**** BEGIN LOGGING AT Sun May 30 02:59:56 2010 May 30 18:13:25 hi, is anybody awake? May 30 18:15:17 I have a question concerning the current hardware development and if the gta03 is really dead. May 30 18:20:47 yes it is May 30 18:24:32 Unless someone comes up with a large slice of a million dollars for dev work May 30 18:27:20 SpeedEvil: what i can't understand is how so many chinese copycat shops manage to produce so many distinct devices, while free software/hardware companies always can't afford even the first step? May 30 18:27:53 I'm unsure how distinct they are. May 30 18:28:05 There are several issues I think. May 30 18:28:39 Possible lack of protection, and ability to buy working phone designs 'off the shelf' - which you just munge a bit, and put a case on. May 30 18:28:43 Cheaper labour. May 30 18:28:55 Already having a factory pumping out phones. May 30 18:30:12 SpeedEvil: the case is what bothers me May 30 18:30:23 Everybody says producing _any_ case is prohibitively expensive. May 30 18:30:49 I'm not sure that's quite true. May 30 18:30:59 There must be toolmaking houses in china. May 30 18:31:16 SpeedEvil: as to the working phone designs, yes, they have some kind of common knowledge base (schematics/software/etc), do you want me to give you a link to an interesting article about that? May 30 18:32:04 I question http://cgi.ebay.co.uk/CLEAR-CRYSTAL-HARD-CASE-SKIN-COVER-POUCH-Nokia-N900-/300430878736?cmd=ViewItem&pt=UK_Mobiles_Accessories_RL&hash=item45f3136810 - uses a really expensive mould May 30 18:32:08 sure May 30 18:34:22 SpeedEvil: bunnie's blogpost (he's the man behind Chumby, and the one who has first hacked Xbox): http://www.bunniestudios.com/blog/?p=284 May 30 18:34:52 Ah - not read that one. May 30 18:34:57 Lots of good stuff on that blog May 30 19:35:29 does anybody know, if the financial calculations for the hardware development are open somewhere? May 30 19:38:46 or other question: how many gta02 do we need to sell to revive 03 May 30 19:40:05 maybe we could lower the costs with a reprap-machine May 30 19:44:24 kardan, more than 100'000 I think May 30 19:45:50 PaulFertser, I can say you why May 30 19:46:14 because most "open source" users, don't want open hardware May 30 19:46:34 they want open hardware with specs like upstream devices May 30 19:46:38 max_posedon: have you read messages on community ML about cost of producing a prototype case? May 30 19:46:49 I think not May 30 19:52:05 I dont think, they dont want open hardware but atm its too hard imaginable to work for it. and its necessary to organize matter which is a bit more than bytes May 30 20:00:44 Is it really impossible to take the financial risk as a community? maybe like this: http://www.cofundos.org/projects.php?tag=openmoko May 30 20:04:00 so the question is, how do we collect 5 million for the first steps (do we really need so much?)? May 30 20:05:22 kardan: do we need any decent things at all? Given how much real hackers Openmoko attracted and how much lusers etc etc i really doubt people really need decent (hacker-wise) smartphones. May 30 20:05:34 Screw that May 30 20:05:42 The community was weak right from the start. May 30 20:08:57 The considerable amount of sane people unsubscribed from the community ML just because of all the useless questions and discussions there. May 30 20:12:29 seems like an organizational problem to separate support and dev mailinglists May 30 20:12:33 No May 30 20:12:41 kardan: there was a support ML May 30 20:14:40 kardan: it's just that people are not interested. Most developers were happy enough with Maemo or iPhone or whatever. I couldn't believe my eyes when i saw how little people are really interested in a free smartphone. I thought there're plenty of hackers around wanting to have such a device in their pockets. Wishful thinking eh, reality proved me wrong, just as it does every time i hope for the best. May 30 20:15:54 i see it like this: even if there is no mainstream market for OH phones there is the a need to have them. so why cant build the hacking community work alltogether in a big support company that collects money for hardware development May 30 20:17:27 * JaMa fully agree with PaulFertser May 30 20:17:29 kardan: calculations were done many times already: to get access to modern components you need to produce at least hundreds of thousands of devices, not some 10 000. And without modern components one is stuck in the past forever. May 30 20:18:39 its always also a question how well information is distributed and: people move very slowly. for me it took months that I visited the OM wiki the second time and even more, that I got my first one. I never had the feeling I need this just now, but I surely know that this project needs to be alive. May 30 20:18:43 everybody talks about OH phones and how comunity would help itself, but we're looking for experienced gta01 owner for really long..ffs May 30 20:19:47 JaMa: and funny thing is that i have gta01 now but i'm totally demotivated May 30 20:19:58 there was so many great hackers involved in OM, but how many people does some development now? May 30 20:20:00 Not funny rather sad May 30 20:21:00 hard to read that but I know that feeling from another scene. for me the solution always is: keep it running. even if you go slow for some time, dont drop it. May 30 20:21:02 JaMa: btw, the fact you agree with me atm is not cool -- i'm in a rather foul mood. May 30 20:21:08 PaulFertser: I know you have gta01 and I know you're experienced much more then it's needed for that testing and also involved in other OM related things.. May 30 20:21:38 PaulFertser: so I didn't have you in mind when I wrote that.. May 30 20:22:12 I'm not involved in anything productive lately, unfortunately. May 30 20:22:36 just switching SHR to 2.6.32+ means no more gta01 images from me (as well as broken packages in feeds) May 30 20:23:26 but I refuse to work on something I have no way to test (and I don't want to buy another gta phone just to test something I'll never use - as I'll prefer gta02) May 30 20:24:49 i dont understand why development is still done for gta01 while the 02 is available May 30 20:25:37 kardan: for users already owning those gta01 May 30 20:26:05 kardan: both gta are quite close, so it's no "extra development only for gta01" May 30 20:27:02 maybe the problem is also that everybody expects community driven development for free which is somehow insane for the amount of work May 30 20:27:26 right now I know that for usable gta01 image we need only: defconfig for newer kernels, xinput calibration confirmation, check sysfs nodes - create patch for FSO, because some sysfs nodes will be probably now different between gta01 and gta02) May 30 20:27:35 kardan: i expected more community members who know basics of *nix systems at least. May 30 20:28:51 can you give a number of experianced unix developers? May 30 20:29:38 kardan: yes.. just a few hackers paid to work on it full-time can do more than 3 times more volunters in free time (after returning from daywork) May 30 20:30:18 PaulFertser: I agree... I get so many messages from people who say they "would help if they could". I always try to tell them it's not as tough as it looks but it doesn't make any difference May 30 20:32:24 Weiss: looks like GNU/Linux systems finally reached those poisoned by the consumerist society. Not sure if that's good or bad. One of my friend promised to switch to *BSD if that happens. Now, as several years passed, he switched from Slackware to Ubuntu instead (/me's confused) May 30 20:32:39 odd.. May 30 20:33:11 hehe May 30 20:33:24 maybe consumerist society got him instead.. May 30 20:33:46 I think a large part of the OM situation right now is that people are scared to tackle difficult projects. making a smartphone work properly is just really hard, and nothing will change that. May 30 20:34:00 JaMa: he moved from RU to JP May 30 20:34:32 for example, whatever happened to all the people who were interested in working on 3D drivers? somewhere I went from being "someone who didn't really know anything but would help if he could" to the "all-knowing Glamo guru", or something May 30 20:35:52 (and, in case anyone didn't notice, I'm not all-knowing) May 30 20:35:56 some clever and interested guys imho moved on after seeing that gta phones aren't so big success as all hopped they will be May 30 20:36:19 Weiss: it's the overall amount of the lowlevel hackers involved that is too low, i guess. Not sure why it happened, probably all the capable folks are already hired by RedHat/Intel/MontaVista/whomever and have no time and passion anymore. May 30 20:36:23 ie Balrog who did first mplayer->glamo hack completly lost his interest later.. May 30 20:37:51 and even those bits were from big parts written originally by openmoko employees and glued together to work with mplayer May 30 20:38:03 yeah May 30 20:38:49 this is also true of, say, regular X.org development as well - few people are really into the scary low-level stuff May 30 20:39:50 Especially given the amount of abuse they face. Even the Intel devs are very unhappy with docs they get for the development. And with the buggy hardware. May 30 20:40:03 indeed May 30 20:40:55 * Weiss enjoys a challenge (and doesn't mean "duh, a challenge I could do!" Simpsons-style) May 30 20:41:01 JaMa: the glamo can't do much more than what is already implemented in mplayer, at least according to the specs May 30 20:41:35 Weiss: of all the X.org hackers i read/heard i like Luc Verhaegen the most. Too bad AMD killed radeonhd development and Novell screwed the devs. May 30 20:41:46 <[Rui]> Weiss: there's a mobiphil who seems willing to learn some more about glamo :) May 30 20:42:10 balrog-k1n: but it would be nice to see this functionality implemented in a non-hacky way, compatible with KMS+DRM blablabla May 30 20:42:11 ie, I would like to dive deeper in some low-level stuff, but after spending all "free" time by poking to all sorts of recipes/project I don't have time to spend more time concentrated on some particular low-level problem May 30 20:42:33 balrog-k1n: hi :), great to see you here May 30 20:42:36 [Rui]: sure.. I wish he'd get involved with the kernel stuff instead of asking questions so often May 30 20:43:21 PaulFertser: well yeah.. though it costs time and it seems it wouldn't be for a huge gain because the hardware is what it is May 30 20:43:29 nicer hardware is available now May 30 20:43:37 balrog-k1n: yep, everything's too late :| May 30 20:43:59 balrog-k1n: I still appreciate your work and belive it can't do much more.. I was thinking more of integrating it with weiss's drm infrastructure or implementing it as vdpau to be more likely accepted upstream May 30 20:44:28 balrog-k1n: nicer in some aspects, though i still do not know about a better educational platform hardware-wise. Everything else is too expensive and closed at the same time. May 30 20:45:15 PaulFertser: true... have you looked at thing like the beagleboard though? May 30 20:45:16 But it doesn't seem many universities are interested in that kind of education i envision. May 30 20:45:40 (btw i've always been advocating switching to omap cpus for the gta02 and i still think the hardware would be less crash happy if it was omap based) May 30 20:45:42 balrog-k1n: sure, i know about it. Not as capable as FR, lacks perepherals and mobility. May 30 20:46:09 But it's a better CPU indeed. May 30 20:46:31 balrog-k1n: have you seen beagle+gta hybrid from golden delicious? May 30 20:46:36 balrog-k1n: OTOH as far as i know it's still not supported properly and fully upstream, i think that tells a lot. May 30 20:47:00 PaulFertser: ah didn't know that May 30 20:47:12 JaMa: nope, let me google it up :) May 30 20:48:01 <[Rui]> The Linux 2.6.3x show the phone is quite capable May 30 20:48:20 <[Rui]> a lot of things are unreasonably slow because of all the debug May 30 20:48:28 balrog-k1n: probably i'm lacking some info... What i really know is that the integrated MUSB core is not documented openly anywhere and that the driver is utter crap. May 30 20:49:27 PaulFertser: the nokia tablets use the MUSB thing too.. is it that bad? or is it a different one? May 30 20:49:56 balrog-k1n: http://www.handheld-linux.com/wiki.php?page=Openmoko%20Beagle&referer=Neo%20Freerunner May 30 20:49:58 balrog-k1n: folks are struggling for fucking several months trying to get USB host support on N900: http://talk.maemo.org/showthread.php?p=658278#post658278 May 30 20:51:20 JaMa: fun idea May 30 20:51:33 balrog-k1n: announcement http://lists.openmoko.org/pipermail/community/2010-May/061829.html May 30 20:53:21 * PaulFertser looks at the weather forecasts and sees he'll have to use the underground instead of a bicycle: 28km through the city in a heavy rain is not something fun enough :/ May 30 20:53:37 * Weiss has heavy rain at the moment May 30 20:53:56 the HVV website also suggested a connection which doesn't physically work, so I spent ten minutes standing in it, as well May 30 20:54:23 balrog-k1n: btw, do you use an MTB or a road bike for utility cycling? May 30 20:54:40 road racing i meant May 30 20:54:53 PaulFertser: MTB, but always wanted to switch :) May 30 20:54:59 raining here as well May 30 20:55:27 PaulFertser: the annoying thing about non MTBs is they're too easy to break May 30 20:55:29 balrog-k1n: aren't you afraid of the speed it gives? ;) May 30 20:58:30 balrog-k1n: my bike computer on my "former MTB" (it got steel fork, slick tires and low handlebar configuration now) shows i reached 59 km/h somewhere on my way from/to work. :) I wonder what i'll get on my road racing bike :) May 30 20:59:02 whoah May 30 20:59:34 PaulFertser: good thing i have no speedometer.. i'd be scared knowing i was going > 40 May 30 21:00:02 balrog-k1n: hehe :) May 30 21:00:04 PaulFertser: do you use the freerunner as the bike computer? May 30 21:01:06 balrog-k1n: no, i have some old Cateye that my pal found in a junk box at a bike shop many years ago. May 30 21:01:28 Attaching a FR not worth the trouble. Especially given how "readable" it's in the daylight. May 30 21:01:50 And that i'd need to connect the sensor somehow. And write a software to process those pulses etc etc. May 30 21:02:14 i knwo people use it for mapping in OSM, special handles are for sale May 30 21:02:15 It would be a funny project if i was in a happy state probably. May 30 21:02:21 <[Rui]> yeah, I wonder how the price would change if the screen was the same transreflexive OLPC uses May 30 21:02:39 <[Rui]> in grayscale mode it would help a lot for the power consumption levels May 30 21:03:18 for gps tracking it is enough to keep it ON in small bike pocket May 30 21:03:47 but I was too lazy to post-process it after returning from trips :/ May 30 21:04:12 Jama thanks for the links. that was the info I came here for. lets make it work! May 30 21:04:46 JaMa: i find most GPSes, including the gta01/02 give better results when they're held in the a constant direction relative to your movement May 30 21:05:08 and in the pockets they shake and turn around May 30 21:05:45 plus i guess some people want to launch tangogps to see what parts they have already mapped on the screen May 30 21:05:50 To go off-topic a bit - I've had OK results with the n900 May 30 21:06:04 well I'm using narrow part of bigger pocket attached to bike.. so it was in almost constant direction for me May 30 21:06:19 http://www.openstreetmap.org/user/SpeedEvil/traces/584852 - traces from my trouser pocket while driving May 30 21:07:05 balrog-k1n: I was using cgps-logger on cmdline and launched tango or navit only in case I wasn't sure where I should go (few times a day) and I had to switch batteries once a day :/ May 30 21:07:25 balrog-k1n: even with display mostly off May 30 21:07:54 balrog-k1n: the biggest surprise for me was rain: the winter was so long that i got used to the snow, and it doesn't really make one cold. Apparently now one needs to carry some raincoat just in case. May 30 21:09:21 Or do you use some membrane clothing? May 30 21:32:37 Anyway, time to try to hide for some hours from this sick world. cya May 30 21:33:03 nn May 30 21:33:14 sleep well :) May 30 21:34:39 gnight! May 30 21:37:47 <[Rui]> PaulFertser: see ya May 30 21:48:26 Thanks folks :) May 31 02:21:13 watisit **** ENDING LOGGING AT Mon May 31 02:59:56 2010