**** BEGIN LOGGING AT Mon May 12 02:59:57 2008 May 12 03:09:43 hey May 12 03:09:54 I am having a problem with another header file May 12 03:09:59 glib-object.h May 12 03:10:12 does anyone know what this is? my make cant find it May 12 03:10:19 probably glib-dev May 12 03:10:21 or something May 12 03:10:27 that's already installed May 12 03:10:31 or is it? May 12 03:10:45 dpkg -l |grep glib-dev May 12 03:11:21 i did that actually May 12 03:11:42 just tried an apt-get and it told me that its up to date May 12 03:11:54 libglib2.0-dev May 12 03:16:37 arrr May 12 03:17:17 so I did a ls -R *.h on my /usr/include directory May 12 03:17:39 there are no glib*.h files in there May 12 03:17:57 wrong May 12 03:18:06 find /usr/include -name glib*.h May 12 03:18:17 or something May 12 03:18:35 on crap May 12 03:18:37 that did it May 12 03:18:58 but you are actually missing `glib-configure --includes` in your makefile May 12 03:19:02 or something May 12 03:19:18 glib-config i mean May 12 03:19:20 ohhhhhhhhh May 12 03:19:21 ok May 12 03:19:24 and --cflags May 12 03:19:27 right May 12 03:19:35 i think thats in the helloworld compilation line May 12 03:19:38 i need to copy that May 12 03:19:43 i didnt understand it before May 12 03:19:47 GeneralAntilles: did nokia password protect the diablo repository? May 12 03:19:56 Looks like it. May 12 03:20:08 lol May 12 03:20:15 good thing i made a copy May 12 03:20:15 :> May 12 03:20:29 *i've May 12 03:22:54 that sucks May 12 03:24:32 KotCzarny: is there a reference somewhere for these build options? May 12 03:24:40 probably May 12 03:24:42 :) May 12 03:24:46 glib-config isnt being found either May 12 03:24:48 lol May 12 03:56:14 ok so the only "glib-" programs that are on my system are glib-genmarshal, glib-mkenums, and glib-gettextize May 12 03:56:22 none of which sound at all like glib-config May 12 03:56:29 may be named differently May 12 03:57:07 i did a find / -name glib-* May 12 03:57:33 it's gotta start with glib, right? May 12 03:59:27 probably May 12 04:00:07 may be for glib1 though May 12 04:04:25 GeneralAntilles: when I just did an apt-get update, it saw the repository and didn't complain at all May 12 04:04:36 GeneralAntilles: what is it like for you? May 12 04:04:49 Haven't looked at it May 12 04:04:56 Some other people were reporting 401s May 12 04:08:32 ohhhhh May 12 04:08:37 maybe is it pkg-config? May 12 04:08:41 probably May 12 04:23:57 so this is my program that wont compile May 12 04:24:12 #include May 12 04:24:15 int main() May 12 04:24:17 { May 12 04:24:21 exit(0); May 12 04:24:22 } May 12 04:24:41 its obvious that there are some rudimentary includes that I dont have May 12 04:24:47 but I cant find this documented anywhere May 12 04:25:14 all of my errors come from within the gps include files May 12 04:46:30 ok May 12 04:46:37 the solution is simple May 12 04:46:50 duh May 12 04:47:16 compile: gcc with `pkg-config --cflags glib-2.0` May 12 04:47:28 i completely GUESSED that line May 12 04:47:52 by looking at the compilation line that is used for the hello world example, which uses gtk May 12 04:50:18 then you haven't guessed completely May 12 04:56:20 lol May 12 04:56:22 ok May 12 04:56:36 so now that it compiles, i tried to add GPS calls to my void main() May 12 04:56:46 and it doesnt have a clue what any of the functions are May 12 04:59:53 well May 12 04:59:58 i dont have a clue what any of the functions are May 12 05:00:02 so if *I* dont know May 12 05:00:05 it sure as hell wont :P May 12 05:00:23 how do i tell my n800 system to boot verbose? May 12 05:01:33 i cant find anything on google about verbose boot May 12 05:03:54 oneword: fbconsole May 12 05:04:24 thats like, 3 words May 12 05:04:31 nope May 12 05:04:34 :P May 12 05:04:54 ok, for a ':P' you're on your own from now :P May 12 05:05:40 *sigh* May 12 05:05:44 but but but May 12 05:05:47 ":P" is "cute" May 12 05:06:22 ok so May 12 05:06:24 fbconsole May 12 05:06:26 is like not on google May 12 05:07:05 i see you talking about it back on 4/26 May 12 05:07:06 and thats it :( May 12 05:07:12 :) May 12 05:07:40 try fb console then May 12 05:09:50 bah May 12 05:09:50 scrwe it May 12 05:09:53 too much trouble May 12 05:55:36 dang i just traced out the berlin wall on a google map and it crashzdor May 12 05:55:42 teh gimp May 12 05:55:58 :) May 12 05:59:54 KotCzarny: care to say how to get a real sb environment up and running? (I've had no luck) May 12 06:00:07 define 'real sb' May 12 06:00:23 someone had a nice info on getting it running on Gentoo May 12 06:00:34 but I lost the url May 12 06:00:54 you mean non-vmware? May 12 06:01:08 yes, as vmware doesn't work :/ May 12 06:01:23 (can't pull the rest of sw) May 12 06:01:28 I've got the latest image to work but I prefer real sb May 12 06:01:34 astorm: actually i have just followed instructions in install.txt May 12 06:01:35 (or install anything) May 12 06:01:41 KotCzarny: hmmhmm May 12 06:01:52 their binary fails here May 12 06:02:01 what's the erro? May 12 06:02:02 for reasons related to ld May 12 06:02:03 error May 12 06:02:08 vdso? May 12 06:02:13 qwerty12: yep May 12 06:02:18 Disable it :) May 12 06:02:21 ;) May 12 06:02:26 and read docs! May 12 06:02:26 huh? May 12 06:02:34 KotCzarny: read these May 12 06:02:50 qwerty12: it's that sysctl, right? May 12 06:03:06 I hate to disable security features May 12 06:03:13 no May 12 06:03:16 No idea on what it is, I just know how to disable it :/ May 12 06:03:16 it's the /proc May 12 06:03:19 a pity sb is broken :P May 12 06:03:26 qwerty: and thanks to who? May 12 06:03:27 :) May 12 06:03:30 KotCzarny: /proc/sys = sysctl May 12 06:03:30 You can do it through sysctl afaik too. May 12 06:03:38 astorm: not quite May 12 06:03:39 KotCzarny, you :). but I did find other method too ;) May 12 06:03:52 mhm, so time to rock :P May 12 06:04:05 first build qill be a real Xfce terminal May 12 06:04:08 i just have a script for starting sbox May 12 06:04:20 Here is vdso sysctl.conf line - vm.vdso_enabled = 0 May 12 06:04:25 KotCzarny: can vdso be disabled per-task? May 12 06:04:33 (or chroot for that matter) May 12 06:04:36 probably not May 12 06:04:40 :/ May 12 06:04:55 but you can reenable it after you finish with sbox May 12 06:05:01 hate to leave my servers exposed to most basic stack overflows :P May 12 06:05:08 mhm May 12 06:05:20 well, could run it in vmware anyway May 12 06:05:29 or in a vserver May 12 06:05:36 then just grab a free p3 box May 12 06:05:43 (latter preferred due to 0 speed hit) May 12 06:06:18 ok, it's vserver then May 12 06:06:50 KotCzarny: btw, mplayer gets real nice when reniced -1 May 12 06:06:55 :) May 12 06:07:01 and esd is reniced to -5 or higher May 12 06:07:09 which indicates scheduler fail May 12 06:07:26 thus I shall speed up 2.6.25 implementation here :P May 12 06:08:01 pity I can't install two kernels here May 12 06:08:20 can't trust kexec either May 12 06:11:11 hrm May 12 06:11:19 pointers are longints on amd64? May 12 06:11:30 or more? May 12 06:13:47 heh, n810 screen is indeed readable in sunlight May 12 06:14:01 w/o backlight even May 12 06:14:03 lucky :( May 12 06:14:08 although loses color May 12 06:14:14 My n800 one sucks badly. May 12 06:14:32 heh, that's the price diff May 12 06:14:38 hehe May 12 06:14:52 true :) May 12 06:14:53 i have to check this transreflectivity someday May 12 06:14:56 we'll see if n900 3,4" format will sell May 12 06:15:12 3.4 ? May 12 06:15:15 KotCzarny: grab N95 - same kind May 12 06:15:27 KotCzarny: smaller, yes May 12 06:15:34 i imagine they leave 800x480 ? May 12 06:15:40 dunno May 12 06:15:45 or even up it to 1024x600 May 12 06:15:56 anyway, right now you need good eyes :P May 12 06:16:02 Why would N900 be 3.4"} May 12 06:16:13 GeneralAntilles: I've heard something May 12 06:16:20 You heard nothing May 12 06:16:30 There's absolutely zero information on the N900. ;) May 12 06:16:32 AStorm: source? May 12 06:16:43 doc|home: here, grep logs May 12 06:16:53 and they definitely wont be dropping the screen size. May 12 06:17:00 ah, a bloke on irc told you? :D May 12 06:17:16 http://www.infosyncworld.com/news/n/8262.html May 12 06:17:19 Especially considering that maemo isn't set up for anything other than 800x480 at the moment May 12 06:17:22 KotCzarny: transflective screen looks like those oldschool lcds with mirror under May 12 06:17:24 3.4" is way too small May 12 06:17:27 The Nokia N900 will be a mini-version of the N800, with a 3.5" display, like the one iPhone has. May 12 06:17:48 That's a joke article. May 12 06:17:56 sure, we will see in june May 12 06:17:58 :) May 12 06:18:07 hopefully, not smaller May 12 06:18:29 curent form could be slightly condensed though May 12 06:18:49 shave off 2 cm width May 12 06:18:49 They're not going smaller. May 12 06:18:58 we will see May 12 06:19:01 You'd have nothing to hold it by if they did that. May 12 06:19:04 they went smaller with n810 May 12 06:19:04 ugh May 12 06:19:09 the article is utter crap May 12 06:19:09 Besides, the N810 is packed tight enough anyway May 12 06:19:13 GeneralAntilles: no, why? May 12 06:19:16 total utter crap May 12 06:19:22 technology is moving forward May 12 06:19:28 GeneralAntilles: no, it's not May 12 06:19:30 I don't want to have to sacrifice more stuff to get the device smaller. May 12 06:19:38 i has lots of free spacee :P May 12 06:19:40 so what was packed yesterday may be lots of space tomorrow May 12 06:19:43 :) May 12 06:19:45 *it May 12 06:20:05 GeneralAntilles: they shouldn't need to May 12 06:20:14 just smarter placement May 12 06:20:23 Easier said than done. May 12 06:20:35 yes, esp. considering cost May 12 06:20:45 saying 'it can't be done! impossible!' is retarded May 12 06:20:51 but it is going to provide virtual 800x1200 with zooming right? May 12 06:21:07 keesj: ask driver devel about it May 12 06:21:12 sure and a kitchen sink May 12 06:21:13 and about 3D May 12 06:21:17 :P May 12 06:21:17 :) May 12 06:21:19 Smaller sucks anyway May 12 06:21:25 Means shitty placement of all the buttons May 12 06:21:33 GeneralAntilles: no, same May 12 06:21:34 Buttons should be on the face, not the slider May 12 06:21:42 I'd much rather a bigger screen, same res, in the same size package. I.e. less bezel May 12 06:21:43 ah, the dir button May 12 06:21:56 I suggest the use of touchscreen May 12 06:22:00 as long screen is multitouch you don't need hw buttons May 12 06:22:02 and they will move the buttons to and multi-rouch pad to the back May 12 06:22:10 buttons are useful May 12 06:22:14 imo May 12 06:22:15 ^ May 12 06:22:18 KotCzarny: you do need the keyboard though May 12 06:22:22 heh May 12 06:22:24 Multitouch is overrated. May 12 06:22:29 I'd like more buttons on the sides May 12 06:22:31 depends on the task May 12 06:22:33 iPhone is a pain to use. May 12 06:22:40 there is space for a few May 12 06:22:54 make it read your brainwaves May 12 06:22:55 :P May 12 06:22:58 naaah May 12 06:23:14 just a few buttons to ease cpnstrolling of canola :P May 12 06:23:29 *controlling May 12 06:23:38 the article is just written as what would some random guy want it to be... May 12 06:23:53 play/pause and forward/backward May 12 06:23:53 yeah May 12 06:23:59 the ovi stuff in it for example May 12 06:24:01 this article is fully hypothetical May 12 06:24:01 volume control is already May 12 06:24:04 there May 12 06:24:14 " May 12 06:24:14 The N900 will be powered by Smart Tablet OS 2008 (based on Maemo), " ? May 12 06:24:15 BS. May 12 06:24:19 lol May 12 06:24:22 also it says it would have opera still May 12 06:24:31 and a lot of other stuff May 12 06:24:33 ok, ignore it May 12 06:24:41 I don't think this guy has ever touched a N8*0. May 12 06:24:43 way ahead of you May 12 06:24:59 it may even be older than n810 May 12 06:25:01 :) May 12 06:25:05 hehe May 12 06:25:12 qwerty12: yes "Smart Tablet OS" is a very nice name IMHO May 12 06:25:21 ok, I'll draw my dream N810 mod May 12 06:25:32 really a minor change May 12 06:25:42 perfomming GPS chip May 12 06:25:53 keesj, but it's not called smart tablet :P. and i'd be disappointed if it still absurdly ran OS2008 (and yes, I know it won't) May 12 06:25:54 GPRS/Phone functionality... (for me that is) May 12 06:26:02 except adding keys is impossible :| May 12 06:26:11 (for me, at least) May 12 06:26:36 astorm: custom usb kb? May 12 06:26:57 KotCzarny: yeah, 4 keys attached with tape? lol May 12 06:26:59 keesj, that would be good for A-GPS. According to Mara, there's a test program in development that uses the GPRS/UMTS of your phone for A-GPS on N810 May 12 06:27:05 something like that May 12 06:27:07 :) May 12 06:27:25 in a form of sleeve May 12 06:27:25 KotCzarny: unsightly and fragile May 12 06:27:41 I'd need to somehow replace current keys May 12 06:27:46 or mod them May 12 06:27:52 and cut more holes May 12 06:28:07 too hard May 12 06:28:39 N810 could use a 2000 mAh battery :P May 12 06:29:04 My N800 could do with a 100000000000000 mAh battery :/ (no joke :() May 12 06:29:06 (that's actually possible to do) May 12 06:29:17 qwerty12: ya, but not feasible May 12 06:29:24 while 2000 is. May 12 06:29:31 your electricity bill would kill you May 12 06:29:39 crank would be nice ;) May 12 06:29:47 olpc or crank charger style May 12 06:29:47 except Nokia likes to reuse battery designs May 12 06:29:48 hehe. Diablo keeps my CPU and Mem usage on high all the time :/ May 12 06:29:53 glass: build one :P May 12 06:29:54 awsum May 12 06:29:55 * keesj is getting a lot of callbacks on the handle_hangover function and INT_COFFER interrupts in the head and goes back to idle mode May 12 06:30:26 btw. anyone with a knowledge about updates/diablo directory 401 ? May 12 06:30:27 :) May 12 06:31:15 ok, feature reqs: May 12 06:31:26 - better gps (but then...) May 12 06:31:36 real separate shift keys May 12 06:31:39 * doc|home couldn't care less about gps May 12 06:32:02 second ctrl May 12 06:32:27 * qwerty12 is now happy because I /finally/ got the 0xffff gui to compile May 12 06:32:41 gaming keys? May 12 06:32:43 :) May 12 06:32:49 naah, why May 12 06:32:54 for some apps May 12 06:33:30 next: 3 more keys on the rim: play/pause, next, prev May 12 06:34:34 maybe: slim it down 1 cm May 12 06:34:46 needed: usb charging May 12 06:34:49 down to 1cm or by 1cm? May 12 06:34:56 KotCzarny: by May 12 06:35:29 ah, software: May 12 06:35:46 make app menus more finger friendly May 12 06:35:55 nah May 12 06:36:00 (I mean those of menu button) May 12 06:36:07 just make drivers more dev friendly May 12 06:36:08 and move menu button to thne cover May 12 06:36:24 and make camera not crash the device :P May 12 06:36:40 * doc|home has never once used the camera May 12 06:36:41 optionally, if not slimming the device May 12 06:36:54 add touchpad for scrolling on the right May 12 06:37:04 (synaptics, hopefully) May 12 06:37:14 another closed driver? May 12 06:37:28 synaptics != closed May 12 06:37:40 new version could be closed :) May 12 06:37:49 why? just use usb May 12 06:38:01 even worse May 12 06:38:04 there is a spare hub inside already May 12 06:38:07 usb isn't so much power friendly May 12 06:38:20 KotCzarny: oh really? it has autosuspend May 12 06:38:32 only if you enable it yourself May 12 06:38:45 not really May 12 06:38:57 new kernels have it May 12 06:39:03 hehe May 12 06:39:08 part of "sleepy linux" project May 12 06:39:10 newer than .21 ? May 12 06:39:18 yes May 12 06:39:23 since .24 May 12 06:39:32 diablo uses .21 May 12 06:39:39 bummer! May 12 06:39:50 at least for now May 12 06:39:51 why can't they just port the darned cx driver May 12 06:39:59 and voltage control May 12 06:41:02 alternatively, pay someone to devel prism54 for spi some more May 12 06:42:53 should I be able to get another n810, I'll put it onto our rev-eng table May 12 06:43:14 * KotCzarny saw 770 for 75$ May 12 06:43:15 and see what the umac does May 12 06:43:23 KotCzarny: same wifi chip? May 12 06:43:26 770 has very similiar vhip May 12 06:43:28 chip May 12 06:43:30 yeah May 12 06:43:38 mhm May 12 06:43:47 so I'll get one May 12 06:43:59 or N800 for that matter May 12 06:44:58 destroy it somewhat :P and analyse traffic to the chip May 12 06:45:26 good for engineer degree :P May 12 06:48:12 ack, can't disable vdso in a vserver May 12 06:48:15 I lose May 12 06:48:18 :> May 12 06:48:35 would need Xen, and that patchset is *oldyc May 12 06:48:35 http://www.msss.com/msss_images/2008/01/31/ May 12 06:48:37 lol? May 12 06:48:40 *old* May 12 06:48:48 as in 2.6.18 May 12 06:48:50 astorm: just disable that vdso May 12 06:48:50 :) May 12 06:48:54 yep May 12 06:51:28 why oh why have I hit a cpu w/o svm... May 12 06:52:15 * acydlord facepalms May 12 06:52:34 my portable hard drive kicked the can, and it had all my pgp and ciphers on it May 12 06:52:43 lol May 12 06:52:45 acydlord: you win! May 12 06:52:52 now you can start real life May 12 06:53:01 your goose is cooked! May 12 06:53:03 ;) May 12 06:53:04 w/o paranoia :P May 12 06:53:07 nah, i issued new keys May 12 06:53:12 wow they have a big zoo there May 12 06:53:18 and emailed just about everyone that has my old key May 12 06:53:20 surprised linuxtag isn't held there May 12 06:53:30 unfortunately i cant revoke the lost keys May 12 06:53:48 acydlord: about everyone = 2 persons? May 12 06:54:02 don't forget to destroy the drive May 12 06:54:03 :) May 12 06:54:12 yeah, flatten it May 12 06:54:14 around 60 persons May 12 06:54:20 use the hammer May 12 06:54:22 etc. May 12 06:54:41 i have some thermite set aside for such occasions May 12 06:55:03 whee May 12 06:55:07 paranoia++ May 12 06:56:03 it's the only way to do it lol May 12 06:56:14 fire. May 12 06:56:15 ;) May 12 06:56:22 I use a flash key May 12 06:56:30 easily destructible May 12 06:56:42 just use a boot May 12 06:56:50 i should encrypt my / May 12 06:56:59 KotCzarny: what for? May 12 06:57:06 just in case May 12 06:57:17 and for education May 12 06:57:18 :) May 12 06:57:21 encrypting your / will cause performance issues May 12 06:57:31 yup May 12 06:57:35 acydlord: it's already on cf May 12 06:57:37 on n, that is May 12 06:57:43 so what performance issues ? May 12 06:57:44 ;) May 12 06:57:52 KotCzarny: you're mad or what? May 12 06:58:01 I'll just put the kernel there May 12 06:58:02 astorm: works for a few years already May 12 06:58:04 the proc will still lose cycles decrypting it May 12 06:58:12 and initramfs with tripwire May 12 06:58:13 acydlord: it's a laptop May 12 06:58:19 not much hdd work May 12 06:58:40 and it has enough mem to cache almost everything May 12 06:58:46 boot time will lose, but nothing more May 12 06:58:51 lol May 12 06:58:53 boot time? May 12 06:58:59 like, once per 3 months? May 12 06:59:00 :) May 12 06:59:07 hmm, any usb key with hardware write lock? May 12 06:59:10 poor laptop, thats how mine died May 12 06:59:23 my usb key has hardware lock May 12 06:59:34 mine is too cheap May 12 06:59:34 acydlord: this one will die from my typing patterns abuse than from uptime May 12 07:00:07 acydlord: heh, the first thing to die is the fan :P May 12 07:00:12 mine was up for about 9 months May 12 07:00:15 what fan? May 12 07:00:17 :) May 12 07:00:26 mine is quiet May 12 07:00:31 must not have an amd powered lappie May 12 07:00:33 and runs fan only when i run epsxe May 12 07:00:41 ibm x32 May 12 07:00:42 :) May 12 07:00:47 and earlier x30 May 12 07:00:53 KotCzarny: mine in hot weather May 12 07:01:04 mine was an ath32 May 12 07:01:18 x30 was so nice.. May 12 07:01:20 makes nuclear fission seem cool May 12 07:01:24 pity it was too abused by me May 12 07:01:46 it ran cpu at 37-45C May 12 07:01:51 without fan May 12 07:01:51 :) May 12 07:02:02 i think my next lappie is gonna be one of the lenovo ideabooks May 12 07:02:07 the y15 looks nice May 12 07:02:14 i don't trust lenovo engineers.. May 12 07:02:24 why not? May 12 07:02:26 and their designers suck.. May 12 07:02:41 the y15 kinda redeems them a little May 12 07:02:42 acydlords: their t60 had issues May 12 07:02:53 As long as it has SSD I would be happy.. :-) May 12 07:02:54 i know it was released in a hurry May 12 07:03:09 yeah, the t60 was just ported from the ibm machining process May 12 07:03:19 but still, you don't release flawed product :) May 12 07:03:24 * nab remembers the SSD talks at the Linux Storage/Filesystem Workshop this year May 12 07:03:45 * KotCzarny runs on cf for many years already May 12 07:03:47 Something on the order of 1/5 power consumption and 10000x performance May 12 07:03:50 no bad sectors yet May 12 07:03:51 :) May 12 07:03:58 Basically, a single drive can do 50K IOPS May 12 07:04:37 but real ssd is pricy.. May 12 07:04:43 :> May 12 07:04:55 indeed May 12 07:05:02 Accessing storage over Wifi/WiMax does not have to worry about exceeding the TCQ depth for the SSD drives May 12 07:05:12 KotCzarny: Sure, but you have alot of new players coming into the market May 12 07:05:18 eg: Intel May 12 07:05:20 yeah May 12 07:05:43 i have read about new techs in flash May 12 07:05:51 still, few years to go May 12 07:05:54 So scaling the Linux kernel to meet 50,000 IOPs per say x12 drive arrays May 12 07:06:15 One of the big culprits is the fact that bdflush kernel thread is single threaded May 12 07:06:28 right now i'm happy with 2xCF to ide adapter May 12 07:06:29 :) May 12 07:06:31 (This was dicussed by one of the SGI fellows at the conference) May 12 07:06:42 But anyways, its a bit OT for #maemo May 12 07:06:43 :-) May 12 07:06:48 a bit May 12 07:06:49 :) May 12 07:07:01 So where it comes into play for OS2008 is the ability to access 900 GB mounts May 12 07:07:07 psh, never May 12 07:07:10 Heh May 12 07:07:11 Uhhhh May 12 07:07:16 well May 12 07:07:17 nab: btrfs will be a win on SSD too May 12 07:07:17 sshfs May 12 07:07:18 That is actually stable May 12 07:07:18 :) May 12 07:07:18 :) May 12 07:07:22 still have to use a semi-full distro for our dev enviroments May 12 07:07:34 lol May 12 07:07:38 It has been tested with a R/O MD RAID0 export over iSCSI May 12 07:07:43 sshfs/nfs/cifs May 12 07:07:52 https://linux-iscsi.org/index.php/Core-iSCSI/Nokia_800 May 12 07:07:58 own May 12 07:08:05 and May 12 07:08:06 https://linux-iscsi.org/index.php/OCFS2/iSCSI_for_OS-2008 May 12 07:08:07 :-) May 12 07:08:10 what about nbd? May 12 07:08:18 unable to verify the identify of localhost.localdomain as a trusted site May 12 07:08:19 ;) May 12 07:08:20 What about it..? May 12 07:08:22 Hehe May 12 07:08:24 *identity May 12 07:08:27 (for cheap people like me) May 12 07:08:27 May 12 07:08:40 I need to fix that. :P May 12 07:08:48 * nab was a sysadmin before a devel May 12 07:08:51 yeah, buy a key? May 12 07:08:54 But anyways... May 12 07:09:08 Yes, you can access a 900 GB read-only mount across iSCSI on the N800 May 12 07:09:20 There are definately limitations with the Maemo file-manager, wrt to scanning the whole archive May 12 07:09:25 why read-only? May 12 07:09:37 Imagine if you had to do a fsck on that large of a device May 12 07:09:46 what abt. building a large file kernel? May 12 07:09:54 and usoing nbd? May 12 07:09:58 *using May 12 07:10:00 There is no reason why with smaller individual access with R/W its not stable for production May 12 07:10:01 btw. it's over wifi? May 12 07:10:09 KotCzarny: Yes May 12 07:10:12 sucks. May 12 07:10:13 :) May 12 07:10:20 then sshfs is a lot better May 12 07:10:23 Using the other filemanager works fine however May 12 07:10:24 over usb should work too May 12 07:10:34 Not to mention it totally works with cluster filesystems May 12 07:10:35 :-) May 12 07:10:37 wifi/usb is slow on n8x0 May 12 07:10:43 ya May 12 07:10:55 and no 10GbE inside :P May 12 07:11:06 eg: when you copy stuff from another node, it updates the dir/file listing in EMFMv2 May 12 07:11:09 my n810 has a tendency to eat files off my windows shares when i try to transfer May 12 07:11:16 acydlord: May 12 07:11:22 NAS kinda sucks over WAN or WIFI May 12 07:11:27 that's a kernel bug May 12 07:11:33 :P May 12 07:11:38 seen that on my laptop May 12 07:11:44 iSCSI can run over GPRS at 10 Kb/sec with R/W mounts (if they are sub 1 GB) May 12 07:11:52 anyway, cifs has horrible performance over wifi May 12 07:11:54 sucks when your connection drops in the middle of an xfer May 12 07:12:03 sshfs runs over anything that's tcp/ip May 12 07:12:04 :) May 12 07:12:19 AStorm: You can easily do 1 MB/sec on b/g if you have a decent access point May 12 07:12:24 KotCzarny: yeah, including IPoAC May 12 07:12:32 ac? May 12 07:12:33 i should try to install freenet on my n810 when i'm feeling less lazy May 12 07:12:36 (from a mobile device with 400 Mhz ARM) May 12 07:12:41 nab: not with cifs May 12 07:12:51 AStorm: I am talking about traditional iSCSI May 12 07:12:57 ya May 12 07:13:00 then maybe May 12 07:13:06 No, that is the case May 12 07:13:06 :-) May 12 07:13:15 but sharing my hdd with that is the overkill May 12 07:13:19 *the* May 12 07:13:23 Actually, the first IBM discussions for iSCSI where for the NAS vs. SAN numbers May 12 07:13:27 sshfs is fast enough May 12 07:13:34 Back in the mid-90s May 12 07:13:42 alternatively, nfs May 12 07:13:48 Sure. :-) May 12 07:13:53 and safer May 12 07:13:54 :) May 12 07:14:00 yep May 12 07:14:05 and eats cpu more May 12 07:14:10 nfs doesn't May 12 07:14:14 But try running 900 GB/sec of NFS over WiMax using multiple communication paths and active-active connectionr recovery May 12 07:14:27 nab: can't - no WiMax May 12 07:14:29 wimax can do 900 gb ? May 12 07:14:34 gb/s May 12 07:14:36 Not to mention that NFS does not have a RFC defined naming convention.. May 12 07:14:36 :-) May 12 07:14:36 and it can't do that much May 12 07:14:45 it can go like 10 Mb/s May 12 07:14:49 :) May 12 07:14:50 at most May 12 07:14:53 KotCzarny: You can get 1 Gb/sec WiMax through towerstream May 12 07:14:55 i have installed Qt4 on chinook, but uable to get qmake May 12 07:15:03 nab: if you're lucky May 12 07:15:04 nab: with any chipset? May 12 07:15:11 or you have to get a better one? May 12 07:15:12 And since we know the N800 works with a 900 GB MD RAID0 R/O export over Wifi May 12 07:15:19 and have 1024 channels around May 12 07:15:20 Yes, it would work over WiMax, probably faster. May 12 07:15:21 :-) May 12 07:15:27 naah May 12 07:15:30 slower May 12 07:15:34 Not for R/O May 12 07:15:35 :-) May 12 07:15:44 blah blah May 12 07:15:45 For 802.11 N it work be faster May 12 07:15:48 locally May 12 07:15:51 readonly = unusable May 12 07:15:55 still, anyone with a wimax experience in tablets? May 12 07:15:56 eg: You can do iSCSI/HD over 802.11N locally May 12 07:16:01 ie. n810wimax ? May 12 07:16:09 AStorm: May 12 07:16:10 KotCzarny: buy me one :P May 12 07:16:16 Well, not unsuable May 12 07:16:20 anyway, I have no N devices May 12 07:16:22 :P May 12 07:16:24 It just means you need multiple mounts for stuff May 12 07:16:25 or WiMax May 12 07:16:41 Having a *SINGLE* shared storage mount it where OCFS2/iSCSI comes into play. May 12 07:16:53 a/it/is May 12 07:17:05 nfs doesn't suck that badly anyway then May 12 07:17:12 Actually, ~50 ms/sec latencies are workable for something like that May 12 07:17:13 unless you get conflicts :P May 12 07:17:27 50ms/sec ? May 12 07:17:28 :) May 12 07:17:30 For multi 100 GB storage object exports May 12 07:17:43 KotCzarny: WiMax network round trip latency May 12 07:17:49 nab: buy me such a disk May 12 07:17:50 i have installed Qt4 on chinook, but not getting qmake May 12 07:17:52 :P May 12 07:18:02 nab: i just can't read the units May 12 07:18:04 :) May 12 07:18:04 ranit: qt4-dev May 12 07:18:12 :-) May 12 07:18:16 KotCzarny: dl/ul May 12 07:18:47 but it comes down to 50 * 0.001 *1/1 May 12 07:18:48 :) May 12 07:19:15 unfortunatly was unable to install qt4-dev May 12 07:19:20 KotCzarny: Oh, there is definately work to be done. May 12 07:19:45 The OCFS2 and LIO folks already have discussions underway.. May 12 07:19:51 :) May 12 07:19:57 it gave me the following err May 12 07:19:59 The shared storage bit is definately a long-term item (espically for WiMax type stuff) May 12 07:20:02 The following packages have unmet dependencies: May 12 07:20:02 libqt4-dev: Depends: libmysqlclient12-dev but it is not installable May 12 07:20:02 Depends: libsqlite0-dev but it is not installable May 12 07:20:08 omg May 12 07:20:09 But it does exist and works. May 12 07:20:15 And you can download it. :-) May 12 07:20:16 mysql for qt4? May 12 07:20:30 yes, "optional" May 12 07:20:34 now that's a nice dependency May 12 07:20:34 :) May 12 07:20:38 as everything in debian May 12 07:20:40 :P May 12 07:20:43 I am waiting for >= v2.6.24 support on OS2008.. May 12 07:20:49 nab: nab astorm May 12 07:20:50 nab: build one May 12 07:20:54 he is working on .25 May 12 07:20:55 :) May 12 07:21:02 haxing the cx May 12 07:21:14 got 2.6.25-omap1 to boot already May 12 07:21:16 So we can use Maemo/OpenMoko devices with Distributed Lock Manager (DLM) and OCFS2 v1.5.0 May 12 07:21:32 Great! May 12 07:21:33 :-) May 12 07:21:36 and have a ... you know what? May 12 07:21:48 who needs that! May 12 07:21:54 sexy party? May 12 07:21:58 (with a tablet) May 12 07:22:05 doc|home: close :P May 12 07:22:12 damn :/ May 12 07:22:25 I am still waiting to hear back from Quim for some devel hardware.. May 12 07:22:37 i would like a tuxonice support May 12 07:22:44 or at least swsusp May 12 07:22:45 :) May 12 07:22:46 KotCzarny: hard port May 12 07:22:50 i know May 12 07:22:52 swsusp should work May 12 07:22:52 :) May 12 07:22:59 unless wifi explodes May 12 07:23:10 The question for me is if the local WiMax provider here (Towersteam) is compat with the N810 Wimax Edition.. :-) May 12 07:23:11 then it'd need a module reload May 12 07:23:29 Or I have to jump on ebay and get a WiMax tower myself. ;P May 12 07:23:39 astorm: not a problem, rc.goodmorning :) May 12 07:23:57 rc.stfu? ;) May 12 07:24:08 fan mod? May 12 07:24:09 :) May 12 07:24:21 I need something less crashy than matchbox for now May 12 07:24:33 go console May 12 07:24:38 or a non-crashy version of it May 12 07:24:50 KotCzarny: no Firefox/Minimo.... May 12 07:25:00 and touchscreen May 12 07:25:01 links -g ? May 12 07:25:10 Or when Xohm is coming to town in my area.. :P May 12 07:25:13 and fairly unusable keyboard layout May 12 07:25:18 Actually May 12 07:25:44 would need to write linux one May 12 07:25:47 * nab considers getting a N810 in a soft-launch town in the US for LIO devel.. May 12 07:25:54 and bind keys too May 12 07:26:00 Xohm soft-launch town May 12 07:26:13 the trick is that fn doesn't work :P May 12 07:26:26 re May 12 07:26:27 Doing remote devel cluster iSCSI for that device would be an interesting setup May 12 07:26:46 nab: did I mention it's overkill? May 12 07:26:53 AStorm: What is..? May 12 07:26:57 What's your opinion about enabling virtual memory on external flash? May 12 07:27:13 :) May 12 07:27:23 jdiaz: good idea May 12 07:27:31 dangerous :) May 12 07:27:32 :-) May 12 07:27:36 unless you're planning on removing it May 12 07:28:01 internal memcard slot not good enough? May 12 07:28:12 tank-man: n810 has none May 12 07:28:12 doesn't matter May 12 07:28:14 :) May 12 07:28:17 :P May 12 07:28:38 unless you are doing critical work May 12 07:28:51 you can stack swap May 12 07:28:57 and mod umount script May 12 07:28:58 I noticed that evince crashed on big pdfs (comic books), and also comix May 12 07:29:08 jdiaz: That reminds me May 12 07:29:08 :) May 12 07:29:08 and these crashes dissappeared using swap May 12 07:29:11 jdiaz: with OOM I guess? May 12 07:29:15 hehehe May 12 07:29:16 On the 900 GB MD RAID test May 12 07:29:17 OOM? May 12 07:29:23 oom-killer May 12 07:29:23 Out Of Memory May 12 07:29:26 :) May 12 07:29:35 One of the LIO devels (brd) has a few GB of photos May 12 07:29:41 I did not see any message... in fact, as far as I remember, the tablet hung May 12 07:29:44 and a reboot was required May 12 07:30:01 Doing the 100+ slideshows was stable using the R/O ext3 mounts May 12 07:30:04 ot didn't May 12 07:30:17 after you wait a longer while, it kills evince May 12 07:30:28 unless killer made a mistake May 12 07:30:32 so I thought that leaving virtual memory enabled should be a good idea May 12 07:30:34 using a b/g AP to LIO-Target x86_64 Debian May 12 07:30:35 and killed X and watchdog May 12 07:30:41 but the warning issued by maemo scared me May 12 07:30:49 about shortening the life of the flash May 12 07:30:51 hmm? May 12 07:30:56 yes, it will May 12 07:31:05 but flash is replacable May 12 07:31:10 but not that much May 12 07:31:13 (shortening) May 12 07:31:17 esp. external May 12 07:31:28 this is why I refered to externa, instead of internal May 12 07:31:32 we need real swap management anyway May 12 07:31:33 I never remove my external card May 12 07:31:43 * KotCzarny boots from external flash May 12 07:31:48 for adding contents I use scp May 12 07:31:48 and has swap there too May 12 07:31:57 it is more convenient than a card reader in the PC May 12 07:31:58 * AStorm has to move system to internal one May 12 07:32:24 well, on n800 bothe cards are removable :) May 12 07:32:27 which fs is the q - I guess btrfs will win May 12 07:32:29 so it doesn't matter much May 12 07:32:35 ya May 12 07:32:41 2G is a lot anyway May 12 07:32:46 depends May 12 07:32:48 :) May 12 07:32:56 if you ever ran maemo-mapper .. May 12 07:32:57 for system, sure May 12 07:32:57 ;) May 12 07:33:01 Idid May 12 07:33:09 Recently I was asking here about my worries about the root filesystem being in flash, and the use of noatime May 12 07:33:12 took like500 mb of maps :P May 12 07:33:22 jdiaz: both fixable May 12 07:33:24 as you can see, I'm a bit paranoid about flash life May 12 07:33:31 mine has ~1.5gb May 12 07:33:38 jdiaz: make an experiment May 12 07:33:48 keep all your critical files on internal flash May 12 07:33:51 KotCzarny: I scraped highest zoom levels May 12 07:33:56 they're pointless May 12 07:34:09 i haven't heard of anyone's internal flash crapping out May 12 07:34:13 and just run os on external May 12 07:34:19 pupnik_: neither have I May 12 07:34:21 pupnik: yet May 12 07:34:25 :) May 12 07:34:35 ok, need to format that flash May 12 07:34:40 (internal) May 12 07:34:45 but qwerty12 is working on it May 12 07:34:47 ;) May 12 07:35:01 btw, what is the proper naming for avoid confusion between the internal flash which holds the root filesytem and the internal flash in /media/mmc2? May 12 07:35:20 KotCzarny, what should be the outcoming of this experiment? May 12 07:35:39 jdiaz: you can name it 256mb flash and 2gb flash May 12 07:35:48 and everyone will know which you refer to May 12 07:35:54 or flash and internal mmc May 12 07:36:04 :) May 12 07:36:10 jdiaz: that if you ever wear out sd card you don't loose anything important May 12 07:36:15 both are the same kind of memory, in fact, right? May 12 07:36:21 jdiaz: no May 12 07:36:22 internal mmc is flash too May 12 07:36:24 :) May 12 07:36:39 but internal mmc has automatic wear levelling May 12 07:36:53 better or worse May 12 07:36:54 :) May 12 07:36:56 while the flash depends on the fs to do that May 12 07:37:08 jffs2 is quite intelligent i think May 12 07:37:12 yes May 12 07:37:19 now, when you are saying flash your are refering to the 256Mb? May 12 07:37:23 and with fs you can upgrade the algo May 12 07:37:25 jdiaz: yes May 12 07:37:25 and internal mmc to the 2Gb? May 12 07:37:30 yes May 12 07:37:33 ok May 12 07:37:44 I find less confusing the sizes nomenclature :-) May 12 07:37:45 KotCzarny: not that it'll help May 12 07:38:07 btw. if you must capitalize units, do it properly May 12 07:38:08 MB May 12 07:38:09 wear levelling can't be improved :P May 12 07:38:09 :) May 12 07:38:13 Mb is mega bits May 12 07:38:14 ;) May 12 07:38:32 astorm: you assume that the best algo is already in use :) May 12 07:38:34 I should make a poll about the use of mmc1 and mmc2 May 12 07:38:37 anyway, btrfs in ssd mode is the way to go May 12 07:38:39 i'm currently dissasembling my portable drive lol May 12 07:38:41 KotCzarny: it is May 12 07:38:51 in my case, I use mmc1 to store music and films whith I replace frequently May 12 07:38:53 there are write units and spare space May 12 07:38:57 and almost never use mmc2 May 12 07:39:07 jdiaz: but the OS does :P May 12 07:39:14 :) May 12 07:39:19 as for the 256Mb, only for storing configuration files and shell scripts May 12 07:39:27 pity there are no sd/flash activity leds May 12 07:39:33 heh May 12 07:39:38 MyDocuments is almost empty May 12 07:39:39 or at least statusbar app :) May 12 07:39:40 oh, you could bind the led May 12 07:39:53 to the mmc driver May 12 07:39:58 :) May 12 07:40:03 unable to install libqt4-dev in chinook May 12 07:40:04 AStorm: May 12 07:40:06 the led uses normal led interfaces May 12 07:40:07 but that would require some hacking May 12 07:40:12 not much May 12 07:40:18 AStorm> jdiaz: but the OS does :P May 12 07:40:21 how is that? May 12 07:40:25 But having the shared storage aspect is important.. May 12 07:40:30 jdiaz: maemo mapper, swap May 12 07:40:30 Morning, all May 12 07:40:31 jdiaz: apt-get for example May 12 07:40:56 yes, that too May 12 07:41:08 ok, I've to clone my system, live :P May 12 07:41:09 I also heard rasterman talking about offline ops May 12 07:41:11 so your advice is better to forget about flash life length and use both mmc as normal hard drives? May 12 07:41:30 jdiaz: let's put it diff. May 12 07:41:39 Which comes from some of the older FSes (eg: Coda), but has not been seen in the next/gen stuff May 12 07:41:55 at least not yet anyways May 12 07:42:03 That is actually the main bit with OCFS2 on Wifi May 12 07:42:13 not forget, but don't make it take your sleep away May 12 07:42:23 yeah May 12 07:42:25 or wake you up in the night May 12 07:42:28 The node fencing the DLM logic does for extended Wifi failure May 12 07:42:30 * KotCzarny has /var as tmpfs May 12 07:42:32 :) May 12 07:43:06 So the OCFS2 Metadata + DLM ensure that all nodes in the cluster see the write updates on remote nodes May 12 07:43:12 I know nothing about filesystems, but guess that jffs2 is specially efficient for flash drivers, making its life longer May 12 07:43:20 but what about a swap partition? May 12 07:43:28 And for production cases using OCFS2, you obvision want the host (usually VM) to fence itself and reboot. May 12 07:43:33 shouldn't it be killer? May 12 07:43:41 it does not use jffs2 May 12 07:43:44 (instead of writing what the logic thinks is the incorrect data) May 12 07:43:48 and a lot of writes are expected May 12 07:43:50 Or due to node timeout, etc. May 12 07:44:21 jdiaz: for flash, yes May 12 07:44:30 but impossible to use on mmc May 12 07:44:46 why impossible? May 12 07:44:51 But its really a manner of tuning timeouts and a mobile/OCFS2 patch May 12 07:44:53 :-) May 12 07:44:55 at least hard May 12 07:44:59 why hard? May 12 07:45:01 and inefficient May 12 07:45:10 inefficient, yes, as all swaps May 12 07:45:11 :) May 12 07:45:13 (if you don't get direct access) May 12 07:45:15 not swap May 12 07:45:20 I meant jffs2 May 12 07:45:21 unable to install libqt4-dev in chinook May 12 07:45:26 swap is another thing May 12 07:45:31 ranit: we've heard May 12 07:45:35 ask on ITT May 12 07:45:50 or wait for nokia to switch to qt May 12 07:45:52 :) May 12 07:45:59 which won't happen May 12 07:46:16 :) May 12 07:46:29 never say never May 12 07:46:29 :) May 12 07:46:34 are there any replacement images i could use for my 770? what the most popular non-maemo os for the 770 is? May 12 07:48:35 AStorm: why it won't happen.... May 12 07:48:46 no alternative OS is popular jaek May 12 07:49:08 it should be some way of asking the flash drivers how much writes has been done, and thus the expected remaining life for the memory May 12 07:49:22 jdiaz: hard May 12 07:49:26 that is a shame May 12 07:49:32 I'm obsessed, I know :-) May 12 07:49:35 very dependent on the write pattern May 12 07:49:50 and no one knows the write count really May 12 07:50:16 battery is other of my concerns. I have to recharge it at least twice a day May 12 07:50:22 so what? May 12 07:50:30 and on wear-levelling algo May 12 07:50:32 li-poly dies with time, not cycles May 12 07:50:43 with temp too May 12 07:50:45 really? May 12 07:50:51 during charging it gets warmer May 12 07:50:52 it dies with deep discharge and temperature May 12 07:51:08 but the temp isn't that high in n May 12 07:51:20 so there is no problem on leaving the charger connected most of the time? May 12 07:51:38 no, except increased power bill May 12 07:51:46 chargers are inefficient May 12 07:52:04 hrm May 12 07:52:10 trickle charging kills it too May 12 07:52:14 not May 12 07:52:16 yes May 12 07:52:28 it kills the ones not designed for it May 12 07:52:42 anyway, N doesn't trickle charge May 12 07:52:53 proof? May 12 07:53:02 your graph :) May 12 07:53:19 my graph was updated once per 60s May 12 07:53:28 fast enough May 12 07:53:38 you'd catch some trickle May 12 07:53:57 not really May 12 07:53:57 how the trickle would show up in the graph? May 12 07:54:00 error margin is big May 12 07:54:13 * jdiaz is fascinated with this conversation May 12 07:54:16 as reads are wacky and inaccurate May 12 07:54:24 heeh May 12 07:54:42 anyway, trickling doesn't reduce life if battery is designed for it May 12 07:55:00 you are assuming that battery is designed for it May 12 07:55:07 it is May 12 07:55:10 link? May 12 07:55:20 did you ever exhausted the life of your battery? May 12 07:55:30 jdiaz: I did :P May 12 07:55:31 i did May 12 07:55:42 how much time did it last? May 12 07:55:51 6-12 months May 12 07:55:52 :) May 12 07:55:54 about a year and half May 12 07:55:58 gasp May 12 07:56:05 but it was noname May 12 07:56:06 :) May 12 07:56:12 jdiaz: like, 12h May 12 07:56:20 not in a nokia phone though - these guy are tough May 12 07:56:21 if you mean a discharge May 12 07:56:26 but according to what said above, battery life doesn't depend on usage patterns/habits, does it? May 12 07:56:36 it does May 12 07:56:37 but if death... yes, 2y May 12 07:56:40 oh, I was talking about the battery of the NIT May 12 07:56:42 in the phone May 12 07:56:56 BP-5L though May 12 07:57:03 2,5y to be exact May 12 07:57:14 jdiaz nit is essentially a phone but the batt usage pattern is more active May 12 07:57:16 astorm: n810 uses bp-4something May 12 07:57:16 lost about half of capacity May 12 07:57:21 n800 uses bp-5l May 12 07:57:23 KotCzarny: yes, the new one May 12 07:57:40 but they are interchangeable May 12 07:57:41 bp-4l is only used in a few phones yet May 12 07:58:00 KotCzarny: not due to size and charging May 12 07:58:04 hrm May 12 07:58:12 that's bad May 12 07:58:12 bp-5l is li-ion May 12 07:58:21 bp-4l is li-poly May 12 07:58:26 nits use li based batteries which means "never let it completely discharge" May 12 07:58:35 ya May 12 07:58:47 I almost managed to do that by accident May 12 07:58:56 well i found out whats wrong with my portable drive May 12 07:59:03 the heads are broke May 12 07:59:13 acydlord: which means byebye May 12 07:59:24 did you have to open it to find out?;) May 12 07:59:42 I've found my NIT turned off in the morning. This occured to me several times, but I assume that the OS switch the unit off before the battery is completely discharged May 12 08:00:15 jdiaz: are you using scrolled rss applet? May 12 08:00:22 jdiaz: yes May 12 08:00:25 I do not know what causes the discharge. It is full when I go to bed May 12 08:00:31 RST38h, no May 12 08:00:34 * KotCzarny found drained battery yesterday May 12 08:00:40 and it was full 2 days before May 12 08:00:41 RST38h: btw, is there a real, working rss client? May 12 08:00:45 then it must be the media scanner May 12 08:00:59 and is OMWeather fixed now wrt weather.com code changes? May 12 08:01:00 I turned off (i.e. renamed) media scanner May 12 08:01:00 i didn't run anything other than ftd in the mean time May 12 08:01:01 astorm: yep. google reader mobile! May 12 08:01:11 RST38h: lol, needs network May 12 08:01:23 but then... loading rss needs too May 12 08:01:28 :P May 12 08:01:29 astorm: and if you are fancy you can even tell it to run in iphone mode May 12 08:01:30 It happens to me when I listen some music while getting asleep May 12 08:01:51 astorm: google gears ported to nit already May 12 08:02:03 jdiaz: yes, it will leave a bit of power in May 12 08:02:07 jdiaz: you answred your question May 12 08:02:25 so that it can boot up May 12 08:02:27 I did not found any "sleep" function in most of media players, so I did a script which sleeps for a time, and then "killall mplayer". I use mplayer for playing May 12 08:02:44 jdiaz: good idea for a canola plugin May 12 08:02:48 :) May 12 08:02:59 astorm: seriously google mobile sites solve most of my net problems on nit May 12 08:03:12 :) May 12 08:03:14 RST38h: same, except for google docs May 12 08:03:17 the script works ok, so mplayer should not be the problem either May 12 08:03:23 need real working abiword May 12 08:03:25 docs mobile works May 12 08:03:30 it gets killed after 20 min of playing May 12 08:03:35 use iphone config May 12 08:03:41 RST38h: naah, I mean, these work May 12 08:03:52 yea, what about abiword? how long? May 12 08:03:53 but aren't good enough at reading doc May 12 08:04:00 RST38h: dunno May 12 08:04:22 probably it is an issue of accidental pressure on the touchscreen May 12 08:04:39 jdiaz: install autolock May 12 08:04:52 it lock nit when light goes out May 12 08:04:53 I have it installed, but I'm not happy with it May 12 08:04:56 RST38h: uhm, isn't it already there? May 12 08:05:12 and make wifi disconnecting after 1h idle May 12 08:05:13 :) May 12 08:05:13 its behaviour is still not predictable to me, after a lot of experimentation May 12 08:05:14 the only problem - it is not configurable May 12 08:05:30 astorm: nope - check May 12 08:05:52 if I slide the lock key, then autolock chaos is assured May 12 08:05:56 shit, this line is long :( May 12 08:06:29 should have bought subway pass BEFORE holidays May 12 08:06:36 ok, time to format /media/mmc1 to btrfs May 12 08:06:41 :) May 12 08:06:46 urgh.. May 12 08:06:49 I like the autolock idea very much, but its interaction with the lock key seems broken May 12 08:06:59 :) May 12 08:07:02 RST38h: you have wifi there? win!a May 12 08:07:13 * RST38h shastorm: gprs May 12 08:07:16 or is it 3g? May 12 08:07:22 ah May 12 08:07:25 no, plain gprs May 12 08:07:27 expensive? May 12 08:07:33 still waiting for 3g May 12 08:07:53 cheap, i clock at around 300-500rur a month May 12 08:08:13 i.e. less than 20 bucks May 12 08:08:22 I think the problem is that the "locked status" is a software flag, not a hardware switch May 12 08:08:36 of course it is May 12 08:08:40 and we arrive to the same status by sliding the lock key or by doing it by software May 12 08:08:40 brb May 12 08:08:53 but there shoukd be a single way to toggle it May 12 08:09:15 maybe there is May 12 08:09:16 :) May 12 08:09:35 so autolock can't differentiate if we are in locked status because the user slided the lock key, or because it was dark May 12 08:09:38 actually ftd allows you to suspend-to-ram May 12 08:10:22 for example, assume the following scenario May 12 08:10:41 I have the unit unlocked. Suddenly, it is pitch dark May 12 08:10:44 autolock locks May 12 08:10:50 then I slide the lock key May 12 08:11:13 in nokia's pressing powerbutton turns on backlight May 12 08:11:14 the unit is already locked, so nothing happens. But autlock detect the key and disables itself May 12 08:11:42 usefull to check the hour without unlocking May 12 08:11:50 you should use flashlight to turn nit on! May 12 08:11:57 wait.. I'm lost May 12 08:12:07 rst38h: only n810 May 12 08:12:07 I have to think better about the scenario May 12 08:12:14 n800 doesn't have light sensor May 12 08:12:15 :) May 12 08:12:36 but there is an scenario in which all fails :-) May 12 08:13:16 I mean, autolock is disabled while the tablet is unlocked, and it is enabled while the tablet is locked May 12 08:13:43 the latter is problematic, because, even if the user has locked the tablet manually, autolock will disable it when it detect light May 12 08:13:51 (will unlock) May 12 08:14:25 oh May 12 08:16:55 jdiaz, there's a similar misfeature on my N70... when the camera lid opens, the keyboard gets unlocked, and the phone is free to do whatever stupid thing the self-pressing keys in the pocket happen to be doing May 12 08:17:15 such as calling people in my address book May 12 08:18:04 and sending them photos? May 12 08:18:05 :> May 12 08:20:02 KotCzarny, possiblyu May 12 08:20:20 I've read about AGPS support in diablo.. How can it be possible? May 12 08:20:40 because it's like winmodem May 12 08:20:42 :) May 12 08:20:58 I'd like the gps driver to store the last known almanach and feed it to the gps device... would make more sense May 12 08:21:39 so not to be expected better fix times? May 12 08:21:56 and kismet to run properly on the n810 ;) May 12 08:22:01 :-) May 12 08:22:17 So what was the URL for the v2.6.25 patches for the N800..? May 12 08:26:51 back May 12 08:35:19 kismet would be nice May 12 08:35:50 i miss kismet May 12 08:36:45 omg May 12 08:36:55 Btrfs sure does eat a lot of cpu on mkfs May 12 08:37:01 :) May 12 08:37:22 prob. calculating csums May 12 08:37:23 serial tablet rapist May 12 08:37:52 I guess this should be an official title given by public vote May 12 08:38:43 to whom? May 12 08:39:13 well, they should know ;) May 12 08:40:38 "I didn't rape her, she wanted it... and said she was 18" May 12 08:40:40 :P May 12 08:40:45 lol May 12 08:40:52 i have read: she was 18" May 12 08:40:54 ;) May 12 08:41:07 most nits are younger than that May 12 08:41:09 midget r4p3 ? May 12 08:41:29 KotCzarny: you perv :P May 12 08:42:35 ehh.. this btrfs doesn't look like finishing May 12 08:42:59 broken? May 12 08:43:02 reminds me of a joke about midgets at a reception in vatican May 12 08:43:21 In the Mütter Museum in Philadelphia, there is a skeleton of a man who had giganticism, next to the skeleton of the dwarf prostitute he impregnated, who died giving birth to his child. May 12 08:43:36 :) May 12 08:43:38 link? May 12 08:43:42 KotCzarny: seems so May 12 08:43:52 AStorm: Greetings May 12 08:43:54 need a non-broken one May 12 08:43:54 Link? Visit Philadelphia. May 12 08:43:57 ah, those pensylvanians... May 12 08:43:59 care to help? May 12 08:44:07 Where where you keeping your patches at..? May 12 08:44:18 nicotine patches May 12 08:44:19 :> May 12 08:44:25 hide them! May 12 08:44:48 ugh, s/where/were even May 12 08:44:49 :-) May 12 08:44:52 nab: my minor patch? right now, on my hdd May 12 08:44:54 (its late where on the west coast May 12 08:44:55 :-) May 12 08:45:12 AStorm: Are you commiting to a git tree..? May 12 08:45:29 yes, but not the public omap one yes May 12 08:45:38 have to send the patch to ML May 12 08:45:52 Can you CC me a copy..? May 12 08:46:08 KotCzarny: could you build testing btrfs-tools for me? May 12 08:46:16 hrm? May 12 08:46:18 or was it btrfs-utils May 12 08:46:51 you have sdk working, haven't you? May 12 08:46:55 :) May 12 08:47:18 Has the btrfs-utils package been debianized..? May 12 08:47:19 not yet, soon May 12 08:47:26 nab: dunno May 12 08:48:22 * nab notes that the userspace stuff for OCFS2 has debian/ May 12 08:48:35 Actually, they apparently have a much better website than the last time I looked: http://www.collphyphil.org/mutter.asp May 12 08:48:49 As in, it exists. May 12 08:48:52 I never did try getting ocfs2console (the gtk UI) up on maemo May 12 08:49:24 KotCzarny: http://btrfs.wiki.kernel.org/index.php/Main_Page May 12 08:49:29 Which is actually pretty useful as it copies /etc/ocfs2/cluster.conf to all of the nodes for a configuration change. May 12 08:49:30 grab the git version May 12 08:50:14 huh May 12 08:50:27 TCP connection to 'packages.debian.org' failed: Operation now in progress. May 12 08:50:38 no, not the debian v. May 12 08:50:41 it's oooold May 12 08:50:57 but easier to get/build, probably May 12 08:51:08 heh May 12 08:51:09 i don't remember git syntax May 12 08:51:12 sorry to ask but why would you want to use btrfs on flash? May 12 08:51:13 :) May 12 08:51:15 but impossible to get working May 12 08:51:22 KotCzarny: not git, hg May 12 08:51:25 mercurial May 12 08:51:30 even worse May 12 08:51:30 :) May 12 08:51:42 because then i don't have utils to get it May 12 08:51:43 :) May 12 08:52:02 install them May 12 08:52:08 the filesystem being very beta and all May 12 08:52:09 only require python May 12 08:52:11 i'm sleepy May 12 08:52:14 RST38h: yes May 12 08:52:28 alpha even May 12 08:52:40 but it's much faster on all flash May 12 08:52:43 but why? May 12 08:52:47 with specific tunings May 12 08:52:51 ok so, since i've been gone for like two+ months... what's the coolest new thing available? :P May 12 08:53:05 rm_you: numpty physics May 12 08:53:10 rm_you: n810 wimax May 12 08:53:11 :P May 12 08:53:14 lol May 12 08:53:19 KotCzarny: neat :P May 12 08:53:26 AStorm: pfft May 12 08:53:36 ah May 12 08:53:54 so... numpty physics. what can we do with that? :P May 12 08:54:00 fun May 12 08:54:02 of course May 12 08:54:11 :P May 12 08:54:12 and brain stimulation May 12 08:54:14 :) May 12 08:54:15 :P May 12 08:54:22 isnt data transfer time prevails over file search time on flash? May 12 08:54:30 s/stimulation/explosion/ May 12 08:54:43 especially if you cache metadata in ram? May 12 08:54:46 RST38h: not exactly, esp. when writing May 12 08:54:52 oh May 12 08:55:20 you need to optimize, so that only one erase block is touched May 12 08:55:32 or as little of them as possible May 12 08:55:49 just mount /var as tmpfs May 12 08:55:51 :) May 12 08:55:57 ... May 12 08:56:00 ;) May 12 08:56:01 and the sd controller on the cad also does wearout management? May 12 08:56:11 RST38h: it does May 12 08:56:21 that's why the optimization is neede May 12 08:56:23 d May 12 08:56:51 otherwise the fs could be wear-levelled directly May 12 08:57:10 and there would be no problem (or much less) with erase blocks May 12 08:57:20 ext* rape the first erase block May 12 08:58:07 also, copy-on-write semantics fit flash well May 12 08:58:15 reducing wear May 12 09:03:19 yawn May 12 09:03:59 KotCzarny: build it, now!!!11 May 12 09:04:11 We need to get some of the oftc.net #linuxfs folks for that question May 12 09:04:12 :-) May 12 09:04:13 but i don't even have the sources May 12 09:04:15 :) May 12 09:04:21 KotCzarny: so get them May 12 09:04:21 Or linux-mm folks May 12 09:04:27 how? May 12 09:04:33 you can grab a tarball of current state May 12 09:04:37 from hgweb May 12 09:04:38 :-) May 12 09:04:46 you want me to do the work, provide me with a link May 12 09:05:06 service start maemo-link-bot May 12 09:05:27 KotCzarny: http://www.kernel.org/hg/index.cgi/btrfs/progs-unstable/archive/tip.tar.bz2 May 12 09:05:27 Starting Maemo-Link_Bot.. [ OK } May 12 09:05:30 :P May 12 09:05:47 nab: what's that? May 12 09:06:14 AStorm: I think its one of those IRC cut/paste web services May 12 09:06:15 :-) May 12 09:06:26 Web v3.0 May 12 09:06:32 Or whatever Cisco calls it May 12 09:06:32 :-) May 12 09:06:33 Web xxx? May 12 09:06:35 :P May 12 09:06:36 hehehe May 12 09:06:39 * nab is just kidding May 12 09:06:55 AStorm, wasn't my build good 'nuff? May 12 09:07:05 inz: hanged May 12 09:07:09 :P May 12 09:07:11 oh May 12 09:07:24 It's not my fault May 12 09:07:28 yep May 12 09:07:28 I just compiled it May 12 09:07:38 and it prob. isn't btrfs unstable May 12 09:07:45 AStorm: May 12 09:07:53 Its moving devel very quickly May 12 09:07:57 astorm: you don't care about -Werror, do you? May 12 09:07:57 So metadata is changing, etc. May 12 09:08:06 KotCzarny: I don't May 12 09:08:11 as long as it works May 12 09:08:14 But from my experience Oracle folks are pretty good for backwards compat May 12 09:08:17 AStorm, I can build you a new deb of the userspace tools, if you want May 12 09:08:21 extent-tree.c:2357: warning: 'bit' might be used uninitialized in this function May 12 09:08:23 :) May 12 09:08:23 Espically if their customers starting using it in production. May 12 09:08:36 KotCzarny: but pastebin the log if possible May 12 09:08:42 what has uuid/uuid.h ? May 12 09:08:49 e2fsprogs? May 12 09:08:52 yes May 12 09:08:55 hrm May 12 09:08:57 or ss May 12 09:08:58 eg: I can mkfs.ocfs2 on OS2008 for max-comapt for the particular struct block_device May 12 09:09:11 Or for performance with newer versions May 12 09:09:34 ok, bbiaw May 12 09:10:08 bath is one place where n810 can't go May 12 09:10:09 http://iou.parisc-linux.org/lsf2008/ May 12 09:10:13 not waterproof May 12 09:10:21 For any Maemo folks interested in LSF topics May 12 09:10:22 :-) May 12 09:10:32 astorm: plastic bag May 12 09:10:45 double if you are afraid May 12 09:10:47 KotCzarny: and use keboard with it? lol May 12 09:10:51 * May 12 09:10:57 *keyboard May 12 09:11:01 touch screen ! May 12 09:11:09 RST38h: even worse May 12 09:11:39 hmm May 12 09:11:41 again May 12 09:11:48 which package has uuid.h ? May 12 09:11:49 :) May 12 09:11:53 KotCzarny: libss May 12 09:12:00 AKA ss May 12 09:12:04 apt-cache search libss May 12 09:12:04 libss2 - command-line interface parsing library May 12 09:12:11 or e2fsproge-dev May 12 09:12:28 uuid-dev May 12 09:12:34 thx May 12 09:12:45 but not found May 12 09:12:53 (in default repos) May 12 09:13:10 oh, uuid-dev, lol@debian May 12 09:14:00 you don't mind package from debian.org? May 12 09:14:01 ;) May 12 09:14:28 hmm. packages.debian.org seems dead May 12 09:14:29 AStorm, the mcr head makes gcc go booboo May 12 09:15:35 Does Quim Gil ever hang out in #maemo..? May 12 09:15:40 sometimes May 12 09:15:42 * nab pings Quim May 12 09:15:44 nab, rarely nowadays May 12 09:15:53 inz: Thanks May 12 09:16:12 I just updated him on the LIO project status via email May 12 09:16:50 So both Maemo/iSCSI and OpenMoko/iSCSI have the same short-term goals May 12 09:16:55 AStorm, http://pastebin.com/d242d27b0 May 12 09:17:26 1) Provide an LIO-Target on Linux/Mac/Win on VMware/KVM/Qemu for export of locally acttached physical and virtual storage May 12 09:17:49 2) Begin to scale the maemo intrastructure to accessing volumes of 100 GBs May 12 09:17:53 inz: what's the compiler? May 12 09:18:17 For both cluster and non-shared storage FSes... May 12 09:18:29 1) Is done May 12 09:18:32 Any Navit devs on here? May 12 09:18:35 AStorm, CodeSorcery "raped" gcc 3.4.4 May 12 09:18:47 AStorm, i.e. the default (and only) compiler for maemo May 12 09:18:47 For #2, there is some userspace work still to be done May 12 09:19:00 inz: wrong May 12 09:19:07 And LIO-devs need WiMax devel hardware for the multi-city shared storage clusters. May 12 09:19:16 :-) May 12 09:19:23 inz: any arm compiler will do May 12 09:19:29 *armel May 12 09:19:32 :) May 12 09:19:45 * KotCzarny has 4.something as a target too May 12 09:19:47 actually, i686-none-linux-gnueabi May 12 09:19:58 doesn't OE build a compiler? May 12 09:20:05 to add another one to the list May 12 09:20:20 I built a cross one May 12 09:20:32 and have one here too May 12 09:20:42 could build that small a package here on n May 12 09:21:03 unlikely without uuid May 12 09:21:12 gimme :P May 12 09:21:25 and e2fsprogs has crazy deps May 12 09:21:28 (as usual) May 12 09:21:33 KotCzarny: Heh May 12 09:21:34 * nab nods May 12 09:21:51 like texi2html May 12 09:21:54 Large configuration space userspace tool May 12 09:21:55 :-) May 12 09:22:25 Which has to run on mmuless SuperH May 12 09:22:36 KotCzarny: I sense mad debian packaging May 12 09:23:11 * KotCzarny tries cheating May 12 09:23:21 Or on the arch/powerpc/cell/ or or arch/s390 May 12 09:23:22 :-) May 12 09:23:45 But anyways May 12 09:23:47 KotCzarny: no, you need libuuid May 12 09:24:01 AStorm: So you are doing your v2.6.25 patches with the CS ARM toolkit..? May 12 09:24:03 i have copied uuid.h from e2fsprogs source May 12 09:24:04 ;) May 12 09:24:16 volumes.c:1211: internal compiler error: in reload_cse_simplify_operands, at postreload.c:391 May 12 09:24:17 lol May 12 09:24:30 Please submit a full bug report, May 12 09:24:30 with preprocessed source if appropriate. May 12 09:24:30 Send email to arm-gnu@codesourcery.com for instructions. May 12 09:24:34 double lol May 12 09:24:53 * nab notes that kernel modules generally need the same minor version (if not the EXACT) for kernels vs. modules May 12 09:24:55 what are you trying to compile? May 12 09:25:05 let's try with gcc 4.x May 12 09:25:09 As this is code that gets loaded into a live kernel memory map. May 12 09:25:20 nab: while compiling? May 12 09:25:21 :) May 12 09:25:31 And is expected to operate as such with the other live code. May 12 09:25:36 kernel modules? which ones? May 12 09:25:41 ESPICALLY for the embedded stuff May 12 09:25:45 johnx: btrfs May 12 09:25:50 eg: non x86 May 12 09:25:50 AStorm, http://inz.fi/btrfs-progs_0.14+mcr20080512-0inz1_armel.deb May 12 09:26:05 inz: :) May 12 09:26:10 let's see May 12 09:26:17 got a fitting module? May 12 09:26:25 What, you want those too May 12 09:26:38 could build one myself though May 12 09:27:07 I replaced a couple of *offset = 0 with memset(offset, 0, sizeof(*offset)) and it compiled May 12 09:27:22 :) May 12 09:27:32 good May 12 09:27:36 pointer code magicians should be shot May 12 09:27:37 should be the same May 12 09:28:01 :) May 12 09:28:03 Awwww May 12 09:28:11 What about function pointer magicians..? May 12 09:28:11 :-) May 12 09:28:20 Or CPP madmen May 12 09:28:20 AStorm, yeah, they should be, but apparently the gcc has some problems with 64 bit integers May 12 09:28:21 :-) May 12 09:28:26 nab: subset of above May 12 09:28:26 :) May 12 09:28:45 At least its portable (to another linux-gcc platform) May 12 09:28:46 :-) May 12 09:28:55 or is it? May 12 09:28:57 ;) May 12 09:28:59 Well May 12 09:29:04 With a little more CPP glue May 12 09:29:05 Sure May 12 09:29:08 Anything is portable May 12 09:29:15 :-) May 12 09:29:18 hrm May 12 09:29:29 how about eniac? May 12 09:29:31 :) May 12 09:29:45 Well, better than a bunch of non-inlined functions with 8 parameters May 12 09:30:00 seems to have hanged at the same place May 12 09:30:05 according to strace May 12 09:30:13 * nab will not start talking about shared memory interconnect APIs May 12 09:30:46 or it's writing May 12 09:30:52 how can I see it is? May 12 09:31:09 :) May 12 09:31:11 Although, making IP based shared memory run over Maemo is a long-term project May 12 09:31:11 :-) May 12 09:31:19 astorm: wanna hack the mmc module? May 12 09:31:20 :) May 12 09:31:28 AStorm, do you remember what was the problematic symbol last time? May 12 09:31:29 KotCzarny: not yet May 12 09:31:35 Or rather, Direct Data Placement of incoming RX packets into application buffers using the iWARP protocol suite. May 12 09:31:44 nab: swap-over-nfs, good times May 12 09:31:45 :> May 12 09:31:49 * nab runs May 12 09:31:54 inz: yes, hanged after 5th pread after open /dev/mmcblk0p1 May 12 09:32:15 s/5th/6th/ May 12 09:32:15 AStorm meant: inz: yes, hanged after 6th pread after open /dev/mmcblk0p1 May 12 09:32:20 The extra layers for file vs. block is most apparent on the client side May 12 09:32:33 bbiaw May 12 09:32:39 extra layer overhead that is.. May 12 09:33:02 actually it's simple to trick kernel into enabling swap using loop :) May 12 09:33:13 over any possible medium May 12 09:33:13 :) May 12 09:33:21 The general rule is that a iSCSI export will scale 2x on the target side compared to initiator using identical hardware May 12 09:34:11 Not to mention that you want to use flags for swap block over network I/O to ensure that userspace network I/O will fail before swap over network does. May 12 09:34:22 s/flags/socket flags May 12 09:34:29 :) May 12 09:34:44 Or some other magic if you are poking struct sk_buff from struct net_device directly May 12 09:34:45 :-) May 12 09:34:52 just buy more memory May 12 09:34:53 ;) May 12 09:35:05 On the RX side datapath for traditional (RFC-3720) iSCSI May 12 09:35:26 On the TX side sendpage() for struct page from struct scatterlist is used. May 12 09:36:10 anyway May 12 09:36:13 it's late May 12 09:36:16 AStorm, did you build yourself an updated kernel module? May 12 09:36:16 night folks May 12 09:36:17 :) May 12 09:36:47 * nab waves May 12 10:13:13 inz: not yet May 12 10:13:41 bbl May 12 10:29:25 <_berto_> solmumaha: I can't write you private msgs May 12 10:29:26 <_berto_> :) May 12 10:29:42 <_berto_> I have to register my nick May 12 10:29:45 <_berto_> wait ... May 12 10:36:12 So what's the deal with this leaked Diablo image then? May 12 10:36:32 it's a leaked diablo pre-release May 12 10:36:35 * lardman goes away for the weekend and misses everything! ;) May 12 10:36:44 I almost missed it too May 12 10:36:45 which has been pulled now? May 12 10:36:52 I guess Nokia password protected the repo today May 12 10:37:04 ah, ok May 12 10:37:17 I saw some mention of AGPS functionality in the release :) May 12 10:37:37 the stuff is in place for it, but that will come in a post-diablo "SSU" May 12 10:37:47 SSU? May 12 10:38:00 "seamless software upgrade" aka update .deb May 12 10:38:04 ah, ok May 12 10:38:29 just trying out Nokia marketing's new acronyms May 12 10:38:54 I saw a log with some component complaing about not being able to access a SUPL server; I was just wondering if we'd be able to provide our own... May 12 10:39:41 * johnx doesn't even know what SUPL is... May 12 10:40:46 Secure UserPlane Location May 12 10:41:23 so basically providing our own AGPS data? May 12 10:41:31 yep May 12 10:41:33 maybe? May 12 10:41:43 Otherwise I suppose it will have to come from an attached mobile phone May 12 10:43:59 then again, it may be quite hard to supply our own data any other way May 12 10:44:56 most important diablo upgrade: modest just works May 12 10:50:47 hi May 12 10:50:57 hello May 12 11:05:11 My bluetooth GPS keeps dropping the connection on OS2008 on N800 - can't find bug reports - anyone know if this is a know problem - it was fine under OS2007 Bora. May 12 11:07:10 Yeah, I got that yesterday, wasn't sure if it was the GPS falling out or the N810 dropping the BT connection and the GPS then going to sleep May 12 11:23:51 http://maemo.mmapps.net/n810_vs_tomtom.jpg May 12 11:24:05 the apparently did not agree on the best possible way May 12 11:24:37 lol May 12 11:24:39 fight! May 12 11:24:50 at least it's not a one-way street, right? May 12 11:25:08 no May 12 11:26:04 rule number one , only listen to one navigation device at the time :p May 12 11:41:14 Do any video players actually support VaAPI? May 12 11:41:24 Or is there no test code yet? May 12 12:04:52 so... Anyone know a way of switching to a minimised non-hildon app? May 12 12:05:05 or do I have to kill it and restart it? May 12 12:05:22 xgalaga? =D May 12 12:05:34 nah, cgoban as it happens :-) May 12 12:06:10 Ah, ok. No ideas though May 12 12:06:28 uhm...I thought I saw some app for that May 12 12:06:30 let me look May 12 12:08:23 here: http://www.internettablettalk.com/forums/showthread.php?p=163431 May 12 12:10:16 thanks! May 12 12:10:20 *tries it* May 12 12:22:20 Doesn't work. Ah well - thanks anyway May 12 12:23:53 it should install wmctrl, you can run that by hand I suppose May 12 13:14:23 whois rst38x May 12 13:24:25 do you think, you can all the things with a FreeRunner with OpenMoko, you can do with a N800 with maemo? May 12 13:24:59 right now? sometime in the future? what kind of things? May 12 13:25:28 the next months.. May 12 13:25:34 probably not May 12 13:25:45 right now, openmoko is unstable and not much software ported May 12 13:26:37 it sounds like you answered your own question :) May 12 13:27:17 hm. display seams nearly the same, wlan both, same mhz cpu... May 12 13:27:27 it's all about the software May 12 13:28:28 both has a no or very ugly cam inside ;) May 12 13:28:54 I'm just glad that my N800's cam hides so I don't have to think about it May 12 13:29:27 hm. how would you people track down battery drains? i can't seem to find anything in htop, but my n800 drains battery pretty badly in idle (but on wlan, no traffic there) May 12 13:30:04 ps auxc and look start killing things that aren't necessary May 12 13:30:19 Stskeeps, offline mode and display black May 12 13:30:23 do you have an ext2/3 formatted sd card with symlinks? May 12 13:30:35 johnx: yes, and metacrawler is disabled May 12 13:31:18 rss reader? some kind of persistent mail check? flash open in browser? a misbehaving statusbar applet May 12 13:31:50 does someone has the same problem: extern sd card formated with ext2, i have to take the sd card out and in again for remounting, after reboot. after reboot, the card is corruped first.. May 12 13:32:38 johnx: installed modest, disabled automatic checks, IMAP, but the new email notification keeps on coming, uninstalled old email ui May 12 13:33:45 i kinda suspect its the email client but haven't found any CPU use to prove it May 12 13:34:09 it barely takes CPU use, just expiring timers that wake it up out of deep sleep May 12 13:35:20 w this is a good song May 12 14:09:49 gah! I am so close to having sound working in debian, but it's driving me insane May 12 14:14:18 johnx, :D May 12 14:14:31 seriously though, this is maddening May 12 14:14:41 alsamixer works and appears to let me set levels May 12 14:14:55 but I think the DSP task isn't actually running May 12 14:20:26 heh May 12 14:20:42 Hit it a few times with a hammer and see if it'll work May 12 14:21:29 problem is I don't even know where the DSP task really lives...what if I accidentally hit init with a hammer or sh? May 12 14:24:15 johnx: how do you mean lives? May 12 14:24:27 the mixer is probably built into the dsp kernel May 12 14:25:06 you should see some output in dmesg to say that the built-in tasks + anything in /var/dsp/dsp_dld.conf has been loaded May 12 14:25:07 It was kind of a lame joke ;) ... and I now know one of my problems May 12 14:25:13 dsp_dld is segfaulting May 12 14:25:29 have you rebuilt it? May 12 14:25:40 there's source for it? May 12 14:25:43 yes May 12 14:25:50 woo! May 12 14:25:59 * lardman tries to remember the name May 12 14:26:05 osso-dsp-* perhaps May 12 14:26:36 http://repository.maemo.org/pool/maemo4.0/free/o/osso-dsp-loader/ May 12 14:26:42 great! May 12 14:27:00 sweetness May 12 14:27:04 see if you can track down the segfult though and let me know where it occurs May 12 14:27:11 right at load May 12 14:27:19 want a strace? it's like 3 lines May 12 14:27:22 how do you call it? May 12 14:27:37 I was trying to call it with -h for help May 12 14:27:50 oh, try "dsp_dld -p" May 12 14:28:02 segfault May 12 14:28:05 hmm May 12 14:28:14 http://pastebin.com/m19f4c58c May 12 14:28:56 ~lart Firefox May 12 14:28:57 * infobot whips Firefox with a wet and grimy noodle just because May 12 14:29:51 hmm, exciting strace output May 12 14:29:55 yeah May 12 14:30:00 I told you :P May 12 14:30:29 :) May 12 14:31:33 lolcar smiles apon you May 12 14:32:52 I don't know if that's dsp_dld not handling well the case where it's not running and you don't pass it the flags it's after? May 12 14:33:39 it should be called with -c and -m at least I guess, to give it the avs_kernelcfg.cmd file location and dsp_dld.conf file locations May 12 14:33:58 I'm basing this on the dsp init.d script for mamona May 12 14:34:08 Ah, ok May 12 14:34:21 has anyone ever seen this runtime error: GLIB CRITICAL ** GLib-GObject - gtype.c2242: initialization assertion failed, use IA__g_type_init() prior to this function May 12 14:36:46 on my N810 I see the init script calls "dsp_ctl kmem_reserve 360000", then "dsp_dld -p --disable-restart -c /lib/dsp/dsp_dld_avs.conf" May 12 14:37:24 haha! May 12 14:37:34 I can't remember but something else runs with --disable-restart. I killed it and guess what? :p May 12 14:37:38 shows I shouldn't borrow mamona's dsp_dld :) May 12 14:37:48 :) May 12 14:37:51 bbiam May 12 14:37:53 qwerty12, dsme probably restarts it May 12 14:38:18 johnx, I know but come on though, --disable-restart and it still gets restarted :p May 12 14:38:36 yeah...os2008 is not your typical linux distro May 12 14:39:45 I just wish it was more open :/. The diablo kernel must be using new .config and maybe some patches to the modules c files (or whatever they are called) May 12 14:40:03 isn't it in svn? May 12 14:40:49 I have no idea, I don't think it's being shared afaik. It's still 21 like the Chinook one but with a newer compile date. Makes me wonder what changed... May 12 14:42:01 It's not in stage. My guess is that it will probably be in diablo scratchbox repo when diablo gets officially released :/ May 12 14:42:11 yeah May 12 14:42:28 that would be fine in keeping with the letter of the GPL...but it is annoying May 12 14:43:41 did you mirror all of the diablo repo? May 12 14:43:57 you might have an "rx-34-kernel-source.deb" lying around... May 12 14:44:04 I'll check :/ May 12 14:44:40 Nay, only kernel-diablo-flasher_2.6.21-200817maemo2_all May 12 14:46:05 Funny thing is that Application Manager doesn't ask for a password and log is fine so I guess I am still in for SSU ;) May 12 14:46:16 I don't think dsme does restart it; if you manage to crash dsp_dld, the whole system goes down sharpish unless you can get it up and running very quickly May 12 14:51:01 qwert: Make sure you save a copy of ftd somewhere safe ;) May 12 14:51:19 'Cause something tells me we will not see it in the official 4.1 release May 12 14:51:33 I got diablo repo backed up on 8gb sd and computer :) May 12 14:52:37 meh May 12 14:52:41 no more updates though :P May 12 14:52:52 Just ask Quim about the kernel tomorrow May 12 14:53:31 Navi, my application manager still updates :/ May 12 14:53:39 O_o May 12 14:53:58 we'll have to see on Friday May 12 14:54:06 Friday? May 12 14:54:17 According to Mara, updates are done on Friday May 12 14:54:24 Mmk May 12 14:54:56 We have to kill all of the evil spies here first May 12 14:58:20 KILL KILL DESTORY May 12 14:58:29 DESTROY, mhm May 12 14:59:32 Get a hitman May 12 15:00:06 No, doing it with one's bare hand is much more fun! =BE May 12 15:00:24 I still rate a glock over hands :p May 12 15:00:58 I can't mirror the diablo repo but if anyone wants certain files from the repo, hit me up :p May 12 15:01:10 immobilize with glock, then use hands... May 12 15:01:23 MUHAHAHAhAHAHAHAh May 12 15:01:43 RIM just announced $150m for funding creative mobile applications for RIM May 12 15:02:00 I.e. they aped Steve Jobs with his "iPhone SDK" May 12 15:02:15 who aped google with their android contest May 12 15:02:17 Blackberries suck. I'd rather have iPhone over blackberry >.> May 12 15:02:31 And I don't like iphone ;) May 12 15:02:35 Last time I checked RIM SDK, it was a Java-only thing and it sucked royally May 12 15:03:03 + RIM's Java VM is so slow that pretty much the only app it can run is Arkanoid May 12 15:03:21 Has anything changed since then? May 12 15:03:23 Ok, I use loop device to emulate a image as a dev file right (loop0, I'd guess?) May 12 15:03:32 RST38h: it sucks that bad? May 12 15:03:37 qwerty12, jffs2? May 12 15:03:41 johnx, yes May 12 15:03:55 just found the answer earlier today :) May 12 15:04:01 RST38h: nokias j2me vm is fast enough for pretty nice fullscreen demo effects and shit May 12 15:04:12 qwerty12, http://maemo.org/community/wiki/modifyingrootimage/ May 12 15:04:18 Cool, I kinda know how but all the images I've made have been damaged :) May 12 15:04:31 Ah, thanks, I saw that page before and didn't realise it properly :) May 12 15:04:47 RST38h: i've only checked rims sdk enough to read the apis to get a grasp whats possible.. the apis seemed nice enough May 12 15:05:23 * qwerty12 needs to update linuxrc file from initfs ;) May 12 15:07:02 johnx, they don't mention it there but mtd-tools is good to have for mkfs.jffs2 to remake the image May 12 15:07:10 it's in standard ubuntu repo afaik May 12 15:07:29 * johnx is poking at debian right now, but I'll keep it in mind May 12 15:07:57 :) May 12 15:08:35 I need to compile mtd-tools for tablet, afaik, fanoush has already done it for his flasher's but I believe they are not full set of tools(?) May 12 15:14:10 * Navi demands debian right now May 12 15:14:27 bah May 12 15:14:45 Navi, I thought you were hater :p May 12 15:14:49 * Navi threatens to shoot johnx's loved one/family/friends/pet May 12 15:14:57 O_o May 12 15:15:02 ok, back of the line May 12 15:15:10 All right May 12 15:15:20 * Navi steps back and threatens to shoot qwerty12 instead May 12 15:15:32 Go on then, try :p. May 12 15:15:35 * qwerty12 pulls out glock May 12 15:15:52 >_> May 12 15:16:02 hi cmarcelo May 12 15:17:45 You got blanked :P. Now you got a reason to shoot :p May 12 15:18:15 * Navi goes ten minutes into the past and pretends that never happens May 12 15:18:44 hello. :-) May 12 15:19:00 Navi did not see that :p May 12 15:19:03 He is in past. May 12 15:19:06 Is the diablo microb in the repos any newer than Chinook's? May 12 15:19:20 Navi, yes but not as new as your svn version May 12 15:19:25 Yeah, I noticed May 12 15:19:30 I posted screenshot of about: May 12 15:20:02 browserd is a nice touch though May 12 15:20:12 what's it do? May 12 15:20:16 if one of my windows causes it to crash, only that window closes May 12 15:20:24 s/window/page May 12 15:20:35 mmk May 12 15:20:48 If your dcc works, I can send you microb debs May 12 15:21:21 Nah May 12 15:21:28 skeen May 12 15:21:28 Modest didn't work either May 12 15:21:35 and I really need email May 12 15:21:40 so I reflashed May 12 15:21:47 My Modest works :/. Just a bit unstable later on May 12 15:21:56 I don't use email though so no worries to me May 12 15:21:59 Mine didn't work at all May 12 15:22:05 modest seems perfectly fine here May 12 15:22:10 I got internal errors just by trying to launch it May 12 15:22:18 retarded shtick May 12 15:22:19 I'm sticking with diablo I think May 12 15:22:23 never was able to open it May 12 15:22:41 johnx, +1 :). Just because it's newer :p May 12 15:23:14 they fixed two of my "favorite" microb bugs and some other stuff May 12 15:23:50 machine-id should be created by default so I can run dbus apps as root minus hildonisations May 12 15:23:54 HOW ARE YA FRIENDS? :D May 12 15:24:08 pretty good May 12 15:24:13 and you? May 12 15:24:22 tnx me 2 May 12 15:25:37 I'm lazy May 12 15:25:45 I need to go up to the uni and check my grades May 12 15:25:49 but I don't feel like it May 12 15:27:08 I have a 1MB initfs.jffs2 file. doesn't sound right to me but **** it, I'm gonna flash. I have chinook initfs extracted May 12 15:28:15 Nay, it hangs on startup. May 12 15:29:06 At least it doesn't shutdown :/ May 12 15:29:15 Woo May 12 15:29:35 qwerty12 any news from max payne? May 12 15:29:45 Italodance, Its never gonna happen. May 12 15:29:57 hola May 12 15:30:02 hi May 12 15:30:26 qwerty12 what a pity :'''( but i have hope May 12 15:33:23 arg, I have no way to tell if this initfs image is damaged or diablo initfs is dodgy by default :/ May 12 15:33:45 get fb_update_mode installed May 12 15:34:02 Yeah, I'm gonna make it insert the modules as the first thing. May 12 15:34:35 johnx, y'know May 12 15:34:42 someone please tell me i need to install this https://maemo.org/downloads/product/OS2008/usbhost/ for external devices or don't need it? May 12 15:35:21 if you get sound working, it would nearly make debian replace OS2008 for me :o May 12 15:35:38 well that's the idea now isn't it? :D May 12 15:35:49 Italodance: or you can install this: http://maemo.org/downloads/product/OS2008/usbcontrol/ May 12 15:35:59 +1 for usbcontrol May 12 15:36:23 johnx, do a rundown test yet? May 12 15:36:28 nope May 12 15:36:31 usbcontrol is better, it shows status and has a gui. That just runs a script afaik May 12 15:36:32 mmk May 12 15:36:36 guys which one is better? May 12 15:36:37 I'll let someone else do it May 12 15:36:41 oh May 12 15:36:43 volume of xmms is low than media palyer May 12 15:36:53 johnx, KotCzarny will do it :D May 12 15:37:00 exactly my thoughts May 12 15:37:10 I want to get this out and encourage people to pitch in May 12 15:37:19 I don't want to pitch in May 12 15:37:28 I'm going to be selfish and horde everything to myself May 12 15:37:32 uhm...thanks May 12 15:37:36 Kidding May 12 15:37:37 :P May 12 15:37:41 ummm external HDD and cd-rom can works with USB Control? May 12 15:37:41 ah :P May 12 15:37:50 or usbhost?? May 12 15:37:51 I'll be handing out custom dwm builds for anyone that wants it May 12 15:38:16 Italodance, external HDDs will work as long as they are FAT16 or FAT32 May 12 15:38:20 is there a other way to raise a volume in xmms May 12 15:38:40 for cd-roms you will need to do install something else May 12 15:39:41 penguinbait made a cd-rom deb. I don't really see the point in hooking up a cd-rom drive to it though :/ May 12 15:41:56 johnx yes i installed cdrom.deb for my n800 so for external HDDs and flash drivers,cool disks....it can read all them without any new install app? May 12 15:42:33 you'll need to install ntfs support if your disk is formatted as ntfs May 12 15:42:51 no for FAT ? May 12 15:42:54 yey ntfs failz May 12 15:43:01 Italodance, Why don't you try it and tell me if it works? May 12 15:43:49 You'll need a powered hub for HDDs May 12 15:44:22 johnx ok sure but i didn't buy usb otg adaptor yet,,,i will go for buy now May 12 15:44:33 And some old flash drives, for whatever reason May 12 15:50:43 How to increase size of mtdblock0 file? I run mknod /tmp/mtdblock0 b 31 0 and then dd if=usr/share/osso/images/initfs.jffs2 of=/tmp/mtdblock0 and "dd: writing to `/tmp/mtdblock0': No space left on device" May 12 15:50:58 eh? May 12 15:51:11 are you using block2mtd or mtdram? May 12 15:51:18 I think you're mixing your instructions May 12 15:51:22 I think both :/. I'm gonna reboot May 12 15:51:34 wait, is this on the tablet? May 12 15:51:38 ah May 12 15:51:39 nm May 12 15:51:44 computer. I'm doing reisub. brb May 12 15:51:54 you don't need to reboot May 12 15:52:10 just rmmod mtdram May 12 15:52:15 and follow the right instructions May 12 15:55:09 sound works May 12 15:55:12 next challenge May 12 15:55:45 what are people's favorite lightweight audio player? May 12 15:55:50 emphasis on lightweight May 12 15:56:04 mocp May 12 15:56:11 well, that's my second favorite May 12 15:56:55 true favorite is mpd + ncmpc May 12 15:57:01 I'm leaning towards mpd May 12 15:57:07 maybe mpd+sonata though May 12 15:57:13 sonata is heavy May 12 15:57:20 it's not *that* heavy May 12 15:57:25 and it doesn't need to stay running May 12 15:57:40 Run it on the N800 long enough to use, and you'll see why I called it heavy. May 12 15:57:52 sounds fair May 12 15:58:08 I've run it on a system with half the RAM and slower swap before though May 12 15:58:13 it took 5-10 seconds for the UI to update with song changes May 12 15:58:19 O_o May 12 15:58:25 It's processor heavy May 12 15:58:28 that sounds like something is just wrong then May 12 15:58:34 I'll give it a shot May 12 15:58:39 Hi dougt May 12 15:58:48 gmpc runs much better May 12 15:58:55 but it's not as nice May 12 15:58:58 being nice and Glade+C andall May 12 15:59:16 the pythonness of sonata bogs everything down May 12 15:59:50 I'll give you that Sonata is better, though May 12 15:59:53 oh, gmpc looks lots nicer now May 12 15:59:55 s/better/nicer/ May 12 15:59:56 Navi meant: I'll give you that Sonata is nicer, though May 12 16:00:01 it's been a long time since I've used it May 12 16:00:05 yep May 12 16:00:15 UI changed completely from .13 to .14 May 12 16:00:16 I was picturing gmpc circa .11 or earlier May 12 16:00:34 they rewrote the glade parts at that time May 12 16:00:39 spiffed it up May 12 16:00:46 and since the tap and hold is more reliable than under os2008 ... *ducks* May 12 16:01:00 :/ May 12 16:01:41 w00t!. I messed up permssions on my home folder so I'm now root for the time being (bad idea but I don't really care) May 12 16:02:02 chown -R user:user ~user/ May 12 16:02:21 uhm...unless you keep unpacked filesystems under it May 12 16:02:32 Not really, thanks :) May 12 16:02:49 heh May 12 16:03:16 that command would really ruin my week if run on my $HOME May 12 16:03:22 let me log out and see :) May 12 16:03:23 hehe May 12 16:03:37 anyways May 12 16:03:51 the tiny buttons on gmpc are hard to hit on the N8X0 May 12 16:04:11 when I was trying it out, I had to up the size a bit May 12 16:04:12 I need to look into gtk theming at some point for zaurus and n800 May 12 16:04:25 Mmm May 12 16:04:41 just 150% would make things nice for a lot of apps May 12 16:04:48 nah, still failing. I'm gonna take a look later May 12 16:04:49 but I don't even know what's possible :/ May 12 16:05:06 heh May 12 16:05:06 fail May 12 16:07:57 I think for diablo initfs, it would be wise to increase totalsize number :/ May 12 16:08:00 audacious looses for not having libmad plugin included in its deb package May 12 16:08:16 qwerty12, you're using the wrong instructions May 12 16:08:25 use the ones that talk about block2mtd May 12 16:08:40 I tried them before and it failed for me, I want to try the other method May 12 16:08:46 it will never work May 12 16:08:54 I'm past caring May 12 16:09:06 I can help you make the block2mtd instructions work May 12 16:09:09 I just used them May 12 16:09:32 I can mount the image fine with block2mtd but my created image is odd :/ May 12 16:09:48 define odd May 12 16:10:02 anyways, the problem with mtdram is it needs a huge chunk of kernel memory May 12 16:10:12 Its 1MB. Smaller than chinook initfs which is wrong. May 12 16:10:50 maybe make your changes then rsync them into a new image? May 12 16:10:59 or maybe the diablo image isn't actually working yet May 12 16:11:20 johnx, maybe. I'm gonna add fb_update_mode to find out. May 12 16:11:30 when I rolled out linux thin clients perfect booting was like the last thing I had working May 12 16:13:55 yes!mtdblock method is brilliant. Expect new initfs soon hopefully ;) May 12 16:14:11 :) May 12 16:14:28 the block2mtd or did you actually get mtdram working? May 12 16:14:29 I pimped up the total_size to 88888 :p May 12 16:14:34 mtdram May 12 16:14:37 heh May 12 16:14:49 how much RAM do you have on that machine? May 12 16:14:52 Diablo initfs size is bigger so it needs bigger size, I had incomplete dump before May 12 16:14:53 2GB May 12 16:15:01 that would explain it :) May 12 16:15:15 only a measely gig here May 12 16:15:21 :P May 12 16:15:41 wait...you needed 88MB for the initfs?! May 12 16:15:49 eh? May 12 16:15:52 No but every little helps :P May 12 16:15:57 I chose random numbers May 12 16:16:25 I gave bad advice earlier actually...what mtdram won't work for is rootfs May 12 16:16:39 can I ignore "tar: .: implausibly old time stamp 1970-01-01 01:00:00 May 12 16:16:39 " message? May 12 16:16:42 yes May 12 16:16:50 Cool May 12 16:19:43 Now I got 1.8MB which is much more better :).Time to flash May 12 16:20:16 YEAH BABY May 12 16:20:21 it lives? May 12 16:20:25 * qwerty12 is running DIABLO initfs ;) May 12 16:20:32 awesome May 12 16:20:38 Time to try out FTD May 12 16:20:47 damn nokia for making me go to such lengths May 12 16:20:55 and thank you for provding the wiki :p May 12 16:21:11 even if I later told you it wouldn't work :) May 12 16:21:28 I guess it balances out or something :/ May 12 16:22:03 jott was wrong :P. Even with Diablo initfs, it still won't retrieve BME values :p May 12 16:22:09 Is testserver needed? May 12 16:22:33 I don't know May 12 16:23:07 I used fanoush remove.lst file to rip it out. Test server goes :/. I'll put back in the battery modules and see what happens :) May 12 16:32:19 * Jaffa had a play with Diablo at JavaOne (an N810W on the Nokia stand), was not immediately impressed. May 12 16:33:16 it's a bugfix release May 12 16:33:22 Agreed. May 12 16:33:32 and it seems marginally more responsive to me at least May 12 16:33:45 Feature wise, I'm not really excited but the rest of it seems to be non-buggy. May 12 16:33:49 apt-get upgrades and modest are the big "features" May 12 16:35:25 + there may be a way to feed agps data now May 12 16:35:44 the lack of a2dp is extremely disappointing May 12 16:35:50 Agreed. May 12 16:35:54 Yep May 12 16:36:05 and the lack of any word one way or the other May 12 16:36:16 if I'd known it wasn't coming I might have worked on it more May 12 16:37:08 I need jott around :( May 12 16:38:21 johnx: bug fix release indeed May 12 16:38:40 ? May 12 16:38:49 running into problems? May 12 16:38:56 Dissapointed by lack of new additions? May 12 16:39:40 I was just disappointed that the microb engine in the repos was old. May 12 16:39:57 yeah...that's not so awesome May 12 16:39:57 and that modest did not work at all May 12 16:40:07 so me May 12 16:40:09 meh May 12 16:40:11 modest is fine here...and for many other people May 12 16:40:16 I assume it's an upgrade issue May 12 16:40:18 Navi, I don't really agree with modest May 12 16:40:24 Me and johnx ran it fine May 12 16:40:28 I upgraded from a clean flash May 12 16:40:29 I even posted screenshot May 12 16:40:41 which email provider? May 12 16:40:51 gmail and my school's May 12 16:41:06 But I wasn't using them anyways :P May 12 16:41:10 I *couldn't* use them May 12 16:41:16 modest wouldn't *open* at all for me May 12 16:41:29 It'd go straight to internal error May 12 16:41:57 * johnx shrugs May 12 16:42:19 * qwerty12 shoulda leans May 12 16:44:00 johnx: re a2dp, we had a confirmation on the list last week that no DSP task was going to be created on Nokia's side May 12 16:44:16 Nokia sucks. May 12 16:44:35 qwerty12, we wouldn't be here if they were completely awful though May 12 16:44:41 If the bluez changes are as good as people say, perhaps it won't be a big issue May 12 16:45:07 but they might or might not in some future release maybe do a2dp, possibly? May 12 16:45:12 johnx, I know but the big things are things we will never see :/ :( May 12 16:45:31 my impression was that as bluez is better now (don't know what version of bluez), it's not worth it May 12 16:45:52 but I've started a Garage project anyway, any help appreciated of course: dsp-sbc May 12 16:46:00 but they still have to do all kinds of crazy routing to do May 12 16:46:12 to make it work with their media-player May 12 16:46:24 the question is when they'll get around to it May 12 16:46:45 heh May 12 16:47:23 I can't see why adding a GStreamer alsa sink won't work May 12 16:48:08 bit of conf file tweakery and the built in player should then output over a2dp; I assume May 12 16:48:50 for some reason the Nokia guy responding to the a2dp bug made it sound more complicated May 12 16:48:59 but maybe things have improved since then in modularity May 12 16:49:17 I don't know; perhaps it is more complicated May 12 16:49:47 something about the fact that system sounds were mixed on the dsp or somesuch May 12 16:50:13 ah, yes that might cause issues, as the DSP is bypassed May 12 16:51:41 But nothing to worry about for us; we just won't hear anything but the music/soundtrack May 12 16:52:19 right, but nokia has some weird ideas about what constitutes a blocking bug May 12 16:52:42 "Can't here the screen-tap sound?! then we can't ship it!" May 12 16:52:51 :) May 12 16:53:56 I wouldn't ship it if the screen-tap was meant to come here ;) :p May 12 16:58:52 omweather locked my box! May 12 16:58:54 * Mousey cries May 12 16:59:15 that's unfortunate May 12 16:59:21 how do you know it was omweather? May 12 17:00:22 because there's a big blank window that says "Other Maemo Weather Settings" open everytime i boot it and otherwise the ui doesn't respond May 12 17:00:54 do you have an ssh server installed? are you booting from sd? May 12 17:00:59 yes, no May 12 17:01:11 well the fix is easy at least May 12 17:01:17 can't ssh in May 12 17:01:22 ah May 12 17:01:31 well that makes it somewhat harder May 12 17:01:35 * Mousey cries more May 12 17:01:44 * lardman heads for home May 12 17:01:47 do you need information off of it? May 12 17:01:57 it'd be nice not to have to wipe it May 12 17:02:46 let me think about it for a couple minutes May 12 17:03:10 I have one way but it involves some amount of command line hackery May 12 17:04:28 johnx: that, i imagine means access to a command line May 12 17:04:33 but the entire ui is unresponsive May 12 17:04:41 well see that's where the hackery comes in May 12 17:05:06 you flash a fanoush custom initfs with builtin telnet and then use usbnet from your desktop May 12 17:05:08 well i've plugged it into usb, maybe i can crowbar omweather out May 12 17:06:25 umm\ May 12 17:06:27 http://www.tablet-guru.com/2007/11/15/a-few-backgrounds-for-internet-tablets/ May 12 17:06:29 :D May 12 17:06:33 wotcher lcuk May 12 17:06:43 hey lcuk May 12 17:06:51 is there a way to transfer files to the IT using a FTP program such as filezilla? May 12 17:06:51 hi folks May 12 17:07:12 Cptnodegard, solmumaha ported proftpd May 12 17:07:14 Cptnodegard, run openssh-server on your tablet and use anything that understands sftp May 12 17:07:34 thx May 12 17:07:35 * Cptnodegard runs off to check May 12 17:07:58 http://www.tablet-guru.com/wp-content/uploads/2007/11/nokia_house.jpg is it really the official nokia house? May 12 17:08:11 oh wait! it is responding on ssh!! May 12 17:08:13 huzzah! May 12 17:08:18 Yes, although I wouldn't call it a house... May 12 17:08:34 Cptnodegard, openssh is on downloads.maemo.org and in extras :) May 12 17:08:46 Italodance, http://en.wikipedia.org/wiki/Image:Nokia_HQ.jpg May 12 17:09:01 qwerty12 tnx dude :D May 12 17:09:15 * qwerty12 thinks Finland looks nice, I have to visit when I'm older :) May 12 17:09:16 np :) May 12 17:09:29 mpd lives! May 12 17:09:34 qwerty12 lol how old are you? May 12 17:09:41 15 :) May 12 17:09:59 ohhh :) May 12 17:14:59 is the openssh installation supposed to take forever? it asked for "new root password", i typed one in and now it just says "installing openssh" with the progressbar not moving May 12 17:15:32 yup May 12 17:15:43 it's generating you a key May 12 17:15:44 :) May 12 17:15:47 ok it finished May 12 17:16:05 ah...it's not supposed to finish :P It's supposed to go on forevvvvaaaa May 12 17:16:07 never used anythign ssh before so im naturally scared of it :p May 12 17:16:16 Cleaner.2007.DVDRip.XviD-VoMiT nice :D May 12 17:16:35 Cptnodegard, ssh is like the most awesome thing ever. nothing to fear :) May 12 17:16:48 but i still have no idea how to use it :p May 12 17:16:56 well, now it's running May 12 17:16:59 no icon appeared in the menu May 12 17:17:07 right, it's a server May 12 17:17:08 johnx, do you use ssh (in diablo repo) or openssh from chinook repo? May 12 17:17:11 Hahaha May 12 17:17:20 * GeneralAntilles beats Cptnodegard with the ignorance stick May 12 17:17:25 hey May 12 17:17:26 qwerty12, chinook I think May 12 17:17:27 im trying :p May 12 17:17:32 :P May 12 17:17:38 GeneralAntilles, cut him some slack :P May 12 17:17:46 so now i do what exactly to access the memory from filezilla? May 12 17:17:48 No quarter! May 12 17:18:14 Cptnodegard, anyways, now that you have ssh user filezilla and connect to your tablet's IP address with user root and the password you chose May 12 17:18:36 local ip or global? May 12 17:18:41 local May 12 17:18:50 if it's inside your own network... May 12 17:19:52 I'm 80% sure that the 1st Nokia image is kept in the bootloader May 12 17:19:52 specific port? May 12 17:20:11 filezilla should know May 12 17:20:15 as long as you pick sftp May 12 17:20:21 or ssh2 or whatever filezilla calls it May 12 17:20:56 * johnx looks at filezilla screenshots May 12 17:21:00 just tell it port 22 I guess May 12 17:22:06 lol nice May 12 17:22:18 quickconnect didnt work because it does ftp only May 12 17:22:21 thx :D May 12 17:22:28 qwerty12: thought it was /mnt/initfs/usr/share/images/logo-nokia May 12 17:22:29 ? May 12 17:22:45 I just replaced it and Nokia one still shows May 12 17:22:54 Also with damaged initfs, Nokia one still shows May 12 17:22:59 interesting May 12 17:23:55 Nokia's got a half-dozen of them sequestered away like a squirrel hides acorns. . . . May 12 17:24:15 I'm not sure if I made image right though :p. I just did chroot and ran the program which displays image and it segfaults May 12 17:24:16 :D May 12 17:24:28 I don't think it's meant to handle 580Kb images though... May 12 17:24:43 it prolly has to be exactly the same resolution too May 12 17:25:00 Which is very small for the Nokia logo. May 12 17:25:00 Nokia default resolution suxs badly May 12 17:25:35 Not sure on the exact size but the resolution is the just the nokia logo. May 12 17:26:35 At least I can try and make it black. Uncompress nokia original one, change white to black and make t2s set black background. I did that with black themed bootmenu. May 12 17:26:52 so this ssh server thingy runs constantly now? i dont have to start anything after rebooting for instance? May 12 17:27:03 nope May 12 17:27:08 comes up on reboot May 12 17:27:12 ^^ May 12 17:27:16 thats actually useful May 12 17:27:23 it's incredibly useful May 12 17:27:35 i wouldnt mind usb if it wasnt for the fact that f***** tablet keeps yapping on about "blabla is in use" every time :p May 12 17:27:47 and you can access the tablet's command line from your computer, too :) May 12 17:29:13 It's incredibly easy to use flash-and-reboot to flash whatever initfs you want. May 12 17:29:41 and now we're back to the days of onboard flashing! all that's old is new again! May 12 17:30:12 * qwerty12 still needs to compile mtd-utils May 12 17:32:07 Is it possible to set hardware and software revision manually using flasher-3.0 or 0xffff? May 12 17:32:46 once again, my information has corrupted and about is May 12 17:38:17 qwert12, yes May 12 17:38:28 I don't know the exact command, though. May 12 17:38:36 Really, I wouldn't worry about it. :P May 12 17:44:34 buttmonkeywindows -__- May 12 17:44:59 Cptnodegard: you talking about my current OS? May 12 17:45:18 technically mine but im sure it would fit yours too :) May 12 17:45:31 so is Windows ButtMonkey edition better than Vista? better than XP? May 12 17:45:49 nearly, mine is actually buttmonkeywindowsxphome May 12 17:46:07 home? i feel sorry for you... May 12 17:46:10 actually I've gone mad, I booted to Ubuntu May 12 17:46:22 XP = xpert problems.... Vista is so bugged not even the abbrivation makes sense :) May 12 17:46:26 im waiting for buttmonkeywindowsxphome service pack 4 to be released before i switch. May 12 17:46:27 qwerty12_N800: cam pre-installed, couldn't be bothered to wipe it May 12 17:46:35 Ah, that's special May 12 17:46:44 They wont add 4.1 to bugzilla because it's "not out yet" May 12 17:46:49 hey General, how are you? May 12 17:46:55 Yet another reason for Nokia to avoid using it. . . . May 12 17:47:00 hey May 12 17:47:03 ah, i'm lucky, mine came with mce May 12 17:47:04 Fine, lardman, yourself? :) May 12 17:47:19 I'm trying to compile a gps hello-world app and I'm getting a compilation error May 12 17:47:20 GeneralAntilles: not bad, about to do some DSP hacking/cursing :) May 12 17:47:23 this thing would pwn with the IT: http://www.advancedmp3players.co.uk/shop/Accessories.7/Speakers.16/ibluon.260/BT-TD03/iBluon_Bluetooth_Stereo_Audio_Transmitter.2441.html May 12 17:47:34 MajMin7, did you get your gps test working? lol obviously not May 12 17:47:48 in gps.h: error: 'PATH_MAX' undeclared here May 12 17:48:03 lcuk: im getting CLOSE May 12 17:48:36 i think i have the program right... and I've had it compile May 12 17:48:38 its ok, once you get the code working you still have to wait 4 hours to run it and get a lock May 12 17:48:43 hahaha May 12 17:48:47 MajMin7: what are you testing? May 12 17:49:01 im just trying to get a program that uses the GPS May 12 17:49:03 hes trying to see if hes on earth May 12 17:49:09 :) May 12 17:49:09 lardman, it's all gonna come to you in one glorious epiphany one day, and DSP enlightenment will be yours. :D May 12 17:50:12 MajMin7, put the whole error log into pastebin and let us have a look May 12 17:50:34 arr May 12 17:50:39 ok May 12 17:50:51 GeneralAntilles: I live in hope :) May 12 17:51:06 lard, is the blitter going to be ready for use May 12 17:51:17 what is pastebin? May 12 17:51:41 http://pastebin.com/ paste the errors into there May 12 17:51:51 and then submit and give us the link it gives May 12 17:52:01 lcuk: I'm trying to work out why it outputs some strange size of image; probably some school-boy error in my pointer arithmetic May 12 17:52:29 i get that ;) ive got a line draw algo thats off by one pixel May 12 17:53:29 Need me to spot for you while you adjust it, lcuk? :P May 12 17:53:33 we will have to do some poking about later in the month and see whats around May 12 17:53:45 "A little to the left—too far, a little back to the right. . . ." May 12 17:54:04 ok May 12 17:54:06 lol GeneralAntilles, ive done all that, and tbh its not important enough to worry about now May 12 17:54:19 just pasted it: http://pastebin.com/m6f57e6bf May 12 17:54:42 hmm, definitely something wrong with my code May 12 17:58:05 MajMin7, ok, before your call to all the gps code, you need to include May 12 17:58:30 04:14.57 VoodooZ__ actually. adding #include did the trick as per http://www.busybox.net/lists/busybox/2001-November/005206.html May 12 17:58:49 apparantly (and ive never seen gpsd before) gpsd included a hard written path in there.. May 12 17:58:55 length May 12 17:59:56 oh May 12 18:01:44 MajMin7, when you are getting errors like this for unknown libraries its an idea to try googling the exact error line, or the key points of the error line. hope that works anyway and you can move onto the next step May 12 18:02:58 yea May 12 18:03:02 i actually did that May 12 18:03:08 just didnt find what you found May 12 18:03:10 dunno why not May 12 18:03:22 thanks tho May 12 18:03:46 thats ok, enough people have helped me in here that im more than happy to try to return the favour :) May 12 18:04:34 help? more like save my ass May 12 18:04:36 lol May 12 18:06:17 lardman, have you seen ftd from diablo repos? the information given may not be new/important but it has "dsp info". screenie: http://i30.tinypic.com/2lu2zuw.png & http://i32.tinypic.com/2j607ci.png :) May 12 18:13:35 ok May 12 18:13:42 so i cheated... May 12 18:13:52 I added to the gps.h file May 12 18:14:02 but i shouldn't do that........ cause its not my header file May 12 18:14:12 that fixes the problem though May 12 18:15:38 just include it before including gps.h May 12 18:15:42 no need to edit gps.h May 12 18:16:00 oh crap May 12 18:16:27 i overlooked the #include in my program (its just something i copied) May 12 18:16:32 qwerty12_N800: I hadn't; am looking at the pics now May 12 18:17:17 qwerty12_N800: Looks like the sysfs info; what does ftd stand for? May 12 18:17:32 field test display May 12 18:17:38 from diablo repo May 12 18:24:31 Any framebuffer gurus about? May 12 18:24:49 scan line go along rows or columns? May 12 18:25:23 scan line is from east to west May 12 18:25:26 qwerty12_N800: hey how goes May 12 18:25:37 EruditeHermit, hi :) May 12 18:26:05 looks like the password protected it May 12 18:26:09 no more updates for us May 12 18:26:27 RX-48 (wimax) is probably gonna have more security... May 12 18:26:29 :( May 12 18:26:52 isn't security a good thing? May 12 18:26:57 or arbitrary restrictions? May 12 18:27:25 lardman, horizontal strips left to right with a stride between each line May 12 18:27:28 not when you like to look around... bb5 is taking a higher role in 810w May 12 18:27:45 echo "Loading OMAP HS Secure Mode Driver module. May 12 18:28:25 qwerty12_N800: you got the new initfs to work? May 12 18:28:50 yes, i dont have a copy on tablet tho May 12 18:28:58 or mtd-tools to dump a copy May 12 18:29:30 ah May 12 18:30:38 nothing really good about it though, except for being newer. it just has stuff on the 810w edition May 12 18:31:12 I only worked on it because I thought battery info in ftd would work May 12 18:31:49 how do you start ftd? May 12 18:32:04 or access it May 12 18:32:28 from the application menu May 12 18:33:10 do you have to install it from somewhere? May 12 18:33:32 yes, the now gone diablo repo May 12 18:33:37 lol May 12 18:34:15 you /may/ be able to get it from app manager in redpill May 12 18:35:34 nope May 12 18:38:46 heh, i just realised the new 810w edition is going to 810w chunks May 12 18:38:58 :P May 12 18:42:29 * qwerty12_N800 may get a 810w, it means initfs would be better (it runs the programmes May 12 18:42:38 on tty1 ;) May 12 18:43:50 wimax module is called pc2400m and the network interface called wimax0 May 12 18:44:36 !proxy May 12 18:53:56 what is spi2.1? in rd_mode, n810w will show extra green text ;) May 12 18:54:20 spi is a bus May 12 18:54:38 I believe it's how the sd card controller is hooked up as well May 12 18:55:06 entirely possible that wimax is attached to spi I suppose May 12 18:55:06 ah, thanks :) May 12 18:56:20 wikipedia knows all: http://en.wikipedia.org/wiki/Serial_Peripheral_Interface_Bus#Advantages May 12 18:57:10 * johnx sleeps May 12 18:57:34 night May 12 19:00:20 johnx's timezone is a lie. May 12 19:00:56 lol, it's only 8:00PM here, gmt +1 May 12 19:01:06 3pm May 12 19:01:22 yawn May 12 19:03:20 to a certain person who likes cpufreq, ondemand is always set at boottime via initfs... May 12 19:03:49 also it's set while compiling kernel May 12 19:04:08 initfs would override it? May 12 19:04:21 could May 12 19:05:34 qwerty12_N800, ive put changes into 2 places: im using a modified acmonitor which switches it to whatever ive configured depending on charge status, and ive got it as a configurable option in liqbase May 12 19:06:02 and sleeping on "powersave" allows me to wake up with a full battery every morning May 12 19:06:07 lcuk, maek ur code suk less n itll werk. ;) :P May 12 19:06:09 cool :). I have to copy your modification soon :p May 12 19:07:02 http://pastebin.com/m5cb3a255 copy it now, its nothing special May 12 19:07:26 thanks :) May 12 19:13:30 ok saving a multiline file. am i wise to check the result of each and every fprintf for an error or does most code just assume if it opens in writable mode the data will get written? May 12 19:13:56 you can always run out of space May 12 19:13:57 :) May 12 19:14:25 yer i know, but syntax for that will lose the flow of the code May 12 19:14:50 im just wondering how other code works, i think i need to do a hunt May 12 19:14:51 just make an addable return code then check after the loop May 12 19:15:32 qwerty: Any luck with appletizing Conky? =) May 12 19:15:53 RST38h, My sbox is broken :( May 12 19:15:59 again? May 12 19:16:00 lol May 12 19:16:10 lol May 12 19:16:20 I killed permissions on my linux account :/ May 12 19:16:28 do backups May 12 19:16:30 :) May 12 19:16:34 i'm loggin in as root May 12 19:17:05 chown -R user:users /home/faheem dont do jack all too May 12 19:17:27 maybe mecause it's kept somewhere else May 12 19:17:33 like in /scratchbox/somewhere May 12 19:17:35 :) May 12 19:18:17 I mean on my main user, not sbox :) May 12 19:18:43 ok, google says nobody bothers to check.. May 12 19:19:21 qwerty: You are still using sbox1? May 12 19:19:26 Umgh May 12 19:19:33 * KotCzarny uses sbox1 May 12 19:19:37 works very fine May 12 19:19:40 :) May 12 19:19:56 RST38h, yes, h8 sb2 May 12 19:20:24 eugh :P scratchbox May 12 19:20:33 eugh :P sb2 May 12 19:20:37 but having a scratchbox set up for faheem is useless if i can't log in as faheem May 12 19:20:43 w0000t \o/ on device compilation May 12 19:20:51 meh May 12 19:21:12 :) May 12 19:21:20 only thing better is no arm emulation :p May 12 19:21:25 having said that im still in vmware May 12 19:22:24 qwerty12_N800: you can fix your homedir by running chown again May 12 19:22:45 * RST38h loves sb2 and hates sb1 May 12 19:22:51 ugly contraption May 12 19:23:05 lcdd, i kept on trying chown :). i'm just gonna do chmod 777 * ... May 12 19:24:24 just mv the whole dir and do mkdir /home/faheem May 12 19:24:30 then do chown on it May 12 19:24:39 and start copying files one by one May 12 19:24:39 :) May 12 19:24:51 yea, may have to, thanks :) May 12 19:27:49 Hit it with a stick May 12 19:30:27 qwerty12_N800: i think you should trace back to what your first mistake was before making any new ones.. May 12 19:31:10 qwerty: remove / reinstall = May 12 19:31:14 namely not owning your $HOME :) May 12 19:31:23 using su and trying to modify an initfs, i didn't touch permissions directly :) May 12 19:31:35 always use su - May 12 19:32:00 damn May 12 19:32:04 thanks :/ May 12 19:32:20 :) May 12 19:32:49 lcdd, thanks for the suggestions, i'll look into everything May 12 19:33:03 lcuk: thanks May 12 19:33:47 * RST38h somehow feels like giving the only useful advice =) May 12 19:34:18 rst: i did that before May 12 19:34:19 :) May 12 19:34:35 heh May 12 19:35:06 * qwerty12_N800 congratulates nokia for putting up empty nokia diablo extras May 12 20:01:53 quick q, how would i find a certain line of code in all *.c files in a directory tree? May 12 20:02:15 find . -name *.c | grep something May 12 20:02:18 grep can do but not sure how May 12 20:02:22 with zsh: grep foo **/*.c May 12 20:02:27 i use scite for that May 12 20:02:31 (if your tree is small enough) May 12 20:02:33 erm, xargs grep something May 12 20:02:43 thanks :D May 12 20:02:55 probable xargs -e May 12 20:02:58 i knew it had something to do with grep, but didnt know how to start it off May 12 20:03:00 but just use mc May 12 20:03:06 it has find string in files May 12 20:03:11 recursively May 12 20:03:12 :) May 12 20:03:31 and allows you to edit directly May 12 20:04:02 i prefer => find . -name \*.c -exec grep -il {} \; May 12 20:04:06 s/il/i/ May 12 20:04:06 gnuSnob meant: i prefer => find . -name \*.c -exec grep -i {} \; May 12 20:04:11 i use a little filesearch i wrote years ago normally May 12 20:04:59 and the winner of simplest solution goes to Paavo_ :D May 12 20:05:05 cos it works May 12 20:05:09 Hello! Anybody remembers which package should be installed in order to have curses terminal description files installed? May 12 20:05:13 What is a CNAME record? May 12 20:05:25 an entry in DNS May 12 20:05:28 alias May 12 20:05:30 zap, ncurses-base May 12 20:05:31 probably May 12 20:05:49 I'm signing up for google apps, and part of the verification process is that I must create a cname record of googles choosing May 12 20:05:51 yes, curses-base, TY May 12 20:08:37 gnuSnob: http://en.wikipedia.org/wiki/CNAME May 12 20:09:06 Sorry, http://en.wikipedia.org/wiki/List_of_DNS_record_types might help a bit more May 12 20:09:25 thanks May 12 20:10:08 it looks like it's identical to what I thought a hostname was May 12 20:14:34 Hostname is just what it says, a hostname May 12 20:15:31 http://www.internettablettalk.com/forums/showthread.php?t=20005 ohhhhhhh May 12 20:16:55 ahhh! no wait, wrong gender... May 12 20:20:29 i want fifa 08 for my IT May 12 20:20:49 just get a psp May 12 20:20:53 +1 May 12 20:21:37 KotCzarny, you voiced the exact words i was thinking :p May 12 20:21:42 :) May 12 20:21:53 lol May 12 20:22:56 I have a psp, dont really use it anymore. the hb is nothing appealing to me like the apps on maemo are May 12 20:23:27 forge, if I have the hostname blah.domain.net, and I also have the CNAME blah.domain.net pointing to something other than the host "blah", then what happens when blah.domain.net is referenced? May 12 20:23:57 b0rkage May 12 20:24:13 Hostnames are followed by .domainname.com - the same convention used by CNAME May 12 20:29:26 GeneralAntilles, you still have the black "themed" initfs_flasher.tar? I want to see if diablo initfs has enough space left over to flash it... :) May 12 20:29:50 kulve: you about? May 12 20:31:36 kulve: not to worry, found it - poll_disable(task); May 12 20:33:05 ah forget it, i just realised linuxrc has too much chhanges. i'll hack up the bootmenu myself for diablo initfs May 12 20:36:28 qwerty12_N800, it's 250 miles north of here. ;) May 12 20:37:16 so, any news on diablo release date? May 12 20:37:30 hehe, no worries. i'm gonna have to get on my computer and start up a file comparing session :) May 12 20:39:55 i'm upset they never fixed fmradio & bluetooth headset bug :( May 12 20:40:28 Unix, but blah.domain.net isn't actually a hostname, blah is the host part and domain.net is the domain part of the whole name May 12 20:40:54 domain.net is a hostname too May 12 20:40:58 or can be May 12 20:41:57 Well yeah May 12 20:46:43 Nokia pulled a sick one with diablo, I have no /etc/hostname o_0 May 12 20:47:09 does it matteR? May 12 20:47:11 :p May 12 20:47:29 just make a S99-local-qwerty and do what you want here May 12 20:47:37 chinnook has it and ssh looks weird :p May 12 20:47:40 (in /etc/rc2.d/) May 12 20:49:22 what are the screen dimensions of the N810? May 12 20:49:31 800x480? May 12 20:49:42 Ya May 12 20:50:20 strange, when I fill a 800x480 image using the framebuffer there's a gap at the bottom May 12 20:50:28 Still waiting for the same concept as 810 but with just a tinytinytiny bit bigger screen and a bit better layout on the keyboard and buttons and more yummy hardware :o May 12 20:50:36 and if I fill 400x240, it certainly doesn't take up half the screen height May 12 20:50:53 lardman: depends on colorspace May 12 20:50:56 :) May 12 20:50:59 oddly as it sounds, making hostname speeded up n800 a bit. hosts still has host name from chinook. placebo? May 12 20:51:01 and actual res set May 12 20:51:20 qwerty: depends on app May 12 20:51:21 KotCzarny: so if I want to use RGB565, for example May 12 20:51:53 that should be 16bit data, so I'm writing 800x480 16bit unsigned shorts to the framebuffer May 12 20:51:57 16bit * res then i think May 12 20:52:14 lardman May 12 20:52:23 lcuk: yep May 12 20:52:24 what mode are you in May 12 20:52:29 rgb565 May 12 20:52:54 are you certain and are you accounting for the stride May 12 20:53:43 hmm, otoh it's interesting if it's x -> fb -> lcd May 12 20:53:49 or another path May 12 20:54:06 lcuk: now that I don't know - stride May 12 20:54:22 KotCzarny: it's DSP -> fb May 12 20:54:40 the stride is not always present, it depends upon the actual setup of your framebuffer May 12 20:54:54 hrm May 12 20:55:02 but from what you have described there is one. is the output still garbled May 12 20:55:04 lardman: hardware fb or emulated one? May 12 20:55:20 lcuk: so that's effectively the gap at the end of each row if you're not filling at the maximum pixel size? May 12 20:55:31 yes May 12 20:55:42 KotCzarny: It's the shared memory framebuffer which is then passed to the LCD controller May 12 20:55:47 its what is used as an interline gap for blitters May 12 20:55:51 uhum May 12 20:56:15 lardman, take a pic of the output (digital camera) May 12 20:56:21 lcuk: so I should check and see if the LCD can do 24bit mode, which I think it can, then add a gap to the end of my lines May 12 20:56:33 and if you have a clean source image that its meant to be a similar pic would be good May 12 20:56:38 lardman, try writing 800x1px red, then 800x1+0+1, green etc May 12 20:56:48 should give a hint of the layout May 12 20:56:48 :) May 12 20:56:48 lcuk: not very exciting, just a block on the screen, correct width, but not height May 12 20:56:58 all one color May 12 20:57:11 yes, as it's supposed to be May 12 20:57:16 i have gotten used to staring at similar things when i was writing my blitter code May 12 20:57:43 ahhh, you need a patterned test - as kot suggests - it will help you immensely May 12 20:58:19 yeah, well my pattern is actually 10 frames of one colour, then 10 of another, then 10 of another May 12 20:58:28 does your entire code run from the dsp or do you have some fb init code May 12 20:58:39 in std arm May 12 20:58:39 but if there's some overlap, it will disappear to fast for me to see I imagine May 12 20:58:55 do one frame then? May 12 20:59:03 lcuk: all on the dsp, on the arm it creates an empty X window so we can see the output May 12 20:59:15 hrm May 12 20:59:18 :D its a whole other world in there May 12 20:59:19 direct rendering? May 12 20:59:20 :) May 12 20:59:21 i am impressed May 12 20:59:37 don't be yet, it's not working as well as I hoped it might May 12 21:00:49 if you can change your algo to output 4 lines normal and 1 whole line black you get minimum pattern May 12 21:01:16 yeah I can do that, am using octave to generate an input file which I then send over May 12 21:01:29 if you are writing too much data to the screen the line will not start/end at the ends, it will spiral May 12 21:01:54 (original goal was to convert & display YV12->YUV420, but have had to take a few steps backwards as only garbage came out) May 12 21:01:56 reminds me of planar fun on ami May 12 21:01:57 :) May 12 21:02:08 i started by creating an incrienting greyscale May 12 21:02:14 ok, hang on a few May 12 21:02:17 KotCzarny, :) i think that every day May 12 21:02:36 lcuk: how about writing memory visualizer? May 12 21:02:37 ;) May 12 21:02:49 should be very simple May 12 21:03:13 wrapped to the screen or a full plane horiz/vertic scrolling May 12 21:03:25 configurable width May 12 21:03:28 with +/- May 12 21:03:41 and movable with dpad May 12 21:03:42 :) May 12 21:03:43 entirely plausable May 12 21:03:57 dpad? feck of - kinetic works fucking wonders May 12 21:04:04 ahm, sure :) May 12 21:04:09 lol May 12 21:04:09 Octave is chugging along May 12 21:04:10 forgot about it for a moment May 12 21:04:23 hi May 12 21:05:30 back soon, my shed is calling May 12 21:05:33 :) May 12 21:05:58 lcuk, hiding fugitive? :p May 12 21:06:03 KotCzarny, one thing, could you just try somethin for me May 12 21:06:16 what happens if you compile liqbase on normal machine? May 12 21:06:19 can it run May 12 21:06:22 back later... May 12 21:06:25 hmm May 12 21:06:32 probably May 12 21:06:36 if you use regular sdl May 12 21:06:47 and not hw/hildon specific things May 12 21:07:03 is the button on the earphones / microphone visible in userspace? May 12 21:07:06 you should ifdef the code May 12 21:07:10 all x, this vmware box wont let me and im not diagnosing now, it was just a side thought May 12 21:07:16 for nokia specific parts May 12 21:07:19 theres nothing specific hildon May 12 21:07:27 touchscreen code? May 12 21:07:27 :) May 12 21:07:31 xsp May 12 21:07:32 fb code? May 12 21:07:37 NULL May 12 21:07:45 then should work May 12 21:07:49 without any changes May 12 21:07:50 it uses xv over x May 12 21:08:01 assuming your x can do 800x480 mode May 12 21:08:02 could you just see while im codin :) May 12 21:08:02 etc May 12 21:08:08 ok May 12 21:08:09 :) May 12 21:08:16 thx :D May 12 21:08:26 changed pass on ftp? May 12 21:10:59 ok May 12 21:11:00 so May 12 21:11:07 [suddenly waking up] Octave runs on Maemo?!? May 12 21:11:18 rst: why shouldn't it? May 12 21:11:31 I am now able to compile the GPS hellworld code. but it doesnt seem to do anything May 12 21:11:36 has anyone ever used this code before? May 12 21:12:05 no May 12 21:12:06 Kot: no,no, just getting excited about it May 12 21:12:11 lcuk: no Xsp on my laptop May 12 21:12:17 it is os2006 only though May 12 21:13:05 MajMin7, no - you baked it yourself from a recipe you found on the net :) May 12 21:13:08 and not available in slack packages May 12 21:13:18 ok kot, ill do it another time May 12 21:13:25 i will commend xsp code May 12 21:13:27 it was just a ponder May 12 21:14:47 hmmm ftp is up and things havent changed May 12 21:14:53 RST38h: yes, but was actually running it on the desktop May 12 21:15:05 530 login or pass incorrect May 12 21:15:23 yer i see, i just pasted it back into the box, try again May 12 21:15:36 ok, worked May 12 21:15:46 0O mixup? May 12 21:15:59 don't know, i just use a bookmark May 12 21:16:02 that worked before May 12 21:16:03 :) May 12 21:16:05 how bizarre May 12 21:16:11 it should still work now May 12 21:16:39 latest is obviously today May 12 21:16:49 lardman: eh... May 12 21:17:10 ok May 12 21:17:32 i hate us date format May 12 21:17:54 http://lifeandhealth.guardian.co.uk/food/story/0,,2279357,00.html May 12 21:17:58 Kot: you are not alone May 12 21:18:00 liqbase-kc_20080511_174524_quickpause.tar.gz May 12 21:18:02 you have to date yourselves in a specific way KotCzarny? May 12 21:18:22 * RST38h constantly finds himself wondering which part is the month May 12 21:18:27 czr, i'm talking about us locales May 12 21:18:28 :) May 12 21:18:35 * czr prefers the chinese date format May 12 21:18:43 It's low in fat, low in food miles and completely free range. In fact, some claim that Sciurus carolinensis - the grey squirrel - is about as ethical a dish as it is possible to serve on a dinner plate! May 12 21:18:46 * lcuk prefers stardate format May 12 21:18:51 british dates for me May 12 21:18:51 i like 2008-05-12 May 12 21:18:55 japanese date format has some basis behind it May 12 21:19:00 easy to sort May 12 21:19:01 KotCzarny, that's the chinese May 12 21:19:13 * RST38h has switched to British locale too May 12 21:19:15 although obviously they use additional characters. May 12 21:19:18 i would love unixtime date format May 12 21:19:25 if i had math skills May 12 21:19:27 :) May 12 21:19:33 KotCzarny, and more than 32-bits ;-) May 12 21:19:41 I have been using Russian one for a while, but then Maemo tries to display days of week and months in Russian- rather ugly May 12 21:19:55 unixtime could easily be converted to 64bit May 12 21:19:56 :) May 12 21:19:57 kot, for the kinetic test, you need to copy liqed.c May 12 21:19:59 Kot: In hex, I hope? May 12 21:20:14 thats the fixed file it uses to view May 12 21:20:28 copy what from where to where May 12 21:20:39 src folder to where your running it May 12 21:20:46 ahm, ok May 12 21:21:44 hmm May 12 21:21:51 also, you depends on fb May 12 21:22:01 or maybe not yet May 12 21:22:02 :) May 12 21:22:44 heh, funny koreans: http://koltsoff.com/pub/lcd_usage.png May 12 21:22:55 (that's from a datasheet I'm reading atm) May 12 21:23:18 hrm May 12 21:23:19 :) May 12 21:23:29 yeah, lcd is poisonous May 12 21:23:30 * qwerty12_N800 smashed up screen on n-gage May 12 21:24:01 yeah but why do they want me to crush them? May 12 21:24:11 i still have the bits in drawer :/ May 12 21:24:15 for easier lcd removal? May 12 21:24:28 KotCzarny, it still doesn't make sense :-) May 12 21:24:35 :) May 12 21:24:41 "it is recommended to crush unnecessary LCDs" May 12 21:24:45 no, they want the liquid out of the crystals May 12 21:24:50 there is probably a method in this madness May 12 21:24:57 context will help though, wheres the rest May 12 21:24:58 ah, lcuk that would explain it May 12 21:25:00 crystals ARE the liquid May 12 21:25:01 :> May 12 21:25:09 that's all that there is to "safety" May 12 21:25:31 a small cell containing something which can expand can also explode May 12 21:25:51 RST38h: talking about Octave; sorry for the delay May 12 21:25:52 under the wrong conditions they could be mini bombs? May 12 21:26:11 dont worry lardman, any light yet? May 12 21:26:14 probably. May 12 21:26:35 was chatting to gf, so not sure :) May 12 21:27:11 ahhh, the mysterious lardwoman May 12 21:27:31 who wants to crush LCDs? Samsung? I know why. May 12 21:27:35 that will certainly get you in trouble :D May 12 21:27:37 :P omg she will kill me May 12 21:27:41 RST38h, yes. May 12 21:27:48 you chose the wrong name May 12 21:27:53 or actually DISPLAYTECH. May 12 21:27:56 'tis true May 12 21:28:21 i think i need to put a bit of a video together May 12 21:28:39 im not gonna get everything i wanted done so will have to sorta stop with what i have May 12 21:28:52 right, so I'm writing one black line when the line number is divisible by 4, and the lines are a bit jagged May 12 21:29:13 ok, look at the first one, does the black stop before the end of the pysical first line May 12 21:29:22 or does it appear to extend over past the end May 12 21:29:31 lardman: photo pls May 12 21:29:34 or just look at one specific line May 12 21:29:36 hmm, hard to tell as I'm writing lots of frames quite quickly May 12 21:29:37 czr: If old LCDs are not destroyed, nobody will buy new LCDs May 12 21:29:38 or screenshot May 12 21:29:44 so for the GPS device, do I need to start it with gpsmgr_start(...) before I can use the device? May 12 21:29:45 RST38h, Har :-) May 12 21:29:47 lardman: write one frame May 12 21:29:47 thus, old LCDs are impediment to commerce May 12 21:29:48 :) May 12 21:30:14 I'd like to use OLEDs actually May 12 21:30:18 but they're too expensive. May 12 21:30:20 OLEDs degrade May 12 21:30:23 so I'm stuck with crappy LCDs. May 12 21:30:24 I know, I know; need to alter the code to not destroy the X window immediately May 12 21:30:42 RST38h, everything degrades! May 12 21:30:44 or just keep it going until you manually stop it May 12 21:31:00 But LCDs with LED backlights are a serious problem - LEDs do not degrade as quickly as old backlight tubes May 12 21:31:07 that requires some sort of X events, wasn't going to bother with those May 12 21:31:29 ok then, repeat frame 0 1000 times or long enough to give you a chance May 12 21:31:29 RST38h, problem for samsung you mean? May 12 21:31:51 lcuk: I know, just can't type that fast :) May 12 21:32:08 :P May 12 21:32:39 czr: yep May 12 21:32:43 how do you run the code you just compiled? are you in an ide with a go button, or do you save and run script May 12 21:32:58 lcuk: me? May 12 21:33:00 yer May 12 21:33:09 RST38h, well, the display I'm looking at atm has no backlight. at all. May 12 21:33:12 I have to compile it for the dsp, then sftp it over, then run it again May 12 21:33:32 lardman: mount your dev dir via sshfs ? :) May 12 21:33:34 though in this case I can just alter the arm side code, so scratchbox + sftp + run over ssh May 12 21:33:36 ill show you the script i use, you may be able to adapt to help you May 12 21:33:58 KotCzarny: I've got a graphical sftp window so works as well May 12 21:34:16 heh May 12 21:34:24 but with sshfs mount you don't have to copy anything May 12 21:34:28 just run as from local dir May 12 21:34:29 :) May 12 21:34:56 + you can write a debug to the same dir directly May 12 21:34:56 :) May 12 21:35:24 ah, I see, well I manage to crash the machine every other go, so not sure mounting it is the best idea May 12 21:35:33 but I will bear it in mind, thanks May 12 21:35:57 http://pastebin.ca/1016276 i just run that, it makes copies executes and forces a killall when i press ctrl+C from my big keyboard May 12 21:35:59 lardman: ssh key with no pass + S99-local in rc May 12 21:36:00 :) May 12 21:40:27 lcuk, pm May 12 21:44:11 right, so I have black lines every 10 lines or so now; the ends are jagged. The curious thing is that it looks like I've not written enough data to each line May 12 21:44:31 i.e. the right hand end of the black lines is one pixel above the rest of the line May 12 21:46:54 is there a screenshot tool for Chinook? May 12 21:47:12 yes but doubful it will work using directfb May 12 21:47:16 t May 12 21:47:43 easiest will be camera if its available May 12 21:47:58 yeah, will see what I can manage May 12 21:51:49 what does openSSH client do? May 12 21:52:37 ~lart Firefox3 for randomly locking up May 12 21:52:37 * infobot judo chops Firefox3 for randomly locking up May 12 21:53:11 http://people.bath.ac.uk/enpsgp/nokia770/dsp/PICT0047.JPG it's pretty big - 1.8Mb May 12 21:53:27 You don't have permission to access /enpsgp/nokia770/dsp/PICT0047.JPG on this server. May 12 21:53:28 :) May 12 21:53:37 hmm, I just got that too May 12 21:53:41 hang on a minute May 12 21:55:23 try now May 12 21:55:34 works :) May 12 21:55:48 it had execute permissions for some reason May 12 21:56:14 Sooooooo May 12 21:56:26 looks like what mplayer does when you press random buttons May 12 21:56:27 hmm May 12 21:56:32 looks like a half res May 12 21:56:34 or something May 12 21:56:35 * qwerty12_N800 thought my w810 camera was bad... :p May 12 21:56:46 It's set to be 400x240 May 12 21:56:57 but screen is set to 800x480 May 12 21:56:57 so that I could see if it overran at the edges May 12 21:56:59 apparently May 12 21:57:03 yes May 12 21:57:17 I'm only writing into the top lefthand corner of the fb May 12 22:02:08 hey May 12 22:02:11 the step position migrates towards the left as new frames are drawn May 12 22:02:17 can anyone tell me why this code doesnt work: May 12 22:02:18 http://pastebin.com/m7333a074 May 12 22:02:28 The signal handler that I register never fires May 12 22:03:33 you've entered your own loop, rather then the gtk event loop? May 12 22:03:45 ahhhhhhh May 12 22:03:46 ok May 12 22:03:52 that sounds plausible May 12 22:03:57 so May 12 22:04:06 is it possible to use the GPS without using GTK? May 12 22:04:18 of course May 12 22:04:34 just without a gui, or at least without a maemo-ized gui May 12 22:04:47 all of the example code uses glib May 12 22:04:59 im not familiar with glib May 12 22:05:03 i know pthreads etc May 12 22:05:34 in that case: s/gtk event loop/glib event loop May 12 22:05:51 what do you want to do? May 12 22:08:38 well May 12 22:08:59 right now im just trying to demonstrate that I can poll for my gps location May 12 22:09:06 and print it May 12 22:09:07 lardman, your initial offset is off, you are either starting -40 bytes too soon (off and above) or you are starting +200, the +200 is less likely because you dont have a bare patch in very top left (both approx) May 12 22:10:10 if your line lengths were off, the black line would spiral, so it feels more like your initial offset May 12 22:10:26 is it your input data? May 12 22:10:46 rm_you, good news about your n800 (mostly :p) :) May 12 22:10:54 lcuk: strange, I'm starting at 0 (i.e. the address of the start of the framebuffer), unless there's something at the start of it? May 12 22:11:10 no, input data should be fine (he says) May 12 22:11:16 try writing smaller chunks May 12 22:11:21 ie. 100px blue May 12 22:11:25 10px black May 12 22:11:27 100px blue May 12 22:11:28 etc May 12 22:11:40 well, 20px black May 12 22:11:40 :) May 12 22:11:53 ok, will tweak the code May 12 22:12:15 or something more dividable May 12 22:12:59 hey, you say it drifts to the left May 12 22:13:43 sleep. May 12 22:13:51 the other confusing thing is that I was trying to make the screen red May 12 22:13:58 lol May 12 22:14:02 writing 0xf800 May 12 22:14:06 that would put a dampner on your day May 12 22:14:26 I only just noticed that was what the Octave code was supposed to be producing :) May 12 22:14:45 how many bytes does octave output per frame May 12 22:14:59 unless I'm being stung by some endianness thing May 12 22:15:11 which I probably am now that I come to think about it May 12 22:15:39 hmm, though 16bits values are supposed to be littleendian May 12 22:15:45 too late in the evening to think that thought May 12 22:15:48 through May 12 22:15:49 are they 2*400*240 and does octave put any kind of header information on them May 12 22:16:14 no, I'm just writing with fwrite as a uint16 value May 12 22:16:34 write it to a file first May 12 22:16:45 and see what's inside May 12 22:16:46 :) May 12 22:17:15 00 F8 ad infinitum May 12 22:17:57 ok May 12 22:18:02 so a different question May 12 22:18:05 first 00 00 starts at offset 0x3840 May 12 22:18:28 the maemo website says that the device which we connect to the GPS has the following members: May 12 22:18:29 * device->online: Whether GPSD is connected to a GPS or not May 12 22:18:29 * device->status: The GPS status May 12 22:18:30 * device->fix: The GPS fix information May 12 22:18:30 * device->satellites_in_view: The number of satellites that the GPS can see May 12 22:18:30 * device->satellites_in_use: The number of satellites used in fix calculation May 12 22:18:31 * device->satellites: An array of satellite details May 12 22:18:48 where can i find this documented formally? May 12 22:18:57 it doesnt tell me the types of these pointers May 12 22:19:18 it's in the gpsd source iirc May 12 22:19:29 iirc? May 12 22:19:44 at least the Nokia patched version (iirc = if I remember correctly) May 12 22:19:54 oh May 12 22:20:06 wow May 12 22:20:10 i can read their source? May 12 22:20:16 KotCzarny: so looks like the file is ok May 12 22:20:17 where's that May 12 22:20:26 one point of failure less then May 12 22:20:33 http://repository.maemo.org/pool/maemo4.0/ May 12 22:20:55 oh May 12 22:21:01 KotCzarny: Will write 10 black, 90 blue/red, repeat May 12 22:21:06 i found this once but it looked too cryptic at the time May 12 22:21:20 try diff color codes May 12 22:21:25 depending on some counter May 12 22:21:37 kind of color cube to expect and see the effect May 12 22:21:41 ;) May 12 22:21:53 i can has colorcube? May 12 22:22:04 MajMin7: When in doubt, or you don't find something, look under o for osso May 12 22:22:10 MajMin7: http://repository.maemo.org/pool/maemo4.0/free/o/osso-gpsd/ May 12 22:22:21 ps, dont try drawing on < 800*480 May 12 22:22:21 ahhhhhh May 12 22:22:30 lcuk: why not? May 12 22:22:31 * KotCzarny throws a rubik's cube for a lcuk May 12 22:23:02 sorry lardman not you, ive got a 256*256 colorcube in liqbase and its a bit frisky about resolution changes May 12 22:23:13 ok :) May 12 22:23:33 your lines remind me of it :) May 12 22:25:58 http://technology.timesonline.co.uk/tol/news/tech_and_web/article3897340.ece something like this would be good for LinuxTag May 12 22:26:10 or indeed in general May 12 22:27:12 i wanted to have a bluetooth radar for linuxtag :( time is against me May 12 22:28:07 i was gonna have an app running on my laptop and on my 810 and have then ping all nearby bts May 12 22:28:26 sounds like it would have been pretty cool :) May 12 22:28:31 using signal strength i was gonna pick them up May 12 22:28:41 yer, time. May 12 22:29:00 at least ill be able to read while im there :) May 12 22:29:07 and draw May 12 22:31:47 Read and draw? laaaammmmmeee May 12 22:32:13 You should be going around with a video camera and asking people what they like to do in their free time May 12 22:33:03 I meant to ask, do we need to buy an entry ticket? May 12 22:33:58 Anyone try out mediabox yet? May 12 22:34:21 The sound keeps resetting to 50% every time a song changes :/ May 12 22:36:17 weird, I did 10 black followed by 90 red/blue repeated 8 times across the screen May 12 22:36:30 pic pls May 12 22:36:40 I then get black lines going down the screen as expected, but the whole lot moves sideways as each new frame is displayed May 12 22:36:44 ok May 12 22:45:03 http://people.bath.ac.uk/enpsgp/nokia770/dsp/PICT0048.JPG May 12 22:45:21 http://people.bath.ac.uk/enpsgp/nokia770/dsp/PICT0049.JPG is better actually May 12 22:45:46 now displaying at 800x480, but note the gap at the bottom of the screen May 12 22:46:03 Ooh May 12 22:46:05 Cool stuff May 12 22:46:05 so maybe you are starting waaay befor screen? May 12 22:46:10 before May 12 22:46:45 try moving starting offset by 800 bytes untill you see black at the top? May 12 22:46:58 worth a go May 12 22:47:22 aargh, bloody Ubuntu taskbar has locked up May 12 22:47:49 maybe the fb is centered on some region May 12 22:48:56 any ideas how to change desktop using the keyboard only? May 12 22:49:09 ctrl-arrow ? May 12 22:49:14 ctrl-alt-arrow ? May 12 22:49:20 alt-arrow ? May 12 22:49:58 good call, ctrl-alt+arrow May 12 22:50:02 switching to console and killing offending app can work too May 12 22:50:13 offending app is the desktop though May 12 22:50:18 acroread hangs everything for me sometimes May 12 22:50:30 but killing it from console is enough May 12 22:50:49 then kill the desktop May 12 22:50:51 :) May 12 22:51:00 ok, so you reckon to change start_x to 800? May 12 22:51:17 buf+offset ? May 12 22:51:21 :) May 12 22:51:30 yep May 12 22:51:50 you are already writing to it directly May 12 22:51:58 then just write at higher position May 12 22:52:47 blit some digits instead of pixels for optical recognition could help too May 12 22:52:51 *blitting May 12 22:52:53 yes, skip the first 10 or 20 lines May 12 22:53:05 or even better May 12 22:53:15 make a code that will react to key presses May 12 22:53:21 lol no xevents May 12 22:53:22 KotCzarny: yeah, I'd need to write some code to get the data for that though May 12 22:53:40 lcuk: but hw keys should be accessible May 12 22:53:41 i guess May 12 22:53:42 :) May 12 22:54:06 as a scancodes from a kernel May 12 22:54:07 etc May 12 22:54:12 Anyway, I'm shattered, time for bed May 12 22:54:29 lardman: :) May 12 22:54:34 at least you know you are writing the correct length lines :) May 12 22:54:35 I'll have another look tomorrow lunch time and let you know what I find; thanks for your help chaps May 12 22:54:39 its progress, good night May 12 22:55:25 i have all this kind of messing to do whenever i change the blitter May 12 22:56:05 strange, closing an xterm took down 4 more + some "explorer" (Galleon?) windows and freed up my desktop May 12 22:56:12 :) May 12 22:56:20 probably started from there May 12 22:56:31 yeah, will do more experimenting tomorrow when I can think straight :) May 12 22:56:32 night May 13 00:09:54 nokia is awesome! May 13 00:12:02 Why? May 13 00:12:17 Just in general, or have they done something noteworthy lately? May 13 00:20:48 well I lost the back battery cover to my N800, and they are going to send me a replacement for free May 13 00:20:57 i was about to drop $21+s/h to buy one May 13 00:41:12 whee that's nice of them May 13 00:41:41 Honestly I was hoping for something revolutionary and amazing, but that's still good to hear. May 13 00:42:54 well that, and they saved me $100 on my car insurance May 13 00:43:02 they saved you $100 on car insurance? May 13 00:43:09 i thought they only could save 15% or more May 13 00:43:21 it's nokia, they can do whatever they want to May 13 00:43:38 except make a decent video chat app for the n800 i guess :-/ May 13 00:44:05 how about a video chat app for the desktop? May 13 00:44:11 nobody i know can see me. hear me, yes, but not see me May 13 00:44:22 and i ask you, ladies and gentlemen, what is the point of video chat, if it only works with other n800s? May 13 00:44:23 skype? May 13 00:44:36 what about it? May 13 00:44:40 nobody i want to talk to uses it May 13 00:45:18 most everyone is on gtalk/jabber May 13 00:48:05 Damn I wish my friends used jabber May 13 00:48:21 Everyone I know uses AIM. *hork* May 13 00:48:59 AIM with its OSCAR "we're going to add random features and change things without sharing the spec" protocol. May 13 00:50:24 While I'm ranting about things, the contrast between the internal link font and the background on the internettablettalk wiki is insufficient. May 13 00:50:49 i never use page colors May 13 00:51:07 i set them to black bg and white fg May 13 00:52:05 Well that works for you, but what about everyone else? I think it's reasonable to ask that the site be designed correctly in the first place. May 13 00:52:16 nah May 13 00:52:21 haha May 13 00:52:31 i just don't want to send email to every html author in the world May 13 00:52:47 Yeah.. May 13 00:52:54 so i just ignore their styles and use mine May 13 00:53:19 some people think light gray on white is nice May 13 00:53:25 but it's completely unreadable May 13 00:53:26 :) May 13 00:54:09 I don't mind slightly off-black on white, but the problem with that is that the contrast varies widely from monitor to monitor. May 13 00:54:21 that too May 13 00:54:26 So what's readable on one monitor might look like ass on another. May 13 00:55:06 I honestly really hate web design and web pages in general. I wish content was provided in a form more like rss. Data is provided for you, and your reader formats it however the hell it wants. May 13 00:55:35 That would solve so many problems. Accessibility and small device support would be provided entirely by the browser. May 13 00:55:39 that's what css is for May 13 00:55:39 :) May 13 00:55:50 Sort of. May 13 00:56:18 Right now, CSS is for making some really awesome designs, and a lot of broken web pages. May 13 00:56:35 Luckily, CSS allows me to fix people's web pages with my own stylesheets. :P May 13 00:56:42 * Grackle makes one for ITT May 13 01:01:53 * gnuSnob wonders where to get cheap pull-out hard drive rails May 13 02:01:39 gnite May 13 02:01:44 nite nite May 13 02:01:51 * lcuk goes to sleep happy **** ENDING LOGGING AT Tue May 13 02:59:56 2008