**** BEGIN LOGGING AT Tue Mar 21 10:59:57 2006 Mar 21 11:00:05 lrg: Good news on both fronts :) Mar 21 11:00:20 RP: i agree Mar 21 11:00:58 I haven't checked if the record only issue is still there on cxx00 but I suspect it is. For now I've let zaurusd leave sidetone enabled Mar 21 11:01:30 RP: it might help to see the amixer settings when you hear the c7x0 tone. I can the check the wm8731 and ask a hw engineer Mar 21 11:02:49 lrg: Interestingly the tone changes if I increase playback volume so perhaps we have a L/R swapped somewhere? Mar 21 11:03:18 RP: interesting... Mar 21 11:04:55 lrg: The problem is we have no idea how the jack connects to the wm8731 - At least its simpler that the wm8750 so we stand some chance of working it out :) Mar 21 11:05:41 RP: true, I'll speak to a hw engineer about this and what could cause such a tone. It may be possible to deduce the connections. Mar 21 11:06:29 RP: btw, boku is almost got audio out of his driver :) Mar 21 11:06:56 lrg: To hear the tone, I need Mic Boost on and the mixer sidetone switch on. Altering the playback volume or sidetone volume changes the tone Mar 21 11:08:54 Bernardo: do nand backup before booting 2.6 on akita Mar 21 11:09:03 too late... Mar 21 11:09:17 RP: i could kiss you !!!!!! Mar 21 11:09:22 lrg: No other mixer settings change things expect the Jack Function - you can hear the tone for both headphones and headset and the tone is different Mar 21 11:09:47 RP: ok, is it possible to dump amixer or alsactl output when this happens on headset and headphones. Mar 21 11:11:29 RP: almost perfect Mar 21 11:12:00 package manager still looks for lists in /var/lib/ipkg instead of /usr/lib/ipkg/lists Mar 21 11:12:43 and no sound in opie (at least screen taps) Mar 21 11:12:55 ade|desk: It works? Mar 21 11:12:59 yes Mar 21 11:13:06 ade|desk: nice :) Mar 21 11:13:18 Bernardo: Does sound work outside of opie? Mar 21 11:13:22 i am looking at a screen configuring locale-base- crap Mar 21 11:13:37 ade|desk: Ah - that takes a while :) Mar 21 11:13:48 I must make some nicer images at some point... Mar 21 11:13:54 Bernardo: do nand backup before booting 2.6 on akita Mar 21 11:14:02 RP: let me see if I have anything here to test... Mar 21 11:14:03 hrw|work: I did... :) Mar 21 11:14:06 Bernardo: ;) Mar 21 11:14:22 Bernardo: ipkg will use /var/lib/ipkg/ because on collie we are out of space Mar 21 11:14:40 RP: i really don't care howlong it takes, this is a non gcc3 build and i am happy Mar 21 11:14:47 but ipk still uses /usr/lib/ipkg Mar 21 11:14:49 hrw|work: Looks like opie sound is tempramental - this is something outside of my control (and I seem to remember this issue from a long time ago?) Mar 21 11:15:04 only package manager points to /var/lib/ipkg Mar 21 11:15:16 RP: do you think its a driver issue ? Mar 21 11:15:18 Bernardo: hm Mar 21 11:15:25 bbl Mar 21 11:15:41 nick ade|afk Mar 21 11:15:55 RP: the QtE OSS implementation is incomplete to name it like this Mar 21 11:16:03 Bernardo: just flashed 3.5.4.1 on c7x0 and have 'lists_dir ext /var/lib/ipkg' in ipkg.conf Mar 21 11:16:32 hmm objdump is a dead end Mar 21 11:16:39 zecke: interesting, i wonder when they will use alsa Mar 21 11:17:02 lrg: Qtopia4 - the holy grail that fixes all issues - Mar 21 11:17:07 ah Mar 21 11:17:32 lrg: I think they will not support it any time soon Mar 21 11:17:40 :( Mar 21 11:17:52 hrw|work: you're right Mar 21 11:17:58 sorry Mar 21 11:18:14 I had that commented on my backup ipkg.conf file I had just restored Mar 21 11:18:25 lrg: I think its opie at fault as sound is sometimes there sometimes not accross a spectrum of devices... Mar 21 11:18:43 ~lart dell for making a crappy touchpad on the latitude Mar 21 11:18:43 * ibot pulls out a ClueBat (tm) and thwaps dell for making a crappy touchpad on the latitude Mar 21 11:18:45 lrg: I've sent you some mixer configs which give the tone Mar 21 11:18:54 RP: thanks Mar 21 11:19:12 * lrg has lost all mouse control on his latitude.. Mar 21 11:20:05 So the question is whether to use 2.6.15 or 2.6.16 for 3.5.4.1 :) Mar 21 11:20:43 * RP is reckless and would go for 2.6.16. 2,6,15 is so yesterday ;-) Mar 21 11:21:09 does the nokia 770 have host usb? Mar 21 11:21:11 RP: hear hear Mar 21 11:21:57 katossi_uni: In hardware yes. Not sure how much its supported though Mar 21 11:23:09 aha, too bad Mar 21 11:27:44 pb__: How do I find my operands in the hexdump? :} Mar 21 11:33:46 Microsoft just paid me to tell them linux is great :-) (Ill wait to see if cheque ever arrives) Mar 21 11:34:22 RP: no probs with 2.6.16 here yet Mar 21 11:35:32 2.6.16 and pcmciautils .. go on ;) Mar 21 11:36:38 swapfile format changed between 2.4 and 2.6? Mar 21 11:37:28 * Bernardo starts the long task installing most libxine plugins... Mar 21 11:38:44 Bernardo: I think they might have developed a new format, yes Mar 21 11:38:58 Bernardo: if you have idea about meta packages for libxine then send infos Mar 21 11:39:14 hrw|work: task packages ;-) Mar 21 11:39:31 hrw|work: if I have the time to look at it, I'll do it Mar 21 11:39:59 Bernardo: thx Mar 21 11:41:01 also the memory leaks in libipkg don't help at all when installing a few packages Mar 21 11:41:15 03hrw 07org.oe.oz354x * rbc5f4431... 10/packages/gnumeric/files/remove-docs.patch: gnumeric 1.6.2: added missing patch Mar 21 11:42:29 Hi Liam Mar 21 11:44:01 join #openzaurus Mar 21 11:44:58 Hey Dirk Mar 21 11:45:43 do13: any chance to look at the backlight build problem yet ? Mar 21 11:46:08 lrg:Yeah, it works. Mar 21 11:46:34 lrg: But I need some tweaks to buld the image Mar 21 11:46:55 do13: ok, i'll work around for the time being Mar 21 11:47:10 lrg: Do you have a working userland? Mar 21 11:48:06 do13: not yet, i's just failed on creating the image - due to missing kernel modules. I'm updating and rebuilding kernel now Mar 21 11:48:44 lrg: which modules are misssing? Mar 21 11:49:37 do13: the new sound and touch, my fault Mar 21 11:50:22 ~hail koen for 'create images from feed' functionality Mar 21 11:50:25 * ibot bows down to koen for 'create images from feed' functionality and chants, "I'M NOT WORTHY!!" Mar 21 11:50:33 didn't zautrix have a patch for zuc to work on 640x480? Mar 21 11:50:56 Bernardo: had but in a way that it does not work on 320x240 Mar 21 11:50:59 lrg: No. These are build in. Wait I'll look for my tosa.conf Mar 21 11:51:52 do13, lrg: If you RRECOMMEND modules you can avoid builds failing when you build them in Mar 21 11:51:54 hrw|work: and can we make it so that patch application is dependent on using a 640x480 device? Mar 21 11:52:32 PR: thx Mar 21 11:52:38 Bernardo: or do zuc, zuc-hires which is ugly Mar 21 11:53:10 RP, do13: It's my fault. I had a half complete build on friday and did an update this morning before resuming the build - so it was abit out of sync. Mar 21 11:54:39 well, more urgent in my personal priorities list is opie-solitaire... :) Mar 21 12:00:35 RP: the usb port is "stuck" in host mode? Mar 21 12:00:59 RP: the usb port is "stuck" in host mode? Mar 21 12:02:03 ~lart isa proxy Mar 21 12:02:03 * ibot decapitates isa proxy conan the destroyer style Mar 21 12:07:49 somehow I've disable the akita keyboard in konsole Mar 21 12:07:54 time to restart opie Mar 21 12:08:37 Bernardo: rmmod ohci_hcd ? Mar 21 12:09:21 Bernardo: Probably fiddling with Sysrq? :) Mar 21 12:10:02 RP: I think I somehow pressed a shift+fn combination while writing "swapon /media/card/.swapfile" Mar 21 12:10:18 had to restart opie to get the keyboard responding again Mar 21 12:10:24 fat fingers... Mar 21 12:10:34 standard qte keyboard problem ;( Mar 21 12:10:40 I keep breaking its keyboard handling too :-/ Mar 21 12:11:02 so I'm not the only one Mar 21 12:14:16 03hrw 07org.oe.oz354x * ra179ed86... 10/packages/sharp-binary-only/ (2 files in 2 dirs): sharp-aticore-oss: added make 3.81 patch from .dev Mar 21 12:14:21 there goes opie again... Mar 21 12:14:39 ~lart libipkg Mar 21 12:14:39 * ibot pulls out a ClueBat (tm) and thwaps libipkg Mar 21 12:16:09 hrw|work: ?? Mar 21 12:16:28 hrw|work: is the keymap broken? Mar 21 12:17:30 RP: now I have sound after this last boot (3rd boot since flashing) Mar 21 12:18:26 Bernardo: Interesting. I'm blaming opie ;-) Mar 21 12:19:47 zecke: some users reported that Fn+shift kills keyboard on c7x0 - iirc it even was in OE bugtracker Mar 21 12:20:32 hrw|work: well, the guy never replied and the patch was wrong Mar 21 12:20:47 hrw|work: the keyboard driver is an interpreter Mar 21 12:21:03 hrw|work: if it is asks to navigate to column X Mar 21 12:21:31 hrw|work: and now VoidSymbol gets released, the driver will act accordignly (in compliance to man keymaps) Mar 21 12:24:06 ok Mar 21 12:28:22 hrw|work: I might be able to commit a updatet kernel-keymap.patch Mar 21 12:28:40 hrw|work: It has a workaround for broken keymaps and handles >7 bit codes correctly Mar 21 12:30:47 zecke: cool Mar 21 12:31:08 it it added support for the zoom and light fn keys, would be even better Mar 21 12:31:09 RP: BOOTSTRAP_EXTRA_RRECOMMENDS += "kernel-module-soundcore ..." should work? Mar 21 12:34:10 do13, RP: it's failed again - cant satisfy deps for task boostrap wrt to audio and touch modules. It looks like it's asking for the i2s driver which is not required for tosa, I'll try and fix up zaurus-clmashell-2.6.conf to remove the i2s dep for tosa Mar 21 12:35:11 lrg: First remove *wm* from tosa-2.6conf Mar 21 12:35:32 do13: ok. Mar 21 12:35:58 do13: yes Mar 21 12:36:10 RP: http://pastebin.ca/46469 Mar 21 12:38:12 do13: Something still has them in DEPENDS by the looks of it. The unsatisfied reccomendation messages are just warnings Mar 21 12:38:24 (RDEPENDS that is) Mar 21 12:38:54 Probably zaurus-clamshell-2.6 - I didn't know tosa was using that Mar 21 12:39:17 and given it should install dependent modules, we probably don't need to list i2s in there anyway Mar 21 12:39:27 RP: fixing that now Mar 21 12:51:23 openslug-native is full of stuff... bit too hardcore for me when it comes to <1G storages Mar 21 12:53:26 it can stand being slimmed down a bit Mar 21 12:59:45 NAiL: I'm thinking about something like task-native-sdk Mar 21 13:00:09 task-native-sdk-{python,perl,x11,gtk2,opie} - that kind of stuff Mar 21 13:00:27 not a bad idea Mar 21 13:01:03 openslug-native was kinda thrown together so we had a quick way of installing build-essentials ;) Mar 21 13:11:36 who has admin powers on OE bugzilla? Mar 21 13:22:52 03nail 07org.oe.dev * r13554b43... 10/packages/linux/ixp4xx-kernel/2.6.16/defconfig: ixp4xx-kernel: Update defconfig Mar 21 13:25:33 RP: Is this ok http://pastebin.ca/46471 ? Mar 21 13:29:59 NOTE: Handling BitBake files: \ (0025/3216) [ 0 %] Mar 21 13:30:14 why need he to do that each time ? Mar 21 13:30:22 s/it/he Mar 21 13:36:02 do13: yes, that's fine Mar 21 13:46:22 lrg: mt pull Mar 21 13:46:23 NOTE: package bootstrap-image-1.0: completed Mar 21 13:47:12 03dirk 07org.oe.dev * r5c016e9d... 10/conf/machine/include/zaurus-clamshell-2.6.conf: zaurus-clamshell-2.6.conf: remove soundmodules from RDPEPENDS for tosa Mar 21 13:47:16 03dirk 07org.oe.dev * r79dcadb7... 10/conf/machine/include/tosa-2.6.conf: tosa-2.6.conf: wm97xx support are buildin, remove RDEPENDS Mar 21 13:59:51 do13: ok Mar 21 14:03:39 lrg: I hope later today, I can update to 2.6.16 Mar 21 14:17:01 morning Mar 21 14:19:55 hi chouimat Mar 21 14:20:23 RP: waking up from suspend got it suspended again. Looks like the "on" button might be repeating? Mar 21 14:21:11 also, removing the power cord takes about 30 secs to get the light to dim Mar 21 14:22:15 putting the cord back in - 10 secs before the backlight brightens Mar 21 14:23:16 checked again, 15 secs to off, 10-11 secs to on Mar 21 14:24:04 RP: http://www.oesf.org/forums/index.php?showtopic=18110&st=0&#entry119576 Mar 21 14:24:21 ibot: terrier is Sharp Zaurus SL-C3200 Mar 21 14:24:22 hrw|work: okay Mar 21 14:24:50 trisoft have sl-c3200 available Mar 21 14:25:28 hrw|work: how much do they cost ? Mar 21 14:25:38 RP: can we provide kind of image for them to test does our 2.6 works on terrier? this way we could be first !sharprom distro supporting it Mar 21 14:26:08 ljp: at trisoft.de 'coming soon' price Mar 21 14:26:41 hrw|work: I think I'll buy 1 Mar 21 14:26:51 lrg: EUR 541 at pricejapan.com Mar 21 14:27:26 hrw|work: thanks Mar 21 14:28:15 541? isnt too bad Mar 21 14:28:35 wrong symbol, bac char chart Mar 21 14:28:39 bad* Mar 21 14:31:52 lrg: conics.net 529 EUR Mar 21 14:33:58 trisoft would probably want >700EUR Mar 21 14:36:08 do13: thanks, btw gpe-image for tosa still failing for the same the deps (sound & touch) for task-bootstrap Mar 21 14:36:31 lrg: rebuild task-bootstrap Mar 21 14:36:55 hi Mar 21 14:36:57 hrw|work: sorry, how do I do that ? Mar 21 14:37:38 hrw|work: I've deleted the tmp/stamps for it and did a bitbake gpe-image Mar 21 14:38:10 lrg: bitbake -i and then 'rebuild task-bootstrap;build gpe-image' Mar 21 14:38:32 hrw|work: ah, ok - thanks Mar 21 14:41:28 np Mar 21 14:42:44 hrw|work: Ask them try the 2.6 spitz image on it. The only thing I can think of to go wrong is the lack of the right pcmcia ID Mar 21 14:43:52 hrw|work: Perhaps first ask them to find out the pcmcia IDs under 2.4 and we can check they're in 2.6 Mar 21 14:44:02 RP: read thread on oesf Mar 21 14:44:30 lrg: Just watch out for import duty - probably about an extra 80GBP Mar 21 14:45:18 RP: that's hefty :( Mar 21 14:46:17 lrg: 20GBP clearance fee and the rest is VAT :-/ Mar 21 14:47:38 hrw|work: Having read the thread, I suspect the spitz image will just work then Mar 21 14:49:18 hrw|work: thanks for the xxd hint Mar 21 14:49:29 hrw|work: it worked - took a bit longer than I had expected though Mar 21 14:50:14 Bernardo: I suspect that's an issue with opie's apm polling taking a while to notice the user has started charging Mar 21 14:50:17 zecke: xxd rule Mar 21 14:50:33 The suspend/resume button repeat is worrying as that should be fixed... Mar 21 14:50:37 zecke: vi file then :%!xxd then edit then %!xxd then :wq Mar 21 15:09:59 hi Mar 21 15:10:06 CoreDump|home: hey Mar 21 15:10:59 someone has idea how to make irk-belkin and irk-targus working? then have to conflict but also have to be installable Mar 21 15:19:58 nope :( Mar 21 15:37:08 RP: also, the opie mixer doesn't "speak" alsa Mar 21 15:38:26 Bernardo: opie does not speak alsa at all Mar 21 15:42:33 hrw|work: ok, will look at it... Mar 21 15:42:56 Bernardo: It should be able to talk via the OSS mixer. I suspect that communication is limited Mar 21 15:47:08 JustinP: thx (we speak about e17 - right?) Mar 21 15:47:23 darn, kernel 2.6.16-r0 too big for spitz Mar 21 15:47:47 "| Size is 1306348" Mar 21 15:48:14 too many things in the kernel Mar 21 15:48:37 it's not the place for driver i find Mar 21 15:49:57 hrw|work: I assume so... Mar 21 15:56:22 kergoth: ping Mar 21 16:02:20 RP: could the defconfig-cxx00 please define cc optimize for size? Mar 21 16:02:50 RP, : you said in your doc that your initrd.bin is gzipped file , but i don't succeed to uncompressed it Mar 21 16:03:13 Genesis: Note is says gzip -c Mar 21 16:03:35 jnc: I'm not sure I like that option although I can see why you'd want it in your case Mar 21 16:03:49 ho Mar 21 16:03:53 what is the size of the root partition for the c750? Mar 21 16:04:22 i'll test somethings else Mar 21 16:04:30 RP: it's standard for x86 kernel configs, reccommended by kernel peoples as i grok it Mar 21 16:04:44 JustinP: ~23M Mar 21 16:04:53 i see how it could break though Mar 21 16:04:59 ack Mar 21 16:05:08 ok, well, e-image-core is probably also too big Mar 21 16:05:15 shows 21M on the site Mar 21 16:05:33 jnc: I'd be interested in seeing if there is a difference in kernel performance Mar 21 16:05:45 (vs. memory usage) **** BEGIN LOGGING AT Tue Mar 21 16:07:35 2006 Mar 21 16:07:48 lardman|work: I'd appreciate it. The e17 stuff has been tested only on spitz & akita so far Mar 21 16:08:06 JustinP: Just the core image from the standard download location? Mar 21 16:08:14 RP: the bitbake performance is totally weird Mar 21 16:08:37 RP: no method takes more than 0.05 seconds on invocation Mar 21 16:08:52 RP: only calling these functions 17.000.000 times accumulate Mar 21 16:09:10 zecke: So the solution is to call less things? :) Mar 21 16:09:57 RP: we have ~3000 bb files but we call execute 33.000 times Mar 21 16:10:16 hrw@bitbake:~/devel/build/oz354x/_test$ ipkg install irk-targus Mar 21 16:10:16 Installing irk-belkin (0.11-r2) to temp... Mar 21 16:10:30 koen|away: could you update python to python2.4.2 (from 2.4.1?) Mar 21 16:10:32 any ideas how to resolve that? Mar 21 16:10:32 lardman|bbiab: core first off, yes. I suggest trying it with a loopback image, I'm not sure if it fits in the flash.... Mar 21 16:11:01 zecke: Do we need to execute all these things at parse time? Mar 21 16:11:17 zecke: There are a limited number of variables we're interested in... Mar 21 16:11:40 RP: well each variable could influence other variables Mar 21 16:12:07 RP: part of it will be to parse each .bbclass only once Mar 21 16:12:44 RP: I accidently closed my terminal. within some hours I will have the hotshot profile loaded :} Mar 21 16:13:11 zecke: I realise the variable dependencies are tricky... Mar 21 16:13:15 that will eat a huge amount of memory Mar 21 16:13:27 I appeciate the reasons behind kergoth's redesign... Mar 21 16:15:11 jnc: It depends how you do it. The idea was to parse into a database Mar 21 16:15:25 RP: which makes things slower at first Mar 21 16:15:45 zecke: but rebuilds should just be able to rebuild the variables that changed Mar 21 16:16:16 RP: yeah you want to store the syntax trees for bbfiles and classes Mar 21 16:16:26 RP: and then you want to assemble those :} Mar 21 16:16:50 I've given that a fiar bit of thought. I did reach some bits of the dependency tracking that seemed difficult to solve though due to bitbake's variable structure Mar 21 16:16:53 RP: and I'm pretty sure there will be efficient tree algorithmis for that already Mar 21 16:17:00 I can't quite remember that that was... Mar 21 16:17:53 zecke: One of these days I'll write this multi threaded builder which will also use some interesting algorithms. I have most of them in my head :-/ Mar 21 16:18:13 would eliminating executable code from bb files be helpful? Mar 21 16:18:46 jnc: Its a rather fundemental part of OE/bitbake Mar 21 16:19:03 i don't find it necessary Mar 21 16:19:26 * zecke watches RP's jnc monologue and plays jeapardy withi his answers Mar 21 16:19:50 there are python-isms which could be abstracted out Mar 21 16:19:55 jnc: bbclass files are just bb files Mar 21 16:20:03 like things referencing FILE Mar 21 16:20:10 RP: I will finish the c++ parser, but currently I think a better data implementation will help us more Mar 21 16:20:16 hm Mar 21 16:20:37 okay i guess i wasn't refering to bbclass Mar 21 16:20:58 RP: would you be interested in the hotshot profiler file? Mar 21 16:21:01 jnc: Internally they're the same thing to bitbake - renamed for users to see what goes on more clearly Mar 21 16:21:22 * zecke googled for "hotshot cuntime" yesterday Mar 21 16:21:31 zecke: I'm not sure my python is elite enough to be able to use it? Mar 21 16:21:45 RP: do you have python 2.4.2 installed? Mar 21 16:21:56 hotshot uses cuntime as abreviation :} Mar 21 16:22:54 zecke: I have on certain machines, yes Mar 21 16:23:06 ewi546.ewi.utwente.nl/OE/bitbake.hotshot Mar 21 16:23:14 from hotshot import stats Mar 21 16:23:24 s = stats.load('bitbake.prof') Mar 21 16:23:29 wait... Mar 21 16:23:32 dir(d) Mar 21 16:23:37 dir(s) even Mar 21 16:24:35 zecke: I guess this would explain why the cache isn't of too much benefit Mar 21 16:25:20 * RP wonders if googling for that expression breaks zecke's ISP's AUP :) Mar 21 16:25:35 jnc: by 'executeble code in bb' you mean all those do_compile(), do_install() etc parts? Mar 21 16:25:56 hrw|work: i mean things like sed statements Mar 21 16:26:05 RP: AUT? Mar 21 16:26:16 zecke: Acceptable Use Policy Mar 21 16:26:20 RP: ISP == Uni ATM Mar 21 16:26:33 RP: I sincerly hope I do not violate that :} Mar 21 16:26:49 jnc: so all do_*() parts Mar 21 16:27:22 hrw|work: those are needed by bitbake and used in many places Mar 21 16:27:41 i meant unnecessary executions of code Mar 21 16:28:00 a sed statement at runtime is unnecessary Mar 21 16:28:09 ping koen Mar 21 16:28:34 using a bit of python code referencing FILE should be unnecessary Mar 21 16:29:53 these things exist because bitbake/OE does not provide a consistent way, or it does and it's not being used Mar 21 16:30:18 it makes optimizing more difficult Mar 21 16:31:39 mickey|zzZZzz: http://blogs.sun.com/roller/page/alanc?entry=security_hole_in_xorg_6 Mar 21 16:32:27 sed makes optimizing difficult? Mar 21 16:32:49 are you saying we should remove the *shell* code from bb files? I don't see how that could even be possible Mar 21 16:33:12 * zecke pets his ignore list Mar 21 16:33:32 * zecke gets some popcorn and continue playing jeapardy Mar 21 16:33:56 "Why shouldn't we remove all methods?" Mar 21 16:34:13 Bitbake Tuning 300 Mar 21 16:34:54 pals, we should regulary play online games Mar 21 16:38:26 :-) Mar 21 16:38:50 RP: wierd, either a 2.6.16 change or a gpe change, but powerbutton is now Calendar + Suspend :-) Mar 21 16:41:42 cu Mar 21 16:47:58 RP: my hotshot profile is still loading Mar 21 16:53:55 RP: we have poodle in UK for shipping costs Mar 21 16:54:51 hrw|work: That sounds good! :) Mar 21 16:55:02 RP: can he send it to you? Mar 21 16:55:07 RP: aa_ user on #gpe Mar 21 17:00:36 cu Mar 21 17:20:15 bye Mar 21 18:18:46 sshhhhh Mar 21 18:19:04 are you trying to sleep? Mar 21 18:20:30 hey pb__ Mar 21 18:21:23 me? nope. just waking up Mar 21 18:21:35 didnt want to disturb anyone sleeping in here Mar 21 18:21:44 pb__: are you going to CELF ? Mar 21 18:27:37 That's a good question - who is going to CELF out of interest? Mar 21 18:27:55 * lrg is going to CELF Mar 21 18:33:16 hi Mar 21 18:33:45 I am about to try things I am not really clever enough to try on my ipaq3715 Mar 21 18:34:35 now the kernel boots with that script thing, and I can log in ssh over usb Mar 21 18:35:08 seems that it is not mounting any drives. Woulkd the next step be to attempt to get something more than the 16 mb ramdisk? Mar 21 18:38:43 lrg: probably not, unfortunately Mar 21 18:39:11 ljp: ah right, very public spirited of you Mar 21 18:39:21 pb__: you're famous, what should I do in general terms? please? Mar 21 18:40:01 or in fact, who should I speak to Mar 21 18:40:06 aa_: what are you trying to achieve? I don't really know what the state of the art is on the 3715. Mar 21 18:40:06 in fact, never mind Mar 21 18:40:25 heh, okay Mar 21 18:40:48 pb__: well, there is a booting kernel, and ssh, nicely, but nothing more than a ramdisk and no mounted partitions Mar 21 18:40:58 but still never mind...I will browse for a while Mar 21 18:41:04 righto Mar 21 18:42:22 zecke: replied. thats fine with me Mar 21 18:42:34 make that infamous :) Mar 21 18:43:33 well, I remember these names from a time when I used to use oz Mar 21 18:43:36 :) Mar 21 18:46:56 kergoth: you saved me to lookup that agreement from my mailbox Mar 21 18:47:10 kergoth: and I can finally take that code from ROAD back to OE ) Mar 21 18:47:20 kergoth: and bring it in shape for Qt4 Mar 21 18:50:22 kergoth: would have been a shame to redo the fixes Mar 21 18:55:45 ok, seems I am stuck asking annoying questions. Really sorry. Mar 21 18:58:03 the 3715 has an internal disk it calls "storage" what are the chances of mounting that? Mar 21 18:59:23 zecke: nod. glad to hear it Mar 21 19:00:36 also the radisk is 12 mb and I think there are 64 mb available, should I set it to use the whole thing? Mar 21 19:05:58 chances are 4,500,546:1 Mar 21 19:06:08 in my favour? Mar 21 19:06:35 i dunno. you tried mounting it? Mar 21 19:06:49 I am going through /dev trying to munt everything Mar 21 19:07:09 my mummy told me to learn shell scripting, but well, I was lazy Mar 21 19:08:55 now, although I am clueless, I am sensing that maybe /dev/ is just wrong Mar 21 19:09:25 I think I need to do things like find out where these things are or something Mar 21 19:10:03 ok, well, the sd card could also be the thing to do it Mar 21 19:17:31 aa_: check dmesg, see if it says anything about mtd devices Mar 21 19:17:54 ok, thanks Mar 21 19:18:11 if so, those would be your internal storage. you would need to be careful about mounting them, though, since your machine might become unusable if wince gets upset. Mar 21 19:18:37 morning kergoth Mar 21 19:20:56 pb_: unfortunately no mention Mar 21 19:22:39 oh wow, I found instructions Mar 21 19:23:58 the instructions end like this "If I were a real hacker, I would download the ARM cross-compiler, download Mar 21 19:24:01 hey Mar 21 19:24:01 the kernel and the patches and start editing and linking Mar 21 19:24:04 in the asic3 SD card module. Or something like that." Mar 21 19:24:15 OT: how can I scroll in a screen? Mar 21 19:24:27 control-a escape Mar 21 19:24:37 g-up pg-down Mar 21 19:25:59 RP: Okay here it is: from 5 minutes parsing, two are spend compiling and executing functions Mar 21 19:27:29 Rewrite it into c(++) and precompile c-based .bbs into .so :) Mar 21 19:28:49 polyonymous: that doesn't make it more efficient Mar 21 19:29:08 polyonymous: compiling each function and executing is almost O(1) Mar 21 19:29:16 Well, in fact, I was kidding, but I'm not sure if it's true. Mar 21 19:29:20 (with the functions we have) Mar 21 19:29:56 kergoth: did you ever used the feature where you downloaded a BBFILE from the internet? Mar 21 19:30:03 You won't have to compile them each time. Mar 21 19:30:15 (but then again, don't take it seriously) Mar 21 19:31:06 zecke: it'd be interesting to see a breakdown of that per function. Mar 21 19:31:07 zecke: hmm, i think i tested it once. the idea behind that was to facilitate having the .bb in the upstream source tree, pulling that down and including it for use in a .bb repo Mar 21 19:31:12 been a long time though Mar 21 19:31:31 and bb isnt really a very good fit for development at the moment anyway Mar 21 19:31:35 zecke: (and with compile and execute separated out) Mar 21 19:32:27 kergoth: it stops where your localpath is a tar.gz ;) Mar 21 19:32:49 ewi546.ewi.utwente.nl/OE/bitbake.hotshot Mar 21 19:33:03 (and again don't try to search for hotshot and cuntime Mar 21 19:33:17 zecke: heh, do you go by the name "hotshot" these days? Mar 21 19:35:02 pb_: I just wanted to get some hints - well I got other 'hints' Mar 21 19:35:19 hehe Mar 21 19:36:32 kergoth: and the 'scoping' of python functions is the one 'wanted'? Mar 21 19:37:01 i dont understand what you'er asking Mar 21 19:40:04 I guess he's asking whether it's correct that "def ...:" pokes the function into the global builtins namespace. Mar 21 19:40:52 ah. well, the problem with it atm is that it could easily be prone to side effects. no isolation Mar 21 19:41:04 moving it into its own namspace might be a Good Thing(tm) Mar 21 19:41:07 yeah Mar 21 19:41:59 tricky to do that now without creating a compatibility headache though Mar 21 19:42:06 I guess you'd need to introduce a new scoped variant of "def" Mar 21 19:43:22 thats true. we really shouldve implemented a means of versioning for our metadata at the start Mar 21 19:43:28 right Mar 21 19:43:33 we suck, but we knew that already Mar 21 19:43:33 Quick question, what is the clean command? Mar 21 19:43:53 -c clean? Mar 21 19:43:54 hehe Mar 21 19:45:02 kergoth: global scoping is nice as well Mar 21 19:45:10 kergoth: just parse the functions once Mar 21 19:45:25 kergoth: specially handlers, and helper functions of .bbclasses Mar 21 19:48:37 not sure what handlers you're thinking of, but helper functions might well benefit from that. Mar 21 19:49:02 package.bbclass is quite an offender in that regard, it has loads of functions that don't really need to be compiled again and again Mar 21 19:50:07 whee.. you know whats fun... a segfault before the mmu has been setup Mar 21 19:53:53 kergoth: eleet Mar 21 19:58:11 gotta love uml though. makes debugging stuff like that so much easier Mar 21 19:58:52 3A Mar 21 20:00:07 mm, yeah. though, xen is my virtualizer of choice these days. Mar 21 20:00:23 i need to try that, have yet to do so Mar 21 20:00:31 seems to perform a lot better than uml from what I've seen so far Mar 21 20:00:57 nice Mar 21 20:01:52 yeah, it's very cool. only problem I've had with it is that it seems to hate my Via C3 machines. Mar 21 20:02:27 pb_: if package.bbclass would define different functions depending on the bbfile Mar 21 20:02:35 that would be horror - eh chaos Mar 21 20:02:38 but again chaos is good Mar 21 20:02:40 hm the hardware virtualizer for the i5/p5/openserver platform kicks ass, but it's a little bit expensive Mar 21 20:03:01 zecke: I don't think it can. afaik, the "def ...:" syntax doesn't go near bb.data, so there is no way for the bbfile to influence the code. Mar 21 20:03:11 woglinde: hehe Mr. i5 certified? Mar 21 20:03:22 pb_: this is what you want to believe :} Mar 21 20:03:30 zecke calm down *g* Mar 21 20:03:40 obviously that isn't true for "python xxxx() { .... }" though: you really do need to parse those every time. Mar 21 20:03:45 zecke let's fix the RMI code Mar 21 20:03:54 zecke: right, this belief is what lets me sleep at night Mar 21 20:04:08 woglinde: is it my time to get involved? Mar 21 20:04:59 NOTE: Parsing finished. 0 cached, 3217 parsed, 0 skipped, 0 masked. Mar 21 20:04:59 ERROR: Nothing provides dependency foo Mar 21 20:05:00 real 0m14.396s Mar 21 20:06:00 for the record that is bitbake, with a 'Null Parser', without any usage of eval and exec Mar 21 20:06:13 but with configuration parsing, and going through each bb file Mar 21 20:06:51 * chouimat is away: dishes Mar 21 20:08:50 very good Mar 21 20:10:03 to satisfy the speed junky my voices say "Ship.IT, Ship.IT" Mar 21 20:25:39 How do i clean the "memory" of a bitbake build? Made some changes to a .bb file and want it to delete the old compiled and compile again Mar 21 20:26:17 bitbake -c clean -b .bb Mar 21 20:27:30 CoreDump|home: will that trigger the bitbake bug? Mar 21 20:27:40 nope Mar 21 20:28:05 which bug anyways? Mar 21 20:28:22 i seem to have misplaced my paycheck Mar 21 20:28:23 CoreDump|home: -c clean cleaning all dependencies as well Mar 21 20:28:24 this cant be good Mar 21 20:28:31 oh :( Mar 21 20:28:35 zecke: ouch Mar 21 20:28:41 kergoth: print a new one? Mar 21 20:29:04 yea, will have to call them and have them put a stop payment on teh current one and send a new one Mar 21 20:29:07 bleh Mar 21 20:38:23 kergoth: suck Mar 21 20:45:44 god i suck with gdb. i should fix that Mar 21 20:46:09 * pb_ cats the gdb manual to kergoth's terminal Mar 21 20:46:23 hi Mar 21 20:46:32 i have probleme with compilation of gpe-image Mar 21 20:46:37 that's very sad. Mar 21 20:46:38 touch: ne peut faire un touch sur `/home/ovea/laradji/emulation/gentoo/os1/build/tmp/deploy/ipk/Packages': Aucun fichier ou répertoire de ce type Mar 21 20:46:38 NOTE: Task failed: /home/ovea/laradji/emulation/gentoo/os1/build/tmp/work/gpe-image-1.0-r19/temp/log.do_rootfs.6577 Mar 21 20:46:48 sounds like you failed to set a DISTRO Mar 21 20:47:04 erf ok Mar 21 20:47:15 if you don't specify a DISTRO, you won't get any packages created Mar 21 20:47:25 for build a i386 work with qemu Mar 21 20:47:29 what distro i use ? Mar 21 20:47:39 look in conf/distro Mar 21 20:47:47 you probably want "familiar-unstable" or some such Mar 21 20:48:03 ok Mar 21 20:48:04 tks Mar 21 20:48:04 or, alternatively, you could add package_ipk to INHERIT directly Mar 21 20:49:06 i have choose familiar Mar 21 20:49:10 unstable Mar 21 20:49:36 i need to rebuild nano and task-bootstrap Mar 21 20:49:38 ? Mar 21 20:50:11 I think it should work to just "bitbake gpe-image" again Mar 21 20:50:20 you don't need to build nano, that's just an example Mar 21 20:50:24 i have a syntax bug with local.conf syntax Mar 21 20:50:39 what's the error? Mar 21 20:50:46 i have comment require sanity.conf in bitbake.conf for resolv it Mar 21 20:50:58 oh right, yeah Mar 21 20:51:07 your bitbake is a bit too old, but I think commenting out that line is OK Mar 21 20:51:16 ok Mar 21 20:51:23 1.3.2 bitbake Mar 21 20:52:36 yeah, the latest is something like 1.3.3.1 Mar 21 20:52:41 ok Mar 21 20:52:56 I don't think it matters, but I'm not completely sure. Ueberhacker Zecke would be your resource for that information. Mar 21 20:53:15 the wiki doc is very usefull :) Mar 21 20:53:25 i have never use bitbake or monotone Mar 21 20:53:58 hi can someone pls help me i am trying to use cyrus-sasl under Openslug i installed only the standard package cyrus-sasl and cyrus-sasl-dev and i always get this error on my screen and in my /var/log/messages: http://pastebin.ca/46537 Mar 21 20:54:21 kami22: you should probably take that up with the openslug folks in #nslu2-linux or someplace Mar 21 20:54:26 i only want to make a image for view opie and gpe on qemu Mar 21 20:55:17 but i think its a package problem Mar 21 20:55:38 nacer_: for 'require' you need to have the latest released version of bitbake? Mar 21 20:55:54 RP: should we dump trunk into a bitbake releasE? Mar 21 20:56:17 03dirk 07org.oe.dev * r67f607ed... 10/packages/linux/ (3 files in 2 dirs): linux-openzaurus_2.6.16: add patches for tosa Mar 21 20:56:40 uml forks off a shitload of processes... how to determine which one to attach my gdb to.. Mar 21 20:56:43 bleh Mar 21 20:56:50 do you know what OS use this pda phone : Mar 21 20:56:52 http://www.imcosys.com/ Mar 21 20:56:54 ? Mar 21 20:57:36 zecke: i have comment the require and i try without the last version Mar 21 20:57:37 kergoth: are they actually forks, or just threads? Mar 21 20:57:45 if i have a bug a i install it Mar 21 20:58:08 "detaching after fork from child process" Mar 21 20:58:09 :( Mar 21 20:58:15 looks like both Mar 21 20:58:18 doh Mar 21 20:58:29 * kergoth shakes fist at uml Mar 21 20:58:46 maybe you can attach a separate gdb to each one Mar 21 20:58:59 use the power of your multitasking operating system, and all that Mar 21 20:59:29 nacer_: http://www.linuxdevices.com/news/NS4475809720.html Mar 21 20:59:55 hmm, you can add c code to make it wait for a gdb to connect to it, cant you? Mar 21 21:00:06 * kergoth really needs to suck less with gdb one of these days :) Mar 21 21:00:45 lrg: pong Mar 21 21:00:51 hey koen Mar 21 21:01:04 hey koen Mar 21 21:01:32 koen: i think I may have fast (100Mb) mirror for OE Mar 21 21:02:24 kergoth: try just putting in a sleep(100) or something. When gdb connects it should cause the sleep to return immediately. Mar 21 21:02:32 ah. Mar 21 21:02:38 thanks Mar 21 21:02:57 i'm such a printf/printk debugger :P Mar 21 21:03:04 koen: I should ssh get access to it in a few weeks and set up some accounts. someone could then set up an OE mirror. Mar 21 21:03:37 get ssh Mar 21 21:03:41 or alternatively, the freebsd folks have a slightly more sophisticated version: see the end of http://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/debugging.html Mar 21 21:04:17 kergoth: heh, yeah. given the sucky state of gdb on arm, it's easy to get used to printf. Mar 21 21:04:32 but gdb on i386 is usually pretty good. Mar 21 21:04:52 pb_: hehe Mar 21 21:04:56 pp_: heh, use printk all the time Mar 21 21:05:04 yeah, you crazy kernel guys Mar 21 21:05:05 pb_: printf is there any other debugging tool for arm? Mar 21 21:05:25 zecke: a flashing LED ;) Mar 21 21:05:26 lrg: you can start with 2.6.16 Mar 21 21:05:34 zecke: yeah, oscilloscope Mar 21 21:05:49 do13: will do, I'll mt pull and rebuild Mar 21 21:06:12 zecke: and if your rich, a logic analyser. Mar 21 21:06:20 they all are pretty good tools Mar 21 21:06:33 LED's when getting the bootldr running on your new board :} Mar 21 21:06:42 oh it is green, it initialized the UART1 Mar 21 21:07:16 zecke: at least mainstone has hex leds for dumping memory addresses :) Mar 21 21:07:33 wow Mar 21 21:08:01 zecke: although, I've not had to do that for a long time as things have been stable for ages. Mar 21 21:08:15 zecke: indeed. and then, when your code gets more sophisticated, you can use the same led to send error messages in morse code Mar 21 21:08:23 heh Mar 21 21:08:25 * zecke remembers writing a seven segment display with gal Mar 21 21:09:00 pb_: but what about if you have a timing proble ;) Mar 21 21:09:07 problem Mar 21 21:09:45 mm, then I guess you're fairly screwed Mar 21 21:11:01 incidentally, you don't even need to be all that rich these days to have a logic analyzer; see http://cgi.ebay.co.uk/HEWLETT-PACKARD-1630D-LOGIC-ANALYZER_W0QQitemZ7602281344QQcategoryZ25399QQssPageNameZWDVWQQrdZ1QQcmdZViewItem for example Mar 21 21:12:35 pb_: wow, that's good value. I've been thinking about one for home..... Mar 21 21:13:26 hmm not responding here Mar 21 21:13:46 I guess you need the equivalent ebay.de url :-} Mar 21 21:14:00 it's item number 7600640080 Mar 21 21:14:19 here is an usb based one http://www.pctestinstruments.com/ Mar 21 21:14:22 pb_: thanks Mar 21 21:15:17 Hmm. I could use a logic analyzer :) Mar 21 21:15:46 Yeah. It's nice Mar 21 21:15:53 do13: very nice (except for the windows front end) 200MHz will definatly be enough for all portable audio busses Mar 21 21:16:20 I2C, SPI and RS232 Interpreters are good Mar 21 21:17:00 My scope is a bit behind the times now as well I guess... Mar 21 21:17:19 RP: old analog one? Mar 21 21:17:38 do13: Yes. Very old... Mar 21 21:18:24 RP: Same here.. Mar 21 21:18:48 * lrg has a http://www.tek.com/Measurement/cgi-bin/framed.pl?Document=/Measurement/logic_analyzers/home.html?wt=257&link=/Measurement/logic_analyzers/home.html&FrameSet=logic_analyzers at work Mar 21 21:20:05 lrg: cool news about the mirror Mar 21 21:20:09 The funny thing is that they sell it with 8Mb of physical RAM and only enable 512k of it. You need to buy a software key to get the rest Mar 21 21:20:28 koen: yes, i'm hoping it will come in handy for other stuff as well Mar 21 21:20:39 I wonder if I can talk my boss into kitting out an electronics lab :) Mar 21 21:20:48 somehow I doubt it... Mar 21 21:21:07 RP: you could alway come up to Edinburgh and borrow some of my stuff Mar 21 21:21:21 wow I 'parsed' all bitbake files within 9 seconds :} Mar 21 21:21:32 zecke: awesome Mar 21 21:21:33 zecke: wow Mar 21 21:22:03 lrg: totoally unusable though Mar 21 21:22:11 koen: who would be the best person to set up the OE mirror ? Mar 21 21:22:16 zecke: :( Mar 21 21:22:26 zecke: not awesome Mar 21 21:22:33 lrg: Thanks, I'll keep that in mind :) Mar 21 21:22:58 pb_: well I now I know that collecting files takes 9 seconds + parsing Mar 21 21:23:27 pb_: eval+compile+exec takes about 1:50 minute Mar 21 21:23:42 zecke: so, what did you actually do in the 9 seconds? is that just reading the files and lexing them? Mar 21 21:24:26 pb_: sleep(8) Mar 21 21:24:40 zecke: ah, I sense an opportunity for optimisation Mar 21 21:25:13 pb_: it parsed local/conf, getting FILES, PROVIDES from the file Mar 21 21:25:27 lrg: mickeyl, hrw or me Mar 21 21:25:28 pb_: so we have 9 seconds python overhead Mar 21 21:25:38 mickey|zzZZzz: ping? Mar 21 21:25:51 koen: ok, I'll send you all an email when I know more. Mar 21 21:31:45 lrg: ok, thanks Mar 21 21:40:21 zecke: oh, crumbs Mar 21 21:41:28 pb_: what are 9 seconds, taking account ewi is not the fastest machine :} Mar 21 21:42:01 oh, I thought ewi was totally the fastest machine Mar 21 21:43:15 pb_: just because it pulls within hours :} Mar 21 21:44:00 pb_: what takes 10 seconds on ewi takes 20 on my ibook :( Mar 21 21:46:37 * chouimat is away: pub Mar 21 21:47:37 zecke: ah, ewi is only twice as fast as your ibook? then it is indeed slower than I thought. Mar 21 21:47:40 zecke: What proportion of data operations are flag based vs operations on the variables themselves? Mar 21 21:48:00 RP: I don't know yet Mar 21 21:48:30 RP: a.print_callers('getVar') Mar 21 21:48:44 RP: do get a list of who called a method looking like getVar* Mar 21 21:49:07 RP: getVarFlag gets called by getVar as well Mar 21 21:50:10 pb_: now I'm at two seconds for bitbake opie-image Mar 21 21:50:32 very good Mar 21 21:50:35 pb_: and bitbake world takes 8 Mar 21 21:50:38 presumably gpe-image is much slower due to x11 bloat Mar 21 21:50:50 pb_: so building the world package list takes 6 seconds Mar 21 21:51:05 pb_: yeah also parsing the massive amounts of badly formated c code Mar 21 21:51:17 pb_: takes at least 2^n the time parsing c++ takes Mar 21 21:51:24 so I heard Mar 21 21:52:08 zecke: is it IO or cpu bound? Mar 21 21:52:17 koen: both, I guess Mar 21 21:52:38 koen: could you install python 2.4.2 on ewi? Mar 21 21:53:32 zecke: done Mar 21 21:53:34 night all Mar 21 21:53:39 koen: thanks Mar 21 21:54:17 damn Mar 21 21:54:24 searching hotshot Profile... Mar 21 21:54:38 Yeah I want to see profiles of 16 year old brit girls Mar 21 21:57:07 File "", line 1, in ? Mar 21 21:57:07 File "/usr/lib/python2.4/hotshot/stats.py", line 3, in ? Mar 21 21:57:07 import profile Mar 21 21:57:07 ImportError: No module named profile Mar 21 21:57:12 hmm still broken Mar 21 21:57:55 RP: who was writting the CE-RH1 driver? Mar 21 21:58:32 do13: CE-RH2 I think and it was JustinP Mar 21 21:58:48 RP: This is nice http://home.arcor.de/relei2004/zaurusfernb.html Mar 21 21:59:31 hehe Mar 21 22:00:26 do13: Nice to know it works the way I thought it did :) Mar 21 22:00:55 RP: Can you read the text? Mar 21 22:02:40 do13: Not exactly but I think the meaning is fairly clear - circuit diagrams are a universal lanuage :) Mar 21 22:03:02 :) Mar 21 22:03:33 Its interesting the values changes between c7x0 and cxx00 Mar 21 22:03:35 RP: I tried explaining that to a few CS students Mar 21 22:03:55 RP: the gave me blank look and said "Dude, just use java for that" Mar 21 22:04:03 i have a error with compilation of glibc Mar 21 22:04:05 | make[2]: *** [/home/ovea/laradji/emulation/gentoo/os1/build/tmp/work/i386-linux/glibc-2.3.5+cvs20050627-r3/build-i386-linux/csu/divdi3.os] Erreur 1 Mar 21 22:04:05 | make[2]: Leaving directory `/home/ovea/laradji/emulation/gentoo/os1/build/tmp/work/i386-linux/glibc-2.3.5+cvs20050627-r3/libc/csu' Mar 21 22:04:05 | make[1]: *** [csu/subdir_lib] Erreur 2 Mar 21 22:04:19 * koen wonders how to do circuit diagrams in java Mar 21 22:06:59 koen: They probably meant you needed coffee :) Mar 21 22:08:43 night all Mar 21 22:09:10 mickey|zzZZzz: I need you :) Mar 21 22:09:43 'night Dirk Mar 21 22:10:08 koen: CS students, for the most part, are idiots... don't bother explaining it ;) Mar 21 22:13:47 zecke: bad luck, it appears that mickey has fallen into some kind of enchanted 100-year sleep Mar 21 22:15:52 hmm without psyco times go from 2 to 6 seconds Mar 21 22:17:45 heh Mar 21 22:17:58 RMK gives an example on how *not* to use git Mar 21 22:18:11 koen: with his external patches? Mar 21 22:18:29 koen: well, I love it. The second question was to get even more from rmk :} Mar 21 22:18:41 hey now you publish patches, gives us your git tree Mar 21 22:19:10 with parsing in python it goes from 6 seconds to 2m26 Mar 21 22:25:45 Trying to compile busybox, Giving me some errors about posix_types.h http://pastebin.com/615183 any suggestions? Mar 21 22:31:55 Anyone? Mar 21 22:32:48 'night all Mar 21 22:42:33 stupid question of the night - the oz354x branch still uses the org.openembedded.oz354fam083 dir? Mar 21 22:42:49 Bernardo|away: that depends Mar 21 22:43:02 Bernardo|away: if you do co --branch oz354x you will have a oz354x dir Mar 21 22:43:15 Bernardo|away: if you did up --branch oz354x the dir name is not changed Mar 21 22:43:27 zecke: thanks Mar 21 22:46:02 now to find out why am I getting a "branch 'oz534x' is empty" Mar 21 22:46:18 534? Mar 21 22:46:24 354 Mar 21 22:46:39 Bernardo|away: did you pull that branch as well? Mar 21 22:47:26 yes, I did "monotone --db=oe.db pull monotone.vanille.de 'org.openembedded.oz354x" Mar 21 22:47:35 good Mar 21 22:47:35 and it got 30 revs at least Mar 21 22:48:53 also, if I do a list branches on that db, it has dev, dreambox, oz354fam083 and oz354x Mar 21 22:49:17 nm Mar 21 22:49:19 if you try to check it out? does it work? Mar 21 22:49:26 seems I did a mistype Mar 21 22:49:48 trying to check it out now again and it seems to be doing something Mar 21 22:50:00 yep, got the dir right now Mar 21 22:50:06 I must have mistyped before Mar 21 22:50:14 thanks zecke, sorry for the trouble Mar 21 22:51:26 i need grunts Mar 21 22:51:50 'night everyone Mar 21 22:51:50 mithro: hi Mar 21 22:51:58 mithro: can one mix WxPython and PyGame? Mar 21 22:52:50 anyone want to be a Grunt for the Thousand Parsec project? Mar 21 22:53:26 mithro: what is a grunt? Mar 21 22:53:46 mithro: I can make sounds Mar 21 22:55:37 zecke: somebody to do the boring work, IE fix up the xml description of the protocol Mar 21 23:07:47 JustinP: ping Mar 21 23:09:18 e-image-core doesn't appear to work (on the c750, though I don;t think that's the issue) Mar 21 23:09:43 I'm presented with a login prompt, and am returned to it every 20-30s or so Mar 21 23:10:53 http://people.bath.ac.uk/enpsgp/Zaurus/temp/ has dmesg.txt and logread.txt output - looks like a problem with entrance Mar 21 23:11:12 Jan 1 00:13:06 c7x0 daemon.crit entrance: Cannot initialize requested display ":0". Exiting. Mar 21 23:12:17 * lardman wonders what timezone JustinP is in Mar 21 23:12:24 ~seen JustinP Mar 21 23:12:28 justinp is currently on #oe (21h 28m 59s) #openzaurus (21h 28m 59s). Has said a total of 29 messages. Is idling for 1h 28m 21s, last said: 'in general, although there are some frontends'. Mar 21 23:12:59 back tomorrow, night-night Mar 21 23:14:25 ot, but if anyone ever wants to use gdb on a user mode linux, disable CONFIG_MODE_TT Mar 21 23:14:32 when its enabled, it hoses gdb Mar 21 23:14:42 burned a few hours on that damn config option :P Mar 21 23:36:53 can i use generic distro ? Mar 21 23:36:59 or its justs a example ? Mar 21 23:37:07 for build oe Mar 22 04:06:05 RP: hey uh, 2.6.16-r0 kernel doesn't seem to boot in Angstrom build for spitz Mar 22 04:06:12 still not EABI capable? Mar 22 04:06:20 or did i have to turn it on somehow Mar 22 04:11:16 i have some probleme with monotone Mar 22 04:11:42 ~lart monotone Mar 22 04:11:43 * ibot brandishes Excalibur! "With this sword, I vanquish thee, monotone!" and lops off monotone's head Mar 22 04:11:52 One down.... Mar 22 04:11:58 http://pastebin.com/615603 Mar 22 04:12:05 So that must be how to deal with multiple heads.... Mar 22 04:13:05 nacer_: try 'LANG=C monotone pull ' Mar 22 04:13:14 ok Mar 22 04:13:32 nacer_: there were some problems at various points with the UI code getting confused by i18n Mar 22 04:14:00 i try in console ? Mar 22 04:14:12 LANG variable dont work Mar 22 04:14:16 upgrading might help too, depending on what version you're using now, I don't remember whether 0.25 had this bug or not. Mar 22 04:14:21 0.25 Mar 22 04:14:25 nod Mar 22 04:15:11 perhaps i can remove the output ? Mar 22 04:15:24 --quiet perhaps Mar 22 04:15:26 solve the pb Mar 22 04:15:31 that might work too :-) Mar 22 04:15:41 nop Mar 22 04:15:46 --quit have no effect :p Mar 22 04:16:27 oh, right, --quiet affects chatter, not the tickers -- try --ticker=none Mar 22 04:16:35 or --ticker=dot might work too Mar 22 04:17:59 nop Mar 22 04:18:02 same error Mar 22 04:19:56 sigh Mar 22 04:20:30 okay, so let's figure out the right environment variable to get you back in the C locale, for monotone :-) Mar 22 04:20:41 yup Mar 22 04:20:46 i make some search Mar 22 04:21:01 do you have any environment variables named LC_ set? most likely would be LC_ALL? Mar 22 04:21:08 LC_ALL Mar 22 04:21:36 LC_ALwork Mar 22 04:21:39 its work Mar 22 04:21:40 tks Mar 22 04:22:06 LC_ALL=C monotone ... Mar 22 04:22:54 yeah Mar 22 04:25:59 do you think its a good id to use generic distribution in local.conf of bitbake ? Mar 22 04:26:21 i want to build a embedded system for x86 Mar 22 04:26:49 only for try opie and GPE Mar 22 04:28:19 for run it with qemu :) Mar 22 04:28:33 i dont have found any livecd for try this software Mar 22 04:28:48 so i try de build image Mar 22 04:31:01 new problem :) Mar 22 04:31:03 http://pastebin.com/615618 Mar 22 04:31:23 afk cmoke Mar 22 04:31:24 smoke Mar 22 04:34:13 nacer_: looks like your binutuls version. try making PREFERRED_VERSION_binutils a lower version Mar 22 04:39:16 sweet, my mom just talked to me Mar 22 04:40:06 for real. not that "hey i need you to hammer some boards together, wash the dishes, and write a zillion thankyou cards" Mar 22 04:40:40 ok JustinP Mar 22 04:41:07 how to 'make config' the kernel? Mar 22 04:42:02 JustinP: where can i found doc about the PREFERERED OPTION ? Mar 22 04:42:10 JustinP: i dont have example on my local.conf Mar 22 04:45:17 conf/distro/preferred-* Mar 22 04:45:20 jnc: i dont understand you question Mar 22 04:45:26 JustinP: ok Mar 22 04:45:47 nacer_: i want to run 'make config' on the linux kernel sources for the cxx00 target Mar 22 04:45:57 not sure how to do that Mar 22 04:46:08 goal is to generate EABI-enabled config Mar 22 04:47:34 jnc: i dont know :) Mar 22 04:47:45 jnc: just go inthe work dir and do it Mar 22 04:47:55 I suggest menuconfig, however Mar 22 04:48:20 :/ Mar 22 04:48:25 some things changed names Mar 22 04:48:29 this isn't going to be clean Mar 22 04:48:39 need help from someone responsible for those defconfig files Mar 22 04:48:55 make oldconfig Mar 22 04:49:22 CONFIG_ZBOOT_ROM_BSS=0x0 Mar 22 04:49:22 +CONFIG_CMDLINE="console=ttyS0,115200n8 console=tty1 noinitrd root=/dev/hda1 rootfstype=ext3 rw fbcon=rotate:1 dyntick=enable debug" Mar 22 04:49:26 # CONFIG_XIP_KERNEL is not set Mar 22 04:49:33 i don't think that should be in defconfig Mar 22 04:49:38 ...as an example Mar 22 04:53:34 where can i found the version of binutils can be used for binutils Mar 22 04:53:41 or do you recomend one ? Mar 22 04:55:01 i have found Mar 22 04:57:20 jnc: is that in the workdir? Mar 22 04:57:40 jnc: yes, it should be...that's how it works AFAIK...see the kernel bb files Mar 22 04:57:47 nacer_: packages/binutils Mar 22 04:57:58 JustinP: yep i have see Mar 22 04:58:01 tks Mar 22 04:59:48 by default bitbake use the last version ? Mar 22 05:00:00 example there is a binutils-cvs Mar 22 05:00:05 he build the cvs one ? Mar 22 05:01:44 JustinP: that is a kconfig option Mar 22 05:02:09 it appears though i am quite certain that it should not Mar 22 06:32:17 anybody still alvie? Mar 22 06:32:39 yup Mar 22 06:35:36 ...and screaming for it Mar 22 06:36:05 looking to buy a atmel avr / pic programmer, any suggestions Mar 22 06:36:28 you can forget about using linux with it Mar 22 06:37:06 heh yea i know, just for hardware hacking purposes Mar 22 06:37:15 start saving for an windows xp license Mar 22 06:37:43 that's going to be the most expensive part Mar 22 06:37:45 oh with the programmer, hah. its late, i thought running linux on the avr Mar 22 06:38:17 really i thought i came across programmers that worked with linux Mar 22 06:38:45 the willem is well talked about Mar 22 06:39:17 there are lots availible Mar 22 06:39:25 depends on how much you wanna spend Mar 22 06:39:35 btw jnc linux has run on some PICs Mar 22 06:39:43 i checked though, and the code is greedily ill-licensed by a guy who links against win16 dlls Mar 22 06:40:00 (willem) Mar 22 06:40:03 like $50-100. something cheap that will work with the popular pics / avrs. i'm leaning towards avrs over pics Mar 22 06:40:27 TheMasterMind1, what language youplanning on programming in? Mar 22 06:40:49 see midibox project for good examples of PICs in use Mar 22 06:42:04 i can tell you from personal experiance that is my biggest problem using then on/in linux Mar 22 06:42:10 them* Mar 22 06:42:48 the opensource and non-windows compilers for C are very ... limited Mar 22 06:43:28 FSF seems to overlook that with the free BIOS initiative Mar 22 06:43:38 gcc-avr no good? C if possible but i'm getting them to tinker and learn so im' not opposed to diving down into asm Mar 22 06:43:41 won't return my emails Mar 22 06:44:14 expect unoptiomised code that uses 30-50% more memory on the chip Mar 22 06:44:46 wow Mar 22 06:44:48 actually one sec Mar 22 06:45:26 ok i own a copy of ccs picmicro linux compiler Mar 22 06:45:51 "Linux that support the PIC10, PIC12, PIC14, PIC16, and PIC18 microcontrollers." Mar 22 06:45:57 i have never used that one Mar 22 06:46:10 so it may be decent Mar 22 06:46:18 it wasn't cheap =P Mar 22 06:46:36 nore is HTC or microchip's Mar 22 06:47:06 yea true. i did my research before buying it though, just don't remember any of the details right now Mar 22 06:47:23 altho i am curious if HTC will release a linux compiler now Mar 22 06:47:26 but anyway, i'm good on that end. need a programmer basically Mar 22 06:47:39 they are using Eclipse for their HiTide IDE Mar 22 06:48:03 so first of all, whats a good pic to start out with Mar 22 06:48:21 depends what you would liek to do Mar 22 06:48:47 the 16f877 is pretty standard fare for new people Mar 22 06:49:06 yea that sounds familiar, should be fine for me Mar 22 06:49:19 simpler than the 18F and still has enough to be interesting Mar 22 06:49:28 programmers generally cover the entire 16f family? Mar 22 06:49:36 mmm Mar 22 06:49:45 3/4 i would say Mar 22 06:49:48 good enough Mar 22 06:50:24 but if your gonna go with PICs i would suggest an ICD Mar 22 06:50:28 so i need usb/serial/parallel/whatever programmer that will work in linux Mar 22 06:50:33 ICD? Mar 22 06:50:52 if you do enough or get interested enough the debugging features are very useful Mar 22 06:51:03 yeah Mar 22 06:51:19 it's microchips own Mar 22 06:51:31 suppoerted by a few tools in linux Mar 22 06:51:55 the ICD2 has 1 or two supportes that i know of Mar 22 06:52:37 but alternatively if your only looking for an intro Mar 22 06:53:00 you can buy a bootloader board and chip preprogrammed Mar 22 06:53:12 for less than $30 Mar 22 06:53:46 you can make the board yourself for ~10 Mar 22 06:54:06 or protoboard it for ~3 Mar 22 06:54:18 hrmf. last time i tried making myself one i was too lazy to follow through Mar 22 06:54:25 lol Mar 22 06:54:36 its just a MAX232 and a couple caps Mar 22 06:54:45 .. and resistors Mar 22 06:54:47 really? Mar 22 06:54:51 yeah Mar 22 06:54:53 maybe it was an avr programmer i was trying to make Mar 22 06:55:32 ok so what do you mean by bootloader board and chip preprogrammed? Mar 22 06:55:53 in order to use the bootloader you have to get it on the chip Mar 22 06:56:25 and a bootloader board just interfaces teh serial cable to your chip Mar 22 06:56:39 hence the MAX232 chip Mar 22 06:56:51 * NAiL hugs his STK500 Mar 22 06:57:17 alrighty gotcha. should be able to google up the rest Mar 22 06:57:25 what about avrs, whats a good one to start out with Mar 22 06:57:29 90S-something? Mar 22 06:57:52 might ask NAiL what he prefers Mar 22 06:58:13 looks like he has a prety fancy dev board Mar 22 06:58:27 Is it fancy? ;) Mar 22 06:58:45 NAiL, compared to what i use, yes Mar 22 06:58:46 Friend of mine works over at Atmel, so I'm getting some stuff kinda cheap :) Mar 22 06:58:54 yeah? Mar 22 06:59:03 yup Mar 22 06:59:13 i'll take cheap/free dev stuff any time :) Mar 22 06:59:23 hehe Mar 22 06:59:54 heh Mar 22 07:00:00 speaking of, chris still working at TI? Mar 22 07:00:21 I'm not sure how much the STK500 costs, but it's a brilliant board to start development... And it's got an AtMega on when you buy it. Mar 22 07:00:40 its 79 USD at digikey Mar 22 07:00:59 http://www.digikey.com/scripts/dksearch/dksus.dll?KeywordSearch?Keywords=ATSTK500%20ATMEL&site=us Mar 22 07:01:23 $79?? Mar 22 07:01:35 yup Mar 22 07:01:53 hrm yea 79 on digikey. none available though Mar 22 07:02:55 Anyway, if you're gonna go with AVR, the STK500 is a very nice way to go :) Mar 22 07:03:03 lol, infact its not available anywhere Mar 22 07:03:23 heh Mar 22 07:03:31 that makes it kind of hard ;) Mar 22 07:04:33 whats a good atmel chip to start out with Mar 22 07:04:39 something like 8 or 20 pins Mar 22 07:04:48 not sure i like the name 'terrier' Mar 22 07:04:55 aka c3200 Mar 22 07:06:13 well there's only so many names to choose from Mar 22 07:06:24 TheMasterMind1: Well, I'm using a lot of ATtiny44 (http://www.atmel.com/dyn/products/product_card.asp?PN=ATtiny44) Mar 22 07:07:54 The STK505 board adds support for the 14-pin and 20-pin AVR microcontrollers. Mar 22 07:07:54 hm Mar 22 07:08:20 the tinys are kinda neat, especially since they hardly use any power :) Mar 22 07:08:31 ah, yeah, the 505 is basically the same as the 500 Mar 22 07:09:07 i was kind of hoping for blacklab Mar 22 07:09:38 or sharpei Mar 22 07:09:39 505 looks like an addon board Mar 22 07:09:41 *cough* Mar 22 07:09:51 quepasa? Mar 22 07:10:28 urgh, I can't check right now, gotta go Mar 22 07:11:18 bug me later about the stuff ;) Mar 22 07:11:23 alright, thanks for the help Mar 22 07:11:29 np Mar 22 07:11:32 *gone* Mar 22 07:20:41 good morning Mar 22 07:21:51 hi Mar 22 07:22:15 Angstrom boots w/ EABI on spitz, udevd crashes though Mar 22 07:22:23 that means i can't log in Mar 22 07:23:31 your better off than i Mar 22 07:23:35 oh? Mar 22 07:23:55 might have EABI support in 2+ yrs Mar 22 07:24:02 which hw Mar 22 07:24:13 ipaq sa Mar 22 07:24:17 hmm Mar 22 07:24:34 hi all ! Mar 22 07:24:48 hey Mar 22 07:25:05 waht is the difference between QTopie(Trolltech) and opie(OE)? Mar 22 07:25:16 QTopia Mar 22 07:25:21 sorry Mar 22 07:25:28 not QTopi(e?) Mar 22 07:26:00 s/QTopie/Qtopia/ Mar 22 07:26:11 opie branched from qtopia some time ago Mar 22 07:26:20 now has more features Mar 22 07:27:06 the main difference is sources Mar 22 07:27:29 there are many closed QT sources used onlyt by Trolltech and thier licensees Mar 22 07:27:43 I just can't build kernel 2.6.15 for the akita, bitbake keeps defaulting to 2.4.20 even though I've set PREFERRED_VERSION and KERNEL_VERSION Mar 22 07:27:55 branch 354x Mar 22 07:28:15 in dev branch it builds the 2.6 kernel Mar 22 07:28:24 Trolltech told me if we want to include QTopia in our boards we have to pay runtime licenses, because of selling hardware with QTopia... when we include opie there could trolltech not do anything ? Mar 22 07:28:46 not sure Mar 22 07:29:21 tt could use some money though Mar 22 07:29:34 not a fan of GPE? Mar 22 07:29:38 I'm not sure if they could ask for runtime licenses for qt/e Mar 22 07:29:38 grma, i wouldnt count on that Mar 22 07:29:53 i think who you want to talk to is eitehr zecke or pb_ Mar 22 07:30:11 one of them do or did work at Trolltech Mar 22 07:30:23 emte: ok better asking zecke or pb_, correct ? Mar 22 07:30:49 they would know troll's licenesing better than anoyone else i would think Mar 22 07:30:56 ok thanks Mar 22 07:31:01 oooh i'm logged into spitz w/ EABI-built Angstrom Mar 22 07:31:01 since they also devel opie Mar 22 07:31:03 yay Mar 22 07:37:19 hi all Mar 22 07:48:21 emte: do you have a atmel programmer? Mar 22 07:48:29 s/atmel/avr/ Mar 22 07:48:58 grma: if you use gpl software, you'd have to abide by the licensing Mar 22 07:49:29 and for the record, I do work for Trolltech Mar 22 07:51:21 hrmm. who was it that moved to australia Mar 22 07:51:29 that would be me Mar 22 07:52:04 thought so Mar 22 07:52:08 still there or move back? Mar 22 07:52:20 still here. married w/ kid :) Mar 22 07:52:25 oh wow, nice Mar 22 07:52:42 dont think I'm getting away too easily, unless I get deported Mar 22 07:52:57 i don't remember you working for TT Mar 22 07:53:06 guess i forgot Mar 22 07:53:09 ya, about 2.5 years now Mar 22 07:53:17 after the move? Mar 22 07:53:26 moved here for the job Mar 22 07:53:33 k, that rings a bell Mar 22 07:53:55 ljp: I meeting now, can i ask you some things later ? Mar 22 07:53:55 been a while since i've been here Mar 22 07:54:09 s/I meeting/ I hav a meeting/ Mar 22 07:54:57 yup Mar 22 07:55:22 it may take me a some time to answer... Mar 22 08:02:48 emte: http://www.ccsinfo.com/content.php?page=compcompare Mar 22 08:42:43 grma: yeah, best to talk to ljp. I have never worked at trolltech and I don't think zecke has either. Mar 22 08:50:26 morning Mar 22 08:53:11 morning Mar 22 08:53:23 morning hrw|work congrats on scoring the Poodle Mar 22 08:53:55 XorA: thx - I had to go home when aa_ told magic word 'poodle' Mar 22 08:55:20 http://pastebin.com/615789 if interested Mar 22 08:56:22 hrw|work: read it last night, was very cool Mar 22 08:56:35 fast shot Mar 22 08:58:11 hrw|work: what are you running on c7x0 at moment Mar 22 08:59:00 3.5.4.1-alpha1 Mar 22 08:59:00 03hrw 07org.oe.oz354x * r5a02f158... 10/packages/base-files/ (6 files in 6 dirs): base-files: cleanup SD/MMC mountpoints - remove 2.4 on 2.6 only machines, comment 2.6 on tosa Mar 22 08:59:07 03hrw 07org.oe.oz354x * r53f9176d... 10/packages/linux/ (36 files in 2 dirs): dropped 2.4.20-crapix support - we use 2.6 on cxx00 Mar 22 08:59:16 opie? Mar 22 08:59:21 yes Mar 22 08:59:27 brb Mar 22 09:00:25 03hrw 07org.oe.oz354x * r4e8be8a8... 10/packages/linux/openzaurus-pxa-2.4.18-rmk7-pxa3-embedix20031107/ (defconfig-corgi defconfig-husky defconfig-shepherd): dropped 2.4.18-crapix support for corgi/shepherd/husky - we use 2.6 on them Mar 22 09:03:24 morning all Mar 22 09:03:47 hi RP Mar 22 09:04:54 hrw|work: I noticed the comments on the forum and see you've dealt with some of them :) Mar 22 09:05:13 Bernardo|away: this time you should get proper kernel built Mar 22 09:05:47 RP: I'm trying to create nice image for myself Mar 22 09:06:24 RP: and it has nice side effects Mar 22 09:06:43 RP: please enable 'tun' module in 2.6 kernels Mar 22 09:07:13 RP: can we enable most of not-yet-enabled modules? Mar 22 09:09:50 is there any volunteer for a test? it involves going to the console (in the handheld) and opening "vi" Mar 22 09:10:11 someone familiar with maemo here? I need to know does *osso* means maemo stuff Mar 22 09:10:22 Mardy: I can Mar 22 09:10:28 I think I spotted a bug, but I'd like someone else to verify :-) Mar 22 09:10:31 hwtech: thanks Mar 22 09:10:48 hrw|work: Which other modules do you think we need? Mar 22 09:10:49 just write two lines in vi: Mar 22 09:10:52 but have to reboot Z Mar 22 09:11:08 hrw|work: np, please let me know when you are ready Mar 22 09:11:13 RP: we like we - users can have misc strange ideas Mar 22 09:11:14 Bernardo|away: In the branch, use ZKERNEL_VERSION = "2.6" Mar 22 09:12:01 hrw|work: More modules slows my builds down. If people use/request them, fine, otherwise... Mar 22 09:12:29 RP: ok - then 'tun' atleast Mar 22 09:12:40 hrw|work: Agreed, that one is fine Mar 22 09:12:48 Mardy: osso does mean Maemo Mar 22 09:13:55 RP: thanks, although that was hrw's question :-) BTW, in italian "osso" means "bone" Mar 22 09:14:03 RP: ok. I will drop all *osso* from feed Mar 22 09:14:35 Mardy: vi ready Mar 22 09:15:15 Mardy: Sorry, too early in the morning. Some of their other names are amusing in English :) Mar 22 09:15:32 RP: btw - google maps have nice cityplans for england - I gave part of your address to it and only left is searching for proper street Mar 22 09:15:34 03hrw 07org.oe.oz354x * r124f9d07... 10/conf/distro/openzaurus-3.5.4.1.conf: Mar 22 09:15:34 openzaurus 3.5.4.1 conf: bumped to alpha2 (we gave alpha1 to users already) Mar 22 09:15:36 - added 2.4/2.6 kernel selectors for Zaurus machines Mar 22 09:16:03 hrw|work: you can see my house in google maps, even the greenhouse that used to be in the garden Mar 22 09:16:04 hrw|work: The post code alone should be enough to have it find my street :) Mar 22 09:16:13 hrw|work: in vi, write two lines, then press ESC (cancel), and move up to the first line Mar 22 09:16:25 You can also clearly see my house... Mar 22 09:16:33 Mardy: echo $TERM, Ill bet its set to vt100 Mar 22 09:16:36 Mardy: and? Mar 22 09:16:51 hwtech: then press the press "i" (insert), and enter Mar 22 09:16:56 ops Mar 22 09:16:57 hrw|work: you get corrupted screen Mar 22 09:17:11 hrw|work: that was for you, not for hwtech, sorry Mar 22 09:17:13 Mardy: no problem on linnux console Mar 22 09:17:14 Mardy: try removing set TERM=vt100 from /etc/profile Mar 22 09:18:50 bonjour obergix[work] :) Mar 22 09:19:03 XorA: it happens anyway Mar 22 09:19:22 I just tried with TERM=linux and TERM="" Mar 22 09:20:09 hi Genesis Mar 22 09:20:32 Mardy: you tried in opie-console? rxvt? on linux console? Mar 22 09:20:56 hi ____law____ Mar 22 09:21:10 hrw|work: only in linux console Mar 22 09:21:13 hrw|work: I'm wondering about "ignorning" unknown SCR structures and trying to use the cards anyway. This might get a few more cards working... Mar 22 09:21:14 <__law__> hi hrw|work Mar 22 09:21:23 I'm trying to impenet hardware accel on the w100, and I noticed that Mar 22 09:21:29 <__law__> ~lart my inet connection Mar 22 09:21:29 * ibot sends a legion of lawyers after my inet connection's head Mar 22 09:21:34 impenet->implement Mar 22 09:21:54 Mardy: You're doing a good job too :) Mar 22 09:22:01 RP: I would like to get one of such cards to check it Mar 22 09:22:20 RP: yes, I see the same corruption, so I guess it's all right :-) Mar 22 09:22:37 Mardy: I delete that line totally so TERM is set by the terminal like it should, and busybox vi is nicer on rxvt and putty now Mar 22 09:22:51 Mardy: but busybox vi is fairly duff, so dont expect miracles :-) Mar 22 09:23:59 btw, even with accels, I don't see a performance improvement in console scrolling Mar 22 09:24:52 Mardy: Really?. Rats :-( Mar 22 09:31:04 ~lart our not existant maemo maintainer Mar 22 09:31:04 * ibot pours gasoline all over our not existant maemo maintainer, ignites the fire, and then enjoys some toasty marshmallows with the glorious blaze Mar 22 09:33:35 god I want that 32Gig flash drive Mar 22 09:33:53 XorA: Yeah, at what cost though? Hell, even 4gb flash isn't cheap. Mar 22 09:34:07 NAbyss: cost is no object in my dreams :-) Mar 22 09:34:18 XorA: True, true :) Mar 22 09:34:27 Translating dreams to reality though, a little hard :) Mar 22 09:34:37 NAbyss: 3 of them would make all my computers HD free Mar 22 09:37:40 XorA: 1000$ :) Mar 22 09:42:25 ~hail ipkg Mar 22 09:42:30 * ibot bows down to ipkg and chants, "I'M NOT WORTHY!!" Mar 22 09:43:07 I have directory on ewi where I do 'ipkg update;ipkg install soemthing' so I can test feed upgrades Mar 22 09:46:05 RP: you have new mail, with an attachment, too :-) Mar 22 09:56:25 Mardy: I think you need to work on error handling for a start :-/ Mar 22 10:00:23 hi ! I'm back from my meeting ... how is that now with QT/QTopia/opie and licenses? QTopia is from Trolltech correct ? Mar 22 10:00:42 and QTopia is under a Trolltech license ? Mar 22 10:01:50 RP: you mean checking the return of w100_fifo_wait()? Mar 22 10:03:28 Mardy: Yes. I can see why you've done what you've done but kernel developers won't understand Mar 22 10:04:28 mmm... but it's the same logic used in aty/radeon_accel.c and mach64_accel.c Mar 22 10:05:08 Hmmm. They might accept it then, I don't know. Depends on the person you submit it to I guess Mar 22 10:05:29 wow... bitbake is having quite a lot parse errors suddenly :( Mar 22 10:06:23 RP: who should be the person? Russell or the FB mantainer? Mar 22 10:06:35 Mardy: You aso need to check your indenting. Other than that it looks good Mar 22 10:07:40 Mardy: Not Russell. You want the FB maintainer on linux-fbdev-devel - Antonino Daplas Mar 22 10:09:00 grma: qtopia is GPL or commercial license Mar 22 10:09:08 morning Mar 22 10:09:24 hi lardman|work Mar 22 10:09:34 RP: Is cpufreq disabled on the 3.5.4 releases for c7x0 (I saw no mention of it in dmesg)? Mar 22 10:09:38 hi hrw|work Mar 22 10:09:47 lardman|work: 3.5.4 yes. 3.5.4.1 no Mar 22 10:10:01 ah, okay Mar 22 10:10:21 grma: if you use qtopia/gpl then your stuff must be gpl. if you pay for commercial license then you can have your apps closed sources Mar 22 10:10:24 hrw|work: ok but what is with the point of selling hardware with QTopia? There were runtime licenses, does someone have experiences with that point ? Mar 22 10:10:24 lardman|work: I have no idea what's in 3.5.4 the kernel was that long ago! Mar 22 10:10:33 RP: :D Mar 22 10:10:47 morning all Mar 22 10:10:51 hi Dirk Mar 22 10:11:04 RP: I must admit I was a bit shocked to see that I'd reverted to 2.6.14-something Mar 22 10:11:14 Hey Richard Mar 22 10:11:19 JustinP: ping Mar 22 10:11:26 grma: iirc hardware with qtopia is sold by companies which get commercial license and add some own propertiary stuff - look at sharp for example Mar 22 10:11:50 grma: or motorola with their a780 phones - they do not use qtopia but use qt/e commercial licensed Mar 22 10:12:01 hi dirk Mar 22 10:12:15 hrw|work: i want to use the GPL version i have problems with astatement from trolltech: if you sell hardware with QTopia you have to pay runtime licences... Mar 22 10:12:22 Hey Marcin Mar 22 10:12:30 Mardy: You do know the w100 can do hardware scolling btw? That would really make the console fly Mar 22 10:12:44 grma: ah. then contact sales at trolltech. Mar 22 10:13:01 grma: here ljp could do some help - he is @trolltech Mar 22 10:13:13 hrw|work: ok we don't want to add commericail things, maybee our custommers, but that is then not my beer :-) Mar 22 10:13:25 i tried to contact ljp, but he is very busy Mar 22 10:13:51 hrw|work: but thanks for your infos Mar 22 10:13:59 grma: then contact sales @trolltech Mar 22 10:14:00 i try to contact ljp Mar 22 10:15:43 hrw|work: already done but sales@trolltech are persons who want to sell something :-) so you don't get that good infos about if you are hurting there licences or not ... or I didn't understand all correct, that could also be ;-) Mar 22 10:16:02 grma: Speak to their legal people? Mar 22 10:16:40 NAbyss: who would that be, legal people ?? what is that ? Mar 22 10:17:23 good morning all Mar 22 10:17:56 hi koen Mar 22 10:18:00 hi koen Mar 22 10:18:04 hi koen Mar 22 10:18:19 RP: I didn't know... What are the registers involved? do you have any sample code? Mar 22 10:18:25 grma: Their department that handles legal enquiries and contracts etc. Mar 22 10:19:10 NAbyss: ahh, thanks Mar 22 10:19:31 uf. cleaned on ewi - 6G freed Mar 22 10:26:38 Mardy: Try adjusting mmGRAPHIC_OFFSET if I remember rightly Mar 22 10:26:48 hi guys Mar 22 10:26:59 having a strange problem here Mar 22 10:27:27 Mardy: There is also potential to support virtual resolutions like 600x400 by increasing the size of the screen margins Mar 22 10:28:17 have bitbake 1.3.2 running here... when i try and use it with the .dev branch of openembedded, it seems to work fine... but if i try and use it with the .oz354fam083 branch, i keep getting EOL on single-quoted string errors for virtually every .bb file... any idea? Mar 22 10:28:34 HuntrCkr: OE and bitbake were made to match, keep them both up to date Mar 22 10:28:39 HuntrCkr: it is in topic Mar 22 10:28:57 RP: if I change the graphic offset, won't fb memory be affected? Mar 22 10:29:10 HuntrCkr: for .oz354fam083 branch I used bitbake-svn r329 without problems Mar 22 10:30:00 Mardy: I don't think so. I'm 100% sure there was a register that let you move the displayed window around... Mar 22 10:30:01 I mean, I fear that direct fb memory access might write to the wrong location... or is GRAPHIC_OFFSET taken into account for accelerated operations? Mar 22 10:30:19 I forgot an "only" Mar 22 10:30:21 agreed... i read the topic... yet, if .dev is the development branch, isn't it more up to date than the release branch? Mar 22 10:30:33 Mardy: I don't know... Mar 22 10:30:50 also, i heard from someone earlier that the .dev branch doesn't work anymore with the newest SVN bitbake Mar 22 10:31:09 Mardy: Change graphic offset and see that happens to the accel ops? :) Mar 22 10:31:57 03xora 07org.oe.dev * r50d71efb... 10/packages/mplayer/ (3 files in 2 dirs): Mar 22 10:31:57 mplayer_1.0pre7.bb, mplayer-atty_1.1.5.bb : Add a gcc4 patch to allow Mar 22 10:31:59 compiling with gcc4.X to work. Thanks to Raymond Danks. Mar 22 10:32:00 RP: that's a good start :-) Mar 22 10:35:06 HuntrCkr: hard to tell for me - I do not use .dev Mar 22 10:35:12 HuntrCkr: I use .oz354x now Mar 22 10:49:26 Good Morning Mar 22 10:52:37 hi Bernardo Mar 22 10:52:46 hey Bernardo Mar 22 10:53:03 Bernardo: For the branch you need to set ZKERNEL_VERSION = "2,6" Mar 22 10:54:38 RP: I don't get bitbake :} Mar 22 10:54:51 RP: I think we can win 20 seconds by improving bb.event.registerCode Mar 22 10:55:06 zecke: good morning Mar 22 10:55:19 hi pb__, zecke Mar 22 10:55:27 yo ph5 Mar 22 10:56:29 hey pH5 Mar 22 10:56:33 RP: thanks Mar 22 10:56:50 zecke, you truly must be a god. yesterday run the 'parsing' in under 10 seconds and today win 20 seconds from that? Mar 22 10:56:51 hey koen Mar 22 10:57:04 I was finding it strange that setting the PREFERRED_VERSION wouldn't work... :) Mar 22 10:57:04 pH5: isn't time relative? Mar 22 10:57:25 pH5: I had 'parsing' done to 2 seconds Mar 22 10:57:32 zecke: 20 seconds would be rather nice. I still can't get the profiling to work :-/ Mar 22 10:57:33 pH5: parsing == this spinning progress of bitbake Mar 22 10:57:53 RP: Debian's python2.4.2 seems broken Mar 22 10:58:01 RP: it complains it can't import profile Mar 22 10:58:21 I thought hotshot was a replacement for profile? Mar 22 10:58:48 RP: it uses the same API for Statistics Mar 22 10:58:49 I think my python skills just aren't l33t enough :) Mar 22 10:59:02 pH5: you don't happen to be at the FU campus today? Mar 22 10:59:39 pH5: I comitted a 10+ seconds improvement when your not bitbaking the world **** ENDING LOGGING AT Wed Mar 22 10:59:56 2006