13 23:34:35 03rpurdie * r561 10bitbake/ (bin/bitbake lib/bb/cache.py): Aug 13 23:34:35 trunk/bitbake/bin/bitbake: Aug 13 23:34:35 * Add some variables to the data cache Aug 13 23:34:35 * Move self.preferred into the same place as other similar cache data Aug 13 23:34:35 trunk/bitbake/lib/bb/cache.py: Aug 13 23:34:36 * Bump cache version number Aug 13 23:51:35 03rpurdie * r562 10bitbake/bin/bitbake: Aug 13 23:51:35 trunk/bitbake/bin/bitbake: Aug 13 23:51:35 * fix references to findBestProvider (closes #1288) Aug 13 23:59:43 03rpurdie * r563 10bitbake/ (bin/bitbake lib/bb/providers.py): Aug 13 23:59:43 trunk/bitbake/lib/bb/providers.py: Aug 13 23:59:43 trunk/bitbake/bin/bitbake: Aug 13 23:59:43 * Move getProvidersRun -> bb.providers.getRuntimeProviders Aug 13 23:59:43 * Make filterProviders build_cache_fail parameter optional, pending Aug 13 23:59:44 rethink/removal Aug 13 23:59:46 * Add NoProvider exception Aug 14 00:02:47 03rpurdie * r564 10bitbake/bin/bitbake: Aug 14 00:02:47 trunk/bitbake/bin/bitbake: Aug 14 00:02:47 * Fix up getRuntimeProviders calls in previous commit Aug 14 00:07:50 kergoth: http://www.rpsys.net/openzaurus/temp/bitbake_multithread-r3.patch if you want to see where its leading :) Aug 14 00:10:23 sweet, looks clean Aug 14 00:11:35 kergoth: Its not quite as functional as the existing bitbake in that it won't build alternate providers when one fails yet but all the structure is there to handle it and the multithreading works :) Aug 14 00:13:02 very cool Aug 14 00:17:06 will be nice to use that.. so much time is wasted for each test iteration waiting for builds, for just about any oe work. frustrating at times Aug 14 00:17:12 * kergoth kicks his builds Aug 14 00:17:52 kergoth: I'm applying some tweaks to poky, then it will work against it - feel free to experiment :) Aug 14 00:18:46 look forward to it. heh, i need a scripot to dump a graphviz style dependency graph for a section of the metadata. would make it a lot oeasier to grasp the flow of large semgents, like X, opie, etc Aug 14 00:19:18 kergoth: bitbake head has dot graph generation Aug 14 00:19:36 cool Aug 14 00:19:57 kergoth: This code I'm working on can also output dependency data but its text based atm :) Aug 14 00:23:27 kergoth: The changes needed are now in poky so you just need a patched bitbake head to work with it :) Aug 14 00:23:31 * RP -> sleep **** ENDING LOGGING AT Mon Aug 14 02:59:57 2006