**** BEGIN LOGGING AT Wed May 25 02:59:57 2011 May 25 04:44:53 ss May 25 04:45:14 Should i enable hardware acceleration hildon desktop? May 25 04:45:32 Hwsync? May 25 05:32:22 hello May 25 05:35:10 can someone help me my after flashing my n900 reboots itself bootreason is sw_rst May 25 05:38:31 sw_rst means "maemo pretends it is smarter than you and decides itself it is time to reboot" May 25 05:40:43 what can i do? May 25 05:42:49 reflash? May 25 05:43:01 can you get to NOLO? May 25 05:43:21 oh, and if you're in a bootloop, remove the battery now May 25 05:43:47 otherwise it'll drain your battery and you'll have to find another way to charge it before reflashing May 25 05:43:54 i flashed it 3 times and this comes every time May 25 05:43:59 hm May 25 05:44:06 what are you flashing? May 25 05:44:17 COMBINED and/or VANILLA? May 25 05:44:20 i'd flash both May 25 05:44:42 make sure you don't reboot the phone (the -R flag) between the two May 25 05:45:22 i flashed without -R both vanilla and combined after after that i removed battery t and wait and this didnt help May 25 05:46:43 what is NOLO May 25 05:47:03 Google? May 25 05:48:02 ok nokia bootloader May 25 05:49:42 i'm stumped then :( May 25 05:49:56 kerio how can i get to nolo May 25 05:50:03 ideally reflashing both should bring the n900 to a pristine state May 25 05:50:12 itschue: are you in a bootloop? May 25 05:50:26 or can you reflash? May 25 05:50:48 are you sure you have the right files for the flashing? May 25 05:50:57 where i can see if iam in bootloop May 25 05:51:46 i typed in my imei for file download May 25 05:56:04 kerio? May 25 05:56:31 itschue: can you get the computer to recognize your n900 in flashing mode? May 25 05:58:47 im on work im only with my n900 online May 25 05:59:19 an have no computer here May 25 06:02:01 im at aprox. 6 pm @home then i will check the .bin files May 25 06:02:37 after 1 flashing this error happens very often May 25 06:04:26 then i flashed both and wait a long time befor rebooting and then shutdown and remove battery and wait turn on wait long time so i restart from yesterday 8pm only one time in night May 25 06:06:10 kerio May 25 06:13:31 someone active ? May 25 06:15:45 some awake May 25 06:17:24 strohhalm german? May 25 06:17:38 a bit May 25 06:17:45 but only idling here May 25 06:18:02 ok you cant help me or? May 25 06:18:56 no, but need to flash my device soon, too :) May 25 06:19:52 and why ? May 25 06:20:25 cause the cam doesn't work anymore May 25 06:21:53 and you know why May 25 06:23:17 not that sure, but i changed to these powerkernel and tried some other -extra/devel stuff May 25 06:44:13 word May 25 06:45:26 hello May 25 06:48:45 hm. i never even tried but does maemo support nfs? May 25 06:49:07 doing it remotely would probably require something seriously out there huh? or some sort of battery sucking thing May 25 06:49:09 not ootb but there's probably addons you can get for it May 25 06:52:07 actually not quite, its not nfs, its cifs May 25 06:54:28 cifs? May 25 06:54:30 heh May 25 06:54:37 that sounds even worse than nfs May 25 06:54:41 but i guess using it remotely would be easier May 25 06:54:51 since i'm assuming smb has some sort of encryption May 25 06:55:02 at the very least user/pw encryption May 25 06:56:35 this rubber cover i got for my n900 totally sucks May 25 06:56:44 the screen protector is good tho May 25 06:56:59 generic brand? May 25 06:57:10 i got 3 of them pretty cheap.. $11 for 3 covers and the flimsy made in china rubber cover May 25 06:57:13 yeah May 25 06:57:16 lol May 25 06:57:34 i guess it works better than nothing does May 25 06:58:02 after how awesome i have gotten this n900 i am considering getting another newer one May 25 06:58:21 maybe try to rsync them heh May 25 06:58:33 one for day to day and one to wear a suit with May 25 06:58:33 I've got otterbox for one of mine, makes it look fatter but despite N900 being dropped a few times it never really suffered anything serious May 25 06:59:11 mine is pretty good too other than a few minor sliding into the pocket scratches on the screen May 25 06:59:28 you would have to really look for them tho May 25 06:59:36 but i still want a good as new one :) May 25 07:00:12 but this rubber cover.. it isnt TOO bad.. but i expected a tighter fit May 25 07:00:21 i mean.. it is latex rubber May 25 07:00:28 it'll be hard to find new ones these days unless you go to shops and hope they still have them May 25 07:00:42 Macer, they're cheap too :) May 25 07:00:59 psycho_oreos: amazon has them May 25 07:01:05 im pretty sure dell too May 25 07:01:14 o.O May 25 07:01:20 i got mine from dell ;) May 25 07:01:49 which was actually kind of a surprise May 25 07:02:09 i could possibly find the parts but i would be a little nervous about that May 25 07:02:39 busting my n900 open to lose like... the camera lens or something even more oddball May 25 07:14:27 any of you guys messed with osso-backup? as in editing the stuff in zip files (backing up without password), repacking it and then getting osso-backup to set it up according to one's modified preferences? May 25 07:16:03 No, I've found no reason to modify a backup: I just modify the original file May 25 07:16:15 and if things break, that's what backupmenu's for May 25 07:17:11 you are sure, that you edit the backup and try to restore, iterative until it works? ;) May 25 07:17:21 *dont May 25 07:18:29 should i enable hwsync for hildon-desktop ? May 25 07:19:35 its the settings that I got osso-backup to backup. I haven't yet installed backupmenu (hence I'm trying to setup a backup clone in case something seriously gets messed up on my main one). The settings if I were to restore that onto my other N900 the Display configurations under Settings automatically freezes Settings app May 25 07:20:30 strohhalm, that didn't quite make much sense :) May 25 07:20:59 ;) May 25 07:21:05 must leave, xy May 25 07:21:06 cya May 25 07:21:07 I guess I'm pretty much on my own now, going to delete some %gconf.xml files and replace them with null length files May 25 07:21:11 bye May 25 07:22:36 good luck May 25 07:23:09 hmm I guess replacing them with null length files shouldn't hurt, I can see some %gconf.xml files with zero length May 25 07:24:07 thanks :þ May 25 07:33:47 kerio May 25 07:37:42 hello docscrutinizer May 25 10:09:20 hello kerio May 25 10:11:51 helloocscrutinizer May 25 10:13:37 hello itschue May 25 10:14:17 ello May 25 10:14:36 itschue: Ant news on the N9? May 25 10:15:29 ant? May 25 10:16:57 typo :þ May 25 10:17:10 dont have news about the n9 i have problems with my n900 May 25 10:17:11 any* May 25 10:17:22 itschue: What is it? Broken usb port? May 25 10:17:25 ok May 25 10:19:13 nodevice is rebooting by itself and battery is draining after flashing vanilla and combined both flashed without -R and waiting long time after every step May 25 10:21:27 probably try it the other way round May 25 10:21:39 first fiasco then vanilla May 25 10:22:23 and this is safe because the nokia page says others May 25 10:23:59 Psycho_oreos both ways should work as long as you don't reboot inbetween May 25 10:24:30 merlin no different between both ways? May 25 10:24:33 merlin1991, apparently not in itschue's case :) maybe just reflash the fiasco and reboot, that might work May 25 10:26:26 psycho_oreos flash fiasco then no reboot and then vanilla or reboot after both steps? May 25 10:27:02 Afaik it's vanilla no reboot May 25 10:27:08 fiasco reboot May 25 10:27:29 ~flashing May 25 10:27:30 hmm... flashing is http://wiki.maemo.org/Updating_the_tablet_firmware May 25 10:27:42 itschue, just flash the fiasco and reboot May 25 10:28:26 ok i try this when iam at home May 25 10:29:17 itschue: also see http://wiki.maemo.org/Updating_the_tablet_firmware#Flashing_the_eMMC_in_the_N900 May 25 10:34:18 the first link says after flashing vanilla i need to flash combined May 25 10:35:28 i flashed vanilla and combined and used the first link May 25 10:35:50 ut i have still the problem May 25 10:37:22 im back later break is over and have to work now May 25 10:38:08 can anyone remember that Nokia Beta labs link for an app which acts like an ad-hoc IM and presence announcer over wifi and bluetooth? May 25 10:38:36 iirc the video was of some people sat in a sports cafe, all very cheesy, but the idea would be cool for conferences I think May 25 10:39:32 lardman, you attending meegoconf? May 25 10:41:12 MohammadAG: nope May 25 10:41:27 otherwise I'd certainly not be up at this time of the morning ;) May 25 10:41:47 I was just thinking about disruptive technologies May 25 10:42:09 lardman, I did make a Bluetooth Messenger app with Qt Mobility May 25 10:42:23 it turned out well May 25 10:42:26 cool May 25 10:42:30 works on Linux/Symbian May 25 10:42:40 but never released it, it was a school project May 25 10:46:25 sounds like a cool basis May 25 10:46:43 I was thinking of being able to wander around a conf with device saying who you are and what your interests are May 25 10:46:53 probably useless, but cool imho :) May 25 11:07:19 morninf May 25 11:07:30 Morning chem|st May 25 11:09:31 good morning guys May 25 11:09:49 morning chaps May 25 11:09:55 Sf people waking up? :D May 25 11:12:52 * MohammadAG wonders if there's a vBulletin API May 25 11:46:17 is there something more powerfull then nokia n900? May 25 11:46:34 Chuck norris May 25 11:46:45 plenty others but none will run maemo OS May 25 11:46:47 yea, my notebook May 25 11:47:44 nokia n900 is already 3years old machine - that is not normal to have nothing. Is Nokia still alive? May 25 11:48:20 does Nokia sound dead to you? May 25 11:48:29 Now that MS has it, yes... May 25 11:48:57 there's ongoing rumours of n9 coming out however, who knows if it'll ever come out May 25 11:49:00 Well, thing is that there are faster bits of hardware out there... But the software doesn't do as well May 25 11:49:00 It's pining for the fjoords. May 25 11:49:08 N950 //will// be coming out. May 25 11:49:25 That's confirmed: FCC even has approved it May 25 11:49:32 Err - no, it's not. May 25 11:49:42 A prototype has been approved by the FCC May 25 11:49:45 maby not n950 or n9 as the /name/, but some device will May 25 11:49:47 Memo seems quite enough good to be. But hardware is slow. May 25 11:49:50 Maemo May 25 11:49:51 This does not mean it will go into production. May 25 11:49:55 Or what OS it will run. May 25 11:50:03 do they still sell n900 in finland somewhere? May 25 11:50:06 I thought that was N9 not N950, the stuff on tmo May 25 11:50:10 even when it goes into poduction its a dead horse :P May 25 11:50:16 Who knows what it's called May 25 11:50:27 If things are too slow... tweak it or OC May 25 11:50:36 nokias future is wp7 ... mine is somewhere else May 25 11:50:39 is it not possible to port maemo to some powerfull android device? May 25 11:50:39 cehteh: Even if it's dead - that's another 2 or so years for someone to pop up with a 'proper' meego phone. May 25 11:50:46 strohhalm: No. May 25 11:50:51 why not? May 25 11:50:58 SpeedEvil: i hope my n900 will last 2 years more :) May 25 11:51:03 Maemo comprises closed elements in a tangled mess - in conjunction with the open eleemnts. May 25 11:51:16 You cannot use the open bits without the closed bits. May 25 11:52:02 hm - are there some serious reasons to have "closed elements"? May 25 11:52:14 okay May 25 11:52:15 SpeedEvil: How much of the main UI(Theme bits, desktop, statusbar etc) are closed? May 25 11:52:29 because that's what nokia gave us May 25 11:52:30 some people think they are serious May 25 11:52:31 Like, what could be dropped on top of a Debian system? May 25 11:52:34 nda, patents, closed firmware i think markinfo ;) May 25 11:53:06 markinfo: The closed elements were chosen and designed to make it hard for another phone maker to use the open elements that they connect to. May 25 11:53:16 This seems to be the only real design rationale. May 25 11:53:17 the whole GSM (and above) is all well founded on security by obsurity May 25 11:53:41 when people start hacking on that then the infrastructure will scatter in pieces May 25 11:53:43 this is the same situatiation as with WIfi cards? there is binary firmware over that is "open sourced" communication possible. May 25 11:57:00 i wonder when the industry get the fact, that inventing working foo is better than crappy foo May 25 11:57:27 markinfo: No. May 25 11:58:00 markinfo: In many markets, it's actually illegal to - or at the least won't get approval - if you have an opensource GSM stack that the user can edit. May 25 11:58:05 RX-51_2009SE_20.2010.36-2 <-- this is ne latest release? May 25 11:58:06 actually illegal to sell May 25 11:58:55 markinfo: Fundamentally, radio is a shared resource. GSM and other cellphone protocols work because the tower control everything. May 25 11:59:38 markinfo: Two users deciding to use their cell-radios to directly talk to each other and not use up any of their data allowances can wipe out calls on that channel for tens of kilometers. May 25 12:01:29 hm - open Source GSM, it looks danger ... but that problem is very similar to the WIFI cards - or not? May 25 12:01:35 kindof design flaw May 25 12:01:43 It's not a design flaw. May 25 12:01:56 It's a fundamental issue. Radio is a shared resource. May 25 12:02:27 Cellphones work specifically because the cell network is engineered so that there are never two devices transmitting on the same frequency at the same time. May 25 12:03:17 thinking flaw then even May 25 12:03:26 Wifi - pretty much - works with shared channels, as it accepts that sometimes it's just not going to work, and that nearby devices on the same frequency will wipe out the channel. May 25 12:03:58 i cant operate a radio transmitter on commercial bands without infriging the law .. so is hacking the gsm transmitter too May 25 12:04:33 or when i code a virus which hacks some machines (and use it) then its against the law May 25 12:05:06 you cant just prevent that by telling "virus coding cant be open source" the bad guys will always try May 25 12:05:07 But with GSM and other cell-devices - the approval has as a condition of use that it can't be used to do that. May 25 12:05:18 The problem isn't so much one-offs. May 25 12:05:42 * RST38h yawns, eats a kitten May 25 12:05:49 Soo, news? May 25 12:05:53 The problem is a 'free data' app popping up on a popular handset that completely wipes out the network in the area. May 25 12:06:12 which is illegal to use May 25 12:06:23 P2P is illegal to use too. May 25 12:06:32 p2p is not illegal May 25 12:06:47 what you do with it might be illegal May 25 12:06:47 Ok - P2P is illegal to use to infringe copyright. May 25 12:07:17 But the above issue arises with 12 year-olds - for example - using the above app. May 25 12:07:33 SpeedEvil: And so is word, email, web, etc. May 25 12:07:45 SpeedEvil: there's nothing different with php. May 25 12:07:49 eh, P2P. May 25 12:08:03 Veggen: The point is the illegality doesn't stop people doing it. May 25 12:08:11 no, that's true. May 25 12:08:19 Hence the fact that locking down the cell-radio has some point. May 25 12:08:43 thats what i meant with design flaw .. technically each device should authenticate properly with each other and the network .. if not its like operating a pirate radio transmitter, when you do and get catched then you get sued and have to pay high fines May 25 12:08:59 there is no need to obscure this by closing the souce May 25 12:09:01 SpeedEvil: there'll *always* be people infringing copright. That has never been the problem, the problem is that the media business has stopped delivering services that people are willing to buy once they actually get money to do so. May 25 12:09:14 cehteh: The source - the GSM protocol - is quite open. May 25 12:09:33 cehteh: It's only the implementations that are locked down. May 25 12:09:34 if something is decided to be illegal by our society it just is .. May 25 12:09:56 yes i think there is no much point in doing so May 25 12:10:03 cehteh: If you have an open-source GSM modem, you can 'file off' the serial numbers. May 25 12:10:30 yes ... and? May 25 12:10:55 if i steal a car i can grind out the serials too .. does that make it better or legal or acceptable? May 25 12:11:21 The fundamental problem is that radio is a shared resource - anyone transmitting on an unauthorised channel on a GSM networks allocation will knock off up to 16 calls. Some of which may be emergency calls. May 25 12:12:15 stolen cars run by drunken / drugged shitheads also share the same street and overrun people May 25 12:12:25 Sure - it's not. The issue is that your average teenager - who is infringing on copyright over P2P is not likely to see a major difference between doing that, and running a 'free data transfer to your friends' app. May 25 12:13:03 yes but thats not a technical problem you can engage with a technical solution (locking the code down) .. its about education and responsibility May 25 12:13:16 Good luck with that. May 25 12:13:27 no, that is something you can stop by locking down the hw May 25 12:13:38 I look forward to hearing how you're going to educate everyone to follow the law. May 25 12:13:53 chances are that a teenager bright enough to develop their own, won't be a teenager anymore by the time it's ready to go May 25 12:13:56 eventually someone will release (if not already happend) a free/open/hacked GSM stack on the black market May 25 12:14:15 cehteh: That's not the issue. May 25 12:14:18 that will require hacked hw too though May 25 12:14:24 which will be expensive May 25 12:14:51 cehteh: The issue is hardware - widely available - being able to be used in ways that will incidentally jam other users in a many kilometer radius. May 25 12:15:07 And with incentives for people to use it that way (free data). May 25 12:15:08 i bet some day they find a way around .. and some day later it will apear on some server just for download for the kids May 25 12:15:42 Hence the locking down of the hardware. There isn't another way round this. Radio is by nature a broadcast and shared medium. May 25 12:16:02 the point of the separation is so that can't happen cehteh, so no amount of sw can alter that May 25 12:16:21 haha .. they saied that about blueray too :P May 25 12:16:47 yes, though that is slightly different May 25 12:16:50 Without rigid control, your cellphone range drops from 15km to 3km, and then only sometimes, and the battery usage goes up 100* May 25 12:17:00 but yes, perhaps there is a weakness somewhere, who knows May 25 12:17:04 i ran the unlock for the fm transmitter in my n900 and the world did not end and the fcc has not showed up at my door May 25 12:17:18 dangergrrl: short range May 25 12:17:28 dangergrrl: Sure - the FM transmitter is still operating at a really tiny power though. May 25 12:17:37 and no-one dies because they can't listen to Radio 1 May 25 12:17:38 and someone told me not long ago that they heard something about a full gsm stack in open source May 25 12:17:48 yes May 25 12:18:01 But you still need the hw to run it on May 25 12:18:18 dangergrrl: If you are 2km from a cell-tower, and transmit to your friend without negotiating with that cell-tower, you will knock everyone else further than 1km or so on the same channel off the network. May 25 12:18:21 and I bet your average teenager's pocket money won't stretch to that May 25 12:18:27 dangergrrl: Including some emergency calls. May 25 12:18:33 now i haven't been teranmsmitting outside the approved fm range, i don't have a receiver for that :P May 25 12:18:46 dangergrrl: GSM -opensource - hardware exists. May 25 12:19:00 It's expensive - though it will get cheaper. It will never be mass market though. May 25 12:19:10 open source gsm bts http://openbts.sourceforge.net/ May 25 12:19:20 yeah - that. May 25 12:19:36 Also - there is a cracked phone, being used for openhandset May 25 12:20:40 i think the problem is not the closed gsm stack, nokia shows that it is possible to provide a free userland with closed parts e.g. for radio, grafic. but why nobody build such stuff with --usable --customizable May 25 12:20:56 i suppose if i had a friend who lived next door May 25 12:21:23 Yeah - there is no reason why the user cares about an open-source GSM stack. May 25 12:21:35 If the stack is done right, it exposes everything the user might care about. May 25 12:21:39 *and* i could transmit bulk data without being charged by my cell provider May 25 12:22:03 if you had a friend who lived next door you'd push a cat6 cable through a wall May 25 12:22:11 i would be happy with a closed gsm stack, too. if i get a device (system) like the n900 May 25 12:22:14 (where GSM = all cellphones) May 25 12:22:16 because fuck radio, 10GigE is the shit May 25 12:22:33 lo9l May 25 12:22:37 indeed May 25 12:23:00 * lardman votes for fibre, Gigabits are old school May 25 12:23:07 but 2km cat6 is a bit expensive x) May 25 12:23:08 so the idea of a teenager breaking the whole GSM network just for fun is a straw man May 25 12:23:52 It's incidental. May 25 12:23:58 besides they had a fake gsm cell demo at DEFCON last year May 25 12:24:05 It's not something you care about - you just want free data. May 25 12:24:21 GSM is trivial to break in some fun ways May 25 12:25:20 a fake GSM cell can just tell your phone: naah, i don't like encryption, don't use it May 25 12:25:36 and no phone will warn the user it's dropped encryption May 25 12:27:52 i don't buy any of the justifications for closed drivers or NDA hardware as valid May 25 12:27:55 I think it's possible to also get the SIM secrets by the network requesting a weak encryption, and then cracking that. May 25 12:28:15 who the hell thought that security through obscurity was a good idea? :| May 25 12:28:40 The fundamental reason for a phone maker is that in many markets, it's actuively illegal to sell an open -GSM phone. In most others you will not get approval. May 25 12:28:42 it's BS for GPUs and it's BC for GSM stacks too May 25 12:28:56 BS even May 25 12:29:00 It's _not_ security through obscurity. The GSM protocol has been published forever. May 25 12:29:27 ati and nvidia still reverse engineer each other's stuff and their closed drivers only inconvenience some users May 25 12:30:28 and do phone makers use faraday cages when testing new proto stacks? May 25 12:30:58 Yes. May 25 12:31:11 And their own network emulators. May 25 12:31:21 i mean sure not everybody lives next to a htc r&d facility May 25 12:31:27 If for no other reason than you can't control the real-world. May 25 12:31:54 moo all May 25 12:32:14 hey RST38h May 25 12:32:44 if someone wants to break something they will May 25 12:34:34 so you make it as hard as possible for them to do too much damage May 25 12:35:28 i worked in a group at IBM that was doing small quantity manufacturing or prototype servers, we had to send each one across the street to the anechoic chamber for fcc certification before shipping it May 25 12:35:40 but we were using it in the lab before it passed May 25 12:35:51 lardman: heya May 25 12:36:27 and they always came back with the data cables passed through toroids etc May 25 12:36:43 RST38h: no jokes today about N9/50s, promise! :) May 25 12:40:17 lardman: Android FTW! May 25 12:40:24 * RST38h hides, just in case there is a riot May 25 12:41:20 RST38h: hw-wise certainly May 25 12:41:42 Tab has just had an updated kernel source code drop, so Linux for the win! May 25 12:41:48 i think the real reason people want to give us closed toys May 25 12:42:32 is because they are autocratic control freaks and should be the first against the wall when the revolution comes May 25 12:42:51 haha May 25 12:43:14 hmm May 25 12:43:30 cheaper and probably faster to not develop in the open May 25 12:43:47 dangergrrl: i like the way you think May 25 12:44:07 hey we had autocratic control freaks in charge of gcc development for a long time May 25 12:44:12 gcc forked over it May 25 12:44:53 and LLVM apparently kicks much more ass than gcc May 25 12:45:09 ugh May 25 12:45:20 wtf. my calendar isnt syncing up May 25 12:45:23 egcs had open development gcc was closed May 25 12:45:50 iirc they remerged and the egcs faction "won" May 25 12:46:41 some people just like to control others May 25 12:47:06 tbh sometimes you need someone to try herding the cats May 25 12:47:17 I don't know about the gcc case mind you May 25 12:47:41 you ever tried to get cats to do what you want? May 25 12:48:27 unless it coincides with what the cat wants, good luck May 25 12:48:50 quite, encouragement and guidance May 25 12:49:02 but we're only talking about metaphorical cats May 25 12:50:39 sadly the us is likely to have only one GSM provider soon May 25 12:53:48 bbiab meeting May 25 12:53:59 unless the fcc commissioners decide to actually regulate something like they get paid to do May 25 12:58:58 i was in the middle of the gcc thing in the 1990s May 25 12:59:27 i made the decision to ship a gcc snapshot as part of a commercial release May 25 13:07:36 omg nokia releases golden symbian phone! May 25 13:23:58 ZOMG May 25 13:25:54 RST38h, is that the first time? May 25 13:25:57 * lcuk recalls seeing similar May 25 13:28:39 anyone know where i can get curl for the n900? do i have to compile from source? May 25 13:28:49 It's in one of the repos May 25 13:29:13 apt-get wont find it May 25 13:29:15 IIRC May 25 13:29:19 Maybe tools? May 25 13:29:31 Would wget work for you? May 25 13:30:49 i have wget, id like curl tho May 25 13:31:10 for compatibility reasons and i dont want to rewrite some scripts May 25 13:36:56 http://curl.haxx.se/download.html ?> May 25 13:37:47 http://brnz.org/hbr/?tag=maemo May 25 13:39:20 http://maemo.org/packages/view/libcurl3/ May 25 13:39:33 doesn't have the binary May 25 13:39:41 Gyjf, SDK repository and was included in SSU PR1.2 on device May 25 13:40:11 well, included from PR1.0 May 25 13:40:14 Gyjf: tl;dr; http://brnz.org/f/maemo5/curl_7.18.2-8maemo6+0m5_armel.deb May 25 13:41:51 thank you vldcnst May 25 13:43:39 thank you SpeedEvil May 25 14:32:19 ( and no phone will warn the user it's dropped encryption ) Old Siemens C25 did May 25 14:32:37 And where is Siemens now? May 25 14:32:39 * RST38h cackles May 25 14:36:22 nice feature, i stand corrected May 25 14:36:38 almost all phones do not May 25 14:37:29 dangergrrl: I've been in charge for the GSM modem of Openmoko. Not even we, the EE at OM, had full source of the GSM stack of the calypso modem May 25 14:40:07 dangergrrl: scenario: N900 comes with FOSS GSM stack, some dude is patching it to fix the "doesn't reregister after long time in bunker" issue, seems to work, everybody is installing it. 3 Months later the newspapers come up with "Mysterious GSM failures that tore down half of USA got spotted - some 'fix' in a Nokia phones SW caused timing issues in the celltowers" May 25 14:40:36 DocScrutinizer: hoi May 25 14:40:48 DocScrutinizer: is killing bme dsmetool? May 25 14:41:02 rephrase plea May 25 14:41:04 se May 25 14:41:34 DocScrutinizer: I was to kill bme for having the device die when the bat is realy empty May 25 14:41:47 and there is no init.d May 25 14:41:56 ok, just do "stop bme" May 25 14:41:58 but dsme does not do the trick either May 25 14:42:18 ah ok May 25 14:44:31 had it to die yesterday and start from (bme death) dead -> charged while off -> running till dead without bme + logging bq27* output May 25 14:44:43 any further advice? May 25 14:47:47 chem|st: not really, just maybe do an occasional *full* bq27200.sh printout, as the loop mode printout is somewhat missing a few bits May 25 14:52:36 how do I loop in bash scripts? May 25 14:53:18 or can I write to the file while I have $process >> $file May 25 14:53:47 is liblocation open source? May 25 14:53:52 to be precise timed loop is the question... May 25 14:53:52 What are you trying to do? May 25 14:54:07 (wthat was to chem|st) May 25 14:54:13 lardman: nope, but the library is a not much complex d-bus wrapper May 25 14:54:20 interesting parts not on it May 25 14:54:33 Lo does anyone use the Ovi Maps extended stuff (voice navigation addon?) May 25 14:54:34 and the interesting parts are also closed presumably? May 25 14:54:45 chem|st: bq27200.sh 5 >log.txt May 25 14:54:51 javispedro: just wondering about offline cell id location lookups May 25 14:55:00 run bq27xx.sh 60 >> bqtest.txt and add a bq27xx.sh >> bqtest.txt every 10mins May 25 14:55:39 lardman: that would be in the server. either way, does it do that, or you mean you're trying to add that feature? May 25 14:55:42 DocScrutinizer: ah build in? May 25 14:56:01 while true; do bq27xx.sh >> bqtest.txt; sleep 600; done May 25 14:56:15 javispedro: I don't think it does that, it would be a useful feature to have and therefore if added would be best as an extension to liblocation imo May 25 14:56:30 chem|st, let's back up, why do you want to kill bme and run bq27200.sh in a loop? May 25 14:56:33 i.e. using a local database of cell id lat/lon positions May 25 14:56:33 lardman: yep, probably doable. May 25 14:56:51 even more special: while sleep 600; do bq27xx.sh >> bqtest.txt; done May 25 14:57:04 DocScrutinizer, btw I think it's mostly eeprom data that's missing from the one-line output May 25 14:57:12 (database of cell id->lat/lon positions= the cause of the recent android/iphone privacy senseless panics) May 25 14:57:18 ShadowJK: flags May 25 14:57:20 I wonder, is there an open database of wifi router ids + lat/lon? May 25 14:58:00 Yeah the flags and cycle counts that only update once per cycle May 25 14:58:35 cycle count is largely boring for loop, flags like CI et all are not May 25 14:58:51 javispedro, google maps on j2me used to collect this data too, and send it back to google May 25 14:59:21 that and google spycar is how they've built up a large database of cellid to location mapping :-) May 25 14:59:42 ShadowJK: yep, we could do likewise though if we really wanted to May 25 15:00:50 CI is updated once every 30 cycles :-) May 25 15:01:03 yeah May 25 15:01:23 wut? May 25 15:01:32 every 30 cycles ?? o.O May 25 15:01:46 WTF?! May 25 15:01:54 There's a open cellid db iirc. But it was kinda weird, iirc.. And interest/momentum was somewhat lost after iOS/android came May 25 15:02:10 there are several May 25 15:02:41 openbmap being just one of those projects May 25 15:02:48 ok normal loop then so the file remains writable while >> is active May 25 15:03:18 I cannot believe that the same spanish l10n error microB has is also present in firefox... May 25 15:03:24 just have the *60 >>* running and every 600 I add a "normal" output to the file May 25 15:03:35 it's basically: if(cycles since last learning cycle > 30) CI=1 else CI=0; May 25 15:03:37 instead of "search history" being translated to, well, "search history", it was translated to "history of searches" May 25 15:03:58 I don't remember if it was 30, 50 or 15, but significantly > than 1 May 25 15:04:04 thanks Doc May 25 15:04:15 ShadowJK: nope I had CI=1 with 12 cycles already (first time I had a look after) May 25 15:05:00 lardman: yw May 25 15:05:44 chem|st, maybe it never ever completed a learning cycle, this is quite likely with bme :-) May 25 15:06:04 cycles are increasing May 25 15:07:06 a learning cycle, not a cycle May 25 15:07:51 21 cylces so there was a learning cylce ended 21 charges ago... May 25 15:09:07 21 since last full reset May 25 15:09:13 21 since learning May 25 15:10:12 ergo: never had any complete learning cycle May 25 15:11:06 counter for "since last learning" got reset with full reset May 25 15:11:20 but CI got set to 1 May 25 15:12:24 so LMD == iLMD now too, then :) May 25 15:13:46 chem|st: btw since the chip's adjusting of actual capacity is limited to */ 1.25 (iirc) it might need several learning cycles till it drops from iLMD to the real value of ~1300 May 25 15:14:29 the chip isn't really designed for getting detached from battery cell May 25 15:14:53 also not meant to work with incorrect iLMD May 25 15:15:56 chem|st: you should read the datasheet, and maybe May 25 15:16:02 ~batteryfaq May 25 15:16:04 [batteryfaq] http://wiki.openmoko.org/wiki/Battery_Questions_and_Answers May 25 15:16:18 and my bq27000 script linked to in there May 25 15:17:05 it predates N900 and is for the 27000 variant, but they are identical modulo the I2C/HDQ interface May 25 15:17:22 is there any way to lift the restriction of downloading Ovi apps via apt-get? May 25 15:17:40 is there a way to pipe microphone input to output? May 25 15:17:51 the ovi web site bugs me, it's slow and boring May 25 15:17:57 MohammadAG: hehehe, a GOOD question May 25 15:17:57 like old people May 25 15:18:01 besides arecord + aplay, too slow May 25 15:18:11 DocScrutinizer: MohammadAG: there is. May 25 15:18:24 it bypasses speaker protection so use at your own risk May 25 15:18:40 javispedro: I gather there is, but it seems to need insane amount of coding a special app to do so May 25 15:18:41 either play with alsamixer (I do not remember the cotnrols atm) or download cfmradio source and play with the advaudio.c May 25 15:18:42 no gstreamer way? May 25 15:18:57 DocScrutinizer: nah, it wasn't very hard. May 25 15:18:57 Lantizia, it bores you? May 25 15:19:13 walking around grocery shopping is boring May 25 15:19:19 but finding apps is fun May 25 15:19:28 lcuk, more fun using apt-cache search :P May 25 15:19:37 MohammadAG: there *should* be a gstreamer way too May 25 15:19:53 but it would be slower than arecord aplay May 25 15:20:36 I believe cfmradio uses what should be the fastest method that also uses pulseaudio (besides hacking pulseaudio itself; that is what module-loopback does) May 25 15:20:51 (remember reason to use pulse is speaker prot.) May 25 15:21:03 chem|st: feel free to ping me (and/or SpeedEvil / ShadowJK ) if any questions related to bq27x00 (-datasheet) May 25 15:21:12 MohammadAG, actually I remember you made a alternate apt-get didn't you on the forum for this? May 25 15:21:19 if i can find it :S May 25 15:21:45 what I do not realy understand is how the learning reset works May 25 15:21:58 http://talk.maemo.org/showpost.php?p=830385&postcount=19 May 25 15:21:59 :D May 25 15:22:06 javispedro: a) I think there's neither arecord nor aplay (surprise, as it's the same binary ;-P) on maemo? May 25 15:22:26 DocScrutinizer: ah, well, I guess not "by default" ;) May 25 15:22:32 javispedro: b) even if there is, I don't think it would work without killing fsckng PA May 25 15:22:32 btw, I wouldn't run n900 to empty without bme. It cuts power to the cpu without warning, so there is danger of filesystem corruption. On Smartdevices' SmartQ5/Q7, this even made its emmc unusable :P May 25 15:23:55 DocScrutinizer: the point was that if he thinks aplay+arecord is too slow, nothing sw-based is going to beat it May 25 15:24:04 javispedro: c) you need proper amixer settings anyway, but it seems there is just NO "scenarios" (they call it scenarios in OM/SHR) to feed to either alsactrl or amixer, to make mic and speaker output work May 25 15:24:04 chem|st, if battery has been charged to full, then discharged to less than 3248mV, without any charging in between, without removing battery, then it adjusts its caibration towards what it measured May 25 15:24:08 ShadowJK: so what is your suggestion? May 25 15:24:58 ah ok that was the plan, May 25 15:25:23 didn't want it to die on it but wait till close to real empty and shutdown May 25 15:25:28 ShadowJK: that's not *entirely* true, aiui the PMU is sending an IRQ to CPU prior to hard shutdown May 25 15:26:00 javispedro, using the code from arecord you could theoretically produce a marginally more optimal set of routines May 25 15:26:17 lcuk: definitely not May 25 15:26:27 (well, you could do something in kernel-land =) ) May 25 15:26:33 that would be faster too =) May 25 15:26:39 of course May 25 15:27:03 for all I know arecord code sahould be "optimum" May 25 15:27:05 but I do believe there's _was_ a "scenario", which was how fmrx was done originally May 25 15:27:13 that is, using one of the bypasses May 25 15:27:20 javispedro, rm_work do you guys remember when I designed: http://liqbase.net/liq.20081102_235650.gary.scr.png May 25 15:27:33 you can tune it by passing weird parameters like mmap etc May 25 15:27:49 yes May 25 15:27:55 lcuk: yeah that was lulzy a bit May 25 15:28:00 I dug out the sketches used for the design :) May 25 15:28:00 but there's no feature of ALSA api that's not supported by arecord afaik May 25 15:28:02 http://liqbase.net/liq.20110525_162059.liqrecentsketches_classic.scr.png May 25 15:28:30 heh May 25 15:29:01 lcuk: i still am not sure exactly how that would work, as the colors displayed would be different depending on the brightness :P so it's a moving target May 25 15:29:02 chem|st, stop bme once you get your usual low battery warning. start "./charge21.sh" and "./bq2700.sh 5", monitor the bw window waiting for EDV1 column to change to 1. There'll also be a warning popup from bq27200.sh. Connect charger immediately. Wait until 15% charge. Disconnect charger. Wait 30s. Stop charge21.sh (ctrl-c). "start bme", wait 30s, connect charger. May 25 15:29:11 i guess it could be done, but seemed odd to me May 25 15:29:15 rm_work, it always is May 25 15:29:18 rather I would just expect a slider May 25 15:29:22 lcuk: javispedro: so the main problem of arecord|aplay is the | ;-) May 25 15:29:28 IE, what ABL did :P May 25 15:29:29 but it is clear that min/max is all the way across the screen May 25 15:29:32 yeah May 25 15:29:38 and low at the left and high at the right May 25 15:29:49 which makes it, essentially, just as good as a slider, but with extra color :P May 25 15:29:50 the slider was listed as first principle May 25 15:30:03 the extra bit is the granuality May 25 15:30:14 yeah May 25 15:30:32 there was so many people (nokia included) saying "you only need 4 steps" May 25 15:30:36 ShadowJK: sounds nifty May 25 15:30:45 and others saying, I need 255 steps in 0.0000001 step increments May 25 15:30:50 that part is intriguing i suppose.... top slider puts you in ranges of 64 or so, and then second puts you in ranges of 8 inside that May 25 15:30:59 yeah May 25 15:31:05 and the bottom slider is all 8 May 25 15:31:10 * lcuk nods May 25 15:31:11 yeah, i like the design actually May 25 15:31:16 but it's SOOOOO BIG May 25 15:31:20 omfg May 25 15:31:25 that is the bit that doesn't matter May 25 15:31:43 i feel like it belongs in a settings menu under "show me the crazy-huge advanced version" May 25 15:32:08 not as the default, but as a crazy config option for the "more concerned" May 25 15:32:16 ShadowJK: where do I get the charge21.sh and what does it do? May 25 15:32:29 rm_work, when it is in the control panel, it is the same size as a brightness icon May 25 15:33:06 yeah, i guess it pops down and onscures the whole screen, but that's ok, since it's temporary anyway May 25 15:33:09 so it was never crazy huge level May 25 15:33:11 *obscures May 25 15:33:12 http://liqbase.net/liq.20090625_222814.lib.scr.png May 25 15:33:40 rm_work, all this just came back and being reviewed because I collated all the sketches ever drawn May 25 15:33:46 lcuk: you're aware the human eye can't tell brightness changes < *2 / *.5 ? May 25 15:33:49 and selected those which had decent info on them May 25 15:33:49 heh May 25 15:34:03 DocScrutinizer, not important May 25 15:34:15 and we are not talking about humans, these are linux maemo petrol heads. May 25 15:34:17 DocScrutinizer: Sure it can. May 25 15:34:20 yeah and most people can't tell the difference between reference monitor speakers and a $20 set of earbuds May 25 15:34:29 DocScrutinizer: If they are relative, it can go _way_ deeper than that May 25 15:34:30 but people still spend $1000 on them and SWEAR they can May 25 15:34:31 so with 255 steps for linear brightness you got a reasonable range of max 8 steps for brightness May 25 15:34:38 HID May 25 15:34:40 DocScrutinizer, choice is more important May 25 15:35:05 "The subsystem can handle 24bit brightness resolution, WHERE are my 16777215 levels???!!!!!" May 25 15:35:22 SpeedEvil: those aren't relative when it's about brightness setting of LCD-BL May 25 15:35:34 DocScrutinizer: I agree with you for the most part, though I would still allow maybe 10-12 just for super-sensitive people and special cases, so at that point it may as well be a slider May 25 15:35:37 dimmer switches in the home vs regular light switch etc May 25 15:36:10 ShadowJK: http://enivax.net/jk/n900/charge21.sh.txt that one? May 25 15:36:11 For me at least - I want brightness levels lower than the '2 of 255' that is the hw minimum on the n900. May 25 15:36:19 I'm somewhere in the middle between the "crazy Nokians with their 4 brightness levels" and the "nutjobs who want MAXINT brightness adjustments" May 25 15:36:20 rm_work: I fully agree May 25 15:36:37 rm_work, the brightness doofer I did happily bridged that gap May 25 15:36:42 without showing any extra ui May 25 15:36:49 * lcuk really liked designing that :) May 25 15:36:54 heh May 25 15:37:23 Lantizia, it's somewhere on mohammadag.xceleo.org/public/maemo/debfiles May 25 15:37:27 btw, the graffiti wall I am using is close to how original liqbase did it :) May 25 15:37:47 yeah, i do like that interface actually, it would work very well as either a fullscreen brightness changer on a handset (the taskbar icon is clicked and it pops up and essentially takes the whole screen) or as a widget on a tablet device May 25 15:37:48 javispedro, is there a quicker way? I'm on the road May 25 15:38:10 I am really looking forward to a real usable device May 25 15:38:14 so i can get back into meego dev May 25 15:39:24 javispedro, and it's for the FMTX, so speaker proetection isn't needed May 25 15:39:53 I wouldn't do it on the move for a first time because you might not be able to reset it May 25 15:39:59 to nroaml settings May 25 15:40:00 chem|st, that's the one May 25 15:40:02 meego/maemo dev May 25 15:40:22 javispedro, don't plan on changing controls May 25 15:40:34 but I remember a method on tmo using gstreamer May 25 15:40:39 and it was quite fast May 25 15:40:49 it will be slower than arecord May 25 15:41:11 it wasn't :p May 25 15:41:21 sth like gst-launch pulsesrc ! pulsesink shall work May 25 15:41:26 arecord + aplay are fast actually, but after sometime... May 25 15:42:01 you need gstreamer*-utils or whatever is called May 25 15:42:14 s/utils/tools May 25 15:42:59 actually that is slow... May 25 15:43:02 otherwise, get pulseaudio-utils and use parec | pacat May 25 15:44:02 everything will be slow (were talking 1sec), I do not know how a "fast" gstreamer method could exist. May 25 15:44:15 1s is acceptable May 25 15:44:24 but arercord goes up to 5s after a min May 25 15:45:03 ShadowJK: ty May 25 15:45:16 javispedro, thanks May 25 15:46:45 might be a pulseaudio/alsa layer bug May 25 15:48:29 MohammadAG: what? 5s skew/latency after 50s of piping arocord to aplay? May 25 15:48:36 it's only to shout at someone in a car May 25 15:48:55 DocScrutinizer: note that I remember that aplay & co would be using the pulseaudio alsa emulation on a n900 May 25 15:48:59 *remembered May 25 15:49:11 javispedro: indeed May 25 15:49:25 and PA shit might mess up rate May 25 15:49:48 so this would be one situation where pacat would work better than aplay =) May 25 15:50:00 so quite easily the aplay via PA alsa plugin might be a few % slower than the rate of arecord May 25 15:50:27 ~nuke pukeaudio May 25 15:50:29 * infobot prepares 100 missle silos, and targets them at pukeaudio ... B☢☢M! May 25 15:50:38 I need to backport the loopback PA module someday May 25 15:50:56 backport ? May 25 15:50:58 it allows you to configure the daemon to do that stuff for you, avoiding two IPC calls per packet.. May 25 15:51:06 DocScrutinizer: yeah, was introduced in a more recent version of PA May 25 15:51:16 omfg May 25 15:51:50 ~kill pukeaudio, with fire May 25 15:51:51 * infobot shoots a ionized graviton gun at pukeaudio, with fire May 25 15:51:57 would do wonders for PA-based radio apps battery life May 25 15:52:48 you cannot use large packet sizes because of increased latency, quite noticeable when tuning.. May 25 15:53:01 so small packets -> IPC overhead -> cpu usage. May 25 15:53:12 nukiing that friggin PA that follows that "APE centric approach" and reinvents the wheel, just without spokes, that would really help save some battery May 25 15:53:21 ~nuke pulseaudio May 25 15:53:23 * infobot prepares 100 missle silos, and targets them at pulseaudio ... B☢☢M! May 25 15:53:47 wait til L.P. is near May 25 15:54:41 just some 30min ago it stroke me startupd is another einvention of the whell of upstart. Will this moron *never* learn? May 25 15:54:54 startupd? May 25 15:54:57 you mean systemd? May 25 15:55:01 err yes May 25 15:55:13 Huge record/play skews are very common. May 25 15:55:24 Because in most devices they use separate clocks. May 25 15:55:28 derf: on PA, you bet they are May 25 15:55:41 I mean in the hardware... way before it gets to software. May 25 15:55:57 derf: sane drivers know how to deal with that May 25 15:56:13 You let me know when you find sane drivers. May 25 15:56:26 PA and L.P. started to come up with a "solution" that simply ignores that fact May 25 15:57:48 derf: a proper concept does NO assumptions on the accuracy of hw clocks for AD and DA, just syncing buffers in a way that fills/yanks frames or buffer segments when needed May 25 15:58:38 derf: you got that problem for syncing audio to a video stream ALL the time May 25 15:59:01 and there are evidently working concepts how to deal with it May 25 15:59:15 Syncing audio and video is completely different. May 25 15:59:30 definitely not May 25 15:59:35 You just measure the audio playout rate and display video frames accordingly. May 25 16:00:12 you got still one master clock (usually systemclock) and a number of not synced clocks for the streams that need close supervision and adjustment May 25 16:01:43 same thing for RTP (VoIP), even more complex there as you not only got your local AD and DA clock and systemclock master, but also those of the far end which you need to sync to May 25 16:02:19 No one actually uses the system clock to sync audio and video. May 25 16:02:32 audio / streaming-at-large isn't for noobs May 25 16:02:53 As for RTP... most things just drop packets and/or invoke PLC to stretch the audio. May 25 16:03:14 derf: I don't even need to start to show you about your prev staement being false May 25 16:03:43 Okay, show me a system that actually resamples to keep things in sync. May 25 16:04:08 pff, look for yourself, you'll find some May 25 16:04:13 Uh huh. May 25 16:05:05 and e.g twinkle definitely uses systemclock to sync the up to 6 audio streams May 25 16:05:33 not to talkj about all the windows commercial video players May 25 16:05:37 hello May 25 16:07:07 derf: btw your initial assumption also is incorrect, in most devices the audio card has only one master clock that's used to scale down to the different AD and DA channels' samplerate May 25 16:08:04 so generally they are "same clock" May 25 16:08:41 at least as rec and pb are same audiocard May 25 16:08:41 Systems with multiple 'audio cards' May 25 16:09:05 ...have hacks you find in internet to sync their clocks May 25 16:09:11 hw mods May 25 16:09:32 for studio grade multitrack recording May 25 16:10:19 nobody's going to implement separate hw clock generator chips on a single soundcard for rec and pb May 25 16:10:52 no sane rationale to spend the additional money for a useless and bad feature May 25 16:11:31 Indeed. May 25 16:13:55 but obviously PA author L.P. had NFC about all that May 25 16:15:20 btw this common clock being the only rationale why ALSA devices always consist of a downstream&upstream May 25 16:16:05 If you're wanting to be truyly arbitrary about I and O, you need to support resampling. May 25 16:16:39 So, I downloaded the twinkle source and spent a few minutes looking. May 25 16:16:42 which ALSA does via "plug" and "resample" plugins, on your discretion May 25 16:16:45 It does not resample to maintain sync. May 25 16:16:51 Instead it drops packets, just like I said. May 25 16:17:07 I never clamed anything else May 25 16:17:20 and indeed I never suggested resampling May 25 16:17:39 12:02:41 < derf> As for RTP... most things just drop packets and/or invoke PLC to stretch the audio. May 25 16:17:39 for dynamic syncing of streams May 25 16:17:42 12:03:04 < DocScrutinizer> derf: I don't even need to start to show you about your prev staement being false May 25 16:18:23 12:03:32 < derf> Okay, show me a system that actually resamples to keep things in sync. May 25 16:18:26 12:03:57 < DocScrutinizer> pff, look for yourself, you'll find some May 25 16:18:41 That seems to be the opposite of "never suggeseted resampling". May 25 16:18:57 I'm going to stop talking to you now, as I don't think you have a clue. May 25 16:19:02 [2011-05-25 18:02:08] No one actually uses the system clock to sync audio and video. [2011-05-25 18:03:03] derf: I don't even need to start to show you about your prev staement being false May 25 16:19:57 I said "audio and video" for a reason... your counter-example was multiple audio streams. Which I already said was a completely different case. May 25 16:20:04 Anyway. May 25 16:20:07 -= The End =- May 25 16:20:24 derf: I'm fine with you stopping talking to me now, as *I* think you think I'm an idiot May 25 16:21:48 and I think you're probably in a relatively small and exclusive club with this notion - at least when it comes to my technical merits May 25 16:30:32 MohammadAG, Lo, I'm using your modified apt package so I can download packages normally restricted to having to use the Ovi Store (handy when you want a one-liner apt-get to get all your apps back)... are all the Ovi stuff in the maemo.nokia.com repo? like I'm found bounce and angrybirds but not the angry birds levels or even firefox May 25 16:31:58 Lantizia, yep May 25 16:34:30 MohammadAG, any easy way of finding the corresponding names? (apt-cache search doesn't seem to be helping me much) also what about the pay-for ovi apps? they in there too? (paid for the levels you see) May 25 16:35:02 brb need to pop to vets May 25 16:36:19 paid ones are on a different server May 25 17:05:07 most (all?) video players use audio clock as "master clock".. System clock is only used to sleep until estimated next video or audio frame May 25 17:06:06 Alsa plugins and pulseaudio type things can often lose resolution and accuracy of reading the audio clock, causing audio/video desync or jittery framerates and/or choppy audio May 25 17:11:07 Considering the architecture for most video players were designed when P3-500 was a reasonable cpu in the "performance" class, where resampling 44.1kHz stereo took 30% CPU, it was never even considered :) May 25 17:12:11 Morning, all May 25 17:12:39 ShadowJK: it's basically irrelevant what's your master clock. The point is you have to nominate just ONE master clock, and sync all the media streams to that, by either inserting/dropping whole audio snippets (packages) or single samples/frames (which is a sort of poor man's resampling), or by dynamically adjusting the slave clock of the media stream running out of sync. Basically same applies to video. Esp if you're playing a 25fps video May 25 17:12:41 to a video-card/monitor that's running at an arbitrary vsync of e.g 77.8Hz May 25 17:15:13 the good thing is that your video clock is (almost) always faster than your video stream, so there's no need to start interpolating or resampling video in time domain May 25 17:15:42 indeed May 25 17:15:57 but with audio you'd get stutter, crackling or overlapping segments if you had such an approach May 25 17:16:29 usually doesn't really matter that much - at least on voip that's the common approach May 25 17:17:39 resample (e.g by alsa plughw plugin) to the roughly correct sample rate for the hw, the insert/drop single (usually) 20ms buffer segments May 25 17:18:21 I think people are less sensitive to distortion in telephony for historical reasons May 25 17:18:43 for that you need a somewhat accurate idea of how much is your audio buffer's fill up May 25 17:19:23 and it seems that's exactly where PA fails on alsa compatibility plugin May 25 17:20:36 it never shows up on playing back usual mp3/wav/whatever as in that case your audio hw is the master clock source May 25 17:20:43 in many cases it works better than alsa's dmix plugin :-) May 25 17:21:03 but it spoils your audio as soon as your app tries to closely manage the audio buffer directly May 25 17:21:03 dmix is a master of losing buffer fill info May 25 17:21:27 never seen this happening on twinkle May 25 17:22:05 there is one alsa function call that sometimes fails, twinkle audiodev takes care to work around this May 25 17:22:49 iirc it's been exactly about get_buffer_fill_percentage() or buffer_reset() or whatever May 25 17:23:27 get_delay too May 25 17:23:48 who does need that anyway? May 25 17:25:00 you want to know how big the buffer (software and hw buffers) is so you can time video to audio May 25 17:25:42 buffer resets are generally so time consuming and messy on alsa you'd be better off making up shit to feed the buffer so it doesn't run empty May 25 17:27:19 ShadowJK: (want to know how big) on ALSA you *define* size of buffer on opening a audiodevice May 25 17:28:27 if you need low latency close timing control audio, you go for smal segment sizes and also few segment sizes of your audio buffer May 25 17:28:52 s/few segment sizes/few segments/. May 25 17:32:13 defininig it only works within the limitations of the harware and the plugins. Every plugin that does something non-trivial adds a buffer. From buffer fill you can calculate a time, but that time is "time until exception", get delay is "time until buffer tail comes out of speaker" May 25 17:33:50 ShadowJK: I'm not sure about that, as in "that's not a buffer that adds latency, it justgets one chunk of data, processes it in max CPU grunt way it can, and passes it down to next plugin" May 25 17:34:33 there's no double commit A/B swap or whatever May 25 17:34:56 afaik May 25 17:35:58 this async "each plugin has a buffer" approach is what PA does aiui May 25 17:36:39 while for all I know the ALSA plugin stack should be strictly sync from bottom up to app API May 25 17:38:58 that's why all the get_delay() rant L.P. started on ALSA is not very coherent to me May 25 17:39:42 I *could* be completely wrong though May 25 17:41:02 So dmix has this single buffer, and has to wait on every client to feed it a single frame, before it can be fed further on? May 25 17:42:03 and if it has to wait and can't accept more data, what does it report through buffer fill, and what happens to apps that tries feed it more? :p May 25 17:43:11 ShadowJK: (has to wait for each client) that'S the idea of an event driven API, yes May 25 17:44:04 what happens to apps that try feed more? duh! they block on write(), no? May 25 17:44:58 (most simply API concept, of course you can also do a nonblock write(), and several other interface designs) May 25 17:45:20 (up to callback) May 25 17:45:40 so while dmix waits, the buffer to the app is full, but audio runs, and you get a difference between buffer fill and get delay :p May 25 17:46:26 ShadowJK: sorry, I have to run. But your between-the-lines concern isn't applying May 25 17:46:36 it's how hw buffered IO works May 25 17:46:41 all hw IO May 25 17:50:22 if you need a delay bbl May 25 17:52:32 http://imprint.printmag.com/daily-heller/a-curious-similarity/ May 25 17:58:30 I want to share my home's wireless connection from my n900 -> a computer with no wifi, but with bluetooth and usb. I've used http://wiki.maemo.org/USB_networking#Host_USB_Network_Configuration to connect the devices via usb, the computer is running Ubuntu 11.04. I'm hoping someone here might know how to do it and is willing to help me set it up. May 25 17:59:36 Right now I'm connected via bluetooth but that isn't as fast a connection as I could have using usb May 25 18:00:14 hello May 25 18:00:21 hello May 25 18:00:39 yigal: You mean you want the n900 to act like a wifi card for your computer? May 25 18:00:47 Not 3g? May 25 18:02:40 SpeedEvil: yes, I mean right now I believe that is what I'm doing via bluetooth May 25 18:03:13 SpeedEvil: there is at least no sign that my 3g is working but a sign shows wifi, and I have the device connected via bluetooth the my n900 May 25 18:03:19 In principle - bring up the wifi interface - then raise the USB interface, then set routes. May 25 18:03:35 SpeedEvil: it's the routing that I'm ???? May 25 18:03:47 SpeedEvil: I'll look into it May 25 18:04:26 Sorry - I'm tired. I've done this before - but I forgot. May 25 18:04:32 Hack - you can use socks and ssh -D May 25 18:05:24 SpeedEvil: true enough :D, I'm glad that you were able to get it going before May 25 18:07:11 gives me hope May 25 18:07:18 :D May 25 18:07:34 My N900 is dead and has been for a few days now. I thought the battery was empty, because when I started it, it suddenly shut down after a few seconds. It now looks like the usb plug may be damaged, but I've connected the USB cable and the LED flashes read very shortly and every second or so. May 25 18:07:43 it seems similar to this issue: http://discussions.europe.nokia.com/t5/Maemo-Devices/OMG-My-N900-won-t-charge-just-a-blinking-RED-light-flashes-when/td-p/715929 May 25 18:08:07 any advise will be greatly appreciated. Specially if it leads to my being able to get it back up and running. May 25 18:09:28 DocScrutinizer, from video player pov, I don't need low delay, delay can be whatever and if the audio subsystems buffers 60 secs that's fine too, but whatever the delay is, I want to know it :p May 25 18:09:28 it is suggested that this might be a security issue caused by turning the device on and off many times and that reflashing it might do the trick. The question, then, becomes: how? I can't get any life in it at all. May 25 18:09:53 the red LED flash is the biggest lifesign I've seen yet. May 25 18:10:00 ShadowJK: that's pure nonsense May 25 18:10:23 red flash every second or so sounds like the charger chip emitting an error, resetting, and retrying May 25 18:10:36 ShadowJK, any ideas? May 25 18:11:20 jo-erlend: get a new battery May 25 18:11:43 ShadowJK: ,,,from a ALSA PoV May 25 18:11:50 DocScrutinizer, I've never had any problems with the battery before though. May 25 18:12:00 DocScrutinizer i think my n900 i now working wihout reboot error after several flashs May 25 18:12:10 jo-erlend: yoh, it just breaks at a certain moment May 25 18:12:21 ~seen timeless_xchat May 25 18:12:23 but I seem to recall that it is possible for the battery to be so drained that it cannot be charged using the phone? May 25 18:12:28 timeless_xchat <~timeless@firefox/developer/timeless> was last seen on IRC in channel #meego, 2d 19h 53m 51s ago, saying: 'sivang: what's that url i need?'. May 25 18:12:35 jo-erlend: nope. May 25 18:12:42 I don't care how big the delay is,because any sensible sized audio buffers will inherently have too big a delay anyway, so I always have to feed the audio much ahead of video anyways May 25 18:12:53 jo-erlend: It charges in hardware to a minimal level, without anything booting up. May 25 18:13:16 ShadowJK: what makes you think "any sensible sized buffer is too big"? May 25 18:13:58 when I try to charge using a USB cable to my computer, then it flashes red. If I connect the charger that uses my wall as power source, then nothing happens. Any ideas why this might be the case? May 25 18:14:14 jo-erlend: connection problems May 25 18:14:53 ShadowJK: ALSA could handle buffers down to 2 segmants of 1 frame each, I guess May 25 18:15:56 which would mean your app has to have realy short latency when the signal arives, whatever it is SIGIOWAIT or dunno what May 25 18:16:00 jo-erlend, no it doesn't indicate a security measure. It indicates that there is a charging problem. The problem could be your dc adapter, cable, plug, port, n900, battery terminals, battery pins, or battery May 25 18:17:02 I mean you'd have issues reliably keeping the buffer filled May 25 18:17:13 no doubt, it is the plug in the phone that's faulty. I wiggled a little bit and I got that orange light indicating that it was charging and that little shake to indicate that it was switching on, but then it died. May 25 18:17:45 jo-erlend: have you tried replacing the USB cable? May 25 18:18:06 cleaning the receptacle cavities May 25 18:18:29 check with a magnifying glass May 25 18:19:33 SpeedEvil, yes. the charger works if I hold it properly. the usb cable only flashes red. May 25 18:20:41 ShadowJK: so use a larger buffer spec when opening the ALSA device, and be sure your delay is N-1 segments * samplerate * frames/segment when your app gets signalled to deliver the next chunk of data - after initial fill up of whole buffer with N segments of data May 25 18:20:43 usb cable connected to a hub, laptop, front usb ports on a desktop, back port on a desktop pc? May 25 18:22:16 ShadowJK: that's how buffering is supposed to work May 25 18:22:45 (of course ALSA has way more nifty things like min-fill and max-fill thresholds etc) May 25 18:23:05 but basically that's the way things work May 25 18:24:24 it's quite impossible and also useless to get current delay to a resolution < one segment worth of audio May 25 18:25:05 All I want to know is how many bytes are in the buffer, and how much delay there is between buffers empty and sound from speakers stopping (the last one is often sufficiently close to 0 that it doesn't matter). get_delay is supposed to be both combined May 25 18:25:32 where delay = actual_real_time - currently_played_back_timestamp May 25 18:26:47 is it normal for N900 to make a ... muted bell sound when it dies from lack of power? May 25 18:27:01 yes May 25 18:27:14 Should i enable HWSYNC? May 25 18:27:32 though for me it sounds more like an alien got beaten to death by a piano May 25 18:28:47 ShadowJK: you fill your audio buffe with segment 1 + 2, you start video and audio same moment. When ALSA asks for segment 3 you know it just finsihed with segment 1 May 25 18:29:09 actually ideal time ref for get_delay would be audio clock ;p May 25 18:29:10 (assuming a 2 segments buffer) May 25 18:31:05 it works perfectly if you have for example frag-size 1024 and fragments 16 in alsa, because you get buffer fill un multiples of 1024 and get_delay in worst case has granularity similar to that May 25 18:31:15 if your video is still in timeframe of segment one when alsa asks for segment 3 then you're late on video playback. when you're a large way into timeframe 2 or even started 3 then you'Re too fast with video -do whatever you need to correct it May 25 18:31:42 same applies for a buffer consisting of 8 or 16 segments accordingly May 25 18:32:01 but then all bets are off if you don't talk to the hw device anymore and have to go through a plugin chain, and you get all sorts of weird buffer sizes and fragment numbers, and granularities May 25 18:32:04 no need for get_delay() *really* May 25 18:32:43 ShadowJK: sorry, don't see any of the issues you just mentioned May 25 18:34:42 your method is fine, but in practice you get fewer segments than you asked for, and much bigger segments than you asked for May 25 18:35:03 is it normal for N900 to make a ... muted bell sound when it dies from lack of power? May 25 18:35:06 anyway bottom line is: PissAudio isn't going to cure *any* of the problems related to audio playback May 25 18:35:10 the splashscreen makes that sound May 25 18:36:40 MohammadAG, oh, ok. Well, after wiggling the plug a little bit, I could see that familiar orange LED and it tried to boot, showing those dots before it made that sound and died. :/ May 25 18:37:21 have any of you people attempted to fix a broken USB connection? How difficult is it to get to and is there lots of electronics close to it? May 25 18:38:30 I'm pretty sure someone did it May 25 18:41:28 jo-erlend, http://mg.pov.lt/maemo-irclog/%23maemo.2011-04-22.log.html#t2011-04-22T17:15:17 May 25 18:41:44 would not work at all if the usb plug has actually come off the board fully May 25 18:41:51 but seems to work nicely May 25 18:41:51 pulseaudio only works or adds a benefit when it has exclusive access to the audio device, because then it can actually get the parameters it asks for, and provide accurate timing and buffer fill info to its clients. The additional benefit is of course that it has documentation, and that the documentation is comprehensible to more people than the ones who wrote it ;p May 25 18:42:53 eldar's rumors are getting boring May 25 18:44:47 ShadowJK: what? May 25 18:44:58 ShadowJK: pulseaudio is source of problems May 25 18:45:04 and that's all May 25 18:45:32 Arguably the effort spent on pulseaudio could've been spent on fixing alsa bugs, alsa api, and alsa docs instead, but that brings us back to to alsa being incomprehensible to everyone except those who wrote it, and a few others May 25 18:46:06 that's why there are things like openAL and other friendly libraries May 25 18:46:21 jacekowski: indeed May 25 18:46:31 * RST38h throws up at the mention of OpenAL May 25 18:46:57 none of the arguments to start PukeAudio holds valid or delivers after 2 (?) years with this shit May 25 18:47:00 Once again, someone should simply update /dev/dsp API to the latest trends May 25 18:47:09 Screw ALSA, PA, OpenAL, etc May 25 18:47:45 add userspace mixing if necessary, use sockets instead of dev nodes May 25 18:47:56 OSS :) May 25 18:48:46 iirc oss comes with libalsa interface these days too May 25 18:48:52 Also, preventively kill or castrate anyone who comes up with his own audio framework May 25 18:50:42 don't need libalsa interface May 25 18:50:53 just install alsa-lib and redirect all output via asoundrc May 25 18:51:24 there's some alsa plugin as well involved iirc May 25 19:09:37 Well, alsa had, what, 5 years, and nothing happened after those years of user and distro despair :) May 25 19:10:21 So I guess it's only fair to let something else try for another 5 years :P May 25 19:10:45 * DocScrutinizer feels so SICK May 25 19:11:26 iirc, the guy who started pulseaudio was also one of those "argh, not yet another audio framework and daemon" guys, amusingly enough :P May 25 19:11:36 most are May 25 19:11:49 (maybe that made him actually try use alsa) May 25 19:11:50 ShadowJK: orly? L.P.?? nah! May 25 19:13:30 I might be thinking of one of the other projects that started at about same time as pulseaudio :) May 25 19:13:34 afaict he just looked at ALSA, said "Meh, how's delay handled there? Now this is something I can do better (and ignore all the rest of really highly qualified devel crowd)" May 25 19:14:32 he also postulated ALSA has no (working) dmix, and no per app volume (aka softvol plugin) May 25 19:15:26 fact is dmix always worked for me, while PA caused nothing but pain in hte ass for its implementation of mixing 0 to 0 and eating 30% of CPU to do that May 25 19:17:56 and the friggin volume-per-app been published by me (rather late after it got available) here: http://tech.groups.yahoo.com/group/twinklephone/message/1731 -- when got PukeAudio started? May 25 19:18:06 dmix never worked for me. the example confs don't work, and dmix's "stealth" sound daemon hangs all the time :/ May 25 19:20:51 You know, alsa might be superior technically in every way, but it doesn't matter if nobody can figure out how to properly set it up May 25 19:21:14 with alsa I mostly wondered why I had to set things up May 25 19:21:21 instead of the thing just working May 25 19:21:58 heh :P May 25 19:22:11 oss4 just worked May 25 19:22:20 per-app, multiple apps accessing it May 25 19:23:23 that btw also how things are on freebsd May 25 19:23:31 oss compatible May 25 19:24:38 ruskie: just like ALSA did for me, since 1.01 the latest May 25 19:25:16 actually mostly since 0.99 May 25 19:25:38 before that dmix wasn't set up correctly by default May 25 19:26:52 ShadowJK: btw how the hell is PA going to offer better data to user than ALSA, when it got direct access to the hw audiocard, which in the end for PA is... ALSA May 25 19:28:26 PA is just hiding the direct access to hw parameters that you get via ALSA API, and gives you info about PA's own borked internals instead May 25 19:29:04 Because it doesn't go through dmix, or any other plugins. That's where the problems start :) May 25 19:29:13 meh May 25 19:29:32 show me the problems introduced by dmix May 25 19:29:51 you can do very easily, by using aplay -v May 25 19:30:06 First link me a friggin conf of working dmix ;p May 25 19:30:08 and pasting the diagnostic output - I doubt PA got sth similar May 25 19:30:49 ShadowJK: you won't find any working on your PA infected system May 25 19:31:38 and sorry, but no - I can't tell you what to do to get rid of PA to test plain dmix May 25 19:31:40 My debian box doesn't have pulseaudio May 25 19:34:50 UTF-8 is unicode right? May 25 19:35:11 In the sense that hamburger is food, yes May 25 19:35:29 MohammadAG, UTF-8 is an 8bit encoding of unicode May 25 19:36:12 thanks May 25 19:38:06 anybody already work with opencv with python? i'm trying "import cv" and "import opencv" and i got "No Module Named opencv May 25 19:38:53 ShadowJK: http://www.xs4all.nl/~mfnboer/twinkle/faq.html#alsa_no_sound /faq -> general 1) May 25 19:39:10 ~rape xs4all May 25 19:39:11 * infobot takes xs4all behind the WallMart and makes a few grunts and screams May 25 19:42:59 ShadowJK: if there's no similar definition anywhere on your debian system, then go kick somebody else's arse May 25 19:43:48 this is really bare standard on all boxes I installed since ~5 years May 25 19:43:57 /etc/alsa is pretty much empty May 25 19:44:12 try usr/hsare/alsa May 25 19:44:29 toldya it's *standard* May 25 19:46:22 since ALSA0.99 May 25 19:46:28 or 1.01 May 25 19:46:53 so no more need to get it into /etc May 25 19:47:53 NB ALSA got sponsored and maintained by SuSE, and this is from a SuSE(8?) box May 25 19:48:27 it's quite possible other distros never got the turn to properly set up ALSA on their shit May 25 19:49:07 and possibly that's also been the *true* reason to start PA May 25 19:51:23 and you bet the ALSA integration into distros won't get any better since PA came over us like plague May 25 19:53:33 It's not like anyone actually understands how to do "alsa integration"... May 25 19:53:58 yeah, way too hard to c&p those files in /usr/share/alsa May 25 19:54:39 I admit frankly I also don't understand all of that, but it "just works" - for me, on OpenSUSE May 25 19:55:31 hi guys May 25 19:55:33 :) May 25 19:55:38 yo MoonTiger May 25 19:55:47 how's life with your new N900? May 25 19:55:50 can you help me get the sdk tools set up ? May 25 19:55:54 loving it May 25 19:55:55 :) May 25 19:56:08 i followed the wiki instructions but got nowhere May 25 19:56:33 i have the vm running on my windows partition but i hate windows May 25 19:56:50 i have an xubuntu partition just for maemo dev work May 25 19:56:59 but cannot get it up and running *sigh* May 25 19:56:59 MoonTiger: join your request with damnshock who should be here as he asked same question in #maemo-ssu and got advised to ask here May 25 19:57:31 i'll give it another go but it didn't end well May 25 19:57:34 :) May 25 19:57:44 i want to wite a sharing plugin for my website May 25 19:57:52 then do a few small aps May 25 19:58:27 laging a bit too i think May 25 19:58:32 * DocScrutinizer suggests to open #maemo-get_sdk_up channel, with fixed schedules when an experienced devel is helping newbies to get that pile-o-shit running May 25 19:59:12 heh i'm not technically a noob on either linux or dev work ... or even maemo to be honest ... so i *should* be able to do it May 25 19:59:20 i'll try again and report back May 25 19:59:22 ;) May 25 19:59:37 I'll ask x-fade to add 50 karma points to that brave devel's account ;-D May 25 20:00:13 MoonTiger: it's for sure a good idea to get help here. It seems a LOT of users fail on that enigma May 25 20:00:47 ok thnx ... i want to try again now anyways and it doesnt matter if i screw this whole install up as its just for this May 25 20:00:52 and there seems to be few to no proper wiki pages to instruct May 25 20:00:52 sb comes straight from hell May 25 20:01:08 indeed merlin1991 May 25 20:01:17 it does that May 25 20:01:51 maybe its just easier to dev on the N900 itself :) May 25 20:02:10 http://wiki.maemo.org/Documentation/Maemo_5_Final_SDK_Installation May 25 20:02:15 covers it enough though May 25 20:02:23 at least I was able to install it that way myself :D May 25 20:02:31 MohammadAG is a really great instructor when it comes to SDK / SB - I am afraid he got fed up by the job though ;-D May 25 20:02:38 merlin1991, yah i went thru that but it leads nowhere May 25 20:02:46 but the whole process takes forever May 25 20:02:50 i'll try it again May 25 20:02:52 MoonTiger: what's your base os? May 25 20:03:03 i tried to use the wizard thing and i think tha might be what srewed up May 25 20:03:21 i installed a xubuntu 11.04 just for this on a spare partition May 25 20:03:40 MoonTiger: hint#1: the script there for installation is worth nuttin May 25 20:03:47 i figured May 25 20:03:55 i'll start over and do it manually May 25 20:04:01 exactly May 25 20:04:04 the wizards suxx May 25 20:04:13 it seemed too nice to be true May 25 20:04:15 nah it just simply FAILS May 25 20:04:25 yah ... it dies half way thru May 25 20:04:29 but this has proven to work: http://wiki.maemo.org/Documentation/Maemo_5_Final_SDK_Installation#Installing_Maemo_5_SDK_on_x86-32_Debian_based_distribution May 25 20:04:39 get the 2 sh chmod them dun them be happy May 25 20:04:45 s/dun/run/ May 25 20:04:46 merlin1991 meant: get the 2 sh chmod them run them be happy May 25 20:04:46 and, as MohammadAG (?) said: there's no resume, so go thru dl-hell over and over again May 25 20:05:00 right May 25 20:05:06 i'll report back :) May 25 20:05:08 thnx May 25 20:05:14 yw May 25 20:07:40 I wonder what sb is going todo when I install meego sdk to the same machine May 25 20:07:57 prob running some hidden rm -rf /megosdk script :D May 25 20:08:26 minimal bootstrap or which install? May 25 20:09:07 runtime and all dev pkgs? May 25 20:09:08 MoonTiger: what are the other options? May 25 20:09:16 yea that sounds reasonable May 25 20:09:21 coolio May 25 20:09:32 prepare for quite some dl May 25 20:09:43 ok well its downloading stuff so thats good :) May 25 20:09:47 especially when you come to the nokia binary stuff May 25 20:09:51 * MoonTiger makes some tea May 25 20:09:54 some 2GB, you shoud have twice that free on / May 25 20:10:06 i have 14Gb for this partition May 25 20:10:44 should fit :-) May 25 20:11:14 scratchbox, the idiot package that installs to / May 25 20:11:43 I actually would suggest installing sb to a vm image May 25 20:11:52 and instantly making a snapshot after finishing the install May 25 20:12:00 I did mess mine up beyond repair May 25 20:12:02 twice May 25 20:12:11 nah, that's level 2 of the adventure May 25 20:12:30 its ok ... this partition is like a vm for me May 25 20:12:43 chromium and sync is neat May 25 20:12:47 yeah but partition != snapshot you can revert to :D May 25 20:12:59 its all part of the fun May 25 20:13:11 How do I identify which device is the sdcard from the shell? May 25 20:13:12 however cp -A /scratchbox /SB-backup is surely a good idea if you got the space May 25 20:13:15 i could clonezilla the partition May 25 20:13:26 DocScrutinizer: nope that doesn't work May 25 20:13:28 tried that :D May 25 20:13:36 darm, rats!! May 25 20:13:40 dies on all the internal weird stuff sb uses fro the emulator May 25 20:14:07 you mustn't cp while logged in to sb! May 25 20:14:15 I did it from outside May 25 20:14:17 it does nasty bindmount magic May 25 20:14:47 but the funniest results I got when trying to git add /scratchbox May 25 20:14:54 doesn't matter, a mere rm -rf /scratchbox kills your system, when you got sb "started" May 25 20:15:12 sb "starts" by default May 25 20:15:15 it runs in the bg May 25 20:15:21 at least some process does May 25 20:15:27 not here May 25 20:15:38 have to do some login or sth May 25 20:15:42 yea May 25 20:16:02 but I still watch some kind of halting sb worker / whatever msg when shutting down @ the end May 25 20:16:21 jeeez, that'd be it, SB running by default, from boot May 25 20:16:48 I think my sdcard doesn't have a filesystem formatted or a partition, how would I go about doing this in maemo? May 25 20:17:02 mkfs? May 25 20:17:04 snappy: from the device? May 25 20:17:07 yeah May 25 20:17:13 snappy: filemanager, click-hold ->context menu May 25 20:17:39 DocScrutinizer: thanks May 25 20:17:52 hth May 25 20:18:03 sometimes it fails, then come back here May 25 20:18:13 worked great, formatted and finally mounted. May 25 20:18:21 \o/ May 25 20:18:33 DocScrutinizer: is there a tool to partition on the n900 May 25 20:18:41 sfdisk May 25 20:18:52 ah yes, sfdisk, forgot about that May 25 20:18:58 though it's not a tool, it's a desease May 25 20:19:18 I suggest cfdisk May 25 20:19:37 last time I did some commandline editing I used fdisk May 25 20:20:12 s/editing/partition editing/ May 25 20:20:12 cfdisk is just "friendly" May 25 20:20:14 merlin1991 meant: last time I did some commandline partition editing I used fdisk May 25 20:20:34 but sfdisk is all we have on the n900 right? May 25 20:20:40 while sfdisk is a bitch May 25 20:21:04 hmm, I got cfdisk, don't ask me where from. Suggest apt-cache search May 25 20:21:22 gnu tools? May 25 20:21:26 prolly May 25 20:21:44 there's actually a cfdisk package May 25 20:22:17 wow a lot of stuff being installed May 25 20:22:26 ya bet May 25 20:22:40 * MoonTiger makes more tea May 25 20:22:44 the trickiest thing about sb is to find out where you're supposed to put your damn source May 25 20:22:49 can I get a cup? May 25 20:23:02 hehehe May 25 20:23:13 I wanted to compile just a hello world and spend 30 mins finding out all the nasty rebinding sb does when switching targets May 25 20:23:44 merlin1991, yah May 25 20:23:56 the whole sb thing is a bit confusing to me tbh May 25 20:23:58 /scratchbox/users/jr/home/jr/starhash-enabler/debian/changelog May 25 20:24:07 I also didn't get it why sb ships it's own git within sb when you're going to use outside code editor and versioning anyway May 25 20:24:13 YMMV May 25 20:24:29 I'm even more confused when I use a VM to run an OS to run SB May 25 20:24:33 (also the git within sb is really old) May 25 20:24:42 ShadowJK, tell me about that one :/ May 25 20:25:09 dev for android is way easier to start at least May 25 20:25:59 MoonTiger: there's been sth called MADDE, never looked into it but aiui it's way more "user friendly" May 25 20:26:18 DocScrutinizer: only as far as you have no external dependencies May 25 20:26:25 mhm May 25 20:26:38 yeah, gather as much May 25 20:26:45 if you have them it's more convienent to get into sb and you apt in there May 25 20:26:50 it's strictly app (Qt?) development May 25 20:27:10 it is, but ofc you can get it to accept your libs too :D May 25 20:27:11 not sure about qt May 25 20:28:11 hello May 25 20:28:25 anyway for /scratchbox/users/jr/home/jr/starhash-enabler/debian/changelog or whatever it's for you, you better get a symlink or even bindmount May 25 20:29:00 unless you just love it to type long paths into your term May 25 20:29:14 to /scratchbox/users/jr/home/jr/YourProject actually May 25 20:29:53 or simply want to use vi in SB native, nuttin else May 25 20:30:26 tbh what's your alternative to vi on ubuntu native May 25 20:30:29 gedit? May 25 20:30:34 whoa! thats done but wtf is that crap about the nokia binaries?? May 25 20:30:49 but then you probably really should consider developing on N900 instead ;-P May 25 20:30:49 the closed bits of maemo May 25 20:31:05 *sigh* May 25 20:31:07 you need them too May 25 20:31:38 so i have to do an apt-get *inside* scratchbox now? May 25 20:31:43 yep May 25 20:31:49 first edit sources.list May 25 20:31:55 and you have todo it for both targets May 25 20:31:56 ok ... did i do the last apt-get in sb too? May 25 20:32:10 hehehe, I cloudily recall the pain I felt when doing all that May 25 20:32:11 my n900 shutdown by itself ant wont start May 25 20:32:39 MoonTiger: everything in the wiki after /scracthbox/login May 25 20:32:40 recharge then May 25 20:32:41 ok in sb now May 25 20:32:43 has to be done in sb May 25 20:32:45 right May 25 20:32:52 got it May 25 20:33:05 except for the dns fix, if you need it May 25 20:33:39 MoonTiger: hint#2 - do a 'mount' in your standard system xterm now, to see what sv really did to your system May 25 20:33:53 no! May 25 20:33:54 :) May 25 20:33:55 sb* May 25 20:34:06 DocScrutinizer: what did you use todo the starhas enabler? May 25 20:34:08 sb_conf not found May 25 20:34:10 *sigh* May 25 20:34:25 because it's sb-conf May 25 20:34:27 merlin1991: nuttin, it's just preinstall and postrm scripts May 25 20:34:28 oooops sb-conf May 25 20:34:31 ;) May 25 20:34:45 damn it May 25 20:35:01 maybe you did another package too? :D May 25 20:35:07 wtf is it asking me "terminal type" for? May 25 20:35:23 but of course I used x86-kate to edit those files May 25 20:36:18 when I first tried doing something in sb I came from a win / visual studio bg May 25 20:36:31 reducing inetraction to SB terminal to the bare mandatory minimum May 25 20:36:53 I was missing all the comfort an ide gives you May 25 20:36:56 merlin1991: haha May 25 20:37:03 nice scary experience May 25 20:37:04 especially when navigating multiple code files May 25 20:37:47 the thing I missed most was the go to definition part May 25 20:38:12 looking at code from other people that's so handy when you see a function call and wonder wtf that one actually does May 25 20:38:27 ummmmm any idea why its asking me for my terminal type? May 25 20:38:37 nope May 25 20:38:42 merlin1991: keep in mind SB is some kinda chroot, and you should look at it like a black (well gray) box, using your normal tools to edit, copy, whatnot inside that lengthy path May 25 20:39:11 just kickof a compile run in SB, then go back to your used tools to continue hunting bugs May 25 20:39:16 DocScrutinizer: I realized that, and then failed in the attempt to set up a working eclipse environment May 25 20:39:41 that's about where I gave up :D May 25 20:40:13 hmm, I never got eclipse to really work for me as well, sth with git(?) always made me scratch head and give up on it May 25 20:41:05 that's basically why I refuse to touch that whole crap May 25 20:41:28 not really fun to use that environment to build anything May 25 20:41:51 I wouldn't mind going to a term for sb and git May 25 20:42:16 yeah, but exclipse seemed to madatorily depend on it May 25 20:42:21 yea May 25 20:42:39 * merlin1991 just wanted the code editing helpers from exclipse, but instead got a bunch of errors May 25 20:43:00 and I had a plain project dir that I never got married to that git over web thing that is meant to contribute to large projects May 25 20:43:23 just booted my dev notebook May 25 20:43:30 I want to finally finish my project :D May 25 20:44:02 in fact it's been twinklephone where this eclipse adventure started and ended May 25 20:44:24 so quite a few years in the past May 25 20:44:55 ffs what did ubuntu doo May 25 20:45:04 wireless card not recognized again May 25 20:45:24 pcmcia? May 25 20:45:35 nah May 25 20:45:42 just a crappy chip May 25 20:46:02 rmmod && insmod (aka modprobe (-r) ) May 25 20:46:16 yea just wondering how the module was called again May 25 20:46:27 lsmod|less May 25 20:46:43 ok so now we download *more* stuff May 25 20:46:48 err, maybe that doesn't help ;-) May 25 20:47:04 MoonTiger: yeah, prepare for more to come :-D May 25 20:47:10 DocScrutinizer: problem is I don't see the card in lspci May 25 20:47:19 hmm May 25 20:47:21 * MoonTiger settles in May 25 20:47:23 * merlin1991 gets a screwdriver May 25 20:47:32 hahaha May 25 20:47:41 less /var/log/boot May 25 20:47:51 empty May 25 20:47:54 gotta love ubuntu May 25 20:48:03 lastbot then, or whatever May 25 20:48:14 actually i rather dislike ubuntu as much as windows these days May 25 20:48:21 or "get a real distro" May 25 20:48:33 * MoonTiger uses arch May 25 20:48:57 I should replace that ubuntu with debian May 25 20:49:02 but then I'd loose my sb install! May 25 20:49:06 ~ubuntu May 25 20:49:07 extra, extra, read all about it, ubuntu is caca from this ass http://img394.imageshack.us/img394/3843/18427557718e22857e9rb.jpg May 25 20:49:21 lol May 25 20:50:18 also: http://xkcd.com/424/ May 25 20:50:25 so sb is a vm basically May 25 20:50:39 that xkcd is a classic :D May 25 20:50:48 MoonTiger: it's more like a ... wait for it .. chroot. May 25 20:51:18 only http://xkcd.com/149/ and http://xkcd.com/435/ can reach the one about ubuntu :D May 25 20:51:29 javispedro, ahhhhhhh i c May 25 20:51:39 the ubuntu one is classic May 25 20:52:16 javispedro, but chroot with a diff arch yes? May 25 20:52:54 yes, using qemu "user mode" emulation. May 25 20:52:56 btw DocScrutinizer guess where the fcked architect of my notebook desided to put the screw for the dvd drive May 25 20:53:52 javispedro, thnx ... it kinda sorta makes a little bit of sense now May 25 20:54:10 below the keyboard May 25 20:54:40 hehe May 25 20:55:08 well, cya folks May 25 20:55:21 bye DocScrutinizer May 25 20:55:24 thnx May 25 20:55:45 yw, hope merlin1991 will help you to get a working SDK May 25 20:56:02 hah, opening the back, pulling the wireless card out, and putting it back in brought it back May 25 20:56:41 seems to be going ok so far May 25 20:57:21 DocScrutinizer: you're _LEAVING_? =) May 25 20:57:52 meh, I'm trying to check if the big bluebox is still out there May 25 20:59:11 they've moved in to the north pole while you were away. it was misparked May 25 21:06:57 ok well thats done May 25 21:07:04 apparently i have sb done May 25 21:07:34 so now... do i have to work *inside* sb to edit and compile code? May 25 21:08:12 editing I use my normal user account/editor for May 25 21:08:16 hi MoonTiger btw \o May 25 21:08:24 hey lcuk!! :) May 25 21:08:29 you go into sb with /scratchbox/login May 25 21:08:36 compilation is done by signing into scratchbox console session May 25 21:08:37 and simply compile away May 25 21:08:41 just use normal tools but make sure to save into the sb home dir right? May 25 21:08:47 yeah May 25 21:09:01 then compile from sb login session May 25 21:09:03 got it May 25 21:09:10 the FREMANTLEX86 you can run in the emulator May 25 21:09:15 right May 25 21:09:18 the ARMEL target obviously is for the phone :) May 25 21:09:22 xephyr May 25 21:09:26 yes xephyr May 25 21:09:28 for x86 May 25 21:09:31 ok cool May 25 21:09:36 * MoonTiger goes to install geany May 25 21:10:37 where can i find the cmd lines for compiling? May 25 21:12:11 ok found it all :) May 25 21:12:15 thnx for the help guys May 25 21:15:47 ok sweet ... looks good ... thnx again ... off to beddy byes for me now May 25 21:15:51 laterz May 25 21:36:37 hello May 25 21:37:06 are matans extended shortcuts available in the cssu? May 25 21:42:32 anyone? May 25 21:46:05 maemo haves any gps navigator like garmin, igo, ndrive.. May 25 21:48:31 i found ndrive =D i have map licences since my other cellphone May 25 21:51:04 The reason why a question mark is called "question mark" is because it marks questions. If it happens that you would like to write a question and you forget the question mark, or even worse: append any other punctuation, no sane brain will recognize your sentence as a question. This is the very reason why the question mark was invented and why it is among the very small 255 large character set that's understood all around the world. It is more likely that answe May 25 21:51:04 rs show up after using the question mark following a question than if not using the question mark at all. May 25 21:52:26 So if you have a question and you really would like to receive answers for it, ask yourself if you shouldn't use a question mark to have a better chance to receive answers. May 25 21:53:18 But in case you don't want any answers, please be more explicit that your sentence wasn't meant as a question so that I don't have to type all that crap. May 25 21:53:30 Or even just shut up. May 25 21:58:06 wam: it's 127 character set May 25 21:58:18 even worse May 25 21:58:26 I was just too lazy to verify that. May 25 21:59:08 ASCII is 7bit May 25 22:00:03 jacekowski: if you really want to know, even 5 bits are enough for a "?" May 25 22:01:09 pretty cheap character. May 25 22:04:26 was that aimed at me? May 25 22:05:13 no May 25 22:05:21 I am trying to enable the keyboard shortcuts detailed in the wiki WRT matan modified hildon desktop. May 25 22:06:00 i want to know if ALL the functionality of MMHD was included in maemo CSSU? May 25 22:06:18 i.e. the global keyboard shortcuts. May 25 22:32:14 damn, could somebody tell vi to stay in channel when he's posting some question? May 25 22:34:29 general advice: don't even think about posting a question when you're not prepared to wait at least 60min for anybody to answer. Or, if it's a question with a short "lifetime" then either clearly announce this, or cancel the question when you're no longer interested in the answer May 25 22:35:47 DocScrutinizer: that should be: general advice: you're in irc post && wait && don't feed the trolls May 25 22:35:48 it's quite rude to let others check to find an answer for your question, just to find you already left the channel May 25 22:37:01 * DocScrutinizer idly wonders why he /ns info vi__ just yesterday or the day before - generally not a good sign for vi__ May 25 22:42:12 the other extreme: May 25 22:42:21 looks like this, you know? May 25 22:42:26 anybody? May 25 22:42:36 PLZZZZZZZ!!!!eleven!1! May 25 22:42:45 peoples, you got me? May 25 22:43:11 (sorry, in a grumpy mood) May 25 22:43:17 DocScrutinizer is using !!eineinseinself I think he needs help ;) May 25 22:44:24 * DocScrutinizer idly wonders if smart_questions link shouldn't go to topic May 25 22:44:34 ~question May 25 22:44:36 rumour has it, question is If you have a question and want people to give useful answers, make sure you have read this first: http://www.catb.org/~esr/faqs/smart-questions.html May 25 22:44:42 seems not enough May 25 22:44:56 you'd just provoke "smart questions" about smart questions May 25 22:45:00 i think i'm going to have to give up on ./maemo-sdk-install-wizard_5.0.py May 25 22:45:10 ouch May 25 22:45:18 dangergrrl: indeed you should May 25 22:45:19 ~maemosdk May 25 22:45:20 rumour has it, maemosdk is http://wiki.maemo.org/Documentation/Maemo5_Final_Installation May 25 22:45:23 i have scratchbox installed but it's failing installing stuff inside scratchbox May 25 22:45:42 DocScrutinizer: we need to add DON'T USE THE GUI INSTALLER to that :D May 25 22:45:52 dangergrrl: the nokia binaries? May 25 22:46:14 it's a wiki :-) I never even tried to use it, so I'm not really qualified to comment May 25 22:46:15 this is all free components so far May 25 22:47:05 just i'm on an unreliable slow 3g connection and it's aborting and restarting with the tarballs when one file fails May 25 22:47:19 ah the fun of the installer May 25 22:47:34 tbh the non gui installer seems to be more stable than the gui one May 25 22:47:52 whenever someone comes in complaining about sb install he/she used the gui May 25 22:48:14 Installing packages on 'FREMANTLE_ARMEL'. May 25 22:48:28 it's trying to install like 449 packages May 25 22:48:34 normal May 25 22:49:37 but if one fails and i restart it starts from the rootstrap files again, if it fails again i'll try to follow another method for the rest :) May 25 22:49:44 the SDK is like 2GB of download volume I guess - at least that's about what it needs on your disk space later May 25 22:50:27 dangergrrl: the installer script is deprecated (by MohammadAG) just because of that. "No resume" May 25 22:50:59 aha May 25 22:51:21 well it did install all the scratchbox packages for me, i guess i'll have to start from there May 25 22:51:21 [2011-05-25 22:04:34] and, as MohammadAG (?) said: there's no resume, so go thru dl-hell over and over again May 25 22:51:25 the GUI one fyi May 25 22:51:50 the sh ones are fine of course May 25 22:51:58 use -c to resume May 25 22:52:15 MohammadAG: could you fix that wiki page? May 25 22:52:24 grrl's May 25 22:52:30 not now, in bed :P May 25 22:52:43 you know what you have to change there, I'm ignorant, otherwise I'd do it May 25 22:52:46 ok, i'll get those, thanks, i don't like python anyway :) May 25 22:52:46 just noticed the xchat notification while setting alarm May 25 22:52:48 fixing wiki pages in bed? May 25 22:53:09 MohammadAG: night pal May 25 22:53:22 DocScrutinizer, remember the black/gray thing at the bottom of my screen? May 25 22:53:24 good luck with your exams May 25 22:53:33 yup May 25 22:53:35 it spread across the whole screen (bottom) May 25 22:53:43 eeew May 25 22:53:53 pics? May 25 22:53:54 panel failure? May 25 22:53:58 yup May 25 22:54:00 not good May 25 22:54:08 I suspect a leak in LCD sealing May 25 22:54:16 merlin1991, pics don't show it well May 25 22:54:25 lemme find the old pic May 25 22:54:33 MohammadAG: SLEEP! May 25 22:54:33 oh nvm, it's on the N8 May 25 22:54:36 DocScrutinizer: you mean lc comes out? May 25 22:54:39 * MohammadAG looks at tinypic May 25 22:54:41 yup May 25 22:54:46 weird May 25 22:54:58 that's what I'd suspect first May 25 22:55:07 your screen runs dry May 25 22:55:09 :) May 25 22:55:13 or oil from touchpanel May 25 22:55:24 err what oil May 25 22:55:40 the screen had a protector on since day 8! May 25 22:55:44 afaik and afaicsee it's an oil-separated type of tp May 25 22:56:38 protector? eeew May 25 22:56:57 ketas: a resistive touchpanel has two foils that need to be separated by a tiny space between them May 25 22:57:04 i know May 25 22:57:17 you mean there's substance there? May 25 22:57:22 there are ones with separator bumps and air, and those with oil May 25 22:57:23 some kind of May 25 22:57:30 oh May 25 22:57:38 i thought it's bumps only May 25 22:58:09 oil type is way more sensitive and otically better May 25 22:58:14 optically* May 25 22:58:24 i'm wondering what 5800 uses May 25 22:58:57 i heard bumps can fail May 25 22:59:02 leading to short May 25 22:59:04 look closely at different angles, if it's an air&bumper type you'll be able to tell by visible bumps May 25 22:59:23 ketas: indeed May 25 22:59:28 bumps then May 25 22:59:37 i see pattern May 25 22:59:47 night DocScrutinizer, and thanks :) May 25 22:59:54 o/ May 25 23:00:16 or well wait May 25 23:00:29 it wasn't that screen that had bumps May 25 23:00:34 how about n900? May 25 23:00:39 i don't remember May 25 23:01:17 N900 is an oil type as far as I can tell May 25 23:03:20 is that type more resistant (what a pun)? May 25 23:03:41 hmm, no diff I guess May 25 23:04:08 but it's for sure one of the most sensitive r-ts I even seen May 25 23:04:36 you can't do that with bump type r-ts I think May 25 23:05:01 better than 5800 one? May 25 23:05:13 i might get used n900, then i'll see May 25 23:05:22 also quite possibly it's more durable as there's no bumps that get worn May 25 23:06:01 ketas: you can literally operate the N900 ts by the tip of a feather May 25 23:06:12 (the soft end ;-D ) May 25 23:07:21 hmm May 25 23:07:31 it's like 10 to 20 times more sensible than e.g Samsung Omnia May 25 23:07:48 (r-ts with bumps) May 25 23:07:54 i really don't like capacitive ones May 25 23:08:04 * DocScrutinizer neither May 25 23:08:24 but can't we have both? May 25 23:08:26 :P May 25 23:08:32 would be interesting! May 25 23:08:44 meanwhile I missionated 90% of this chan as well ;-D May 25 23:09:21 ketas: c-ts and r-ts don't work together nicely. The tech requirements are too conflicting May 25 23:09:45 ketas: what's the c-ts property you miss in r-ts? May 25 23:09:47 r on top of c May 25 23:10:09 ketas: doesn't work, as r-ts is like a shielding May 25 23:10:18 whoops May 25 23:10:21 right May 25 23:10:24 it conducts May 25 23:10:30 doh, it's required May 25 23:10:47 how about other way round May 25 23:11:04 might work, but not really good May 25 23:11:15 requires flexible c May 25 23:11:19 ye May 25 23:11:21 yep May 25 23:11:22 ah, sounds like crap May 25 23:11:35 DocScrutinizer: is the freerunner resistive too? May 25 23:11:46 iirc yes May 25 23:11:48 what for do you need c-ts anyway? May 25 23:11:54 cehteh: indeed May 25 23:12:01 easy swipes May 25 23:12:02 my apple newton too :P May 25 23:12:03 bump type May 25 23:12:05 well, i ride, i like that my n900 works with motorcycle gloves on :) May 25 23:12:25 ketas: ok, there's no solution for that May 25 23:12:32 yes, and it works with a fingernail, with a stylus May 25 23:12:39 which is good May 25 23:12:59 dangergrrl: i bet Steves IStore has some motorcycle gloves for capacitative touchscreens too :) May 25 23:13:07 * GAN900 yawns. May 25 23:13:22 How's everyone this fine afternoon? May 25 23:13:26 ketas: to swipe a r-ts, you need a minimum pressure. That's system immanent May 25 23:13:57 ~nfc May 25 23:13:59 somebody said nfc was No Fucking Clue May 25 23:15:10 ketas: anyway I always found swiping on N900 works just great May 25 23:15:46 well sometimes not *that* good .. and also clicking and stuff, but i think thats rather a software problem May 25 23:15:51 well, don't feel tired so I'll sleep in 30 mins May 25 23:16:06 the ts is fine May 25 23:16:12 GAN900: How's the afterlife? :) May 25 23:16:14 definitely May 25 23:16:25 modify libhildon to get "faster" scrolling May 25 23:16:38 javispedro, sunny and pleasantly cool. May 25 23:16:45 afterlife? May 25 23:16:53 do they have internet there really? May 25 23:16:56 you can practically "remove" friction and a list scrolls indifenitly May 25 23:16:57 not faster .. but more consistent registering of clicks, right clicks, double clicks .. May 25 23:17:05 :P May 25 23:17:05 and well 'near' clicks May 25 23:17:28 Qt's right click sucks May 25 23:17:44 you have to hold the exact point you tapped May 25 23:17:54 Gtk has some radius/threshold May 25 23:17:55 if you tap on a link in microb and wait for the menu to show up you have to hold your finger very calm May 25 23:18:08 that's a software thing yes May 25 23:18:12 MohammadAG: has it? is that configureable? May 25 23:18:21 i would like to increase it May 25 23:18:29 cehteh, probably, libhildon is open, so's Qt May 25 23:18:36 you just have to find the right place May 25 23:18:54 i hoped for some config entry May 25 23:18:55 if it's not libhildon it's definately gtk May 25 23:18:59 yes May 25 23:19:09 no, compiled May 25 23:19:10 ketas: when I swipe with "upside" of my pointing finger's nail, I need like ~10g pressure May 25 23:19:31 pressure is in newtons May 25 23:19:37 meh May 25 23:19:39 * MohammadAG is an SI nazi May 25 23:19:39 pond May 25 23:19:48 for the xorg synaptics driver i made a patch to register a FingerPress .. thats a harder press than just a tap May 25 23:19:53 configureable May 25 23:20:04 used for the trackstick emulation May 25 23:20:06 oh, it could also be xorg... May 25 23:20:13 DocScrutinizer: of course May 25 23:20:18 prolly nobody uses it because it hardly discoverable :P May 25 23:20:34 newtons is apples per minute May 25 23:20:41 (but its in mainline since a few years now :P) May 25 23:20:52 ketas, i lol'd May 25 23:21:23 uApl/min May 25 23:21:34 i think the pressure (or area) detection on the n900 could be also used for 2 or 3 levels of pressure sensitivity (touch, tap, press) .. no more May 25 23:21:52 cehteh: I wish synaptics would work for my tp in laptop, recently I can't get it do disable during typing anymore. Hell I can't disable it at all May 25 23:22:11 rmmod psmouse May 25 23:22:30 hmmmmm May 25 23:22:31 well meanwhile i have a nipple on my laptop and a hardware switch to turn the touchpad off May 25 23:22:31 s/rmmod/modprobe\ -r/ May 25 23:22:32 MohammadAG meant: modprobe\ -r psmouse May 25 23:22:47 infobot, read about escaping, please May 25 23:22:54 so my work on the synaptics driver is a bit unmaintained G*G May 25 23:23:24 infobot is not a shell, she only collects ice shells ;-D May 25 23:23:34 nice even May 25 23:23:35 i like the trackpad way more (thats also why i added the emulation to the synaptics driver) May 25 23:23:41 * MohammadAG can't wait for the Microsoft boob mouse device May 25 23:23:43 cehteh: I assume you have no clue how to get the third button on an external IBM keyboard with trackpoint connected over PS2 - to work? May 25 23:23:47 boob? May 25 23:23:59 boobiemouse May 25 23:24:14 they must be planning one with all these nipples on laptops these days May 25 23:24:15 Does it come in a pack of 6? May 25 23:24:18 o.O May 25 23:24:19 or is it 8. May 25 23:24:28 SpeedEvil: no idea but i guess they extended the ps/2 protocol and send some extra events May 25 23:24:38 How many breasts does a mouse have? May 25 23:24:39 SpeedEvil, buy 2 get 1 free May 25 23:25:08 one May 25 23:25:14 girlfriend six-pack May 25 23:25:24 Ah - mice have 10 nipples. May 25 23:25:29 so we've had mice as nipples May 25 23:25:38 and mice with balls (that sounds wrong) May 25 23:25:42 SpeedEvil: biologist's rule of thumb: twice the number of an average litter May 25 23:25:42 that May 25 23:25:49 slightly unrelated but someone pasted this some days ago on irc: http://www.globalpov.com/images/triple.jpg May 25 23:25:49 that's not nipple on laptop May 25 23:25:53 it's clit May 25 23:25:55 Hmm. May 25 23:25:58 Odd. May 25 23:26:05 Wikipedia says male mice have no nipples. May 25 23:26:25 yes, they have balls on the bottom side May 25 23:26:33 LOL May 25 23:26:39 haha May 25 23:26:48 that's why laptops behave so bad sometimes May 25 23:26:52 they are girls May 25 23:26:54 :P May 25 23:27:11 * javispedro hates students May 25 23:27:18 well, sometimes. May 25 23:27:24 * DocScrutinizer cackles May 25 23:27:26 I rmember a women == pc link in here May 25 23:27:27 cehteh, I can't find that image attractive May 25 23:27:40 javispedro, that's racist against me and offends me May 25 23:27:42 i didnt say you should May 25 23:27:43 javispedro, but they've got nothing but good things to say about you. May 25 23:27:46 I demand an apology! May 25 23:27:47 :p May 25 23:28:07 MohammadAG: but 3 is better than 1 May 25 23:28:21 GAN900: heh =) May 25 23:28:26 cehteh, not sure I've seen women with 1 :P May 25 23:29:11 s/with 1// May 25 23:29:25 there's no mammal with one brest, according to biologist's rule of thumb May 25 23:29:33 heeeeeeello there are women on this channel. May 25 23:29:40 where? May 25 23:29:42 MohammadAG: i am pretty sure google image search will find some May 25 23:29:44 * ketas looks around May 25 23:30:05 http://everythingaboutplasticsurgery.net/wp-content/uploads/2008/05/mstc.jpg May 25 23:30:12 ketas: they left when you entered May 25 23:30:18 sfw :P May 25 23:30:32 DocScrutinizer: good May 25 23:30:35 cehteh: sfw? where do you work? May 25 23:30:44 home office :) May 25 23:30:48 hahaha May 25 23:30:53 lol May 25 23:31:58 * javispedro ponders for a moment wheter to ban everyone... May 25 23:32:13 damn it May 25 23:32:16 that'd been fun May 25 23:32:23 including yourself? :) May 25 23:32:28 if you banned only yourself that would've had the same effect :p May 25 23:32:36 yeah May 25 23:32:43 regardless, /mode +i and /mkick ftw May 25 23:32:51 ok .. lets do something serious .. bbl May 25 23:33:27 I never knew javispedro had ops here May 25 23:33:47 * javispedro didn't either May 25 23:34:14 * javispedro goes back to grading.. May 25 23:34:29 you're a teacher? May 25 23:34:47 No, I TA. May 25 23:34:51 he hates students May 25 23:34:55 gotta be ateacher :D May 25 23:35:09 I thought he hates fellow students :p May 25 23:35:22 MohammadAG: I was a fellow student.. until previous week that is... May 25 23:35:29 javispedro: at university? May 25 23:35:46 what's a "TA" btw? May 25 23:36:03 DocScrutinizer: yep May 25 23:36:49 you must be hating students ;-D May 25 23:37:01 Conversion of the Nipple to Hair-Bearing Epithelia by Lowering Bone Morphogenetic Protein Pathway Activity at the Dermal-Epidermal Interface May 25 23:37:03 argh May 25 23:37:25 SpeedEvil: I am seriously starting to consider your channel fsckups are intentional =) May 25 23:37:33 you read interesting stuff SpeedEvil May 25 23:39:54 Sorry - I was copying and pasting between stuff and hit alt-tab once too many between ctrl-v's. May 25 23:40:54 SpeedEvil, sure sure. May 25 23:42:15 GAN900: I gather your lightning reporter intermezzo on #eegoconf is over? So I can remove the special highlight on your nick? May 25 23:42:26 Yeah May 25 23:42:30 Conference is done May 25 23:42:43 Sorry I didn't report anything useful. :P May 25 23:42:43 ah, and no devices May 25 23:42:53 * javispedro grabs popcorn and goes to see tmo May 25 23:43:13 Hello all. I have Modest grab my email from my dovecot@home. Modest only alerts me when I have new mail in my inbox. It'd like to have new mails for some other boxes too. How do I get that? May 25 23:43:50 GAN900: well, when there's nothing usefull to report, what can you do? May 25 23:43:56 Tried to start a TMO rumor. May 25 23:44:31 GAN900: omg tell me, I have popcorn to spare so far May 25 23:45:05 jonaskoelker: I doubt modest can do that May 25 23:45:52 javispedro, tweeted about Something big coming May 25 23:46:01 hahaha May 25 23:46:06 only with nokia messaging May 25 23:46:07 Uploaded a blurrycam of some "devices" to Flickr. May 25 23:46:24 Qole tweet about them coming for everyone at the conference May 25 23:46:28 GAN900: BOFH May 25 23:46:55 Then we uploaded a picture of the buses that came to take us to the Exploratorium last night May 25 23:47:07 "They've got windows!" May 25 23:47:21 hehe May 25 23:47:23 wtf exploratorium? May 25 23:47:42 sounds like abuse by aliens May 25 23:48:02 I bet exploratorium was at least better than that old badtasting-beer factory May 25 23:48:09 * javispedro hides May 25 23:49:50 if you want to find where the latest trolling is on TMO, just follow the latest posts abill_uk thanked May 25 23:51:39 lol May 26 00:01:02 btw what does the n900xray app do? is it a virus? :) May 26 00:01:35 * cehteh thinks its a good way to sneak in malicious software with such kinds of apps :P May 26 00:08:02 cehteh: I think it's a walpaper, no? May 26 00:09:13 If I were to sneak in malware, I'd chainload it from postinst script May 26 00:09:32 nobody seems to ever check complex postinsts May 26 00:09:40 does anyone check anything? May 26 00:09:54 not really, except SpeedEvil May 26 00:10:23 I've read I think install scripts twice. May 26 00:10:27 In practice. May 26 00:10:56 If yuou think people are reading even popular maemo apps for malware ... May 26 00:11:10 pah May 26 00:11:52 javispedro, they had a cloud chamber. May 26 00:12:13 I bet if you hide some maleware in the code you could even get it into extras May 26 00:12:18 I faintly recall some underground guerrilla action to demonstrate testing is worth nuttin May 26 00:12:24 merlin1991: Sure. May 26 00:12:30 no idea what became of it May 26 00:13:40 quite possible it turned out to be no exciting challenge after all May 26 00:14:11 plus the expected educational effect most likely would backfire May 26 00:14:26 Also: 80s cover band. May 26 00:14:48 sjgadsby may or may not have gotten video of part of the shadow council dancing. May 26 00:14:58 the shadow council May 26 00:15:05 o.O May 26 00:15:22 you lost me May 26 00:16:41 * DocScrutinizer ponders to start some tmo trolling just for fun, and to cure boredom May 26 00:17:09 until chem|st bans me :-P May 26 00:17:26 DocScrutinizer: banned, yay :D May 26 00:17:45 ;) May 26 00:18:30 javispedro, ssssh. They don't exist. May 26 00:18:36 yay for insomnia May 26 00:20:07 DocScrutinizer: oh :( can any other mail client do that? May 26 00:20:10 * javispedro remembers something May 26 00:20:16 DocScrutinizer: thanks, by the way May 26 00:20:49 jonaskoelker: dunno, seems there are few other MUAs that are really hildonized May 26 00:21:34 for cli muas i think you can find several that work and are easily or already ported May 26 00:22:00 DocScrutinizer: that do the hildon-alert-thing?? May 26 00:22:12 for sure not May 26 00:22:17 * nox- just ssh's to an *ix box and runs mutt May 26 00:22:52 * DocScrutinizer suggests $random-webmail May 26 00:23:14 great solutions to the problem I'm not having May 26 00:23:18 those web frontends are usually quite nice nowadays May 26 00:23:46 and if there's a valid usecase then it's on a phone May 26 00:23:53 for IMAP especially May 26 00:24:11 err, when IMAP is the alternative - I meant to say May 26 00:24:18 develop your own email client May 26 00:24:31 again, Nokia Messaging... May 26 00:24:40 i think i still sort thru mail faster using mutt than i could on webmail... May 26 00:24:40 nokia messaging doesn't exist May 26 00:24:43 nokia doesn't exist May 26 00:25:12 javispedro, living in denial eh? :P May 26 00:25:20 javispedro: red or blue pill? May 26 00:25:23 this student doesn't exist May 26 00:25:31 damn it doesn't disappear. May 26 00:25:41 is that supposed to be a paradox? May 26 00:26:24 jonaskoelker: please describe "hildon-alert-thing" May 26 00:27:05 as I guess there are some CLI MUAs that can execute arbitrary cmds when new mail arrives May 26 00:28:29 jonaskoelker: so if it's just about a hildon-notfication, you could implement that via dbus-send. If it's about vibrator or indicator blue flashing: same thing, dbus-send May 26 00:28:56 don't think there's hildon-notification dbus method May 26 00:29:10 http://static.arstechnica.com/2011/05/24/pain-civilized-internet.jpg May 26 00:29:21 errr, I meant that age old yellow thingie May 26 00:29:40 email style? no dbus for that :/ May 26 00:30:05 for sure there's no simple way to get proper "you missed 3 calls/mails/sms" with that blinking of upper-left May 26 00:30:41 * MohammadAG has a binary for that May 26 00:30:49 duh! May 26 00:31:14 with arbitrary text for the thumbnail in taskswitcher? May 26 00:31:27 Usage: /root/notify123 <subtitle> <icon> May 26 00:31:31 <DocScrutinizer> that'd be probably friggin useful May 26 00:32:00 <DocScrutinizer> jonaskoelker: see? ^^^ May 26 00:32:33 <MohammadAG> of course, the time is hardcoded May 26 00:32:46 <MohammadAG> since I only did it for tests May 26 00:32:49 <DocScrutinizer> you even get that from the cli mailer of your choice, as long as you can make it start a binary May 26 00:32:55 <MohammadAG> but I'll post sources tomorrow May 26 00:33:06 <DocScrutinizer> cool, now SLEEP! May 26 00:33:36 <MohammadAG> can't May 26 00:33:44 <DocScrutinizer> or get up and wash and brush your teeth as you gotta go to school in 2 hours May 26 00:33:55 <MohammadAG> http://mohammadag.xceleo.org/notify12 May 26 00:34:00 <MohammadAG> http://mohammadag.xceleo.org/notify123 a May 26 00:34:02 <jonaskoelker> DocScrutinizer: huh? May 26 00:34:05 <MohammadAG> grr, drop the a May 26 00:34:16 <MohammadAG> DocScrutinizer, nah, 6 hours May 26 00:34:22 <DocScrutinizer> ooh, ok May 26 00:34:49 <MohammadAG> hmm May 26 00:34:52 <jonaskoelker> DocScrutinizer: what were you saying to me? There's no way to get the blinking upper left? May 26 00:34:56 <merlin1991> DocScrutinizer sending kids to school since 1985 May 26 00:35:00 <DocScrutinizer> jonaskoelker: you asked for a mailer that does imap and hildon-notification-thingie May 26 00:35:09 <jonaskoelker> yep May 26 00:35:25 <MohammadAG> lol merlin1991 May 26 00:35:34 <DocScrutinizer> I told you there's a plethora of commandline MUAs to run in xterm May 26 00:35:59 <DocScrutinizer> jonaskoelker: and you get that hildon-notification-thingie via one of the suggested methods May 26 00:36:00 <MohammadAG> umm, miscalculated, 8 hours May 26 00:36:09 <MohammadAG> but I'm waking up in 5 May 26 00:36:50 <jonaskoelker> DocScrutinizer: I think I missed the suggestions -- my internets were down May 26 00:36:53 <DocScrutinizer> javispedro: wb May 26 00:36:59 <DocScrutinizer> ooh May 26 00:37:17 <javispedro> is it morning already? May 26 00:37:25 <DocScrutinizer> jonaskoelker: http://mg.pov.lt/maemo-irclog May 26 00:37:32 <MohammadAG> in some places yes May 26 00:38:09 <DocScrutinizer> javispedro: good (UGT) morning :-) May 26 00:38:25 * javispedro goes back to bed May 26 00:38:30 <javispedro> gnite folks May 26 00:38:37 <DocScrutinizer> o/ May 26 00:38:37 <MohammadAG> night javispedro May 26 00:40:21 <jonaskoelker> DocScrutinizer: thanks May 26 00:40:25 <DocScrutinizer> yw May 26 00:40:29 <jonaskoelker> MohammadAG: you had a notifier binary? May 26 00:40:45 <DocScrutinizer> [2011-05-26 02:33:46] <MohammadAG> http://mohammadag.xceleo.org/notify123 May 26 00:40:45 <MohammadAG> http://mohammadag.xceleo.org/notify123 May 26 00:40:46 <jonaskoelker> By the way, anyone knows if modest advertises its mail receipts anywhere (i.e. dbus?) May 26 00:41:04 <MohammadAG> no idea, but dbus-monitor --session would help May 26 00:41:13 <jonaskoelker> got source? May 26 00:41:15 <MohammadAG> it helped me code the mediaplayer clone :p May 26 00:41:23 <MohammadAG> notifier? on the laptop yeah May 26 00:41:43 <jonaskoelker> pastebin? May 26 00:41:56 <MohammadAG> can't, not on my laptop May 26 00:42:14 <DocScrutinizer> [2011-05-26 02:32:41] <MohammadAG> but I'll post sources tomorrow May 26 00:42:18 <jonaskoelker> ah, fair May 26 00:42:32 <MohammadAG> or in 5h May 26 00:43:29 * DocScrutinizer swears as can't sit nor stand nor lay down anymore, back hurts like hell May 26 00:45:03 <DocScrutinizer> MohammadAG: please fix the hardcoded timer May 26 00:45:21 <MohammadAG> DocScrutinizer, sure thing May 26 00:45:25 <DocScrutinizer> :-D May 26 00:45:40 <MohammadAG> gimme 5h till I wake up May 26 00:45:46 <MohammadAG> well, sleep and wake up May 26 00:45:49 <DocScrutinizer> lol, NOOO May 26 00:45:56 <DocScrutinizer> ;-D May 26 00:46:13 <MohammadAG> I wish I could IRC while I'm asleep May 26 00:46:16 <MohammadAG> or code May 26 00:46:51 <MohammadAG> my facebook app got most features in one sleepless day May 26 00:46:55 * DocScrutinizer waves May 26 00:47:15 * MohammadAG waves back, tries to sleep again May 26 01:56:24 <Pavlz> hello May 26 01:57:13 <Pavlz> how to update the nokia tablet 770 to the last OS ? May 26 01:57:33 <Pavlz> i got again OS2006 May 26 01:57:56 <Pavlz> i need something more recent May 26 01:58:55 <Pavlz> for the simple situation: SSL's certificate opera expired May 26 01:59:14 <Pavlz> SSL'S certificate google expired May 26 01:59:29 <Pavlz> old flash May 26 01:59:57 <Pavlz> how to do it ? May 26 02:42:44 <andre__> Pavlz, http://wiki.maemo.org/Updating_the_tablet_firmware#Updating_.26_Flashing_your_device **** ENDING LOGGING AT Thu May 26 02:59:57 2011