**** BEGIN LOGGING AT Tue Jun 13 02:59:56 2006 Jun 13 07:15:06 morning Jun 13 07:16:11 any idea how to debug udev ? I tried 089.bb and i get timeout errors when booting, and no network interface or login Jun 13 07:41:51 morning Jun 13 07:57:20 morning boys and girls Jun 13 08:05:09 hey ade|desk Jun 13 08:08:48 damn i wish my headache would go away Jun 13 08:09:09 ade|desk: take head out of desk :-) Jun 13 08:09:21 ade|desk : morning Jun 13 08:09:33 XorA : morning Jun 13 08:10:04 ade|desk : Too much footballities maybe ? Jun 13 08:11:35 ~lart udev Jun 13 08:11:35 * ibot nukes udev with a single large nuke Jun 13 08:12:07 too much sun i think Jun 13 08:13:38 XorA: the desk and I are at one, remove one and the other dies making a squelchy sound Jun 13 08:14:19 ade|desk: :-D Jun 13 08:17:41 XorA: are you supporting Trinidad & Tobago? Jun 13 08:18:56 ade|desk: no way Jun 13 08:21:01 XorA: moinmoin Jun 13 08:21:42 kergoth: hi, latest LTP (linux test packages) can crosscompile on arm platform now Jun 13 08:21:50 minipanda: ??? Jun 13 08:22:02 kergoth: and could you please add ltp.bb into current OE.db? Jun 13 08:22:43 XorA: ooops, i used to chat a guy named "xorAxAx" in #moin, so i think you were him, isn't it? Jun 13 08:23:03 minipanda: not me, that some young pretender :-) Jun 13 08:23:45 XorA: oh, sorry for that, and 'moinmoin' is a quite common greetings way in #moin Jun 13 08:24:10 minipanda: heh heh, no probs, Im sure its polite Jun 13 08:38:53 ade|desk: Although I live in Scotland, I am English :-D Jun 13 08:50:47 XorA: thank goodness for that :) Jun 13 08:53:47 ade|desk: much to my inlaws disgust :-D Jun 13 09:03:19 Build of glibc fails for me. http://rafb.net/paste/results/Vrn9wl70.html Can somebody take a look? Jun 13 11:25:51 hi all Jun 13 11:27:48 I have a problem while I am trying to build a bootstrap-image (http://pastebin.com/7062799) . Do I need to change the PREFERRED_PROVIDERS section in the config file? Jun 13 11:30:17 that file is empty Jun 13 11:30:36 so just touch one? Jun 13 11:32:43 have I just to build a defconfig-simpad file with $touch defconfig-simpad in the given directory? Jun 13 11:35:02 How do I get a better output of what bitbake is doing? I've tryed it with the BBDEBUG = "yes" option, but there is the same output as without it. strace is a bit stressed with bitbake. Jun 13 11:40:02 mr_nice: the pastebin is empty Jun 13 11:40:08 no file, nada... Jun 13 11:40:25 oh sorry Jun 13 11:40:30 just a moment Jun 13 11:47:16 http://pastebin.com/706305 - I think i have to define a Kernel. which I shuld choose? I thought that the machine simpad variable had this defined. I have changed the simpad config file in the machine directory and changed it from 2.4 to 2.6. Jun 13 12:07:28 Sorry but I am a bit confused. Do I have to set the {TARGET_PREFIX} parts in the local.conf or is this one set by MACHINE = simpad ? And if I have to set the TARGET_PREFIX where I could find some documentation about the possible targets? Jun 13 12:10:27 here is my local.conf file http://pastebin.com/706299 Jun 13 12:24:13 sorry my connection to my provider was down Jun 13 12:42:54 mr_nice: something else is wrong for you. MACHINE=simpad should make bitbake include conf/machine/simpad.conf which in turn sets PREFERRED_PROVIDER_virtual/kernel to opensimpad. run bitbake -D -D -D to see whether simpad.conf gets included and run bitbake -nv virtual/kernel to see more input Jun 13 12:43:26 hey RP, you alive? Jun 13 12:44:19 mickey:thx Jun 13 12:46:18 l8er... Jun 13 12:48:24 <[g2]> mickey|swimming have a good swim :) Jun 13 12:48:31 thanks :) Jun 13 12:48:53 <[g2]> mickey|swimming just don't take a PDA with you :) Jun 13 12:49:07 no chance. my wife is going with me and she forbids all eletronics in spare time Jun 13 12:49:10 :) Jun 13 12:49:15 :-D Jun 13 12:53:49 hey [g2], any chance you might come to LCA 2007? :) Jun 13 12:53:52 bitbake -D -D -D brought 2 Errors (i686-linux.conf and INVALID-INVALID.conf are missing) http://pastebin.com/706388 . bitbake -nv virtual/kernel seems to be ok (http://pastebin.com/706390). Jun 13 12:54:16 <[g2]> mithro where and when is it ? Jun 13 12:55:06 Australia, Jan 2007 Jun 13 12:55:07 XorA: Fixed the c7x0 speaker bug, and I think I've fixed the noise issue. will send new patch later .... Jun 13 12:55:50 lrg: cool Jun 13 12:56:25 lrg: cant test this evening, but as long as brown stuff and fans dont meet tomorrow should be able to test then Jun 13 12:57:36 * chouimat is back. Jun 13 12:57:52 heh Jun 13 12:58:11 mithro: I somehow want to get an invite to give a Slug presentation at LCA 2007 ... Jun 13 12:59:27 rwhitby: well when you say invite, you mean a fully paid trip? Jun 13 13:00:00 mithro: not necessarily - but I wouldn't say no. It's only from Adelaide. Jun 13 13:00:51 rwhitby: where's LCA being held? Jun 13 13:01:58 NAbyss: I thought it was Melbourne. Or was that Canberra. I forget. Either is about the same distance from Adelaide flying-wise. Jun 13 13:02:17 rwhitby: your in adelaide? Jun 13 13:02:22 yup Jun 13 13:02:30 wtf? I'm in Adelaide too Jun 13 13:02:34 rwhitby: Hmm, I thought last year was melb.. not sure about this year Jun 13 13:02:45 small world :) Jun 13 13:02:58 mithro: how come I haven't seen you at LinuxSA meetings? Jun 13 13:03:19 rwhitby: i never really like the LinuxSA meets back 5-6 years ago so stopped going Jun 13 13:03:37 NAbyss: in Sydney @ 2007 Jun 13 13:03:37 I gave my nslu2-linux presentation there back in Feb this year. Jun 13 13:03:51 i only got back to Australia from Sweden in mark Jun 13 13:04:00 Sydney. Even better :-) Used to work there at Canon. Jun 13 13:04:29 mithro: ah Jun 13 13:04:50 mithro: Anyway, who do I talk to to present an nslu2-linux talk (including many OpenEmbedded references) ? Jun 13 13:05:08 rwhitby: well I'm trying to organise an Embedded Miniconf Jun 13 13:05:49 which would defiently be intrested in your nslu2-linux talk Jun 13 13:07:19 well, keep me appraised of the dates ... Jun 13 13:07:34 but talkers at the Miniconf don't get any privledges sadly Jun 13 13:07:58 Not even registration waived? Jun 13 13:08:35 ciao Jun 13 13:08:52 there will be CFPs for the main conference however Jun 13 13:09:16 the speaking at the main conference rocks, they treat you like royalty Jun 13 13:09:37 I might have to take a shot at that gig then :-) Jun 13 13:09:44 i would recommend Jun 13 13:09:49 it Jun 13 13:10:08 I'm getting the following error when trying to compile glibc-2.3.5 (required by opie): error: iwmmxt requires an AAPCS compatible ABI for proper operation. I found this error in th config.log Jun 13 13:11:22 rwhitby: where in adelaide are you situated? Jun 13 13:11:35 <[g2]> mithro sounds like fun. Unlikely that I'll go (due to distance/cost) but if I hit the lottery I'd love to go (and pickup dyoung/rwhitby on the way over) Jun 13 13:12:00 [g2]: good networking opertunity :) Jun 13 13:12:03 mithro: work in the city now, live out east. Jun 13 13:12:36 * mithro is at Adelaide Uni, live at Welland Jun 13 13:14:08 rwhitby: is LinuxSA less of flamewar these days? Jun 13 13:14:44 mithro: I've only attended since Feb, and no flames at meetings at all. There seems to be some historical thing on the mailing list, but I just ignore that. Jun 13 13:15:18 when are they? Jun 13 13:16:03 http://www.linuxsa.org.au/meetings/ Jun 13 13:16:45 I'll probably give another talk later this year on "Setting up a new small technical consulting business using open source techologies" Jun 13 13:18:18 or "How to tie together Linux, OpenSSL Client Certificates, OpenLDAP, Apache, Subversion, Trac, TWiki, ...", including a period of time where our main mail server was an Iomega NAS100d running OpenEmbedded. Jun 13 13:18:30 rwhitby: i might turn up to one sometime Jun 13 13:19:12 ok, I've sent an email off to Michael Davies (who I worked with at Motorola) about LCA 2007 ... Jun 13 13:19:55 i know Michael Davies :) Jun 13 13:20:15 meet him at LCA 2004 and 2005 :) Jun 13 13:20:20 yeah, he's a good bloke. Jun 13 13:21:27 i was a speaker at LCA 2005, it was great fun, got to hang out with Rusty and Michael at Rusty's house Jun 13 13:23:13 hopefully I can do another talk at LCA 2007, pondering doing something like "Developing OpenSource Hardware, My experiances developing the CFXS" Jun 13 13:23:58 good luck. Anyway, I'm off to bed. Night all. Jun 13 13:45:09 mithro: hello Jun 13 13:52:44 hi all Jun 13 13:53:13 What's wrong if this happens on pull: monotone: warning: discarding revision data packet ffe371fd98a644914142c69f87ad2396196ab184 with unmet dependencies Jun 13 14:03:01 florian_kc: It sounds like you have a branch those revisions depend on masked out Jun 13 14:03:43 RP: hum well... i didn't change anything - just pulling two public branches Jun 13 14:03:58 florian_kc: Its not normally a problem Jun 13 14:04:36 RP: ah okay, so i just ignore this :-) Jun 13 14:04:48 yes Jun 13 14:29:28 monotone sure makes going away from OE for a while and coming back unpleasant Jun 13 14:29:38 * gb2 taps his foot while monotone pulls Jun 13 14:29:59 heh Jun 13 14:31:44 gb2: You need a cron job ;-) Jun 13 14:34:05 heh Jun 13 14:34:46 heh. I find that monotone makes going away from OE for a while quite pleasant. Jun 13 14:34:50 coming back is indeed a bitch though Jun 13 14:35:11 heh Jun 13 14:39:35 hey RP Jun 13 14:40:04 gb2: schurig brings the kids to bed, goes shopping, paints his garden house whole monotone pulls :-) Jun 13 15:24:17 <[g2]> mithro are you developing open source hw ? Jun 13 16:10:31 hey mallum Jun 13 16:12:10 hey zecke Jun 13 16:55:15 goxboxlive: you are at danger to get your OE license cancelled, because you do not push changes to upstream Jun 13 16:55:28 goxboxlive: so push changes, or we are forced to cancel it Jun 13 16:55:30 just kidding Jun 13 16:55:48 dont scare me here Jun 13 16:56:31 goxboxlive: hehe, the point is we can't cancel the license at all Jun 13 16:56:52 sorry Jun 13 16:56:56 zecke: I am trying to get it to wrork perfect before i push it in to the OE. It seems like it works now, but i have to test it a little more. Jun 13 16:57:21 goxboxlive: there is no such thing as perfect (I'm a perfectionist though) Jun 13 16:57:29 We will also try to gather all the HTC devices in only one machine file if that is possible. Jun 13 16:57:36 :-) Jun 13 16:57:49 goxboxlive: if it compiles and runs, push it to dev --> make florian_kc find bugs for you ;) Jun 13 16:58:05 hey pH5 Jun 13 16:58:08 hey zecke Jun 13 16:59:28 pH5: I the images is buildt, but i have to manually install the modules. It seems like OE doesnt install all the modules in the defconfig file in the cvs linux kernel Jun 13 17:01:19 goxboxlive: just do it like in ipaq-pxa270.conf. put modules to be installed into BOOTSTRAP_EXTRA_RRECOMMENDS and add module_autoload_... lines for those that have to be loaded Jun 13 17:01:21 goxboxlive: right Jun 13 17:04:05 I am doing that Jun 13 17:05:23 evening all Jun 13 17:06:16 "As an adolescent I aspired to lasting fame, I craved factual certainty, and I thirsted for a meaningful vision of human life - so I became a scientist. This is like becoming an archbishop so you can meet girls." Jun 13 17:06:20 morning Jun 13 17:06:45 :) hi kergoth, Ifaistos Jun 13 17:07:23 Is anyone having problems with the recent udev ? Jun 13 17:07:25 kergoth: well if you do alias girls="young boys" ;) Jun 13 17:07:36 kergoth: it will perfectly match for the arch... Jun 13 17:08:08 hahah Jun 13 17:08:31 zecke : :) Jun 13 17:09:40 kergoth: I won't say what "South Park" says about Archbishops and 'girls' Jun 13 17:11:11 g;day kergoth Jun 13 17:11:18 hey Jun 13 17:49:10 Ifaistos: What kind of problems? I'm busy wondering about adding udev 092 to oe... Jun 13 17:50:18 jo mickeyl Jun 13 17:50:21 evening Jun 13 17:50:25 hi mickeyl Jun 13 17:50:41 mickeyl: hola, que tal? Jun 13 17:51:03 mickeyl how was the water? Jun 13 17:51:36 pretty good Jun 13 17:51:40 a whole day of doing nothing Jun 13 17:51:44 good for my health Jun 13 17:52:06 bad for the conscience Jun 13 17:52:08 yeah! Jun 13 17:52:35 mickeyl thats good Jun 13 17:52:43 we should meet for a one-week-all-day do nothing Jun 13 17:52:50 heh right Jun 13 17:52:56 just hanging around somewhere Jun 13 17:53:00 making plans Jun 13 17:53:00 He I am in Jun 13 17:53:01 :) Jun 13 17:53:13 woglinde: you have way too much energy Jun 13 17:53:25 woglinde: I doubt you are the right guy for such a meeting Jun 13 17:53:46 zecke nono ask my wife Jun 13 17:53:49 I am really lazy Jun 13 17:54:44 woglinde: I know that. She is pregnant and doing all the work Jun 13 17:54:53 woglinde: even serving the alcohol... Jun 13 17:55:35 zecke: are double override supposed to work? Jun 13 17:55:40 simpad.conf looks very strange to me Jun 13 17:55:58 mickeyl: like SRC_URI_append? Jun 13 17:56:01 PREFERRED_PROVIDER_virtual/kernel_kernel24 ? Jun 13 17:56:15 i would recommend splitting that up like we do on zaurus models Jun 13 17:56:22 bbl Jun 13 17:56:22 mickeyl hm yes simpad.conf is a little bit crappy Jun 13 17:56:24 require simpad-${KERNEL_VERSION] or so Jun 13 17:56:28 re hrw Jun 13 17:56:30 hi Jun 13 17:56:32 yo hrw Jun 13 17:56:48 mickeyl: It works, I had vetoed the change though Jun 13 17:56:57 mickeyl: it was comitted anyway, it died for me Jun 13 17:56:58 it doesn't work for mr_nice Jun 13 17:57:02 I have to digg for other laptop vendor... Acer service support is awful... Jun 13 17:57:20 zecke: who was maintaining simpad last time? Jun 13 17:57:40 mickeyl: dunno, no one officially. Frederic comitted the last bits Jun 13 17:57:46 ah frederic Jun 13 17:57:51 ~seen frederic Jun 13 17:57:56 mickeyl: he needs to have kernel24 and virtual/kernel in the OVERRIDES Jun 13 17:57:58 frederic was last seen on IRC in channel #debian, 242d 11h 45m 39s ago, saying: 'brb'. Jun 13 17:58:04 hehe Jun 13 17:58:06 hmm 242 days??? Jun 13 17:58:09 brb Jun 13 17:58:10 ~seen frederic_ Jun 13 17:58:10 lol Jun 13 17:58:12 mickeyl: i haven't seen 'frederic_' Jun 13 17:58:17 ~seen fred Jun 13 17:58:25 fred was last seen on IRC in channel #debian, 166d 11h 29m 3s ago, saying: 'wireless sorry'. Jun 13 17:58:31 *shrug* Jun 13 17:58:37 I' Jun 13 17:58:45 I'm going to make it look better next time i have some breath Jun 13 17:58:59 mickeyl thanx Jun 13 17:59:02 go ahead, make sure you test it on your SIMpad first! Jun 13 17:59:10 speaking about that... Jun 13 17:59:13 i have none Jun 13 17:59:14 I will test it Jun 13 17:59:17 ok Jun 13 17:59:20 actually, You all do a mistake Jun 13 17:59:27 it is obviously a ZIMpad Jun 13 17:59:34 ZimtPad? Jun 13 17:59:36 I have asked the guys in the Research lab Jun 13 17:59:47 *hmm hunger* Jun 13 17:59:52 ~bon appetit Jun 13 17:59:57 rumour has it, bon appetit is smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Jun 13 18:00:02 45 minutes to next game Jun 13 18:00:03 mickeyl: haha, my mom just left Jun 13 18:00:15 mickeyl: will you send a recipe for some basic foods? Jun 13 18:00:24 mickeyl: e.g. 1.) find the fridge Jun 13 18:00:34 mickeyl: it is somewhere what your mom calls "kitchen" Jun 13 18:00:40 it would be appreciated :} Jun 13 18:00:41 sabine is type setting her cookbook Jun 13 18:00:45 it looks great Jun 13 18:00:47 cool recipes Jun 13 18:00:57 will send, when it's finished :D Jun 13 18:01:11 yeah Jun 13 18:01:19 nothing basic though Jun 13 18:01:24 pretty advanced recipes Jun 13 18:01:29 like glibc.bb Jun 13 18:01:30 heh Jun 13 18:01:43 ;)))) Jun 13 18:01:59 mickeyl: let me guess... she wrote pine_*.bb recipes? Jun 13 18:02:03 mickeyl: Well. I have one issue with calling our BitBake stuff 'recipes' Jun 13 18:02:04 hehe Jun 13 18:02:09 mickeyl: are they guranteed to terminate? Jun 13 18:02:19 zecke: I'm afraid we can Jun 13 18:02:24 't prove it formally Jun 13 18:02:37 mickeyl: using the definition of an Algorithm from this Arabic guy from the 7th Century Jun 13 18:02:55 one criteria is that it takes finite steps/time -> terminating Jun 13 18:03:16 I fear we can't gurantee that, we should use primitve recursive functions Jun 13 18:03:16 to avoid relabelling it :) Jun 13 18:05:11 heh Jun 13 18:05:17 we really don't want that Jun 13 18:06:00 mickeyl: this would avoid this pro lua crap ;) Jun 13 18:06:20 mickeyl: primitve recursive functions are EASY to embed in any language Jun 13 18:06:22 powerful Jun 13 18:06:39 too fine granular Jun 13 18:06:48 i love side effects Jun 13 18:06:49 :D Jun 13 18:09:40 is CIA still broken? Jun 13 18:09:45 i'm afraid so Jun 13 18:09:50 interestingly only channel reporting Jun 13 18:09:51 yay, mysql is fixed Jun 13 18:09:54 on web site it works Jun 13 18:10:11 * kergoth kicks CIA-9 Jun 13 18:10:38 hey kergoth Jun 13 18:10:40 stupid log rotate for binary logs was breaking and emailing me every fucking night. Jun 13 18:10:50 udev: Add 092. Included is a custom rule which adds a symlink to the touchscreen as /dev/input/touchscreen0 if present (this does not work under udev 089 due to bugs). Several udev conf files required cleanups to work with the new udev version. Jun 13 18:11:08 hey Jun 13 18:11:19 I hate physics :} Jun 13 18:11:23 hi kergoth Jun 13 18:11:36 Now I can start killing off detect-stylus :) Jun 13 18:11:37 RÜ: cool Jun 13 18:11:39 RP even Jun 13 18:12:17 me waits until his imaginary rss reader for the poky tinderbox turns the red light on (again) Jun 13 18:12:54 zecke: I already made these changes to poky earlier today ;-) Jun 13 18:13:19 ah it is red! Jun 13 18:13:26 red! poky is broken, use OE ;) Jun 13 18:13:59 RP: Have you noticed you are an admin for the tinderbox as well? Jun 13 18:14:08 zecke: no :) Jun 13 18:14:15 RP: you can change stylesheet, change the header, footings to point... Jun 13 18:14:25 zecke: That error means you need to rebuild task-bootstrap Jun 13 18:14:26 RP: you can use your OE bugtracker account Jun 13 18:15:06 RP: can zaurusd get revisions other then 0.0-will-be-released-somewhen? Jun 13 18:15:07 RP: to login into the tinderbox (stuff will be send in clear text) Jun 13 18:15:38 zecke: I'm in, thanks :) Jun 13 18:15:56 * RP wonders what he can break in here... ;-) Jun 13 18:15:56 RP: do not place too much viagra spam on top of the site Jun 13 18:16:19 RP: I have not audited the code base yet :} Jun 13 18:16:23 *perl* Jun 13 18:16:47 RP: should I manually clean task-bootstrap? shouldn't the PR have been bumped? Jun 13 18:17:28 zecke: I changed nokia770.conf so it wasn't something I immediately thought of... Jun 13 18:18:03 hrw: I'm not sure if releases are going to happen. At least its proving relatively stable atm :) Jun 13 18:18:23 as the machine influences the task, it should be able to influence the PR string as well Jun 13 18:18:57 zecke: right. I once thought about that but it gave me a headache... Jun 13 18:19:08 yes Jun 13 18:19:13 zecke: poky.conf influences every .bb file... Jun 13 18:20:03 We should checksum the key variables for a given .bb file and use that as the PR Jun 13 18:20:05 you could set PACKAGE_ARCH... well Jun 13 18:20:24 RP: it should be increasing as well :} Jun 13 18:20:28 i dont think i have oe write access anymore. can someone modify generic.conf and generic-uclibc.conf to make TARGET_FPU conditional on arch? should probably have TARGET_FPU_arm and TARGET_FPU_armeb. otherwise usin those distros on an x86 target explodes Jun 13 18:21:16 kergoth: I'm fairly sure you could get write access :) Jun 13 18:21:19 kergoth: generate a key Jun 13 18:21:26 kergoth: good idea. will do asap Jun 13 18:21:28 kergoth: send it to me or mickeyl Jun 13 18:22:36 k Jun 13 18:22:56 I prefer mickeyl Jun 13 18:23:22 woglinde: http://www.spiegel.de/netzwelt/technologie/0,1518,421188,00.html Jun 13 18:23:31 woglinde: lol Raul in a ... Jun 13 18:23:54 hmm.. interesting things can be found in used laptops... IBM T30 with 1G RAM, p4-m/1.9 for 1940 pln Jun 13 18:24:06 zecke hehe Jun 13 18:24:16 woglinde: 'suess' Jun 13 18:24:53 woglinde: haha, He disses Phillips... Jun 13 18:25:16 woglinde: you could have a 'hard shooter that can shoot across the complete field' Jun 13 18:25:31 ja Jun 13 18:25:33 woglinde: or you can play nice, like the brazillians Jun 13 18:25:34 lol Jun 13 18:25:36 aber roul hat hochschuss Jun 13 18:25:43 diese jahr Jun 13 18:25:59 ups Jun 13 18:26:02 sorry Jun 13 18:26:02 hehe Jun 13 18:26:15 roul has highkick this year Jun 13 18:26:34 once again. I hope how many cameras and motors will be broken... Jun 13 18:27:01 woglinde: I loved that he was too shy to specify the youth hotel as destination and used the Hilton instead :} Jun 13 18:30:22 tslib: Remove hardcoded input event devices or detect-stylus calls for machines with modern 2.6 kernels which can use udev Jun 13 18:31:27 RP: Feel free to add other people as maintainers as well (for tinderbox) Jun 13 18:32:11 zecke: I'm not sure who I'd add :} Jun 13 18:32:48 RP: that is not my problem! Jun 13 18:32:57 heh Jun 13 18:33:09 :) Jun 13 18:39:27 Why does task-gpe-base RDEPEND on detect-stylus? Jun 13 18:39:59 argh, forgot my usb stick at home again Jun 13 18:40:06 now i cant copy anything to the dev box on the isolated network Jun 13 18:41:11 hm, /etc/init.d/populate-volatile.sh takes rougly 11s here (with NFS), and is a major contribution to the boottime which i want to improve. It seems to mainly consist of subshell expressions (like TTYPE=`echo ${LINE} | cut -d " " -f 1`;), each of them generates a lot of "GETATTR"-nfs-requests. a.) is that normal (that it takes that long)? b.) how to improve? is this an nfs problem or a busybox problem? Jun 13 18:41:52 the volatile script can probably be made more efficient Jun 13 18:41:56 tmbinc_: busybox. and the init script itself Jun 13 18:41:56 tmbinc_: coredump did some optimisation there recently Jun 13 18:41:58 tmbinc did you realy need nfs on the dreambox? Jun 13 18:42:07 woglinde: yes, i don't want to reflash all the time ;) Jun 13 18:42:15 tmbinc oh okay Jun 13 18:42:17 plus, boottime from flash is also a problem. not sure if it's the same script, though. Jun 13 18:42:47 ok, i agree that the script is not very efficient. but each subshell takes roughly 100ms, which i believe is way too much. Jun 13 18:43:05 tmbinc_: rewrite it in C? Jun 13 18:43:19 tmbinc_: IIRC we needed to undo a CELF patch to busybox Jun 13 18:43:27 well, i would have to rewrite all bootscripts then to get a decent boottime performance Jun 13 18:43:30 tmbinc_: which among other things should have improved performance Jun 13 18:43:35 since it relies on subshells all over the place Jun 13 18:44:00 zecke: hm Jun 13 18:44:21 evening Jun 13 18:45:02 tmbinc_: yeah Jun 13 18:45:33 running the scripts generates about 8000 nfs packets, most of them GETATTR call / GETATTR reply Jun 13 18:45:53 i agree that shellscripts are terrible inefficient, but i cannot believe it's really that bad Jun 13 18:46:10 after all, back in the 386 days, init.d worked the same and wasn't that slow iirc. Jun 13 18:46:20 tmbinc_: well, nfs isnt exactly efficient either. i've found that adjusting the buffer sizes and switching it from udp to tcp helps a bit Jun 13 18:46:26 * kergoth shrugs Jun 13 18:46:40 entrance has a DEPENDS on detect-stylus? :-/ Jun 13 18:46:59 i wish qemu had something like hostfs Jun 13 18:47:04 thats hte only reason i end up using uml anymore Jun 13 18:47:07 :) Jun 13 18:47:21 kergoth: my experience is that when switching from UDP to TCP (or decreasing the buffer sizes) helps, you get overruns on your NIC (i've spent ages optimizing network drivers for NICs with way to small buffers to prevent that :/). however, this time this is definitely not a problem, as the NIC can busmaster. Jun 13 18:47:23 kergoth: Have you tried it with an nfsroot? Jun 13 18:47:48 tmbinc_: ahh Jun 13 18:47:59 RP: i dont have root on this box, so i havent tried to get networking working :P Jun 13 18:48:11 * kergoth is really getting tired of the overhead involved in trying to develop at his current position Jun 13 18:48:44 kergoth: be glad you can bring your own USB stick?! Jun 13 18:48:52 hehe Jun 13 18:48:55 kergoth: ah :-/ Jun 13 18:49:40 zecke: can you point me to the patch which was removed? Jun 13 18:49:51 tmbinc_: I will try Jun 13 18:49:59 kergoth: I posted some qemuarm images to oe@ the other day which included an X image, a few apps and an arm toolchain etc. Jun 13 18:50:11 nice Jun 13 18:50:48 http://www.o-hand.com/~richard/qemu.html Jun 13 18:51:11 lots of folks have been doing qemu related projects lately - I had no idea it was so mature Jun 13 18:51:14 on one of these machines i have a qemux86.conf. should probably consolodate bits into a qemu.inc/conf Jun 13 18:51:48 Was it based on qemuarm? Jun 13 18:52:19 yeah, just slight changes that are arch specific. kernel, ipkg archs, the serial device name.. Jun 13 18:52:33 nice :) Jun 13 18:53:02 We should get that into OE as I keep meaning to make an x86 development target... Jun 13 18:53:05 actually using that same image on both my uml and my x86 qemu Jun 13 18:53:17 since i didnt feel like waiting for a second build Jun 13 18:53:46 I've used spitz and other zaurus builds under qemu for the same reason Jun 13 18:54:48 tmbinc_: http://www.openembedded.org/viewmtn/file.psp?id=b149a1bf73cca2d159bbfc0dc118908df558a8a3&path=packages/busybox/busybox-1.00/celf-ash-builtins.patch Jun 13 18:54:53 tmbinc_: I think it was this one Jun 13 18:55:36 lpotter: hey! Jun 13 18:55:46 lpotter: feel better? Jun 13 18:55:57 hi Jun 13 18:56:06 hi lpotter Jun 13 18:56:13 ya, almost back to normal Jun 13 19:01:21 lpotter: Any ETA to a response to the public questions? Jun 13 19:01:33 zecke: thanks. i see, it replaces a number of functions by internal ones, right? hm. Jun 13 19:01:35 lpotter: e.g. Nokia is fostering Free Software projects Jun 13 19:01:45 tmbinc_: yes, but they broke stuff :( Jun 13 19:02:34 tmbinc_: but we do not apply it to busybox1.01 Jun 13 19:03:43 tmbinc_: e.g. install a full blown bash, and see if it makes a difference Jun 13 19:04:01 you would have had one, if I hadnt gotten sick Jun 13 19:04:21 zecke: yeah, that's a good point Jun 13 19:04:47 :( Jun 13 19:05:06 back to the coulumb force Jun 13 19:13:16 If I have a small (2.2") color LCD and I want to display Bitmaps and text on it is microwindows (nano-x) the best approach? Jun 13 19:13:38 fbdisplay? Jun 13 19:13:48 Gerrath: evas? Jun 13 19:13:49 heh Jun 13 19:13:55 yay Microwindows! Jun 13 19:14:10 zecke, yes I was planning on writing a fb driver for the display. Jun 13 19:14:16 evas? Jun 13 19:15:09 CosmicPenguin, do you really like Microwindows or your just ribbing me :-) Jun 13 19:15:23 I used to be one of the primary developers for Microwindows Jun 13 19:15:41 * Philippe is away: visual contact - melancholic dreams Jun 13 19:16:00 ahh, then you would recommend it. Jun 13 19:16:11 not really - I'm just happy to see the name Jun 13 19:16:24 there are better and more supported framebuffer apps out there Jun 13 19:16:29 notice I said "I used to be" Jun 13 19:16:33 there's a reason for that Jun 13 19:16:47 I see, what would you recommend? Jun 13 19:17:11 You just want to show bitmaps on it? Like a digital picture frame? Jun 13 19:17:43 gpe-bootsplash :) Jun 13 19:17:53 it take PNG or SVG Jun 13 19:17:56 maybe slow low res video at some point but the original release will just be alternating bitmaps with text overlaid. Jun 13 19:18:03 hrw: yeah, but it also takes GPE.. :) Jun 13 19:18:10 Gerrath: have you considered DirectFB? Jun 13 19:18:11 directfb would be a choice if you don't want to handle ugly color format conversions, but don't need crazy window stuff Jun 13 19:18:14 Gerrath: fbdisplay Jun 13 19:18:16 :) Jun 13 19:18:32 yeah directFB has this kind of slideshow application already Jun 13 19:18:34 you could just draw on it yourself, but text is difficult Jun 13 19:18:48 And yes, Microwindows would do the job, but its probably suffering from more bitrot then it should Jun 13 19:18:49 but if you want to do image+text you can use fbdisplay... Jun 13 19:18:57 it can only do these two things... Jun 13 19:19:37 future feature will be video but I don't think we will have it for the first release. Jun 13 19:20:17 then use DirectFB Jun 13 19:20:28 has Video Providers already Jun 13 19:21:08 I'm inclined to agree Jun 13 19:21:22 only ugly part is that DirectFB requires multithreading, even if you don't want to use it. Jun 13 19:21:39 but that should be no real issue Jun 13 19:22:06 is it pretty light weight in size? Jun 13 19:22:13 Gerrath: try it Jun 13 19:22:20 ok. Jun 13 19:22:45 wow - peer pressure Jun 13 19:22:47 Gerrath: this depends on so many things where "Breadbox" would be the best fixpoint Jun 13 19:23:09 Gerrath: but DirectFB was meant for settop boxes with ~4mb of flash Jun 13 19:23:25 ok, then I should be ok. Jun 13 19:23:31 it's way better than X. so if you don't want to handle everything yourself (which would be ultimately light weight), then just use it. Jun 13 19:23:55 zecke: *meant*, yes. The company doesn't exist anymore, after all ;) Jun 13 19:24:02 (but that was of course not dfb's fault) Jun 13 19:24:23 tmbinc_: hehe, did you ever work there? Jun 13 19:24:45 tbminc dont worry zecke ask this all developer he meets Jun 13 19:24:53 lol Jun 13 19:25:00 woglinde: well, in Berlin this is the case :} Jun 13 19:25:07 woglinde: it is either Digital City, or Convergence Jun 13 19:25:20 woglinde: and my "Schubladen" overrun Jun 13 19:25:27 no, i kinda worked (well, i still do) at their competition ;) Jun 13 19:25:47 woglinde: Steve showed me the seminarist video... Jun 13 19:25:50 woglinde: I liked it Jun 13 19:27:03 I have question about toolchains. Jun 13 19:27:34 zecke whats the seminarist video? Jun 13 19:27:43 my nfs seems broken. even with "actimeo=300" (should say, if i understood correctly: cache attributes for 300s) it bombs the server with GETATTR calls, over and over for the same filehandle. Jun 13 19:27:58 woglinde: check the linuxtv wikipedia article Jun 13 19:28:26 OZ 3.5.4 libc6-dev/usr/lib/libc.so has: "GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a AS_NEEDED ( /lib/ld-linux.so.2 ) )" - I had to remove "AS_NEEDED ( /lib/ld-linux.so.2 )" block to get helloworld.c built on Z Jun 13 19:28:30 any ideas? Jun 13 19:28:46 hrw: old binutils? Jun 13 19:29:14 AS_NEEDED is relatively new Jun 13 19:29:19 zecke: same as used to build 3.5.4 Jun 13 19:29:42 zecke: 2.15.94.0.1 Jun 13 19:30:09 .o(2.15 are old) Jun 13 19:30:53 so I should upgrade OZ 3.5.4 to binutils 2.16 to get it working? Jun 13 19:30:54 woglinde: 94 are snapshots :) Jun 13 19:31:06 hrw: check when --as-needed was added to binutils Jun 13 19:31:16 ok Jun 13 19:32:28 directfb compiled with out any issues and the footprint looks small enough :-) Thanks all for the info, now I just need to wrestle writing the LCD FBdriver :-P Jun 13 19:32:59 your LCD is RGB(A) right? Jun 13 19:33:01 "The suggested version that should solve all the ld-related problems with --as-needed is 2.16.92" Jun 13 19:33:24 * zecke gives food to the flowers Jun 13 19:33:28 cya later Jun 13 19:34:00 zecke, yes but it has a controller on board that I need to interface with. Jun 13 19:35:05 Gerrath: but it is framebuffer mapped? Jun 13 19:36:25 tmbinc_, yes, it is frame buffer mapped. Jun 13 19:36:51 it basically just does the refresh feature and has some overlay capability. Jun 13 19:51:11 hi CoreDump Jun 13 19:51:20 hey hrw Jun 13 19:51:43 Gerrath: overlay, perfect match for DirectFB Jun 13 19:57:05 hm, by using the semi-documented "nocto" option in nfsmounts the amounts of getattr calls is reduced dramatically, so it now takes 9s instead of 11s. seems to be yet another problem Jun 13 19:57:15 The LCD is a pretty nice low cost LCD and small (2.2") for any one who is interested, here it is: http://www.sharpmeg.com/Page.aspx/americas/en/part/LQ022B8UD04/ The main saling point for me is it is Transflective. The website says its obsolete but that is only because they are releasing a brighter version in the same package :) Jun 13 19:57:45 lol selling not saling :-P Jun 13 19:58:55 hi all Jun 13 20:00:53 Gerrath: but erm, the lcd has an internal framebuffer, right? Jun 13 20:01:20 is the bus fast? Jun 13 20:01:41 (well, probably fast enough for the relatively small resolution.. which is of course ok for that size.) Jun 13 20:02:05 My old problem again: bitbake bootstrap-image fails with this (http://pastebin.com/707250). My config is here (http://pastebin.com/707256). I deleted (tmp and sources) and builded it new but the same problem as this afternoon :(. BTW is there a thing like make clean for bitbake or is the tmp and source deletion the common way? Jun 13 20:02:29 mr_nice: do not remove DL_DIR content Jun 13 20:03:26 heh, doing a bitbake build in the background sure makes world of warcraft laggy on my home box Jun 13 20:04:06 mixing work and pleasure again? Jun 13 20:04:15 not today, but last night, yep Jun 13 20:04:15 :) Jun 13 20:04:27 hrw:I had the same problem before doing this. I thoght eventually there is corrupted file and I just wanted to kill this error possibility. Jun 13 20:04:35 mr_nice: set a preferred_provider as bitbake asks to? Jun 13 20:04:48 line 19 and 20 Jun 13 20:05:04 mickl sad that the machine thing just do this Jun 13 20:05:15 zecke:ok i will try Jun 13 20:05:18 PREFERRED_PROVIDER_virtual/kernel = "the kernel you prefer" Jun 13 20:05:40 yes but isn't it a overwrite variable in simpad.conf Jun 13 20:06:08 mr_nice: well, I love to repeat myself: This guy made this bastardized change (not mickeyl), which I didn't agree with (see mail archives) Jun 13 20:06:12 mr_nice: and now it is broken Jun 13 20:06:27 mr_nice: ask the guy who did the bastardized change, which I asked him not to commit, to fix it ;) Jun 13 20:06:46 tmbinc_, yes it has an onboard frame buffer. Jun 13 20:07:06 zecke:so shuld I just add the same kernle as in simpad.conf? I will try. thx Jun 13 20:07:49 zecke: I don't think anyone would object to you fixing it ;-) Jun 13 20:08:15 RP: lol, no way I'm going to touch this file Jun 13 20:08:43 I feel abused to give support... ;) Jun 13 20:08:55 mr_nice: don't take this personally, it is late and hot over here Jun 13 20:09:20 If I find problems, I kind of feel obliged to fix them :-/ Jun 13 20:09:33 zecke: too bad. over here it is... raining. Jun 13 20:10:34 zecke: no problem I will just add it in the local.conf file. Jun 13 20:11:16 RP: Well. I clearly said. Do not commit, create a branch Jun 13 20:11:41 RP: and as we have seen this guy who was willing to maintain this ugly hack wasn't seen for +200 days ;) Jun 13 20:11:45 and I'm still around :} Jun 13 20:11:53 lesson learned Jun 13 20:12:01 zecke: So it should have been reverted... Jun 13 20:12:28 RP: well his argumentation was "My gal takes H, so it must be good, I will take even more H" Jun 13 20:12:39 :) Jun 13 20:13:03 pb_: does it really rain? is it one of these days Jun 13 20:13:26 zecke: yah, it is raining quite heavily right now Jun 13 20:13:35 mr_nice: feel free to patch simpad.conf, we will apply the patch for sure Jun 13 20:13:36 we had thunder and lightning a little while ago Jun 13 20:13:53 pb_: don't you need to turn your PCs off? Jun 13 20:13:59 zecke: installing a bash vs. busybox seriously improved performance Jun 13 20:14:14 zecke: no, I just move them to high ground when the water rises Jun 13 20:14:53 tmbinc_: can you give my marketing department numbers? Jun 13 20:14:57 zecke: I just have to learn bitbake and phyton first ;) but I will try it. It could take a long time and it could be that I am never being able to fix it. Jun 13 20:15:01 tmbinc_: 10% faste? 200% faster Jun 13 20:15:19 mr_nice: ah, it will be easy, it will require no knowledge of python Jun 13 20:15:28 mr_nice: and we are friendly, even me Jun 13 20:16:03 zecke:BTW why istn't the usage dictory readable at berli.os= Jun 13 20:16:24 mr_nice: because it shows the ultra secret usage statistics Jun 13 20:16:30 mr_nice: e.g which browsers hit the site Jun 13 20:16:31 ~lart pcmcia subsystem in 2.6 kernel and orinoco/hostap drivers too for making my prism2 card using orinoco just because it is 0156, 0002 Jun 13 20:16:31 * ibot takes large quantities of Krispy Kream donuts and stuffs them one after another down pcmcia subsystem in 2.6 kernel and orinoco/hostap drivers too's throat until pcmcia subsystem in 2.6 kernel and orinoco/hostap drivers too puts on 150lbs for making my prism2 card using orinoco just because it is 0156, 0002 Jun 13 20:16:32 zecke: 4s with bash, 9s with busybox and that nfs option, 11s with busybox without that nfs option Jun 13 20:16:49 worst lart ever Jun 13 20:16:49 cu all Jun 13 20:16:51 tmbinc_: just this script? Jun 13 20:16:54 zecke: yes Jun 13 20:17:04 zecke:ah! yes now it is clear. first I thoght it could be another manual Jun 13 20:17:18 zecke: probably even better when i remove the kernel debug stuff i've inserted to find that "secret" nfs option ;) Jun 13 20:17:34 tmbinc_: 4s vs. 9s sounds like normal variation :) Jun 13 20:17:38 tmbinc_: do some more runs Jun 13 20:17:46 mr_nice: hehe, I need to move the manaul around (one time) Jun 13 20:19:07 zecke: 11s was pretty much damn stable :). but yes, i can do some more runs. Jun 13 20:19:39 hehe, if I woldn't suck so much at stochatisc I could give you the number of runs you need to do Jun 13 20:19:52 inf times probably ;) Jun 13 20:19:52 and yes, I hope that this number is stable (as only init runs) Jun 13 20:20:01 inf times-1 Jun 13 20:20:16 that doesn't REALLY help much Jun 13 20:20:20 zecke:just make a .usage and it wouldn't be so confusing ;) Jun 13 20:20:46 mr_nice: this is provided by berlios, I have zero influence about it Jun 13 20:20:50 zecke: do you know how i can run something n times with "time"? ("time for x in ..." doesn't work) Jun 13 20:21:12 tmbinc_: could you try a more recent busybox version? Jun 13 20:21:19 tmbinc_: no, use '(' ')'? Jun 13 20:21:43 actually to meassure shell perfomance you should avoid using shell? Jun 13 20:22:15 "real 0m3.899s, real 0m3.892s, real 0m3.887s" - is that stable enough? (bash-3.0 with removed debug again.) Jun 13 20:23:02 tmbinc_: We would need to ask CoreDump for what was broken with this celf patch Jun 13 20:23:25 how exactly are you replacing busybox with bash? Jun 13 20:23:34 hmm. scons's flexbility can be somewhat overwhelming at first, but i love it Jun 13 20:23:36 just using bash for /bin/sh or something else? Jun 13 20:23:43 * kergoth pokes at its crosscompilation bits Jun 13 20:24:20 zecke: it broke some or other shell behaviour that his altboot stuff depends on. check the mailing list archive, I think. Jun 13 20:24:32 I could never reproduce the problem myself with a standalone testcase Jun 13 20:24:47 ist the local.conf file chached? Jun 13 20:25:39 zecke: "real 0m6.026s, real 0m5.996s, real 0m6.000s" with busybox version 1.01, debug stuff removed. (ok, the advantage shrinked together, probably because echo calls are bash-internal and not busybox-internal) Jun 13 20:26:00 raduga: yes, i've just changed the symlink Jun 13 20:26:16 mr_nice: it will be recognized that this file changed, and everything will be reparsed Jun 13 20:26:38 pb_: that is bad, a test case would have been my next question Jun 13 20:26:58 tmbinc_: change the busybox config a bit, e.g. using more or less builtins Jun 13 20:27:08 tmbinc_: alternatively try 1.1.x, or settle with busybox Jun 13 20:27:15 bash Jun 13 20:27:16 ... Jun 13 20:27:29 yeah we have a pretty big busybox Jun 13 20:27:32 (~900k) Jun 13 20:27:40 so i imagine the VM overhead for the fork() is the real problem Jun 13 20:27:44 one might say its a busy busybox Jun 13 20:27:50 hehe Jun 13 20:28:35 zecke: "real 0m 7.23s, real 0m 7.21s, real 0m 7.22s" without that "nocto" nfs option Jun 13 20:28:40 and busybox Jun 13 20:28:51 (just for completeness) Jun 13 20:30:10 can?t you try to use a profiler to determine what is eating most cpu time? Jun 13 20:30:53 TheCan: the profiler will eat most of the time :) Jun 13 20:30:58 with a good kernel profiler probably, but i can't do that. are there great tools for that? oprofile doesn't work for mips... Jun 13 20:31:03 oprofile doesn't Jun 13 20:31:16 ah :-/ Jun 13 20:31:35 tmbinc_: you use mips? I thought that it was ppc? Jun 13 20:31:50 lttng runs on mips :) Jun 13 20:31:51 zecke: we have both. newer generation is mips now Jun 13 20:31:52 * RP hides Jun 13 20:32:02 ah Jun 13 20:32:19 LTT, good point. does that work? Jun 13 20:32:22 OT: tmbinc_: what card interface do I need to watch premiere (legally) Jun 13 20:32:23 tmbinc hm I think ralf works a little bit on oprofile Jun 13 20:33:44 zecke: you can't do that 100% legal because using the card protocol alone is not allowed. Anyway, if you still want that, you would need a smartmouse reader. and some "softcam". and of course a descrambling device, if you hav a full featured tt card you can do that. (you're on PC, right?) Jun 13 20:34:13 woglinde: hm, last time i checked it worked for the real mips cores, but not for the toy stuff (4KCe and all that consumer mips cores ;) Jun 13 20:34:16 tmbinc_: no, I'm considering getting a dreambox Jun 13 20:34:24 tmbinc ah I see Jun 13 20:34:25 as woglinde will not give me one of his dbox Jun 13 20:34:38 zecke he you have dbox1 Jun 13 20:34:49 zecke: oh. then you only need a "softcam" and you can use the internal cardslot. still, legality is questionable. Jun 13 20:34:50 woglinde: does it run linux? Jun 13 20:34:55 uclinux maybe ;) Jun 13 20:35:09 freistoss Jun 13 20:35:10 or mklinux or how that was called. Jun 13 20:35:11 tmbinc_: oh, why is that questionable? Jun 13 20:36:14 zecke: because you need a license to support certain CA systems. and getting such a license involves prooving that your system is "secure" (question is, secure against what - after all, the security should be in the smartcard and card sharing is a generic problem...). and prooving that an open system is "secure" is, well, impossible. Jun 13 20:36:24 plus you need stuff like macrovision Jun 13 20:37:09 ah okay, I will put a key sticker on the box then... Jun 13 20:38:14 zecke: these days, you are needing a license for the descrambler (core) in the chipset. and you won't get a license unless you have a closed system. thus, it means that you can't get *any* dvb chipsets without having a closed license. goodbye, linuxtv. Jun 13 20:38:50 wonderful Jun 13 20:38:59 okay, going to bed now Jun 13 20:39:02 thanks for the info Jun 13 20:39:06 'night zecke Jun 13 20:39:13 night zecke Jun 13 20:39:42 woglinde: give me your dbox! Jun 13 20:39:55 zecke: for your amusement, http://bugs.openembedded.org/show_bug.cgi?id=47 Jun 13 20:40:43 i remembr that one....nasty Jun 13 20:40:55 I'm NOT amused :) Jun 13 20:40:58 tomorrow pals Jun 13 20:41:02 n8 zecke Jun 13 20:41:28 tmbinc: I think you do regular builds for the dreambox already? Jun 13 20:41:47 tmbinc: otherwise, I could do regular builds as well... Jun 13 20:42:00 regular builds? Jun 13 20:42:10 http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=poky Jun 13 20:42:19 s/pokxy/YOUR_NAME_HERE/ Jun 13 20:42:21 oh, great, yeah Jun 13 20:42:35 i should finally start putting back patches into .dev :/ Jun 13 20:42:47 if i just had more time... *insert other excuses here* Jun 13 20:42:53 tmbinc :( Jun 13 20:44:12 next bug weekend, we can look into merging Jun 13 20:44:54 first thing i want to do is to merge .dev into .dreambox Jun 13 20:45:15 and fixing the dreambox-only packages to the many changed which occured since.. (CVSDATE -> SRCDATE etc.) Jun 13 20:45:22 then merging back should be easier Jun 13 20:48:47 yeah, we - the OE project - should keep a ChangeLog... Jun 13 20:48:51 nite *finally* Jun 13 21:00:46 it compiles :). I will try to fix the simpad.conf file. ciao Jun 13 21:25:17 how can i prevent a file in SRC_URI to be unpacked? (i want to add a .tar.gz into a package) Jun 13 21:25:49 dunno if theres an existing way, short of providing a custom do_unpack method Jun 13 21:25:54 s/method/task/ Jun 13 21:30:58 there should be an option though Jun 13 21:31:00 that would be a good addtion Jun 13 21:31:04 ;nounpack Jun 13 21:33:16 crap, cant get to the original version of my own damn resume, its on my fileserver in MN which is out of commission Jun 13 21:39:42 doh Jun 13 21:39:50 looking to get out of the bionic warrior business? Jun 13 21:41:03 i need it cause they're hiring me on directly instead of contract, $7-8k pay raise. but yes, also look around some, the amount of politics and red tape and overhead involved with government stuff gets tiresome Jun 13 21:41:04 If he told us, he'd have to shoot us Jun 13 21:41:09 and he could call it testing Jun 13 21:41:12 hehe Jun 13 21:42:24 ggilbert: I was hoping that even kergoth's uber-troopers would have trouble shooting me from this far away. Jun 13 21:42:41 I guess you never know though Jun 13 21:42:50 kergoth: is there any combination of video out/hud-goggles or similar, you'd recommend for a wannabe cyborg-Zaurer? Jun 13 21:43:20 hmm, been a while since i looked at what hardware was available to hobbyists for that stuff, not sure. google :) Jun 13 21:43:40 kergoth: heh. google has.... stuff. ranging from $200-$1000 or more Jun 13 21:43:58 trying to get a handle on what, if any, is worth bothering with, and what might just suck Jun 13 21:44:12 have you used video goggles at all? Jun 13 21:44:16 i can ask around, see if anyone here has messed with commodity type hardware Jun 13 21:44:23 only the ones on the hardware here, i'm fraid Jun 13 21:44:27 s/fr/afr/ Jun 13 21:44:40 at least on the hardware you have, does it seem to work? Jun 13 21:44:59 or is it still just toys? Jun 13 21:45:23 i'm pretty impressed. its not exactly as light as a pair of glasses, but the video quality is much better tahn i expected Jun 13 21:45:31 definately viable Jun 13 21:45:50 the commodity off-the-shelf products are very likely to be less good than what you have, but if the high-end stuff is ok, there might be some room for cheaper stuff that's moderately ok Jun 13 21:45:58 * kergoth nods Jun 13 21:46:23 with iodata/zaurus I'm imagining I may be able to get 5-10fps in 320x240 Jun 13 21:46:44 depending on the card, obviously Jun 13 21:46:50 * kergoth looks forward to snowcrash style gargoyles :) Jun 13 21:47:22 so, maybe the cheaper goggles (in low res, but with decent-width display) may work ok Jun 13 21:47:36 though the cheaper ones tend to use AV input, not vga :/ Jun 13 21:48:28 kergoth: that's what I have in mind. To the best of my knowledge, there aren't any zaurgoyles floating around, but it seems like it... almost could be a viable platform Jun 13 21:49:08 most of the CF vga cards seem to be rated at 1-5fps; but 640x480. presumably 4x that for 320x240, but its not altogether clear Jun 13 21:49:41 1fps would not be very useful, for a text console. 5fps would be close to tolerable, but laggy Jun 13 21:49:44 one based on an Oqo could be cool, lots of computing power. battery life would be an issue though. course, the Z doesnt exactly have spectacular life either as is, would need a battery pack Jun 13 21:49:50 * Philippe is away: visual contact - melancholic dreams Jun 13 21:50:11 some of the goggles have battery packs Jun 13 21:50:24 I'm guessing, that walking around for more than a few ohours with one of these would get tiring pretty quick Jun 13 21:50:43 mind you... the ones you are using, are probably designed for comfortable extended use Jun 13 21:50:56 under... less than ideal environmental conditions Jun 13 21:51:08 imho, its all in the weight distribution. even a fair bit can be comfortable if placed right Jun 13 21:51:18 throw the power supplies at the waist level and it wouldnt be too bad Jun 13 21:51:50 kergoth: if you can ask around, and find out what commodity goggles are... any good, I'd really appreciate it ;) Jun 13 21:51:59 will do Jun 13 21:52:06 i'm not in any hurry to buy something... but i'm very curious Jun 13 21:52:22 i'm sure people were researching them back when they made the original purchasing decision, so /somebody/ must know :) Jun 13 21:52:32 there seem to be quite a few vaporware products that are due on the market "real soon" but many of them seem not to have happened Jun 13 21:53:59 hm... Xybernaut used to sell goggle/pc/etc kits for industrial users, Jun 13 21:54:19 did so for a few years (very expensive) but I don't see any HUD-type displays on their site now Jun 13 21:54:45 they all look like tablets Jun 13 21:55:27 and looks like they don't sell to end-users anymore Jun 13 21:56:46 s/hud/hmd/ Jun 13 21:57:33 looks like their business is on the skids, too Jun 13 22:47:10 ibot: botmail for mr_nice: i fixed the simpad.conf. try setting ZKERNEL_VERSION = "2.4" and check if that improves things. Jun 13 22:47:13 g'night Jun 13 22:59:09 ok probably not the right channel but does anybody know where the opensug now slugos lot are Jun 13 22:59:36 gave up on my ipaq in the end and decided to concentrate Jun 13 22:59:54 on somathing more generic Jun 13 23:00:18 iirc there are a number of channels for it... i havent touched my nslu2 in ages. i remember #nslu2-linux being one of them, back in the day Jun 13 23:00:21 * kergoth shrugs Jun 13 23:02:24 cheers, just thaught somebody might know as they use monotone and bitbake; thanks Jun 13 23:05:00 Does anyone have an idea if 1280x1024 looks better on 17" or 19" tft panel? (isn't it OE-related question?:)) Jun 13 23:05:35 polyonymous: I'd say 17".. smaller pixels Jun 13 23:05:59 NAbyss, That's what I was thinking. Hope not too small, though ;-) Jun 13 23:06:22 Currently I have this resolution on 18" visible area (crt). Jun 13 23:06:22 polyonymous: I've got a 17" 1280 panel at home, another one at work, and a 19" 1280 panel at work.. the 17s win out easily Jun 13 23:06:31 if the 17 can do it i would go for that Jun 13 23:06:43 Mind you.. the 20" panels.. they're rather nice.. 1600x1200 Jun 13 23:07:00 And more expensive too... Jun 13 23:07:25 Yeah.. Jun 13 23:07:29 depends on what your card can do to Jun 13 23:07:33 If only they released decent-res panels in 19" Jun 13 23:07:47 All of the insane-DPI LCDs seem to be taken up by laptop manufs Jun 13 23:07:55 Thank you guys. Jun 13 23:08:29 My old crt hurts my eyes now, so I have to buy a new monitor... Jun 13 23:08:33 just get a GOOD projector Jun 13 23:09:08 I love the resolution of Zaurus display ;-) Jun 13 23:09:38 and 17" are cheaper too. I'll be able to upgrade RAM as well ;-) Jun 14 00:35:42 evening. Jun 14 00:40:56 Ello Jun 14 00:41:22 wassup ? Jun 14 00:41:57 Not a whole lot, yourself? Jun 14 00:42:24 quite same here, I'm learning more on bitbake / oe . Jun 14 00:42:27 to make bb files. Jun 14 01:32:45 * v8jlene needs coffee... Jun 14 01:33:56 magnet: You might find my doco useful... I've been documenting everything about oe as I learn it, a lot is about .bb files. See http://oe.twibble.org **** ENDING LOGGING AT Wed Jun 14 02:59:56 2006