**** BEGIN LOGGING AT Wed Jun 25 02:59:57 2008 Jun 25 03:00:51 Was zhone a from-scratch project from the Openmoko folks, or did it have some earlier incarnation on another platform? Jun 25 03:01:13 it had Jun 25 03:01:15 i wrote it Jun 25 03:01:41 Really! What was it originally targeted for? Jun 25 03:01:46 http://neo1973-germany.de/wiki/pyPhone Jun 25 03:01:50 lol the same of course Jun 25 03:02:05 but it didnt use the framework at that time as it was non existant Jun 25 03:02:33 So it originalted with the "underground" folks? Jun 25 03:02:40 completely Jun 25 03:02:59 Cool. I like your pyPenNotes! Jun 25 03:03:12 this is actually the work of kristian-m ^^ Jun 25 03:03:23 and zad is from emdete Jun 25 03:03:57 Is zhone still pure python? Jun 25 03:04:05 why not? Jun 25 03:04:09 100% python Jun 25 03:04:15 Just asking... Jun 25 03:06:22 suspend/resume is pretty reliable in asu these days Jun 25 03:06:47 be wary of python Jun 25 03:07:08 unless of course you like your 400mhz phone returning to the land of a 10mhz PIC. Jun 25 03:07:08 Why be wary of Python? Jun 25 03:07:25 python is slow Jun 25 03:07:34 you pay a hefty price for using it (being lazy) Jun 25 03:07:37 myths i would call them Jun 25 03:07:42 a very hefty price Jun 25 03:07:48 we at neo1973-germany do everything in py Jun 25 03:07:49 josch: i've done actual benchmarks Jun 25 03:07:56 of course Jun 25 03:08:07 the hard work is still done in c Jun 25 03:08:22 python clocked in about 290 times slower than c Jun 25 03:08:32 this depends on what you do with it Jun 25 03:08:37 for doing the exact same logic/operations Jun 25 03:08:46 this was apples vs appls Jun 25 03:08:58 the more u do in python Jun 25 03:09:05 the more of your execution stack is in python Jun 25 03:09:09 and the bigger price you pay Jun 25 03:09:12 thus - beware Jun 25 03:09:12 if you write gui apps like e* apps then you probably wont see any difference between c and python since in both cases the e* libs do all the work Jun 25 03:09:32 i have heard much trumpeting of "do everything in python! everything must be python" Jun 25 03:09:48 hrm... Jun 25 03:09:50 and it makes me go "well there goes that project. good luck." Jun 25 03:09:58 yuou do only the most minimal layers in pythoin Jun 25 03:09:58 this should be impossible Jun 25 03:10:01 if you use it Jun 25 03:10:09 good luck with writing 2d grahics in pure python :P Jun 25 03:10:16 and maybe with the highe rlevel app - yes Jun 25 03:10:29 never said anything else Jun 25 03:10:33 but a perfect example is there is a "list widget" the python etk bindings implement Jun 25 03:10:36 python is for high level work Jun 25 03:10:37 and its implemnented IN python Jun 25 03:10:45 or for prototyping things that are later written in c Jun 25 03:10:52 now that has been dropped for using core evas objects and etk directly Jun 25 03:10:55 instead oft the python widget Jun 25 03:11:00 there is a noticable speedup Jun 25 03:11:21 one would use the etk bindings of course Jun 25 03:11:24 the problem is that some people want everything in python so "Designerts" can do everythign they like Jun 25 03:11:28 things like processing Jun 25 03:11:31 all those flash demos Jun 25 03:11:35 haha Jun 25 03:11:37 where they do "physics engines" Jun 25 03:11:40 in "flash" Jun 25 03:11:44 you know what i mean Jun 25 03:11:47 noone ever spoke of doing everything in python Jun 25 03:11:55 YOU didnt Jun 25 03:11:59 but others around have Jun 25 03:12:30 others forcing python into "everything must be python" Jun 25 03:12:33 * josch just used ctypes yesterday to profit from some real c-libs :D Jun 25 03:12:33 for us Jun 25 03:12:54 the same people point at things like processing Jun 25 03:13:04 and say "we want designers to use processign to make ui's" Jun 25 03:13:18 and in the same breath say "everything must be python" Jun 25 03:13:22 i know exactly what they want Jun 25 03:13:27 and i know what they dream of Jun 25 03:13:34 and i know just how much that is going to suck. Jun 25 03:13:47 it works on desktops because you have 200x the processing power Jun 25 03:14:26 That's why mobile platforms are fun. Efficiency is still important. Jun 25 03:14:43 so to have that flexibility you need more and more 'python' doing lower and lower level things Jun 25 03:14:43 on the neo it even works for a bunch of dbus daemons in the back but those will probably be reimplemented in c someday Jun 25 03:14:52 implementing whole pyhsics "engines" or animation loops Jun 25 03:14:55 widgets Jun 25 03:14:56 etc. Jun 25 03:15:20 raster, do you see this movement in openmoko? Jun 25 03:15:25 yes Jun 25 03:15:29 o0 Jun 25 03:15:38 nah - no way Jun 25 03:16:07 thats my point Jun 25 03:16:21 there is a place for python Jun 25 03:16:33 and it sure as hell is not one of an actionscritp replacement on a phone Jun 25 03:17:14 for simple high level logic that is not part of a core ui or mainloop Jun 25 03:17:21 its good for high level logic glue Jun 25 03:17:52 it is NOT good at stuff u call in animatin loops to calculate the positions/properties of 100's of objects dzens of times per second Jun 25 03:17:53 :) Jun 25 03:18:17 just the cost of going from c to pythgon and back (marshalling/unmarhsalling) all the parameters and returns is nasty overhead Jun 25 03:19:10 * mwester-trip wishes everyone would just punt the stupid fancy graphics to the back of the line, where it belongs, so that people can focus on making the device do what it was designed to -- like make phone calls!! THEN we can argue about what language fancy wihizzy graphics should use! Jun 25 03:20:30 I call it "GUI-first" development -- when developers mimic marketing types, who (of course) design products GUI-first... it's rather like wallpapering a building, and then trying to erect the walls behind the paper flapping in the wind. Jun 25 03:20:51 dont look at me Jun 25 03:20:58 haha Jun 25 03:21:03 those who pay me insist on having their mockups done first Jun 25 03:21:09 so they can ":see how it feels" on a phone Jun 25 03:21:22 I'm not looking at anyone on this channel, I expect -- the guilty don't show up here! Jun 25 03:21:24 long before any 43 thoguht/design is put into it to make it a long term sustainable system Jun 25 03:21:35 the MOSt important thing is the mockup being pixel-perfect on the device Jun 25 03:21:52 i seed mockups Jun 25 03:22:09 and the first thing i do is tryy and identify what is a common theme and requires a common subsystem or featur Jun 25 03:22:13 and that may need a lot of back-end work Jun 25 03:22:23 but some people just dont care and want quick hacks Jun 25 03:22:34 because "look- this demo app here does it - you can do it for everything too" Jun 25 03:23:24 (even tho the demo app just opens a fullscreen window and does everything itself as opposed to having an actual SYSTEM where multiple processes interact and share the screen - in an organised way, that gets you the same effects, but requires a mountain more infrastructure to happen) Jun 25 03:24:25 Clearly the PHB's have mocked up the wrong things then -- the purpose of a mockup is to ensure that the design is correct before committing to it -- we now have three completely different images for the phones, none of which is really complete... that should never have happened. ( But I'm "preaching to the choir", so to speak) Jun 25 03:25:14 asu is meant to supsercede the old om images Jun 25 03:25:17 so that doesnt matter Jun 25 03:25:31 fso is enytirely separate to the phb designs. Jun 25 03:25:37 so as such asu is its own bubble Jun 25 03:25:40 as is fso Jun 25 03:25:52 and the gtk+matchbox etc. stuff is the old legacy bubble Jun 25 03:25:57 But it sounds like ASU doesn't use the D_Bus structure, and is a dead-end. Jun 25 03:26:00 so that isnt really an issue as such Jun 25 03:26:06 rtm: no it doesnt Jun 25 03:26:39 what should happen is a mockup as an idea Jun 25 03:26:49 raster: You mean, no it doesn't use D-Bus, right? Jun 25 03:27:01 then engineering come bnack with "nide idea - but 80% of it is not possible without massive infra work, so the folowing features you use must be cut out" Jun 25 03:27:05 raster: you can't waste time on mockups! Jun 25 03:27:08 :/ Jun 25 03:27:10 rtm: well dbus exists - of course Jun 25 03:27:12 ASU is a tangled mess of code, almost unmaintainable, and is IMO unlikely to survive under the crushing weight of that Qtopia bloat that barely works. Jun 25 03:27:23 but its not using dbus for making calls or stuff Jun 25 03:27:27 qtopia does that Jun 25 03:27:41 exactly. Jun 25 03:27:46 SpeedEvil: face it. the phb's are designers. not engineers. Jun 25 03:27:48 they want to do design Jun 25 03:27:51 :/ Jun 25 03:27:53 they want to see their designs on the phones Jun 25 03:27:53 So it's a dead-end. Jun 25 03:28:02 simple Jun 25 03:28:10 me, you and everyone else is irrelevant to that Jun 25 03:28:21 their designs (and designs of their designer friends) are what is important Jun 25 03:28:21 Agreed. Jun 25 03:28:26 raster: and changing horses in mid-stream shows you're dynamic. Jun 25 03:28:30 That's why I'm throwing my support behind FSO. Jun 25 03:28:34 and the community, their designs, or experience are not a factor in that. Jun 25 03:28:46 they are designers Jun 25 03:28:52 Let a thousand flowers (partially) bloom! Jun 25 03:29:01 and their designs are what is the core thing to implement Jun 25 03:29:11 SpeedEvil: re: changing horses comment -- are you sure you're not in marketing ;) Jun 25 03:29:25 and that is what we are implementing Jun 25 03:29:27 and we get paid to do Jun 25 03:31:24 So what's the current direction? Qtopia + other stuff? Jun 25 03:31:30 imho the idea of multiple widget sets is fine Jun 25 03:31:34 x is a common factor anyway Jun 25 03:31:56 imho its better than qtoipia (pure/qws) or android Jun 25 03:32:10 u dont have a ":brand new" windowing and toolkit system to have to port to Jun 25 03:32:15 you can use whatever works best for you Jun 25 03:33:35 the dbus apis in fso are just that Jun 25 03:33:40 back-end dbus apis Jun 25 03:33:46 they can slide in almost anywhere Jun 25 03:33:57 * jeffszusz wishes he would have come in at the beginning of the discussion: is there a Spark's Notes version? Jun 25 03:34:27 i see qtopia as a temporary solution in asu Jun 25 03:34:27 jeffszusz: I'm gathering that raster is ready to explode. Jun 25 03:34:33 :/ Jun 25 03:34:34 i personalyl didn't want it Jun 25 03:34:38 it just doesnt feel right Jun 25 03:34:52 but it seems to partially work Jun 25 03:34:53 well make calls Jun 25 03:34:56 kind-of Jun 25 03:35:02 and kind-of answer them Jun 25 03:35:04 qtopia on the Neo was actually openmoko? I thought someone had put qtopia on it for laughs Jun 25 03:35:13 theres been a hefty list of gettign qtopia to behave well under X11 Jun 25 03:35:35 like still having double-draw issues where the qpe apps lay out and raw their window teice on show Jun 25 03:35:46 all sorts of assumptions on the windowing system Jun 25 03:35:52 ie wqs vs x11 Jun 25 03:35:55 qws Jun 25 03:36:10 things like the qpe softmen being a hybrid of x11 and qcom communication Jun 25 03:36:26 whihc all in all has race conditions designed into it as such Jun 25 03:36:35 so these are still being fixed by zecke Jun 25 03:36:59 right now qpe takes about 15 seconds to respond to the modem going "incming call!" Jun 25 03:37:06 and actually showing the incoming call window/screen Jun 25 03:37:11 wich is just unusable Jun 25 03:37:22 Does FSO use any of the Qtopia stuff? Jun 25 03:37:29 If they want to call you, they'll hang on. Jun 25 03:37:33 It's a positive! Jun 25 03:39:28 yeah Jun 25 03:39:33 zecke is looking into this Jun 25 03:39:46 but it has a lot of features as such Jun 25 03:39:50 so maybe its just teething problems Jun 25 03:40:46 hmm Jun 25 03:41:42 * jeffszusz speculates about and hopes for an eventual OpenMoko PDA Jun 25 03:45:20 how is the freertunner different to a pda? Jun 25 03:45:30 other than u get wifi and bluetooth and gsm/gprs too? Jun 25 03:46:32 I think software. Jun 25 03:46:34 i suppose i should have said "internet tablet" Jun 25 03:46:38 Ah Jun 25 03:46:56 think Nokia N810 with open hardware, more open software Jun 25 03:47:10 so other than software Jun 25 03:47:13 though the Freerunner's GPS + GSM = awesomer Jun 25 03:47:17 hos is the freerunner not that? Jun 25 03:47:24 (other than smaller screen) Jun 25 03:47:26 bigger screen, Raster. and a slide-out keyboard so you don't have to use the stylus. Jun 25 03:47:27 ? Jun 25 03:47:35 well n800 didnt have a kbd Jun 25 03:47:41 and its an internet table/pda Jun 25 03:47:42 :) Jun 25 03:47:44 I've been contemplating a PDAish. Jun 25 03:47:47 u werent very specifci :) Jun 25 03:47:54 N810 has a keyboard... also, the bigger screen allows more of a desktop-view instead of a list Jun 25 03:48:11 openmoko: 03zecke 07org.openmoko.asu.dev * r498fca7d6218 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Build a Qtopia version with Theme performance improvements Jun 25 03:48:20 8 bit color screen, 4.3" display, GSM, GPS, 120*25*80mm or so. Jun 25 03:48:25 right now it makes much more sense to do a phone Jun 25 03:48:32 as it just has a much bigger market out there Jun 25 03:48:46 an open device is already a niche enough market Jun 25 03:48:52 may as well try the biggest niche there is Jun 25 03:48:53 :) Jun 25 03:48:57 and I would personally give a public commendation to anyone who ported the N810-specific Canola2 media player to it. Jun 25 03:49:22 aaah canola Jun 25 03:49:23 :) Jun 25 03:49:32 yup Jun 25 03:49:44 shouldnt be too hard Jun 25 03:50:26 my line of thinking goes "mmmmm Canola. I want an n810 just for Canola. No, wait. I want an OpenMoko phone... and internet tablet would be redundant...Mmmm...Canola on a Neo with a bigger screen) Jun 25 03:50:52 well canola should be able to work in the fr's screen Jun 25 03:50:57 just it wont be that snappy Jun 25 03:51:03 it's be usable Jun 25 03:51:06 hm Jun 25 03:51:09 You could put a bugger screen on the neo quite easily. Jun 25 03:51:12 it'd be Jun 25 03:51:15 you could? Jun 25 03:51:15 Bigger Jun 25 03:51:19 All you really need is the case. Jun 25 03:51:21 but the n800/810 have significantly better video hw Jun 25 03:51:38 'quite easily' is assuming you're happy with a SMT rework station. Jun 25 03:51:48 SMT? Jun 25 03:51:59 In other words, not you :) Jun 25 03:52:06 * jeffszusz = not a developer Jun 25 03:52:08 Surface Mount Technology. Jun 25 03:52:27 Basically if you are scared of attacking working hardware with hot bits of metal, no. Jun 25 03:52:36 i support open source...and advocate it, and use it, etc. but I can't code worth a damn yet. Jun 25 03:52:51 aaah Jun 25 03:52:57 well then just buy stuff that uses open source Jun 25 03:53:01 and i haven't touched a soldering iron since grade 9 Jun 25 03:53:05 lol Jun 25 03:53:06 and that at least is true to oss's ideals Jun 25 03:53:08 or mostly Jun 25 03:53:20 ie mstuff that ships source Jun 25 03:53:26 gives back to the community and upstream devs Jun 25 03:53:34 mhm Jun 25 03:54:03 In some ways I'm leaning towards the elegance of the GTA01 shipped image. Jun 25 03:54:19 how so? Jun 25 03:54:42 No root filesystem at all. Just a working kernel. Jun 25 03:54:58 lol...ouch Jun 25 03:55:09 hahaha Jun 25 03:55:20 * pjz playing with writing an android app. Jun 25 03:55:23 SpeedEvil: "working" kernel was a bit of a stretch Jun 25 03:55:35 raster: sort-of, yes. Jun 25 03:55:36 pjz, wrong channel Jun 25 03:56:08 pjz: #android Jun 25 03:56:13 raster, you will laugh - i rly manage to do useful stuff with this kernel :P Jun 25 03:56:21 josch: ahahahah Jun 25 03:56:26 josch: Come on, we could have converted pjz. Jun 25 03:57:58 jeez, chill, I wasn't asking for help or anything, just noting what I'm doing at the moment. Jun 25 03:58:20 besides, FR could run android Jun 25 03:58:40 yeah probably also windows ce Jun 25 03:59:00 anyone played with the FSO images? Jun 25 03:59:36 if playing is more than using zhone, then yes Jun 25 04:04:53 josch: I installed FSO, and while I could get it to make calls (most times) I couldn't hear either way after picking up :-( Jun 25 04:05:04 I suspect I just got a bad image, but it was one that someone had recommended Jun 25 04:05:19 I also heard that it supports suspend/resume, but I didn't see any way in the UI to suspend Jun 25 04:05:51 hehe Jun 25 04:05:54 it supports it Jun 25 04:05:58 as it does sms Jun 25 04:06:17 but fso is not so much about the ui as it is about the things behind it Jun 25 04:08:06 yeah, I know, but when I'm in a position to actually need a mobile phone, access to features through the UI is nice :-) Jun 25 04:08:22 then you have to use asu Jun 25 04:08:23 I'm not at all knocking fso - it's early, and I really like the direction Jun 25 04:08:39 josch: er, afaik you can't make or receive calls with the asu Jun 25 04:08:41 for the first time the direction is to NOT put the ui first Jun 25 04:08:48 and still people complain :P Jun 25 04:08:49 josch: I use scaredycat's gtk imgaes Jun 25 04:09:10 I didn't see anyone complain Jun 25 04:09:15 sry - i barely have experiences with asu Jun 25 04:09:29 i'm mostly with debian, zad and fso Jun 25 04:09:38 debian on the neo? Jun 25 04:09:43 -.- Jun 25 04:09:53 i wrote this comprehensive tutorial: Jun 25 04:10:00 http://wiki.openmoko.org/wiki/Debian Jun 25 04:10:09 i also have some ready to use tar archives there Jun 25 04:10:54 I see how to install Jun 25 04:11:00 But not how to use it to do anything :-) Jun 25 04:11:07 huh? Jun 25 04:11:22 e.g. make a call, listen to music, type stuff in terminal Jun 25 04:11:22 hahaha Jun 25 04:11:34 for making a call either use AT commands via serial terminal or Jun 25 04:11:46 install some openmoko phone stack of your choice Jun 25 04:11:49 this is only debian Jun 25 04:11:58 Why would I spend 2 hours overwriting my neo install to put it in a state where I can't do anything? Jun 25 04:12:00 nothing i tweaked to be usable as a phone Jun 25 04:12:09 not 2 hours Jun 25 04:12:19 you just put the ready to use tar on your sd card Jun 25 04:12:23 and boot from it Jun 25 04:12:33 I thought so, but my point is to get something I can primarily use as a phone, media player, + gps Jun 25 04:12:33 then open up the console and xkbd Jun 25 04:12:40 and apt-get some good stuff Jun 25 04:12:41 then if I can hack on it, that's great Jun 25 04:13:04 well for hacking debian is great because you no longer have to use bitbake to have your stuff compiles Jun 25 04:13:05 d Jun 25 04:13:12 :-) Jun 25 04:13:16 i always natively compile on the neo now Jun 25 04:13:50 Yeah, it's great work... if I had a 2nd neo to leave connected to my linux machine, I might install it to play with Jun 25 04:14:42 but my goal now is to find something that works better than the unsupported GTK to do daily phone stuff with, and so far that's not there Jun 25 04:14:55 correct Jun 25 04:15:12 well, actually, Qtopia would be it, if I could find a Qtopia image now :-) Jun 25 04:15:20 For 01? Jun 25 04:15:24 there are many projects working on that but so far none is usable as an everyday phone Jun 25 04:15:25 yeah Jun 25 04:15:30 kdean06: yes Jun 25 04:15:45 I've got one... Jun 25 04:15:46 josch: I'm putting my money on fso Jun 25 04:15:56 kdean06: linksies? Jun 25 04:16:12 fso is about the framework and NOT the gui in the first place Jun 25 04:16:23 josch: sure Jun 25 04:16:24 you can already do stuff like phone, sms, contacts with fso Jun 25 04:16:52 so what you bet on is that Jun 25 04:17:00 josch: You mean the framework can do it, or *I* can do it, while I'm out and about with my mobile? Jun 25 04:17:23 framework can do it and you can do it with a keyboard and a bash console Jun 25 04:18:55 josch: it might be worthwhile for me to throw together a crap UI to do SMS and contacts Jun 25 04:19:06 josch: Do you know what back end it uses to store contacts? Jun 25 04:19:12 i.e. datastore format? Jun 25 04:20:20 i think four now you may only read/write sim contacts Jun 25 04:20:28 ah Jun 25 04:20:34 still handy Jun 25 04:20:37 but this database thingy is targeted for milestone 2 Jun 25 04:20:51 abraxa__, is working on stuff like this Jun 25 04:20:56 am I correct that there is no UI for hanging up? Jun 25 04:21:14 read here: http://neo1973-germany.de/wiki/pyPimd Jun 25 04:21:22 wurp2, where? Jun 25 04:21:32 in zhone Jun 25 04:21:40 there should be Jun 25 04:21:47 but do not bother about the ui Jun 25 04:21:55 it is only for demonstration Jun 25 04:23:09 having someone put together a UI that shows off what the framework can do can make a big difference in the attention the project gets... Jun 25 04:23:32 I don't think people on the framework team should get distracted with this Jun 25 04:23:34 this is what openmoko did wrong the whole time Jun 25 04:23:43 I don't think so Jun 25 04:23:53 I think what they did wrong was to not pay any attention to infrastructure Jun 25 04:24:02 okay Jun 25 04:24:21 but you guys can pay attention to infrastructure and still have people build a UI that uses your infra Jun 25 04:24:35 Just don't let those people drive the project Jun 25 04:26:57 well, I gotta go to bed Jun 25 04:27:08 g'night all! Jun 25 04:27:14 gotta get up Jun 25 04:34:04 to develop anything openmoko related, one will _need_ gcc 3.3? Jun 25 04:35:21 and is it right that there are.. several source code repositories Jun 25 04:35:27 and I'll need mtn, svn, and git on my system? Jun 25 04:35:56 didn't what used to be mtn switch to git? Jun 25 04:36:04 I have no clue Jun 25 04:36:07 the wiki doesn't seem to be that up to date Jun 25 04:36:19 I checked out the Git repo, and the top level Makefile wants to run mtn :P Jun 25 04:36:28 (openmoko.git, that is) Jun 25 04:37:39 and according to the wiki, everyone still hates Qt/opia Jun 25 04:38:55 kelvie: No to needing gcc 3.3 (although you probably need 3.something to build qemu). Yes to "several repositories". You shouldn't need mtn any longer. Jun 25 04:39:50 mmontour: ah.. do I need qemu though? Jun 25 04:39:56 can't I just run it in Xephyr? :p Jun 25 04:40:36 You don't need qemu. I don't know anything about Xephyr. Jun 25 04:40:42 ah Jun 25 04:44:13 qemu is used by default during the build, to generate locale files for glibc. You can disable this with an 'ENABLE_BINARY_LOCALE_GENERATION = "0"' option in your build/conf/local.conf file. Jun 25 04:52:36 mmontour: alright.. I'll see how far I can get :p Jun 25 04:52:44 I do have all the source so I should be able to figure stuff out along the way Jun 25 04:53:40 This isn't a good time for beginners. There are several possible software images that you can build, none of which works yet. Jun 25 04:54:02 hmm Jun 25 04:54:21 Feel free to try, just expect some pain :) Jun 25 04:54:43 hehe Jun 25 04:54:57 I like using broken software sometimes Jun 25 04:55:06 to help fix it at least Jun 25 05:08:39 openmoko: 03raster 07org.openmoko.asu.dev * rfef3a528c3c6 10/conf/distro/include/sane-srcrevs.inc: update config to make use of aux ANd power as desired and enable lock screen Jun 25 05:08:40 openmoko: 03raster 07org.openmoko.asu.dev * r44339e945cba 10/conf/distro/include/sane-srcrevs.inc: Merge branch 'org.openmoko.asu.dev' of git+ssh://git@git.openmoko.org/var/cache/git/openmoko into org.openmoko.asu.dev Jun 25 05:54:36 w00t! Just got packet radio transmission working from my Neo. Jun 25 06:22:02 freesmartphone.org: 03charlie 07framework * rbdefce25fd1e 10/ (20 files in 11 dirs): Merged preferences service into framework Jun 25 06:32:06 time to try asu Jun 25 06:32:17 an I right to thin asu is a tag? Jun 25 06:32:21 or a branch Jun 25 06:32:44 how does it compare to the trunk? Jun 25 06:38:06 hehe, april turns into august :P Jun 25 06:39:40 ahven, both are gender names as well Jun 25 06:40:00 yep Jun 25 06:40:20 or autumn ;) Jun 25 06:40:46 look at least its not acember ;) Jun 25 06:40:57 openmoko: 03zecke 07org.openmoko.asu.stable * r0a6ead2d1572 10/conf/distro/include/sane-srcrevs.inc: [srcrev] We want the experimental callhandling fixes from Qtopia Jun 25 06:40:57 openmoko: 03zecke 07org.openmoko.asu.stable * rd5188d6389fd 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Build a Qtopia version with Theme performance improvements Jun 25 06:49:32 can somebody tell me what asu.stable is and what asu.dev is and hwo it relates to utopia please Jun 25 06:49:54 erm Jun 25 06:49:56 qutopia Jun 25 06:50:07 is asu qutopia based? Jun 25 06:51:46 yes Jun 25 06:52:15 afaik one of the main tasks was to port it use x11 insteaed of the framebuffer Jun 25 06:52:30 of course i could be pathetically wrong :p Jun 25 06:53:29 is the gtk interface no longer supported? Jun 25 06:53:33 "stable" and "dev" are two branches of ASU (packages are developed on "dev" then moved to "stable" when they seem to be working correctly) Jun 25 06:55:15 The libraries will still be available for 3rd-party gtk apps. The core OM stuff will not use gtk (AFAIK) Jun 25 06:56:45 mmontour: so what exactly is openmoko.dev then? Jun 25 06:56:56 is asu a branch of that? for gta02 Jun 25 06:56:57 ? Jun 25 06:58:40 Something like that. openmoko.dev has newer versions of some of the upstream packages, and will be the basis for whatever comes after ASU. I think. I'm somewhat confused myself by all of these branches. Jun 25 06:58:51 And then there's FSO... Jun 25 06:59:22 erm? Jun 25 06:59:55 freesmartphone.org - another in-development software stack. Jun 25 07:02:43 morning Jun 25 07:03:25 mmontour: that does not seem to be very active Jun 25 07:04:31 so openmoko active stack in now qt based... Jun 25 07:06:30 http://downloads.openmoko.org/framework/milestone1/ <- FSO (see also the recent discussion on the community ML) Jun 25 07:07:24 I am so flipping confused Jun 25 07:07:31 *readign the lists now* Jun 25 07:07:35 You're not the only one. Jun 25 07:07:48 they got qt, gtk+ and e17 Jun 25 07:07:52 what the hell is going on Jun 25 07:08:58 "my god. they have ls, rm and mv! i'm so confused!" Jun 25 07:09:14 *blinks* Jun 25 07:09:16 you should be Jun 25 07:09:27 "crap this distribution comes with zsh, bash AND tcsh! eek!" Jun 25 07:10:04 raster: the real question is what the main stack is based on ... Jun 25 07:10:20 why does there have to be a main stack Jun 25 07:10:27 and why does it have to be based on a widget set? Jun 25 07:10:34 they should ship the phones with empty flash :) Jun 25 07:10:44 so that everyone is forced to choose their own main stack Jun 25 07:11:05 They did that with GTA01. Jun 25 07:11:06 gentoo like Jun 25 07:11:17 raster: mainly because its completely unclear whats going. I used to be GTK+ based...is it now qutopia based? Jun 25 07:11:28 its x11 based Jun 25 07:11:29 I=it Jun 25 07:11:47 i'm trying to make a pioiunt Jun 25 07:11:51 why does it HAVE to be based on one Jun 25 07:11:57 it does not Jun 25 07:11:58 why do you HAVE to limit yourself to 1 toolkit Jun 25 07:12:05 all I am askign is what the main focus now? Jun 25 07:12:05 why do you HAVE to put up boundaries and walls? Jun 25 07:12:09 to have it qutopia based? Jun 25 07:12:16 main focus is python and dbus Jun 25 07:13:26 ok Jun 25 07:13:26 well it's visually a bit messy using different toolkits in the same desktop Jun 25 07:13:35 raster: what balrog-kun said Jun 25 07:13:50 raster: is asu qutopia based? Jun 25 07:13:51 but it's all up to the user Jun 25 07:13:59 we'd better get rid of opengl then Jun 25 07:14:02 geez Jun 25 07:14:07 we must use only gtk Jun 25 07:14:15 NO! Jun 25 07:14:20 hahaha Jun 25 07:14:22 we must only use ncurses! Jun 25 07:14:24 i'm making a point Jun 25 07:14:27 oh! can I start a flamewar too? Jun 25 07:14:28 everyone has his preferences Jun 25 07:14:36 hmmm Jun 25 07:14:41 we should have more closed source stuff Jun 25 07:14:43 of course doc|home!! := Jun 25 07:14:45 stop thinking that the world must focus and limit itself to 1 toolkiit Jun 25 07:14:45 just so we can have 3d games Jun 25 07:14:48 to do everything with Jun 25 07:14:50 very important these days Jun 25 07:14:57 what point? I am merely trying to understand what tookit is the main interface using Jun 25 07:15:01 * doc|home wants crysis on his phone Jun 25 07:15:17 sure gnome can use QT, but its main toolkit is gtk_ Jun 25 07:15:20 gtk+ Jun 25 07:15:27 different things use different toolkits Jun 25 07:15:34 the qtopisa phone apps use qt - guess why? Jun 25 07:15:44 theres a config app - and it uses python + efl Jun 25 07:15:45 * balrog-kun wants 3DStudio Max R6 and Maya on his Jun 25 07:15:54 balrog-kun: hah Jun 25 07:15:59 theres a package installer that is in c + efl Jun 25 07:16:09 we could use the accelerometers to rotate the model! Jun 25 07:16:11 there's nothing on there using gtk by default currently Jun 25 07:16:19 actually, that's a pretty cool idea :o Jun 25 07:16:25 schizophrenia? Jun 25 07:16:37 doc|home: yes! and when someone rings.. something else would happen with the model :p Jun 25 07:16:41 no one WANTS to use qt inside of OM Jun 25 07:16:44 well no enigneering do Jun 25 07:16:51 its there because management wanted to use qtopia Jun 25 07:16:56 ah Jun 25 07:17:06 u're lucky we didnt just dump x11 and everything and go raw qtopia + aws Jun 25 07:17:07 balrog-kun: if it's a face it should adopt a surprised look and have a speech bubble that reads "hey, look, the phone is working, weird!" Jun 25 07:17:13 trhen u would only have qtopia apps and nothing else Jun 25 07:17:20 doc|home: http://phun.cs.umu.se/wiki Jun 25 07:17:34 you're not going to change anything by suggesting anything here Jun 25 07:17:36 hah Jun 25 07:17:50 raster: ripping into me does not help anything either Jun 25 07:18:03 zash: cool, but old, I've had something similar to that on my n800 for months :) Jun 25 07:18:23 goodwill: i'm trying to stop this attitude of "my god! you didnt align with 1 toolkit! you must be idiots" attitude Jun 25 07:18:33 or as if just bncause we use more than 1 suddenly we "change direction" Jun 25 07:18:34 its wrong Jun 25 07:18:49 different apps use different toolkits - because the authros WANT TO Jun 25 07:18:52 raster: I never said that. All I asked is what is the main set up using now Jun 25 07:18:57 well, to be honest there was a bit of direction change Jun 25 07:19:00 progammers will ALWAYS have their too,lkits of chice Jun 25 07:19:03 because they like it Jun 25 07:19:10 or becuase it does something better than other toolkits Jun 25 07:19:12 same as languages Jun 25 07:19:20 raster: sorry if I hit a sore spot Jun 25 07:19:22 should we now write everything in 1 language only? Jun 25 07:19:28 lets drop c and c++ Jun 25 07:19:31 do everythjing 100% in java Jun 25 07:19:33 yes, assembler! Jun 25 07:19:36 raster: I was just trying understand the current situation Jun 25 07:19:37 the gtk stack stopped being maintained Jun 25 07:19:39 of wait no - java';s not trendy anymore Jun 25 07:19:44 lets to it all in python! yes! Jun 25 07:19:47 but that's the management decision Jun 25 07:20:07 fact is development will happen in multipl languagese Jun 25 07:20:09 and multiple toolkits Jun 25 07:20:12 thats life Jun 25 07:20:15 it has been for decades Jun 25 07:20:18 and will continue to be Jun 25 07:20:24 toolkits is the same Jun 25 07:21:35 the point of making interfaces to access features via dbus Jun 25 07:21:47 is that u have a common high-level access protocol (dbus) to do things Jun 25 07:21:50 which is a good approach I'd say Jun 25 07:21:59 you then can c hoose whatever language, toolkkit or whatever that tickles your fancy Jun 25 07:22:01 but before you saying that, I had no idea that was the case Jun 25 07:22:10 continue to rant though ;) Jun 25 07:22:15 but right now we dont have much of that Jun 25 07:22:27 because we are busy creating pixel-perfect reditions of designers ideas Jun 25 07:22:41 if u must have a common toolkit - it's x11 Jun 25 07:22:45 as it doesnt define policy Jun 25 07:22:51 just mechanism by whihc to share a display Jun 25 07:23:47 raster: thank you for the information Jun 25 07:24:12 I think dbus approach kicks ass Jun 25 07:27:46 balrog-kun: so to summarize...it seem OM management decided to go with Qutopia as a shortcut. Where is dev wants to now run dbus interfaces to create api. seems correct? Jun 25 07:30:11 raster: " pixel-perfect reditions of designers ideas" - so you have a visual designer in OM? Jun 25 07:31:23 hey guys.. has anyone tried to record from gta02's internal microphone? Jun 25 07:31:35 i'm trying to find the correct mixer settings Jun 25 07:32:02 cekay: I've done it on gta01 Jun 25 07:32:34 goodwill: yea, more or less Jun 25 07:32:37 mudyc: we have people who show us flash demos and ask fo them to be implemented pixel-perfectly Jun 25 07:32:45 so it should theoratically work :) Jun 25 07:33:08 i hope not all developers want to use python in OM :) Jun 25 07:33:28 raster: that sucks...I recently had rip one up to pieces for doing exactly taht Jun 25 07:33:47 balrog-kun: python would kill all the ramn on that poor phone Jun 25 07:35:24 raster: sounds bad Jun 25 07:36:06 the result isn't all that bad Jun 25 07:36:34 implementing from flash demos? Jun 25 07:36:45 cekay: take a look at http://members.shaw.ca/mmontour/neo/rec.state as a starting point. It's intended to record both sides of a GSM call (incoming on one channel, outgoing on the other). Jun 25 07:37:04 ah thanks mmontour Jun 25 07:37:13 that's for the gta01? Jun 25 07:37:15 goodwill: do you have any statistics that you can base your argument? Jun 25 07:37:18 is there a usb carcharge out there for neo Jun 25 07:37:30 zash: the one in asu, now that i remember there was a flash mockup some time earlier Jun 25 07:37:37 Yes, gta01. Jun 25 07:37:45 cool Jun 25 07:38:29 mudyc: well I am basing that on seeing people write python application crappily Jun 25 07:38:49 mudyc: its just easier to do in python Jun 25 07:39:01 to write badly designed apps Jun 25 07:40:29 goodwill: well, that means no Jun 25 07:41:35 mudyc: whats your point? Jun 25 07:42:37 mudyc: for example people doing imports left and right in python can easily bloat up the footprint Jun 25 07:43:20 goodwill: anyone can share their opinions but your opinion was formed as an statement of language x being memory hungry with no arguments Jun 25 07:44:09 mudyc: I was not writing a article. It was a sentence Jun 25 07:45:19 mudyc: people think python garbage collector does it all. barely anyone uses PySizer or hell even pychecker to check for basic inefficiencies. Jun 25 07:46:21 mudyc: you get a lot more shmucks in python liek that cause the language is easy Jun 25 07:46:36 mudyc: a good developer will write good code regardless of the language Jun 25 07:47:11 mudyc: anyhow that's just my opinion, I can be totally wrong Jun 25 07:47:48 goodwill: yep. anyway, i was actually interested in a tool to actually measure the memory usage Jun 25 07:48:50 mudyc: there is a good article I read a while back, taught me a lot....lets see Jun 25 07:49:28 mudyc: http://wingolog.org/archives/2007/11/27/reducing-the-footprint-of-python-applications Jun 25 07:50:18 mudyc: using gc will help too Jun 25 07:52:44 mudyc: also this: http://guppy-pe.sourceforge.net/ Jun 25 07:52:56 never used it, but I am goign to try at some point Jun 25 07:53:18 goodwill: thanks. seems to be usefull Jun 25 07:53:59 mudyc: no problem Jun 25 07:54:13 goodwill: they hand over flash demos and want to see them implemented as-per the demo Jun 25 07:54:36 zash: thats how its working. for better or for worse Jun 25 07:54:52 raster: asinine!!! so asinine!! I am sorry dude Jun 25 07:55:20 the asu image is basically us implemented the flash demo. Jun 25 07:56:29 raster: you guys do good work, I am very impressed Jun 25 07:58:16 we do the best we can Jun 25 07:58:17 :) Jun 25 07:58:22 :) Jun 25 07:58:38 dbus is an excellent approach in my mind Jun 25 07:59:13 if done right Jun 25 07:59:21 its good for gluing things together Jun 25 07:59:51 yup Jun 25 08:00:03 python is fine when perfromance is not relevant Jun 25 08:00:12 and when the python is kept to the hihgest levels of logic Jun 25 08:00:16 ie - much like dbus - it's glue Jun 25 08:01:02 raster: true...true Jun 25 08:04:20 dbus is horrible to debug and the implementations has pitfalls Jun 25 08:04:52 like any form of ipc Jun 25 08:05:07 it is painful to debug as control exits that of the debugging tools Jun 25 08:05:12 offto another process Jun 25 08:06:40 mudyc: whats a better alternative? Jun 25 08:06:53 mudyc: something nicely decoupled Jun 25 08:08:47 for ipc u can do anything Jun 25 08:08:50 roll your own Jun 25 08:09:01 goodwill: better alternative to what? i think dbus is good enough. for example dbus-monitor is many times useless when working with system bus (debug problems) Jun 25 08:09:14 dbus is just a packetisation format Jun 25 08:09:35 with a data encoding scheme Jun 25 08:09:42 mudyc: dbus-monitor is no good at all...as I recall Jun 25 08:09:51 and defined sockets with parent server process to route messages Jun 25 08:13:37 and my experience with python dbus bindings, gobject dbus bindings (C) and plain dbus api (C) is that plain dbus api is the best implementation. other implementations seem to have weird stuff how they are used and it also depends if you are working on the "server" or "client" side. Jun 25 08:15:02 and the document which says the "dbus - good practices for daily usage" does not exist Jun 25 08:21:12 * goodwill misses python Jun 25 08:38:01 morning Jun 25 08:38:27 mudyc: i sticl to dbus from libdbus and in c. i do not wish to foul my fingers with other religions.. err.. i mean languages :) Jun 25 08:53:16 openmoko: 03zecke 07org.openmoko.asu.dev * rcb548cc96ea4 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Upgrade xglamo, it got a fix for extended keycodes. Jun 25 08:53:31 good morning Jun 25 09:02:54 openmoko: 03zecke 07org.openmoko.asu.dev * rc4ad5b97f18b 10/packages/connman/connman_git.bb: [connman] Make sure that connman gets started, do that after dbus is running Jun 25 09:09:38 mornin Jun 25 10:17:19 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r7ca1b037b35f 10/odeviced/src/plugins/wifi/wifi.c: Fix segfault while loading wifi plugin in the Freerunner Jun 25 10:17:19 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r1c21ec5e46ae 10/odeviced/Makefile.am: Don't install the init script Jun 25 10:17:20 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rc4a41a8353c2 10/odeviced/recipes/odeviced_git.bb: Recipe works now Jun 25 10:17:21 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rb76d59b75f1b 10/odeviced/src/plugins/power/ (power.c power.vala): GetEnergyPercentage: read from capacity if it is present Jun 25 10:28:27 wow. openmoko is running on my htc wizard! Jun 25 10:32:49 great Jun 25 10:32:59 ugh, no fun reading mobile phone news online. most pages i have found are so usa centric its not even fun... Jun 25 10:34:41 i'm using linwizard kernel. i extracted the openmoko jffs2 rootfs image onto an SD card formatted with ext2, and booted it from windows mobile with haret. Jun 25 10:52:36 FSO includes qtopia / Jun 25 10:52:36 ? Jun 25 10:53:04 nope Jun 25 10:53:14 python and efl Jun 25 10:53:28 !! Jun 25 10:53:34 I thought we moved to Qtopia already ? Jun 25 10:53:51 ASU ? Jun 25 10:54:07 currently its just a reference implementatation Jun 25 10:54:14 I don't get it Jun 25 10:54:20 CVirus, yeah ASU uses qtopia on x11 Jun 25 10:54:27 FSO images are different Jun 25 10:54:40 why are the developers working in 2 different pathes ? Jun 25 10:55:17 you mean fso? Jun 25 10:55:32 FSO and ASU Jun 25 10:55:32 its just a framework for others to pick up Jun 25 10:55:47 i guess ASU would use FSO once its done Jun 25 10:56:00 ahhh Jun 25 10:56:02 now I get it Jun 25 10:56:10 what does FSO stand for ? Jun 25 10:56:38 freesmartphone.org Jun 25 10:57:58 thanks Jun 25 10:58:09 CVirus, http://www.vanille-media.de/site/index.php/2008/05/05/openmoko-framework-initiative/ Jun 25 11:28:05 hello Jun 25 11:28:36 i have a problem building openmoko for my pc. Jun 25 11:29:22 i use the mokomakefile to build like it is descript in http://wiki.openmoko.org/wiki/Getting_Openmoko_working_on_host_with_Xephyr Jun 25 11:29:57 the error i get is: | /home/moko/build/tmp/staging/i486-angstrom-linux/usr/lib/libecore_con.so: undefined reference to `curl_easy_setopt@CURL_OPENSSL_3' Jun 25 11:30:32 and lot motr of thr undefined refrenece stuff with *@CURL_OPENSSL_3 Jun 25 11:30:59 the package that fails is edje Jun 25 11:31:11 does somebody know how to fix this? Jun 25 11:35:29 openmoko: 03julian_chu 07org.openmoko.asu.dev * r0a3d0b0cf904 10/packages/openmoko-projects/exposure_svn.bb: [exposure] Marek needs python-edbus for exposure. Jun 25 11:39:26 openmoko: 03zecke 07org.openmoko.asu.dev * r758f39146203 10/packages/qtopia-phone/qtopia-phone-x11_git.bb: [qtopia] Use the gcc options that Trolltech might used for their gta01 build Jun 25 11:39:27 openmoko: 03zecke 07org.openmoko.asu.dev * r5f878534e0f3 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Upgrade exposure, getting more close to a usable version again Jun 25 11:44:32 <`pwgen`> noon: maybe take a look at http://oe.linuxtogo.org/wiki/GettingStarted: if oe build works, openmoko build should also work Jun 25 11:54:57 `pwgen`: SD seems to be a kernel problem Jun 25 11:55:10 my FSO image mounts the SD that comes with the FreeRunner fine Jun 25 11:55:18 however it does _not_ mount the SD that came with the 1973 Jun 25 11:55:30 so I suspect upstream or our patches containing a bug Jun 25 11:55:36 not the fso-image though Jun 25 11:58:28 hey guys.. has anyone tried to record from gta02's internal microphone and has the correct mixer settings? Jun 25 11:58:50 sean reported some sd-card issues on the kernel list recently; I wonder if that is the same thing? Jun 25 12:02:54 <`pwgen`> mickeyl: what is the name of the bb file, that had to used in gta01 kernel build for fso image, linux-openmoko-2.6.24_git ? Jun 25 12:03:35 milestone1 contains linux-openmoko-devel_git.bb Jun 25 12:04:45 <`pwgen`> cekay: mmontour has created last nige a statefile http://members.shaw.ca/mmontour/neo/rec.state Jun 25 12:04:59 <`pwgen`> s/nige/night/ Jun 25 12:05:00 `pwgen` meant: cekay: mmontour has created last night a statefile http://members.shaw.ca/mmontour/neo/rec.state Jun 25 12:05:25 yup, .. he told me. but it's for gta01 Jun 25 12:05:51 i was wondering if someone did that for gta02 already Jun 25 12:06:50 <`pwgen`> mickeyl: ahhh, ill try some kernel builds this night to figure out the problem with the sd card, when i not at work Jun 25 12:07:16 ok, cool Jun 25 12:07:22 it's the andy branch i used Jun 25 12:07:26 perhaps it's better in stable? Jun 25 12:07:42 <`pwgen`> ( office, job, work ) - here i have limited time and access to my build system .. Jun 25 12:07:45 it's not completely broken, just doesn't work with some cards Jun 25 12:07:50 sure Jun 25 12:08:54 <`pwgen`> i had some ( maybe identical card , same distributor/size ) , one could be read by moko , the other not .. Jun 25 12:09:46 <`pwgen`> mickeyl: whats the andy branch ? Jun 25 12:09:57 andy's kernel branch Jun 25 12:10:05 git.openmoko.org -> branches -> andy Jun 25 12:12:58 <`pwgen`> mickeyl: the stuff from andy_green , right ? Jun 25 12:13:29 yes Jun 25 12:13:35 nah - Andy Kauffman -- it's the humor branch! Jun 25 12:13:40 heh Jun 25 12:15:17 <`pwgen`> ahh, i understand now why my kernel images alway had problems with resume/suspend .. he is doing a lot of work for openmoko-kernels . Jun 25 12:16:05 <`pwgen`> linux-openmoko-devel_git.bb will build a kernel based on his work, thats right ? Jun 25 12:16:10 mickeyl: BTW, the fail-safe stuff that attempts to "automagically" avoid the overruns on the GTA01 was removed, so kernels based on the -andy branch recently can be expected to behave rather worst on the GTA01 until the odeviced understands how to hold off the GSM itself. Just FYI. Jun 25 12:17:23 mickeyl, hi, are you responsible for the vala recipe?...if yes, fyi,..its rather old...i have changed it to fetch 0.3.3 and it works without issues Jun 25 12:17:33 ooops..wrong channel Jun 25 12:17:44 mwester-trip: what exactly do i need to do on suspend/resume w/ gsm? i'll add it in, then Jun 25 12:18:24 Sup3rkiddo: yes i am. i did not have a chance to update yet, since i didn't have time updating moko-terminal to the new syntax. i'll give it a go when i have a chance Jun 25 12:18:49 cool... Jun 25 12:21:40 mickeyl: Write "1" to /sys/devices/.../neo1973_pm_gsm/flowcontrolled after the final communication with the GSM is complete when preparing to suspend, then write "0" to the same sysfs file as soon as the device resumes and is ready to talk to the GSM. Jun 25 12:22:00 ah, that's easy Jun 25 12:22:03 thanks Jun 25 12:22:14 will this device node be present on both 01 and 02? Jun 25 12:22:19 Yes. Jun 25 12:22:30 excellent. will add this to the device specific code asap Jun 25 12:22:32 And it behaves the same on both, now. Jun 25 12:23:46 ok Jun 25 12:23:51 while we're talking URAT Jun 25 12:23:53 UART, even Jun 25 12:24:16 multiple processes can talk to one tty at once, right? Jun 25 12:24:28 but only one process actually receives the answers Jun 25 12:24:32 how is that determined? Jun 25 12:24:40 is it the last one that opened it for reading? Jun 25 12:24:50 and if that one closes, will the previous one get answers then? Jun 25 12:25:26 feel free to point me to a tty introduction, i'd like to read one that answer all that, but did not really find something good Jun 25 12:27:14 of course my real interest is pty's Jun 25 12:29:27 Yes, multiple processes can open a tty, and send to it. AFAIK it is non-deterministic as to who gets the rx data -- if both are reading, I think the data will be split between the two. Jun 25 12:30:17 I'll search about to see if I have any current references; my info on this goes back to when SunOS was the *nix of choice... :( Jun 25 12:30:23 hehe Jun 25 12:30:34 (But I have to go off to work now, I'll be back later) Jun 25 12:30:38 k, cu Jun 25 12:33:15 i think you should just try to avoid a situation where two fd's are the same tty Jun 25 12:33:24 in one process or in separate processes Jun 25 12:33:42 hmm Jun 25 12:33:43 well Jun 25 12:33:49 i'm thinking about a way to deal with ppp Jun 25 12:34:02 i'd like to take over the complete init until CONNECT Jun 25 12:34:06 and then hand over to pp Jun 25 12:34:18 ah, hmmm Jun 25 12:34:26 however since my pty will disappear once i close the channel (MUXed line) Jun 25 12:34:28 i need to keep it open Jun 25 12:34:40 and think about seamless handover Jun 25 12:34:45 <`pwgen`> cool lets port opensolaris to openmoko to get the tty working (;-)) Jun 25 12:35:09 you should probably have a specific pppd plugin, Jun 25 12:35:38 hello Jun 25 12:35:40 or your daemon should open the pty, talk to it and when it's ready, create a new pty and let pppd talk to it Jun 25 12:35:52 forwarding data between them Jun 25 12:36:11 oh! that's an interesting approach Jun 25 12:36:29 so i have a chance to intercept stuff Jun 25 12:36:32 i like that Jun 25 12:36:47 yeah, but on the other hand that's yet more rounds for the data between modem and kernel Jun 25 12:37:33 ya, it is for sure, but since we need to handle all the cases where you have gprs on one line and have to handover when a voice call comes in (without closing gprs), i don't see a lot of other chances Jun 25 12:38:08 at least for devices that can not simultaneously gprs and voice call Jun 25 12:39:18 <`pwgen`> is this possible setting ppp on hold, doing voice calls, and getting ppp back ? Jun 25 12:39:26 yes Jun 25 12:39:32 but this needs a bit of clever logic Jun 25 12:39:48 it goes to kernel from modem, then through line discipline to the tty, then to the muxer, then gets written to the pty master, goes through a line discipline again to pty, gets read, gets written to another pty master, through line discipline to the pty and then back to the kernel to decode pty and make a net interface Jun 25 12:40:14 ya, it's a bit scary i know :) Jun 25 12:40:17 well Jun 25 12:40:24 <`pwgen`> i thought there were problems, cause gprs and voice calls are diffrent operational modes for the gsm-modem Jun 25 12:40:26 the other option is to teach ppp some more tricks Jun 25 12:40:31 err, s/pty/ppp/ Jun 25 12:40:47 mickeyl: this may be a better idea i think Jun 25 12:40:52 how complicated is ppp anyways? Jun 25 12:40:53 it has a nice plugins interface Jun 25 12:41:06 looks there's just one implementation available, which is rare in oss land Jun 25 12:41:34 yeah strange, i don't know though Jun 25 12:42:34 ok, well, i first try to get it working at all, before pondering about more complicated case Jun 25 12:42:35 s Jun 25 12:43:27 hmmm, on the other hand..... Jun 25 12:43:33 yas? Jun 25 12:43:39 pppd doesn't really do much, mostly the authentication stuff Jun 25 12:43:50 the decoding and encoding and stuff are done in the kernel Jun 25 12:44:11 it might make sense to write your own little pppd instead (which would be same as making a plugin for pppd) Jun 25 12:44:20 yeah Jun 25 12:44:22 that's what i was thinking Jun 25 12:47:24 unfortunately there doesn't seem to be a minimal implementation around Jun 25 12:55:41 In creating a package in moko , I got do_install error ,compilation is success full then I changed source files and changed setup-env, then I make the package again but It skips compilation phase Jun 25 13:01:34 Any one help me Jun 25 13:04:54 <`pwgen`> ppp does handschake protocols for communicaton stack, i think its not so easy to write an stable 'buugless' pppd Jun 25 13:33:09 openmoko: 03zecke 07org.openmoko.dev * r7349495a3f5d 10/packages/qtopia-phone/qtopia-phone-x11_git.bb: [qtopia] Remove -debug from the build and hope things get faster Jun 25 13:33:10 openmoko: 03zecke 07org.openmoko.dev * rb71ead40febb 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Upgrade kernel to get the a6k netdev fix Jun 25 13:33:11 openmoko: 03zecke 07org.openmoko.dev * r8cfbd97333ed 10/packages/xorg-xserver/xserver-kdrive-glamo_1.3.0.0+git.bb: [glamo] Update the origin of glamo, now hosted on git.openmoko.org Jun 25 13:33:11 openmoko: 03zecke 07org.openmoko.dev * r0829af699f46 10/ (2 files in 2 dirs): [glamo] Put OE patches "upstream" and additonal keyboard fixes Jun 25 13:33:14 openmoko: 03raster 07org.openmoko.dev * r83e1056b4a96 10/conf/distro/include/sane-srcrevs.inc: illume update - new hold to shutdown config and using e features as well as Jun 25 13:33:17 openmoko: 03raster 07org.openmoko.dev * ra81a516677e6 10/conf/distro/include/sane-srcdates.inc: efl update to make sure illume updates work Jun 25 13:33:20 openmoko: 03julian_chu 07org.openmoko.dev * rafe1721035e2 10/conf/distro/include/sane-srcrevs.inc: [opkg] Bump up the reversion of opkg, opkg-native, opkg-sdk to 4491 Jun 25 13:33:23 openmoko: 03zecke 07org.openmoko.dev * r466501466e7a 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Upgrade Qtopia to get qLog changes and be able to set the time Jun 25 13:33:26 openmoko: 03zecke 07org.openmoko.dev * ra43fdcccb861 10/packages/dbus/ (dbus-1.0.3/fedora-compile-fix.patch dbus-native_1.0.3.bb): Compile with recent glibc,gcc,headers (e.g. fedora 9) Add a patch to dbus to properly compile on fedora 9. Jun 25 13:33:32 openmoko: 03zecke 07org.openmoko.dev * r4150cc00c0a4 10/packages/ipkg-utils/ (3 files in 2 dirs): Jun 25 13:33:35 openmoko: Make it work on fedora 9. "control.tar.gz" got listed as "`control.tar.gz" The Jun 25 13:33:37 openmoko: root cause is somewhere else in the parser. So this is a workaround that will Jun 25 13:33:39 openmoko: help to find the right file. opkg on the device should use the proper ar Jun 25 13:33:41 openmoko: executable and that should be able to extract the right files. Jun 25 13:33:45 openmoko: 03mwester 07org.openmoko.dev * rcdb0884e057f 10/packages/ipkg-utils/ (3 files in 2 dirs): ipkg-utils: fix ipk (ar) file processing to handle odd-length data blocks. Jun 25 13:33:48 openmoko: 03zecke 07org.openmoko.dev * r603da2de15ff 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Upgrade exposure to get faster startup speed Jun 25 13:33:51 openmoko: 03zecke 07org.openmoko.dev * r519c9ad31ee2 10/conf/distro/include/sane-srcrevs.inc: [srcrev] We want the experimental callhandling fixes from Qtopia Jun 25 13:33:56 openmoko: 03zecke 07org.openmoko.dev * re8bbe3f1d5c7 10/ (2 files in 2 dirs): [srcrev] Upgrade exposure to be launchable again Jun 25 13:33:59 openmoko: 03zecke 07org.openmoko.dev * rc78722a8ad7d 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Build a Qtopia version with Theme performance improvements Jun 25 13:34:02 openmoko: 03raster 07org.openmoko.dev * rfef631fb0c8a 10/conf/distro/include/sane-srcrevs.inc: update config to make use of aux ANd power as desired and enable lock screen Jun 25 13:34:07 openmoko: 03zecke 07org.openmoko.dev * red00a6094649 10/conf/distro/include/sane-srcrevs.inc: [srcrev] Upgrade xglamo, it got a fix for extended keycodes. Jun 25 13:34:10 openmoko: 03zecke 07org.openmoko.dev * r9328b3420f00 10/packages/connman/connman_git.bb: [connman] Make sure that connman gets started, do that after dbus is running Jun 25 15:11:30 what do i need to do to get rid of : "ls: cannot access ../build/qemu/uImage-2.6.*-om-gta01.bin: No such file or directory" Jun 25 15:11:50 i tried make qemu , make qtopia-x11 but it didn't mad any diffrence Jun 25 15:43:22 freesmartphone.org: 03daniel 07zhone * r28cd6e7c4509 10/src/zhone: zhone: Fixed bug where transition_to was called on the wrong object Jun 25 15:48:02 freesmartphone.org: 03mickey 07zhone * r618a0b1e7abd 10/src/zhone: zhone: catch up with framework dbus bus name changes Jun 25 16:04:24 freesmartphone.org: 03mickey 07framework * rb52aefb934ce 10/framework/ (3 files in 3 dirs): odeviced: fix bug in info module Jun 25 16:12:16 hmm, if anyone has tested zhone, does the battery indicator show status immediately or does it take a while? Jun 25 16:12:35 Sup3rkiddo, immediately Jun 25 16:12:57 josch, thanks... Jun 25 16:13:41 Sup3rkiddo: Was fixed lately Jun 25 16:13:46 commit 317d58d95cee88ba953295f4068cdb2c92508a88 Jun 25 16:13:46 Author: Daniel Willmann Jun 25 16:13:46 Date: Sun Jun 22 16:58:58 2008 +0200 Jun 25 16:13:46 zhone: Update battery status as soon as the DBus APIs are initialized Jun 25 16:14:20 alphaone, well, i am running the latest code, but its running the other odeviced Jun 25 16:14:34 hmm, okay Jun 25 16:14:35 so i was wondering if i am the one at fault Jun 25 16:14:44 but atleast it works :D Jun 25 16:14:53 Sup3rkiddo: cool Jun 25 16:15:04 oh ups I updated two days ago ^^ Jun 25 16:15:20 summatusmentis, yo Jun 25 16:15:26 lol Jun 25 16:15:27 bye Jun 25 16:15:34 Sup3rkiddo, you scared him away! Jun 25 16:15:36 he is scared of me :( Jun 25 16:15:49 * Sup3rkiddo is a nice guy Jun 25 16:19:38 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rfd28dddfa743 10/odeviced/recipes/odeviced_git.bb: Add CONFLICTS and fetch sources from git.fs.o Jun 25 16:31:26 Sup3rkiddo: hey, sorry, was making lunch Jun 25 16:36:15 I didn't see any battery indicator on zhone... Jun 25 16:36:51 Nor a network indicator, nor a signal strength indicator Jun 25 16:38:32 And when I tried to access dbus from python, dbus.SessionBus() returned some weird error like "AutoRun requires X11, and module was compiled without X11 support" Jun 25 16:46:07 Sup3rkiddo: where is the battery indicator? Jun 25 16:51:03 anyone here good with glp/copyleft licenses? If I write a python binding for a c library licensed under gpl. can I then license my bindings as whatever I want or does it count as a "derived work"? Jun 25 16:51:37 You want ##bondage. Jun 25 16:52:06 wurp2, top right :P?..sorry, which battery indicator? Jun 25 16:52:31 As I understand it, you would have to license your code GPL. Jun 25 16:52:42 Not LGPL. Jun 25 16:53:13 ok, yeah maybe Jun 25 16:53:21 * SpeedEvil is not an expert. Jun 25 16:53:36 Is there a ##gpl channel? Jun 25 16:53:47 There is. Jun 25 16:55:14 Alas, it's only me. Jun 25 16:56:17 XorA|gone: phone works on htckaiser Jun 25 16:57:59 XorA|gone: i.e. at communication is flawless and audio works just fine Jun 25 17:07:39 wooooooo Jun 25 17:08:25 wheeeeee Jun 25 17:08:55 /op doesnt work =( Jun 25 17:09:04 sad Jun 25 17:10:23 Sup3rkiddo: I guess it's top right - on my zhone install, that was blank most of the time, then briefly was about half full, then was blank again Jun 25 17:10:24 What is the top left bar? Jun 25 17:10:42 wurp2, thats the signal strength Jun 25 17:11:00 wurp2, as alphaone said, its fixed in the later images Jun 25 17:13:17 Sup3rkiddo: OK, thx Jun 25 17:20:24 josch: Do you know what I'm doing wrong re: access to dbus via python in fso? Jun 25 17:21:00 import python; session = dbus.SessionBus() seems too simple to screw up... Jun 25 17:21:12 our services are on SystemBus Jun 25 17:21:51 wurp2, just look it up in the reference implementation Jun 25 17:22:02 and mickeyl is right ;) Jun 25 17:26:58 in the freerunner and neo /dev/input/event0 is the aux button..isnt it?...or does it change? Jun 25 17:27:23 wurp2: quick question. Do you develop using neo/freerunner, qemu or 686 compiled code running under xnested server Jun 25 17:27:58 Sup3rkiddo, no it's this buttons Jun 25 17:28:06 i once tried to get it to work with evdev Jun 25 17:28:09 but failed :( Jun 25 17:29:25 josch, it's this button? Jun 25 17:29:33 i didn get that Jun 25 17:29:38 both buttons Jun 25 17:29:44 have a look into this source file... Jun 25 17:29:46 one moment... Jun 25 17:29:49 sure Jun 25 17:29:53 that would be cool Jun 25 17:30:31 Sup3rkiddo, http://svnweb.openmoko.org/trunk/src/target/OM-2007.2/daemons/neod/src/buttonactions.c?rev=4471&view=auto Jun 25 17:31:13 josch, :P, i am supposed to write a replacement for that as my SoC project Jun 25 17:31:24 this code opens /dev/input/event%d where %d is from 0 to 10 Jun 25 17:31:33 and reads from it Jun 25 17:31:45 the keycodes can be seen above Jun 25 17:31:59 169 and 116 Jun 25 17:32:20 Sup3rkiddo, i know ;) Jun 25 17:32:47 these were just my two cents as i struggled with evdev and this devices Jun 25 17:32:51 josch, i was looking into that Jun 25 17:32:59 in theory you shouldn't assume that some /dev/input/ node is some device, instead you should loop through all these devices and ask them their names Jun 25 17:33:20 josch, thanks for sharing...so these devices nodes are dependent on the device Jun 25 17:33:28 balrog-kun, so this is why they open 10 devices :) Jun 25 17:33:39 balrog-kun, names as in the keys which they stand for Jun 25 17:33:49 i am just beginning to attack the input part Jun 25 17:34:18 cat /dev/input/event%d seems a bad idea Jun 25 17:34:20 :P Jun 25 17:34:29 with my kernel/udev combination it's event0 for the keys and event1 for the touchscreen Jun 25 17:34:41 josch, and where can i get the code for evdev? Jun 25 17:34:51 Sup3rkiddo, hexdump /dev/input/event0 is much better Jun 25 17:34:54 josch, aha, Jun 25 17:34:56 yeah Jun 25 17:35:02 just learnt that Jun 25 17:35:06 it will print the binary hex equivalent Jun 25 17:35:28 josch: hmm, but this code doesn't check what the devices are Jun 25 17:35:36 balrog-kun, right Jun 25 17:35:37 goodwill: neo, gta01 Jun 25 17:35:57 wurp2: thank you Jun 25 17:36:02 there's some ioctl to ask the device's name and it will say something like "GTA01 power button" or something Jun 25 17:36:09 and "s3c2410 touchscree" or similar Jun 25 17:36:36 so is the output of hexdump a standard one?...i mean each column standing for different things Jun 25 17:37:03 Sup3rkiddo, if this were the case evdev might have worked :/ Jun 25 17:37:16 wow Jun 25 17:37:18 i see Jun 25 17:37:40 looks fun :D Jun 25 17:37:44 what's wrong with evdev though? Jun 25 17:38:02 well i tried it Jun 25 17:38:10 and all i get is either nothing or Jun 25 17:38:13 random keycodes Jun 25 17:38:26 question: what is evdev? Jun 25 17:38:34 humm, straneg Jun 25 17:38:42 Sup3rkiddo: x11 device driver Jun 25 17:38:47 an xorg driver to get keyboard/mouse/touch input Jun 25 17:38:55 Qwell, ah thanks Jun 25 17:39:07 event device ? Jun 25 17:39:10 evdev.. Jun 25 17:39:14 it's the X11 driver for evdev :P Jun 25 17:39:20 hehe Jun 25 17:39:23 evdev is also the name of the whole /dev/input/ interface Jun 25 17:39:28 aha Jun 25 17:39:33 i am SMART Jun 25 17:40:26 *cricket* *cricket* =( Jun 25 17:42:02 http://www.frogmouth.net/hid-doco/c537.html <--- somehow related? Jun 25 17:43:31 i'd use EVIOCGNAME, if you want to find a particular device Jun 25 17:43:56 or if it's something like neod, it makes sense to listen on all devices Jun 25 17:46:10 josch: OK, I check the fso home page, and see a link to reference model, which sends me to a page that includes no info on coding for fso Jun 25 17:46:29 hehe Jun 25 17:46:35 i think event2 event3 are for accels Jun 25 17:46:58 wurp2, zhone is written in python...so you might want to check that Jun 25 17:47:05 DBus + python is easy Jun 25 17:47:06 wurp2, sry i only can send you to guys like mickeyl, alphaone and shoragan Jun 25 17:47:11 :) Jun 25 17:47:21 they are the active devs right now Jun 25 17:47:35 so..event2 and event3 are for accels?...since the hexdump seems to dump em constantly Jun 25 17:47:54 i had this issue with event0 also Jun 25 17:47:59 Sup3rkiddo: Yeah, I know Jun 25 17:48:07 (python + dbus is easy) Jun 25 17:48:11 so on using evdev it constantly printed random keycodes Jun 25 17:48:16 wurp2, hehe Jun 25 17:48:25 That's why I was so surprised when my most basic 2 line python dbus script blew up Jun 25 17:48:36 wurp2, what did it contain? Jun 25 17:48:42 I will try it with SystemBus instead of SessionBus Jun 25 17:48:50 wurp2, of course its SystemBus Jun 25 17:48:54 aah Jun 25 17:48:56 Sup3rkiddo: import dbus; session = dbus.SessionBus() Jun 25 17:49:03 thats SystemBus() Jun 25 17:49:05 Well, I am just exploring the bus Jun 25 17:49:29 use systembus ^^ Jun 25 17:49:36 And the failure looked like a low level problem with the python dbus code, not "there is no SessionBus" Jun 25 17:49:44 wurp2, you might want to check out d-feet Jun 25 17:49:52 graphical debugger for dbus Jun 25 17:50:12 dbus-send does its job too Jun 25 17:50:15 Sup3rkiddo: Cool, I will Jun 25 17:50:27 I need to look into e_dbus, too - I dunno what that is Jun 25 17:50:38 josch, bleh, you nerd! Jun 25 17:50:47 haha Jun 25 17:50:48 wurp2, dbus bindings for EFL Jun 25 17:51:02 ah, cool Jun 25 17:51:18 integrates very nice with the rest Jun 25 17:51:38 josch, so anyway, thats strange since event0 in my case seems to dump when i hit aux....so is this numbering based on some act of god? Jun 25 17:51:57 Sup3rkiddo, i asked this myself when using evdev :( Jun 25 17:52:08 oh... Jun 25 17:52:12 so unfortunatly i have no solution for this Jun 25 17:52:59 josch, hmm, got any links where i can learn more on this? Jun 25 17:53:12 hrm... barely but... Jun 25 17:53:28 http://wiki.openmoko.org/wiki/Kernel#AUX_Button Jun 25 17:53:39 and then the neod source Jun 25 17:53:42 this was all i had Jun 25 17:54:44 man, strange isn't it...the lower you reach, the lesser the docs Jun 25 17:56:02 All that code that uses the low level stuff is the docs, man Jun 25 17:56:32 hehe Jun 25 17:56:34 too true ^^ Jun 25 17:57:03 so read the source luke! :P Jun 25 17:57:11 :/ Jun 25 17:57:37 laziness is a virtue..:P Jun 25 18:00:41 really, is 28 days really the last word on warranties? Jun 25 18:01:11 28 days of hardware warranty? Jun 25 18:01:13 what? Jun 25 18:01:25 thats pretty short if you ask me... Jun 25 18:01:46 indeed Jun 25 18:01:57 Qwell: 28 days and that's for 10 packs Jun 25 18:02:10 individual units are 14 days Jun 25 18:02:11 mmm...yeah Jun 25 18:11:30 ouch Jun 25 18:11:35 josch, behold my uber googling skills, http://www.linuxjournal.com/article/6429 Jun 25 18:11:44 is there an extended plan? Jun 25 18:12:05 * josch beholds Jun 25 18:12:08 :D Jun 25 18:13:33 I want to see an event that when you press the aux button, at random, and only very inappropriate times, it flashes up an image from your .stuff folder Jun 25 18:13:45 rofl Jun 25 18:13:56 hey thats a cool idea Jun 25 18:14:14 the difficult part being determining when's an inappropriate time Jun 25 18:14:27 "Phone call from: Mum" Jun 25 18:14:34 :P Jun 25 18:14:38 "Nooooo!" Jun 25 18:14:55 "I am not mad at you...Just disappointed" Jun 25 18:15:00 hahaha Jun 25 18:15:20 given I've never brought a woman home to my family, she'd probably be relieved Jun 25 18:15:23 * Sup3rkiddo has caused infinite number of disappointments :P Jun 25 18:15:35 Sup3rkiddo: nice... tmi :) Jun 25 18:16:53 but i wonder where i can get info about the hexdumps....on a green terminal it feels hax0ry Jun 25 18:17:39 I have seen the news about Neo FreeRunner, and I'm wondering if it'll have an SD card slot or the amount of storage is fixed? I'd love it with 8GB :D Jun 25 18:17:50 microsd Jun 25 18:18:00 angasule, it does have a microSD slot Jun 25 18:18:07 thank you :) Jun 25 18:18:46 does it support SDHC? Jun 25 18:18:49 and yes it is 8GB compatible Jun 25 18:18:50 yes Jun 25 18:19:00 * SpeedEvil stabs teh stupid SD association. Jun 25 18:19:05 * angasule helps SpeedEvil Jun 25 18:19:18 (they only specify SDHC up to 32G) Jun 25 18:19:33 Sup3rkiddo: i.e the format of the data read from /dev/input/*? Jun 25 18:19:35 oh, so yet another SD acronym will be forthcoming, fun :P Jun 25 18:19:46 angasule: microsd, not sd Jun 25 18:19:56 balrog-kun, yeah, those columns...obviously they mean something... Jun 25 18:20:14 like by say which area of touchscreen was pressed..can i get that info? Jun 25 18:20:17 Sup3rkiddo: there's some struct type in , you just cast the array of bytes to that struct and read the fields Jun 25 18:20:25 doc|work: yes, all I care about is having several GB so I can load all kinds of junk in it Jun 25 18:20:31 :) Jun 25 18:20:32 balrog-kun, oh ok.... Jun 25 18:20:39 for a screen touch i think you get three such events Jun 25 18:20:40 I'll be back after lunch to annoy you all, hope you don't mind :) Jun 25 18:20:44 balrog-kun, yeah Jun 25 18:20:45 one for X, one for Y and one for pressure Jun 25 18:21:23 i see Jun 25 18:21:25 for a button, only one for press and one for release iirc, and one of the fields is the keycode (the KEY_ defined in input.h) or a coordinate Jun 25 18:21:34 yes...thats what i see Jun 25 18:21:37 *defines Jun 25 18:21:48 i think i should get a kernel book Jun 25 18:22:02 or a set of kernel headers :) Jun 25 18:22:14 so, is it just me or is 28 days doa a very lame warranty? Jun 25 18:22:27 It's probably just you. Jun 25 18:22:29 :) Jun 25 18:22:32 :/ Jun 25 18:22:44 balrog-kun, :), but books make for good bed-time read Jun 25 18:23:28 http://linuxdevices.com/news/NS9978560959.html Jun 25 18:23:38 Sup3rkiddo: you mean you kernel headers don't? :P Jun 25 18:23:39 so is it out? Jun 25 18:24:26 xming: out months now Jun 25 18:24:31 you're late Jun 25 18:24:36 balrog-kun, na, but books have a charm of their own....i can read it during power cuts Jun 25 18:24:46 which happens regularly this season Jun 25 18:25:04 Sup3rkiddo: you can see in the dark? Jun 25 18:25:17 doc|home, hey i got candles Jun 25 18:25:24 * doc|work has batteries :) Jun 25 18:25:26 and one of those emergency lights Jun 25 18:28:23 I'm actually surprised that got no response yet... Jun 25 18:28:30 * kdean06 twiddles his thumbs Jun 25 18:33:44 kdean06: which got no response? Jun 25 18:34:06 The months one. Jun 25 18:34:11 kdean06: and it's ok, I knew what you meant :) Jun 25 18:34:50 warranty? Jun 25 18:35:28 xming, Linked to a Linux Devices article asking about (I assume) the Freerunner and was told he's late, been out months now. Jun 25 18:35:34 ah Jun 25 18:35:40 kdean06: yeah, me too :) Jun 25 18:35:44 But explaining that kind of ruined the amusement. :P Jun 25 18:35:46 (surprised) Jun 25 18:44:18 kdean06: but it's not where on sale, except that shipment to VIP Jun 25 18:44:46 Damn engadget for that one. Jun 25 18:44:55 I think some shipments left the factory Jun 25 18:45:06 really? great Jun 25 18:45:14 xming, The latest word thus far is that Freerunner would go on sale in early July. Jun 25 18:45:23 So "very soon" now. Jun 25 18:45:32 that's what I understood from our commercial contact at openmoko Jun 25 18:45:42 in 9 days Jun 25 18:45:46 in the US? Jun 25 18:45:48 shipments were expected to start this monday Jun 25 18:45:57 and they apparently did Jun 25 18:46:00 And some VERY select groups have been shipped, some University students in China, I think. Jun 25 18:46:17 any of working for OM? I am going to .tw for 1 month in a week, hoping to pick one in .tw Jun 25 18:46:17 yep, but I think they were shipped earlier Jun 25 18:46:41 And a handful of early production models went to some people. But they will go on sale for "everyone" in July, assuming nothing messes up. :) Jun 25 18:46:50 * doc|work shakes his fist at students Jun 25 18:46:52 fingers crossed Jun 25 18:47:24 kdean06: even in the US? Jun 25 18:47:36 bahadunn, From what I understand, yes. Jun 25 18:47:40 sweet Jun 25 18:47:54 That once the distribution centers have units, the webstore will open for sale. Jun 25 18:47:57 Worldwide. Jun 25 18:48:22 worldwide has a nice sound to it Jun 25 18:48:26 :) Jun 25 18:48:30 Of course, the data the retailers in europe get them may be different, but that's a different situation. I don't think the US has official resellers. Jun 25 18:48:51 the US has their Fremont hub Jun 25 18:49:07 lukhas, I know they ship from there, but that's still Openmoko and not a reseller. Jun 25 18:49:14 right Jun 25 18:49:25 think it will be available through amazon type sites? Jun 25 18:49:27 as an European reseller I sure hope to get it around the same time than US :) Jun 25 18:49:53 lukhas, Personal opinion... I think you guys may get the advantage. :) Jun 25 18:49:59 But only a matter of a few days, if at all. Jun 25 18:50:06 lukhas: where are your shop located? Jun 25 18:50:16 xming: on the Internet :) Jun 25 18:50:24 Paris, France: bearstech.com/shop Jun 25 18:50:26 lukhas: a bit spam now would be nice Jun 25 18:50:30 ty Jun 25 18:50:44 we already started spamming the French press and some geek-oriented sites Jun 25 18:51:08 I meant spamming me with an URL :) Jun 25 18:51:19 lukhas, As an EU retailer, the retailer, not the manufacturer has to provide a warranty or soemthign? Do I remember that being one of EUs regulatory restriction thingies? Jun 25 18:51:23 oh, ok Jun 25 18:51:35 320 euro ca cout cher Jun 25 18:51:36 kdean06: yes, we provide a 1-year "commercial warranty" Jun 25 18:51:52 My question was more "Do you have to?" Jun 25 18:51:53 :) Jun 25 18:51:54 plus the standard european "2 year conformity warranty" Jun 25 18:51:58 yes, we have Jun 25 18:52:01 Freerunner is being distributed to retailers currently correct? Jun 25 18:52:02 EU rules, French rules, etc. Jun 25 18:52:04 But not Openmoko? Jun 25 18:52:05 kdean06: YES he *has* to Jun 25 18:52:30 not openmoko Jun 25 18:52:39 Cause that right there would make a TON of sense why there's EU retailers but no American resellers. Jun 25 18:52:41 until a French/European buyer sues them Jun 25 18:52:51 kdean06: everything sold in EU *HAS* to have at least 1 year, preferably 2 years (or something like that) Jun 25 18:52:56 Eww. Jun 25 18:53:06 Glad Openmoko decided to do hub distribution then. Jun 25 18:53:46 that why everything in EU is 2x more expensive than US (import tax + VAT + after sales service) Jun 25 18:54:01 320 euros is not that expensive, considering you pay around 50 EUR in VAT on this price Jun 25 18:54:05 Yeah. Which is why I'm glad. ) Jun 25 18:54:19 europe is a funky place Jun 25 18:54:34 Selling directly would increase the cost for the entire "pool" of devices. Jun 25 18:54:42 lukhas: I know, still I always get shock by the price in EU Jun 25 18:54:47 yeah, me too Jun 25 18:54:49 It's already inflated enough by customs, taxes, tarrifs... Jun 25 18:55:09 but I calmed down when I noticed that the 399 USD public price of the FR is without taxes Jun 25 18:55:32 lukhas, You're American, no? Jun 25 18:55:40 kdean06: I'm French Jun 25 18:55:44 Oh, sorry. Jun 25 18:55:46 Duh. :) Jun 25 18:55:50 I work for Bearstech, the French reseller Jun 25 18:56:01 Yeah... I was just asking you about EU retailer rules. :P Jun 25 18:56:06 yeah :) Jun 25 19:00:30 http://mobile.slashdot.org/article.pl?sid=08/06/25/1751228 Jun 25 19:00:30 heh Jun 25 19:00:41 "Mass-market" Jun 25 19:01:47 Yeah... Jun 25 19:02:00 it's not fucken loading Jun 25 19:02:02 That term has, IMO, hurt Openmoko's image. Jun 25 19:02:13 is it alive ? Jun 25 19:02:15 freerunner is out ?? Jun 25 19:02:29 loaded Jun 25 19:02:31 How many people expected "mass market" and "the day it goes for sale" to be the same thing? Jun 25 19:02:57 Very few people understood that after the devs get freerunners there will be a while before the device is actually usable. Jun 25 19:02:57 the 4.3" touhcscreen in that summary is also not quite accurate I think? Jun 25 19:03:10 2.8", no? Jun 25 19:03:14 4.3 ? woah Jun 25 19:03:20 hehe Jun 25 19:03:47 and that should be a 400 MHz processor, not a 500 Jun 25 19:04:01 Why do stupid articles get slashdotted Jun 25 19:04:15 it's a 500 MHz processor running at 400 MHz :P Jun 25 19:04:53 afk, food Jun 25 19:05:44 so ÂâE320 is the Frech price? Jun 25 19:08:04 i saw two german resellers, giving E300 and E300 as the price Jun 25 19:08:14 E300 and E350 Jun 25 19:08:59 what retard wrote this summary? Jun 25 19:09:23 me? Jun 25 19:09:57 please correct where i am wrong Jun 25 19:10:58 zlfita: really? 4.3inch screen? 500mhz(it's running at 400mhz)? processor Jun 25 19:11:26 and it's not selling yet :) Jun 25 19:11:34 I think linuxdevices.com pulled their article Jun 25 19:12:24 hmmm Jun 25 19:12:37 i didn't say anything about the screen Jun 25 19:12:41 still on their front page Jun 25 19:12:41 only prices Jun 25 19:12:49 doc|work: click it Jun 25 19:12:54 zlfita: hmmm, someone else screwed that up Jun 25 19:12:58 Qwell: yep Jun 25 19:13:28 zlfita: apologies for the retard comment :) Jun 25 19:14:50 doc|work: no prob,maybe you'll be a bit more careful:) Jun 25 19:14:59 zlfita: maybe you too ;) Jun 25 19:15:20 * doc|work passes zlfita a chocolate brownie Jun 25 19:17:11 doc|work, Do they make other kinds? Jun 25 19:17:19 I love blondies. :) Jun 25 19:17:24 But they're not brownies. :) Jun 25 19:17:31 **yaay brownie! Jun 25 19:17:32 kdean06: hah Jun 25 19:18:15 shit, 30th minute and it's 1-1 Jun 25 19:18:19 they should use brownies for international diplomacy Jun 25 19:18:26 nice idea Jun 25 19:18:56 "here, george, have a brownie." "Thanks mr ahma...ahma... bud, so, sorry about those nuke accusations. You know, we need oil." Jun 25 19:24:35 guys, slashdot says freerunner is out? http://mobile.slashdot.org/article.pl?sid=08/06/25/1751228&from=rss Jun 25 19:26:57 Store says not. Jun 25 19:27:12 not out, just shipping to stores Jun 25 19:27:25 see channel topic Jun 25 19:27:30 The article is TECHNICALLY correct. Jun 25 19:27:40 It HAS begun shipping to resellers. Jun 25 19:28:14 "mass market" though, *cough* Jun 25 19:28:15 But YOU can't buy it yet. First week in July as the mostly-semi-officialish date for the Openmoko store. Jun 25 19:28:58 hmm, the openmoko AGPS doesn't /require/ cell tower support, does it? it just uses it if available? Jun 25 19:29:15 i wish it it was before 4th Jun 25 19:29:31 angasule: To get *A*GPS, you must have some outside source of your approx location Jun 25 19:29:47 But I've read suggestions that it might ask for city/state/country if no GSM availalbe Jun 25 19:29:56 Nope Jun 25 19:29:57 and you can always get GPS (no A), it just takes longer Jun 25 19:30:09 LMFAO Jun 25 19:30:10 It can boot from cold with no info Jun 25 19:30:13 with the ASU will you be able to switch off gsm? Jun 25 19:30:23 I love one of the comments, someone saying they've got 5 pounds a month to spend on blah blah... Jun 25 19:30:33 Someone responded with "Yes you can use VoIP if you so desire. If you only have £5 though, your money might be better spent on other things." Jun 25 19:30:34 wurp2: ok, but an Assistance Server isn't required? it'll work as GPS anyway? Jun 25 19:30:41 kdean06: heh Jun 25 19:30:52 SpeedEvil: AGPS can't boot from cold w/ no info, unless I'm badly misunderstanding Jun 25 19:30:54 doc|home, Yes, I think you already can... BUt don't quote me on that... Jun 25 19:30:56 angasule: yes Jun 25 19:31:01 kdean06: cool, thanks Jun 25 19:31:15 the a in gps of the freerunner means you can feed the chip with data about the satelites from outside. this data normaly comes over internet. Jun 25 19:31:18 I know for sure there's a button, just not sure if it actually does it. :) Jun 25 19:31:18 wurp2: you are Jun 25 19:31:27 hah Jun 25 19:31:29 wurp2: A Jun 25 19:31:37 GPS in most implentations is optional. Jun 25 19:31:43 400 / 5 = 80 / 12 = 6.6 years of not buying bubblegum Jun 25 19:31:43 /if/ you know your location a little this may make the data smaller and make the calculation of the fix faster Jun 25 19:32:00 let's repeat that AGPS will work without A another 100 times! Jun 25 19:32:02 cpp; Jun 25 19:32:05 wurp2: in few do they actually require it. Jun 25 19:32:13 wurp2: and not in the neo or freerunner Jun 25 19:32:17 err, that was 'cool', not weird c++ Jun 25 19:32:37 hrm, are you really using AGPS if you're not assisting it? Jun 25 19:32:45 angasule: what if you already don't buy bubble gum? :( Jun 25 19:32:45 No. Jun 25 19:32:57 wurp2: http://en.wikipedia.org/wiki/AGPS Jun 25 19:33:06 (unless some bastard has vandalised it again) Jun 25 19:33:07 doc|work: then you have to stop buying chocolate Jun 25 19:33:21 * angasule vandalises it Jun 25 19:33:33 angasule: nevar! Jun 25 19:34:14 * angasule laughs evily in croatian, which makes it a lot more evil, since we all know slavic laughs are eviller Jun 25 19:34:36 maybe I could sell my sister into slavery and buy a neo :D Jun 25 19:34:45 angasule: how much? Jun 25 19:34:55 I guess the price of a neo... Jun 25 19:35:12 plus shipping Jun 25 19:35:18 sister or neo? Jun 25 19:35:27 both, I guess Jun 25 19:35:29 and VAT Jun 25 19:35:31 angasule: where from? Jun 25 19:35:51 SpeedEvil: That description looks an awful lot like the position I was arguing from Jun 25 19:35:53 argentina, she's in pretty good condition, I have two sisters, interested? Jun 25 19:36:15 wurp2: I may have been misunderstanding your position. Jun 25 19:36:17 I wasn't aware that it provided orbital data, though... that should be pre-calculable, and already in the GPS software Jun 25 19:36:21 zlfita: I'll teach them dutch, one is blonde like maxima! :P Jun 25 19:36:28 wurp2: orbits drift over time. Jun 25 19:36:36 Well, I was calling it normal GPS if you're not getting assistance Jun 25 19:36:43 wurp2: they are on low orbits Jun 25 19:36:45 wurp2: rapidly enough that you need to refresh them every week or so at minimum. Jun 25 19:36:51 angasule: no they're not Jun 25 19:36:57 angasule: they're halfway to GEO Jun 25 19:37:01 angasule: 12 hour orbits Jun 25 19:37:17 SpeedEvil: Wow, I wouldn't have thought there would be anywhere near that amount of unpredictable variance in an orbit... Jun 25 19:37:43 SpeedEvil: halfway to GEO != 12 hour orbit... Jun 25 19:37:47 angasule: thanks but i am not dutch Jun 25 19:37:48 SpeedEvil: wow, you're right, I always figured they were below 700 Jun 25 19:37:53 wurp2: chronologically Jun 25 19:38:00 mkay Jun 25 19:38:12 wurp2: 12 hours is still very far from LEO, he is right :) Jun 25 19:38:33 LEO is 90 miniutes ish. Jun 25 19:38:46 With your satellite decaying into the atmosphere in a few decades tops. Jun 25 19:38:56 SpeedEvil: #space ? :D Jun 25 19:39:03 MEO - 12 hours - takes thousands or millions of years. Jun 25 19:39:24 ok, people, you have to be careful with this stuff Jun 25 19:39:34 'meo' is spanish for 'I piss' Jun 25 19:39:52 'moko' sounds like spanish 'bugger' (spelt 'moco') Jun 25 19:40:03 * SpeedEvil contemplates a satellite constellation companion to OM. Jun 25 19:40:09 MEOMoko. Jun 25 19:40:40 SpeedEvil: that's disgusting heh Jun 25 19:40:48 I like the name Neo FreeRunner Jun 25 19:41:37 Moko freerunner. Jun 25 19:41:42 moco = booger (just incase it confused anyone) or snot. :) Jun 25 19:41:44 The fun never ends! Jun 25 19:41:50 Snot a bad name! Jun 25 19:42:01 oh, booger Jun 25 19:42:03 yes, snot Jun 25 19:42:10 Open Snot, great name Jun 25 19:42:18 That said... IF Openmoko decided to come up with a name that didn't offend SOMEONE it would be called the Wii... Jun 25 19:42:24 Oh, wait, that's already being used. Jun 25 19:42:24 SpeedEvil: your "decaying in decades" looked like it referred to the 12 hour orbit Jun 25 19:42:30 With the theme-song by Baccara. Yes Sir, I can Bogie. Jun 25 19:42:41 wurp2: ah - no - LEO only. Jun 25 19:42:43 which sounded totally bogus to me, so I've been searching atmospheric densities... I see now that you meant LEO Jun 25 19:42:59 wurp2: for exaple, ISS will crash in a couple of years if not reboosted. Jun 25 19:43:20 I still don't see how a satellite w/ a 12 hour orbit can drift enough in months that you can't know pretty much exactly where it is Jun 25 19:43:34 SpeedEvil: years? much less, I think Jun 25 19:43:40 wurp2: You don't care about pretty much Jun 25 19:43:46 wurp2: probably perturbing bodies like the moon Jun 25 19:43:47 wurp2: you care about +-1m Jun 25 19:43:48 If you don't have any atmospheric drag, orbits should be permanent Jun 25 19:44:04 wurp2: if the earth were along in the universe, sure Jun 25 19:44:24 angasule: That's still just a 2 body problem Jun 25 19:44:24 angasule: it depends on the point in the cycle where it's reboosted. Reboosts happen every 12 months IIRC normally. Jun 25 19:44:37 wurp2: it's not that simple. Jun 25 19:44:40 angasule: Unless you're transfering momentum to another body, orbits should be permanent Jun 25 19:44:42 A graph of altitude over time shows that it drifts down almost 2.5 km per month. Jun 25 19:44:49 wikipedia dixit Jun 25 19:44:59 and of course, the lower it is, the faster it drops Jun 25 19:45:01 angasule: What drifts? Jun 25 19:45:07 wurp2: ISS Jun 25 19:45:12 wurp2: random stuff like thrusts on solar panels from the sun not being perfectly accounted for. Jun 25 19:45:54 angasule: It vairs over a wide band - 100Km + though Jun 25 19:47:21 SpeedEvil: hmm... why the hell isn't the satellite itself transmitting info about its current position, then? Needing to communicate with a server about global information about a device that's already transmitting to you seems pointless Jun 25 19:47:26 angasule: I think this was from a discussion on how long it could last between reboost optimally Jun 25 19:47:29 wurp2: it does Jun 25 19:47:40 wurp2: it transmits a complete cycle every 12 mins Jun 25 19:47:42 I want one. :) Jun 25 19:47:48 Where can I get one? Jun 25 19:47:55 wurp2: but, you need to recieve that. - or the intermediate 30s info Jun 25 19:48:09 jeward: check your shoes on january 5? Jun 25 19:48:15 wurp2: if you have that info, you can get a fix in well under 30s Jun 25 19:48:21 I'm barefoot. Jun 25 19:48:36 SpeedEvil: why does the FR take 1-2 minutes with AGPS, then? :-/ Jun 25 19:48:50 wurp2: It'[s not using AGPS Jun 25 19:49:02 SpeedEvil: It was in the info I was reading Jun 25 19:49:07 wurp2: just conventional GPS, with satellite orbit data stored on flash. Jun 25 19:49:24 SpeedEvil: It took this fellow 12 mins to get a lock with just GPS, 1-2 min with AGPS Jun 25 19:49:30 wurp2: well - if it is - it's doing it wrong. I've got any number of GPSs that will get _way_ better than 2 min locks Jun 25 19:49:43 wurp2: IIRC there was a problem with early FRs Jun 25 19:49:58 wurp2: something in the anntenna mux being improperly driven. Jun 25 19:50:03 Yeah, every report I've heard of FR GPS was of it taking an almost unbelievably long time to get a fix :-( Jun 25 19:50:04 wurp2: It really depends. Jun 25 19:50:29 wurp2: Usually when standing still outside my FR gets a fix in under a minute. Jun 25 19:50:44 alphaone: With or without assistance? Jun 25 19:50:46 angasule: They'll become available January 5th? Jun 25 19:50:48 Without any information whatsoever Jun 25 19:50:56 alphaone: That's great news! Jun 25 19:51:20 alphaone: what hardware is this? Jun 25 19:51:23 wurp2: Well... Jun 25 19:51:32 alphaone: "without any information whatsoever" means you cleared its almanac too? Jun 25 19:51:45 The thing is as soon as I'm moving and conditions are not so optimal it can take AGES. Jun 25 19:51:59 * SpeedEvil stabs the HH people. Jun 25 19:52:11 * SpeedEvil wishes for datasheets so he could code a nice GPS solution. Jun 25 19:52:14 mmontour: The almanac and other stuff is cleared as soon as the GPS is powered down. Jun 25 19:52:23 jeward: no, on the 5th of january the three hmm, wizards? bring gift, it's a christian thing, I just like the gifts Jun 25 19:52:35 SpeedEvil: There are datasheets on the u-blox page Jun 25 19:52:42 alphaone: for HH Jun 25 19:52:44 I'm both barefoot and non-christian. You've lost me. Jun 25 19:52:46 alphaone: hammerhead Jun 25 19:52:52 alphaone: the GPS in GTA01 Jun 25 19:52:55 ah, I see Jun 25 19:53:20 jeward: I'm also not christian, but I like getting gifts :D Jun 25 19:53:25 There seems to be a way to get the raw GPS data from the u-blox chip as well Jun 25 19:53:33 with the UBX protocoll Jun 25 19:55:09 SpeedEvil: Hardware ist GTA02v5 IIRC Jun 25 19:55:15 or v6 Jun 25 20:06:16 alphaone: not nearly as raw as HH Jun 25 20:06:37 alphaone: with HH, it's basically a software-defined radio - with the hard-realtime only done in the HH Jun 25 20:06:46 SpeedEvil: Might be, I haven't looked too closely Jun 25 20:06:52 alphaone: the UBX stuff gives you rather more 'cooked' info Jun 25 20:07:24 still, I'll take u-blox over HH any day Jun 25 20:08:00 Without drivers, I agree. Jun 25 20:08:04 You can do some pretty cool stuff if you speak UBX Jun 25 20:08:11 s/t d/t open source d/ Jun 25 20:08:11 SpeedEvil meant: Without open source drivers, I agree. Jun 25 20:11:27 Can someone with a FR speak to this?: http://mobile.slashdot.org/comments.pl?sid=595147&cid=23939209 Jun 25 20:11:48 I know some of it is true, and some other bits aren't true, but I know nothing about stereo headsets on the FR Jun 25 20:12:07 heh - I just (seconds ago) replied to that one. Jun 25 20:12:31 why is slashdot saying that openmoko is out but the topic has not changed? Jun 25 20:12:37 am I missing anything? Jun 25 20:12:40 apol: because it isn't out Jun 25 20:12:41 apol: slashdot are idiots? Jun 25 20:12:54 apol: let's put it this way - linuxdevices.com has pulled their article. Jun 25 20:13:03 apol: while it may be technically true (or not - nobodies commented) that they're shipping to vendors. Jun 25 20:13:12 mmontour: Good point about the headset - if you use the nokia compatible one, it *is* mono Jun 25 20:13:13 They are not purchasable. Jun 25 20:13:32 ok Jun 25 20:13:37 then I'll forget about that :P Jun 25 20:14:21 wurp2: how could he get a freerunner so soon? maybe a beta? Jun 25 20:14:44 angasule: or reading mailing lists for apple, and the power of lies! Jun 25 20:15:54 SpeedEvil: steve jobs is a sith lord! but the question is, is he the master, or the apprentice? Jun 25 20:16:41 * SpeedEvil is betting on a Gungan in disguise. Jun 25 20:17:04 hahaha Jun 25 20:17:10 that was mean :P Jun 25 20:24:51 Some people at universities already got Freerunners Jun 25 20:25:16 I haven't got one Jun 25 20:25:21 A few people have drifted through here talking about their new Freerunners Jun 25 20:25:21 I'm at a university Jun 25 20:25:27 and my bachelor project is next year :p Jun 25 20:25:28 CVirus: You're obviously at the wrong university Jun 25 20:25:32 lol Jun 25 20:25:33 Indeed. :P Jun 25 20:25:54 actually my bachelor project is in 2 years ... that's why I need the FreeRunner a bit earlier than others :-p Jun 25 20:26:00 It really shocked me that some apparently random folks at a university were getting them, when people doing GSoC didn't have one yet. Jun 25 20:26:44 slashdot is reporting that the freerunner is shipping, is that true? Jun 25 20:26:50 No. Jun 25 20:27:14 it's shipping from one continent to another, perhaps, not to homes Jun 25 20:27:24 hehe Jun 25 20:27:44 :) Jun 25 20:28:09 much better :-) Jun 25 20:28:36 haha. i guess if i read more closely (http://mobile.slashdot.org/mobile/08/06/25/1751228.shtml) it says 'to distributors' Jun 25 20:28:53 Has that topic ever actually prevented someoen from asking "When's freerunner coming out?" Jun 25 20:28:59 Dunno. Jun 25 20:29:10 we need a bot Jun 25 20:29:15 * SpeedEvil ponders changing it to 'freerunner sold out'. Jun 25 20:30:21 GOAALLL!! Jun 25 20:41:58 haha, subject ;-) Jun 25 20:42:21 There has been a question or two Jun 25 20:42:55 I bet it's been non-stop Jun 25 20:51:08 lol, just read in slashdot, and then i read the topic :) Jun 25 20:56:45 kdean06: See! Jun 25 20:57:00 Wow. Jun 25 20:57:13 TAsn, Is the first human being to ever read a topic before posting. Damn. Jun 25 20:57:15 They do exist! (people that read the topic) Jun 25 20:57:36 BRB, I need to go collect a bet, seems snowballs DO have a chance in hell now... Jun 25 20:58:06 kdean06, worship me :) Jun 25 20:58:21 No thanks, not into that whole "authority over others" thing. Jun 25 20:58:43 well at least buy me a moko :) Jun 25 20:58:52 damn, I want one! Jun 25 20:59:19 We get now - let's say - 5 freerunner questions per day, maybe it would be like 20 without the topic ;) Jun 25 20:59:20 Where can I find the companies that will sell the freerunner I have only found pulster :P Jun 25 20:59:55 As a total side note... Somehow I was just reminded of the Sega GameGear... Jun 25 21:00:20 So I looked it up on Wikipedia and I thought "those specs have to be a typo". Jun 25 21:00:25 Yay for advancement. Jun 25 21:00:35 openmoko: 03werner * r4499 10/trunk/src/host/envedit/Makefile: Jun 25 21:00:35 openmoko: - remove dependencies in "make clean" Jun 25 21:00:35 openmoko: - actually make use of dependencies Jun 25 21:00:37 Yeah. Jun 25 21:00:57 And at one time, 48K/8 bit was a valid game platform. Jun 25 21:01:43 Now I'm contemplating manufacturing a 32 bit 16M computer on a 2cm*2cm PCB Jun 25 21:03:41 boo, no 3D accel? :( Jun 25 21:06:01 balrog-kun has some aspirations towards that, or had anyway, but not officially Jun 25 21:07:38 I wonder what kind of 3D can the chip do Jun 25 21:08:07 the answer is somewhat little Jun 25 21:09:51 does any openmoko variant have stereo headphone support? Jun 25 21:10:02 because I have an idea :D Jun 25 21:10:12 basically same stuff the other 3D accels can do (full OpenGL ES iirc) Jun 25 21:10:21 a potentially amazing idea that will get me all the chicks Jun 25 21:10:25 just scaled down in max resolutions a bit Jun 25 21:10:28 balrog-kun: that's quite a bit Jun 25 21:10:35 yeah Jun 25 21:10:39 not gonna run crysis Jun 25 21:10:49 but some fun stuff, and useful stuff, too Jun 25 21:11:08 hmm, sorry for the stupid question, but is there an emulator? Jun 25 21:11:31 for the freerunner hardware, nope Jun 25 21:11:38 for any openmoko variant? Jun 25 21:11:41 not that i know of Jun 25 21:11:45 for the neo1973 yeah Jun 25 21:11:48 if I develop for one, porting shouldn't be a big problem, right? Jun 25 21:12:10 qemu is an emulator Jun 25 21:12:10 yeah, no porting needed, even if you develop on a different ARM device or a different ARM emulator Jun 25 21:12:32 well, I'll develop on AMD64 :) Jun 25 21:12:55 good idea :) Jun 25 21:12:57 once I finish this OpenAL thingie and the distcc thingie, I was planning on a little VoIP thing Jun 25 21:13:31 nothing new or special, except: it would support positional sound (with OpenAL) and would have an interface to take advantage of it Jun 25 21:13:46 so a call with multiple people doesn't sound like The Voices are angry Jun 25 21:13:59 what do you think? :) Jun 25 21:14:27 with stereo headphones it's possible to do 3D sound, btw, obviously it'd be awful on the mono speaker Jun 25 21:15:14 Well. You don't really need to do full 3d processing for that Jun 25 21:15:41 * bagpuss_thecat wonders if the OpenMoko A2DP support will also send callerid information to his headphones :-) Jun 25 21:15:46 just mix differently along the left-right-axis. A useful idea, to be sure. Jun 25 21:16:05 mjr: ever heard an HRTF demo? Jun 25 21:16:37 it can do front/back distinction Jun 25 21:16:46 angasule: Regarding the stereo headset, 1973 & Freerunner both support stereo headset Jun 25 21:16:53 cool Jun 25 21:16:54 I know, I'm just saying you don't need to do that to be useful Jun 25 21:17:09 angasule: FR doesn't have two external speakers, but who gets stereo from speakers 2" apart anyway Jun 25 21:17:17 mjr: well, I'd be developing the thingie for PCs Jun 25 21:17:20 and I do wonder if the neo has the horsepower for full 3d aural imaging; I dunno? Jun 25 21:17:27 wurp2: you do get a meaningfull amount of stereo. Jun 25 21:17:30 wurp2: yes, I meant for headphone use only Jun 25 21:17:34 hello mjr :) Jun 25 21:17:54 wurp2, I do! I walk down the street with the 1973 upside-down stuck at the bridge of my nose so the speakers are between my ears. Jun 25 21:18:13 kdean06: Now if I see you walking down the street, I'll know who you are. Jun 25 21:18:15 The lady with too many dogs and a really bad odor looks funny at me when i do that. Jun 25 21:18:19 what graphical toolkit does openmoko use? and what sound backend? alsa? Jun 25 21:18:36 angasule: Many, and alsa + pulseaudio Jun 25 21:18:39 the toolkit thing is kinda a mishmash now Jun 25 21:18:45 angasule, Openmoko is built on X, it uses all of them. :) Uses Alsa, plus Gstreamer and Pulseaudio Jun 25 21:18:47 angasule: Although pulseaudio may go away Jun 25 21:19:05 but there is no preference for toolkit? I like qt... :) Jun 25 21:19:10 WE're not sure. Jun 25 21:19:14 Honestly. Jun 25 21:19:24 angasule, the qtopia phone stack is the current choice for phone functions Jun 25 21:19:27 My understanding of what's happening was muddied yesterday. :S Jun 25 21:19:46 But EFL for non-phone apps. :S Jun 25 21:19:53 gtk has been previously used, and efl seems to be the hip trend now... Jun 25 21:19:53 And GTK for non-phone, non-EFL apps. Jun 25 21:20:01 Meh. Jun 25 21:20:07 mjr: neat Jun 25 21:20:11 what's EFL? never heard of it Jun 25 21:20:24 why have 3 graphical toolkits? O_O Jun 25 21:20:31 Enlightenment FOundation Libraries. Jun 25 21:20:40 ah Jun 25 21:20:48 angasule: > 3 ;-) Jun 25 21:21:07 Each toolkit has strengths and weaknesses, and because the aim of openmoko was to NOT make that choice on an Open platform. Jun 25 21:21:10 Let the user decide. Jun 25 21:21:11 angasule: So you can port arbitrary apps, I guess Jun 25 21:21:26 angasule: But I do agree that it would be nice if they chose a default one internally Jun 25 21:21:32 angasule, well, qtopia was taken mostly because they supposedly have good working phone stack (turns out less than advertised and in need of further work, but meh, it seems to mostly work) Jun 25 21:21:46 ok, I just asked the main dev of openal-soft but he hasn't done profiling, no idea how much power is required Jun 25 21:21:48 Perhaps I am simpl becoming more and more cynical, but I don't think "giving users a choice" will ever be appealing in a mass market device. Jun 25 21:22:20 I have used qtopia-core for x86 apps (running on the framebuffer) Jun 25 21:22:32 It's becoming more and more clear to me that people would rather not think about what's on their phone, TV, computer and simply muddle on. Jun 25 21:22:35 we are sort of developing a DE for old PCs Jun 25 21:23:27 hmm, does the openmoko CPU work well with floats? or it's better to go with integers? Jun 25 21:23:36 angasule, no hard floats Jun 25 21:23:43 so or, yes Jun 25 21:24:06 ah, that would definitely have to be changed in openal-soft, then, since positions are specified as floats Jun 25 21:25:12 OpenGL ES is integer oriented? Jun 25 21:26:45 anyway, I seem to have cut short on the different toolkits; gtk was the traditional openmoko toolkit, and efl came with raster, apparently eg. for its ability to do nifty stuff pretty fast or something ;) Jun 25 21:27:08 so that's how we got here, where we go, nobody knows Jun 25 21:27:33 one hopes with theming the user doesn't have to care overly much Jun 25 21:42:42 * mmontour sees another discussion of toolkits, and notes that he's running a couple of lesstif-based apps :) Jun 25 21:43:19 * doc|work wants punch card support Jun 25 21:43:50 Openmoko Neo Difference Engine Jun 25 21:44:26 ONDE, we can do better than that, NODE Neo Openmoko Difference Engine? Doesn't really work but the acronym is so much better :) Jun 25 21:45:12 Yoda agrees! Jun 25 21:45:17 what's the native integer size on the OM CPU? Jun 25 21:45:30 32 Jun 25 21:46:39 cool Jun 25 22:03:09 If some poor EE student puts an S-R latch into the forbidden state, then Q=Q', and all the basic assumptions of logic fail. All the laws of mathematics and physics break down, 0 equals Infinity, atomic nucleii repel their electrons, matter ceases to exist, and the universe disappears in a puff of logic. Jun 25 22:18:50 wurp2, Did you have the AT&T firmware thing fixed? Jun 25 22:24:51 is it true? the phone is available? Jun 25 22:25:12 topic Jun 25 22:25:23 heh. Jun 25 22:25:31 * russellb has one Jun 25 22:25:32 * russellb runs Jun 25 22:26:32 more precisely, shipments are reportedly being hauled to some reseller points Jun 25 22:26:34 kdean06: Yes Jun 25 22:26:45 kdean06: If you mean the one that made some AT&T SIM cards not work in the 1973 Jun 25 22:26:56 Yep, that's what I meant. :) Jun 25 22:27:14 they will presumably announce availability when it is available Jun 25 22:27:24 ahh, oops Jun 25 22:27:30 kdean06: You can check the revision of your firmware... mine is 8 Jun 25 22:27:59 I would have to search around for the gsm command to do that check Jun 25 22:30:27 AT%VER was one Jun 25 22:31:21 Actually, I just wanted to confirm a statement I was sending to the list. :) Jun 25 22:31:47 At one point I filled out the Wiki page with my firmware version. Jun 25 23:05:22 <`pwgen_64`> mickeyl: i tested the linux-openmoko-devel_git . resume/suspend is working. phone call ... tomorrow Jun 25 23:05:36 excellent. thanks. keep me posted Jun 25 23:06:17 <`pwgen_64`> ah, i can mount the sd card ! Jun 25 23:06:23 yay Jun 25 23:09:00 <`pwgen_64`> mickeyl: what is the best way to give the 'Demo Only' some function Jun 25 23:09:33 hook into the 'transition to' Jun 25 23:09:39 and add some own code Jun 25 23:09:48 e.g. an edje Jun 25 23:09:54 or something completely different Jun 25 23:10:13 take a moment and study the .edc Jun 25 23:10:16 <`pwgen_64`> are there plans for the function behind the buttons Jun 25 23:10:21 yeah Jun 25 23:10:25 for now: Jun 25 23:10:42 DIAL | PHONEBOOK | SMS | PREFERENCES Jun 25 23:10:52 later more, of course Jun 25 23:11:07 a whole universe of cool apps waiting for us to finish the basics Jun 25 23:13:29 mickey, I like how the phone vibrate-ring works even when I've killed the UI to do dev stuff. Jun 25 23:13:52 hehehe Jun 25 23:13:57 yeah, the power of daemons Jun 25 23:14:04 actually, one of my goals is Jun 25 23:14:10 to enable people to do a ncurses UI Jun 25 23:14:12 they also have an article on linuxdevices.com saying the Freerunner is out Jun 25 23:14:30 so it's not really out I take it? Jun 25 23:14:37 so i will take care that everything is independent of any UI Jun 25 23:14:53 mickeyl: cool Jun 25 23:15:09 <`pwgen_64`> i have written a simple gtk gui for pyneod .. seem to update this ... Jun 25 23:15:36 mickeyl, great idea -- i was thinking of a cmd-line answer, should all be easy with the framework Jun 25 23:16:08 erko: yep. dead-easy. and we invite all people doing cool stuff. the only requirement is dbus Jun 25 23:16:15 perl answering machine? be our guest Jun 25 23:16:18 :) Jun 25 23:16:27 :) Jun 25 23:16:29 bash sms firewall? no problem Jun 25 23:16:31 ... Jun 25 23:17:28 <`pwgen_64`> dbus-send --system --..... will be enough to do calls ,so a bash cmd interface is already there . Jun 25 23:17:57 yeah, although that's cumbersome Jun 25 23:18:03 qtdbus got it right Jun 25 23:18:24 <`pwgen_64`> dbus-inspector .... is extrem cool ... Jun 25 23:19:06 <`pwgen_64`> with x-11 enabled dbus-launch its possible to use it Jun 25 23:20:39 What is Sean responsible for ? Jun 25 23:21:08 which Sean? Jun 25 23:21:36 Sean Mozz-Pultz Jun 25 23:22:31 ah. Sean Moss-Pultz is the president of Openmoko, Inc. Jun 25 23:22:43 he is responsible for steering the whole ship Jun 25 23:22:54 * doc|work completely didn't see the l there :o Jun 25 23:22:56 on a high level Jun 25 23:23:04 I see Jun 25 23:41:54 openmoko.com has links yays!! Jun 25 23:55:08 openmoko: 03werner * r4500 10/trunk/src/host/envedit/ (tests/Common tests/Makefile xtests/Makefile): Jun 25 23:55:08 openmoko: - tests/Common: invoke "echo" as /bin/echo to avoid backslash interpretation Jun 25 23:55:08 openmoko: - tests/Makefile, xtests/Makefile: set PATH to include . Jun 25 23:55:52 linuxxr: that's nothing, qtopia will have webkit :) Jun 25 23:57:19 openmoko.org and openmoko.com are both down ? Jun 25 23:57:51 what happened to openmoko.com !! Jun 25 23:58:09 I Jun 25 23:58:19 ''m guessing they may be down for updates. Jun 25 23:58:39 it's up now Jun 25 23:58:44 it has completely changed Jun 25 23:59:18 Oh. It now points to the iphone page. Jun 25 23:59:22 (kidding) Jun 25 23:59:31 lol Jun 26 00:00:53 It's been like that for a hwile. Jun 26 00:01:03 https://www.truebox.co.uk/trueboxportal/index.php?wk=OpenMoko Jun 26 00:01:11 First batch shipped 25th June, we are now ordering more Jun 26 00:01:57 Err - what? Jun 26 00:02:13 that's what the link says Jun 26 00:02:19 looks like distributors started shipping alreadt Jun 26 00:02:23 already* Jun 26 00:02:32 Sure - but how did they have any orders? Jun 26 00:03:39 no idea Jun 26 00:08:46 hello,what is the most usable gui for openmoko as now? qtopia? the matchbox gui? another one...i want to be able to make phones calls without typing at commands when i get one... Jun 26 00:09:00 s/at/AT Jun 26 00:09:38 FSO I think Jun 26 00:10:16 and what do you mean by not out yet? do you mean that the online store is not ready? but what about other distributors? Jun 26 00:10:51 http://www.vanille-media.de/site/index.php/2008/05/05/openmoko-framework-initiative/ Jun 26 00:11:09 I haven't seen a site with a purchasable neo. Jun 26 00:11:19 Though I'm looking through the list of distributors Jun 26 00:11:42 CVirus, thanks i didn't think the new stack initiative was already ready...wow!!! so i can choose my GUI Jun 26 00:12:16 no problem Jun 26 00:12:24 you can choose... but it isn't really ready :) Jun 26 00:15:48 SpeedEvil, i saw bearstech's website and it seems that you can buy it...i wonder if they sell it only online...(it would be great to buy it not trough the internet...) Jun 26 00:16:05 SpeedEvil, it's in Europe/France/Paris Jun 26 00:16:26 I can't at the moment anyway. Jun 26 00:16:38 Needed to buy more bits for roof :( Jun 26 00:16:47 SpeedEvil, ah "pre commande..." so not ready yet Jun 26 00:17:14 GNUtoo: same with other two stores I looked at Jun 26 00:17:26 SpeedEvil, ok Jun 26 00:17:41 GNUtoo: well - the UK one claimed to have shipped some - but they may not have had a 'waiting for initial stock' state for their items. Jun 26 00:18:22 SpeedEvil, as my brother is in france it's easier for me to buy in france... Jun 26 00:41:08 ? Jun 26 00:47:27 Tanya: yeah .. this is the interbutt Jun 26 00:47:37 it's a bit confusing at the beginning Jun 26 00:51:54 openmoko: 03raster 07org.openmoko.asu.dev * r4eed2db4e169 10/conf/distro/include/sane-srcrevs.inc: up to 116 for illume - new config defaults for aux btn Jun 26 00:51:54 openmoko: 03raster 07org.openmoko.asu.dev * rc95af4cd8cc2 10/ (3 files in 3 dirs): Merge branch 'org.openmoko.asu.dev' of git+ssh://git@git.openmoko.org/var/cache/git/openmoko into org.openmoko.asu.dev Jun 26 00:53:16 |= Jun 26 00:53:52 :| Jun 26 00:54:13 raster: hey there Jun 26 00:54:34 8] Jun 26 00:55:52 http://trisoft.de/en_openmoko.htm Jun 26 00:56:35 https://www.yoss.de/openmoko/en_preorder.htm Jun 26 00:57:05 let the chaos begin Jun 26 00:57:22 :) Jun 26 00:57:39 yeeehaaa Jun 26 00:58:57 http://linuxfr.org/2008/06/17/24221.html Jun 26 01:02:58 as i said... Jun 26 01:03:01 let the chaos begin Jun 26 01:03:02 :) Jun 26 01:03:25 let the pious vanish for all time Jun 26 01:03:27 any spare one for me ? Jun 26 01:03:38 need to port Haiku to that :) Jun 26 01:05:22 mmu_man: enjoy the porting. :) Jun 26 01:06:07 gnnh. Jun 26 01:06:15 i-hate-tha-web Jun 26 01:06:45 All of it? Jun 26 01:07:06 We should go back to gopher. Jun 26 01:07:11 yeah Jun 26 01:07:36 bah.. just wasted the last few hours fiddeling with urls and details of html and such Jun 26 01:07:50 da weeeb Jun 26 01:07:57 'making it work' when one gets a 'new site' as zip ;) Jun 26 01:07:58 lets stop using software Jun 26 01:08:02 lets make everythign web-based Jun 26 01:08:08 who needs an os Jun 26 01:08:08 *shoot* Jun 26 01:08:11 all u need is a browser Jun 26 01:09:16 I see raster has gone over to iPhone Jun 26 01:09:24 yes Jun 26 01:09:30 i'm all about apple poroducts Jun 26 01:09:37 they are the bestss0rz Jun 26 01:09:45 steve jobs is my hero Jun 26 01:09:50 everything i own is apple Jun 26 01:10:04 and the future is the weeeeb Jun 26 01:10:13 web spreadhseets Jun 26 01:10:15 web mail Jun 26 01:10:18 web irc Jun 26 01:10:41 all on my shiny iphone with multitouch Jun 26 01:10:51 :) Jun 26 01:12:50 shoot that thing Jun 26 01:13:34 fuck the web Jun 26 01:13:37 fuck http Jun 26 01:13:41 fuck all browsers Jun 26 01:13:42 * raster slaps some tags around the place Jun 26 01:14:06 Watch out! Jun 26 01:14:12 The ends of those are pointy! Jun 26 01:14:35 O_o Jun 26 01:14:47 * jeffszusz wonders why CVirus is so irate Jun 26 01:14:51 firefox 3 is eating 20% of my fucken ram because of 5 opened tabs .. fuck it Jun 26 01:14:58 lol Jun 26 01:15:09 hmm.. the headphone jack is on the side of the phone? Jun 26 01:15:15 hehehe Jun 26 01:15:17 kelvie: yes Jun 26 01:15:20 wouldn't it be better at the top or bottom? Jun 26 01:15:26 kelvie: it's not. Jun 26 01:15:27 or ahs this been discussed before? :p Jun 26 01:15:35 CVirus: dude - don;'t u know. u need 4gb of ram to browse the weeeeb these days. Jun 26 01:15:38 kelvie: where it might or might not have been is kind of a moot point. Jun 26 01:15:41 thats why i got 4gb Jun 26 01:15:45 for da weeeeb Jun 26 01:15:50 raster: yaeh .. I noticed Jun 26 01:15:53 yeah* Jun 26 01:15:56 * SpeedEvil hs 192 on this box. Jun 26 01:16:02 http://www.pulster.de/ ... 299euros freerunners Jun 26 01:16:11 * SpeedEvil is contemplating getting out the soldering iron to upgrade. Jun 26 01:16:17 CVirus: note the '0' price. Jun 26 01:16:27 CVirus: next to 'buy' Jun 26 01:16:30 hehe Jun 26 01:17:51 Preis: EUR 19.00 Openmoko Freerunner Tasche Jun 26 01:18:09 I guess that's just a charger. Jun 26 01:18:21 SpeedEvil: ah.. so there was discussion about it? Jun 26 01:18:22 that's the pouch Jun 26 01:18:26 CVirus: I was just looking at that. Time to put mine on eBay. Jun 26 01:18:30 oh Jun 26 01:18:51 kelvie: No. Not externally at least. That part of the design was frozen _long_ ago Jun 26 01:18:51 it's 10euros more expensive than the earphones ! Jun 26 01:19:07 kelvie: 2006? Jun 26 01:19:11 oh wow Jun 26 01:19:25 kelvie: you can use bleutotoh phones too Jun 26 01:20:04 interesting Jun 26 01:20:35 wait.. in stereo? Jun 26 01:22:41 yes. Jun 26 01:22:46 Google a2dp Jun 26 01:24:12 SpeedEvil: has anyone done a2dp on the Neo, or is it just theory? Jun 26 01:24:23 Not that I know of. Jun 26 01:24:36 I tried for a bit, but couldn't get it working Jun 26 01:24:50 Not related to the neo - just my poor understanding of the stack Jun 26 01:31:29 CVirus Tasche means 'bag' Jun 26 01:32:15 mmontour same game as on a notebook. that part is default bleuz Jun 26 01:32:18 eh bluez Jun 26 01:32:54 afiak there is a gsoc about tie-ing up bluez and pulseaudio to do proper a2dp including pairing stuff Jun 26 01:33:57 http://code.google.com/soc/2008/bluez/appinfo.html?csaid=2218999748B418AE Jun 26 01:34:02 roh: I was thinking more of audio-routing issues through the wm8753 chip. Jun 26 01:34:21 mmontour a2dp doesnt pass the wm at all. Jun 26 01:34:39 the pcm link on the csr bt is only of use in sdp mode afaik Jun 26 01:35:16 OK. I don't know much about BT audio, except what I've learned while trying to get a headset to work with the Neo. Jun 26 01:36:03 was it sdp? i think so.. the 'mono fullduplex voice quality thingie for calling' Jun 26 01:36:16 the stereo-hifi mode is higher latency and called a2dp Jun 26 01:39:13 It appears as another alsa device Jun 26 01:39:17 As I understood it Jun 26 01:39:18 SpeedEvil: they put 1gb on the 3g iphone and the softbank 922sh because of the weeeeb. :) Jun 26 01:39:23 But I never got it working. Jun 26 01:39:50 (on neo or desktop) Jun 26 01:46:03 openmoko: 03werner * r4501 10/trunk/src/host/envedit/ (10 files in 2 dirs): Jun 26 01:46:03 openmoko: Let's try this again. The previous solution didn't solve the problem when Jun 26 01:46:03 openmoko: manually running test, e.g., ./blanks Jun 26 01:46:03 openmoko: Thus we have to give the relative path for each invocation of Common. **** ENDING LOGGING AT Thu Jun 26 02:59:57 2008