**** BEGIN LOGGING AT Sun Mar 25 03:01:03 2007 Mar 25 03:01:04 hmm. why on Earth would it use an SVR4 option..? weird Mar 25 03:02:10 I stated another build with a generic .bashrc (my default one has lots in it) a couple hours ago. Mar 25 03:02:26 I shoud get to the point where it dies soon. Mar 25 03:02:32 Ocibu: good plan Mar 25 03:03:20 Heh, it just started gcc-cross-4.1.1-r12 Mar 25 03:04:34 fingers crossed Mar 25 03:07:22 Ocibu: are you using the OpenMokoMake? Mar 25 03:07:36 Yes, that is how I invoke things Mar 25 03:07:49 I have tried using bb from scratch and get the same error Mar 25 03:08:01 Too bad it takes so long to do a test Mar 25 03:08:14 yeah Mar 25 03:09:16 have you had any trouble getting updates for the OpenMokoMake? I've been having issues today. Mar 25 03:10:30 I did a make update-makefile earlier and didn't get any errors Mar 25 03:10:37 I didn't get any updates either Mar 25 03:11:13 I've been getting time-outs pulling stuff from svn.nslu2-linux.org Mar 25 03:11:50 have you had a look at the config.status files generated for the gcc-cross builds? Mar 25 03:12:25 no. is that in the build/tmp dir? Mar 25 03:13:05 yeah - build/tmp/work/ Mar 25 03:13:17 looking Mar 25 03:13:35 actually, it'll be interesting to see where gcc-cross is getting built Mar 25 03:17:19 . Mar 25 03:18:53 Its getting built in the work/arm4t-linux dir Mar 25 03:19:11 seems normal Mar 25 03:19:30 The config is there but it will take a while to see if anything looks amiss Mar 25 03:20:57 I think it worked! Mar 25 03:21:25 openmoko: 03rwhitby * r58 10mokomakefile/trunk/patches/openmoko-HEAD/ (gtk-provider.patch series): Added patch for PREFERRED_PROVIDER_gdk-* problem Mar 25 03:21:34 the other build? Mar 25 03:21:36 dotmef: try that for the gtk+-directfb problem. Mar 25 03:22:12 Ocibu: never seen your problem before. Mar 25 03:22:24 Yup. I think something in my .bashrc has messed up my environment so that the bitbake dies Mar 25 03:22:47 rwhitby: I'm a bit of a n00b. can I get that through an OpenMokoMake update or will I have to manually patch? Mar 25 03:22:54 rwhitby: Thanks. Just knowing its not a common problem with an obvious answer helps Mar 25 03:23:01 dotmef: make update will get it for you. Mar 25 03:23:06 I'm feeling a little stupid right now Mar 25 03:23:36 rwhitby: Ta Mar 25 03:25:00 Ocibu: dude! you found something that someone else could run into Mar 25 03:25:15 and fixed it yourself Mar 25 03:25:23 nothing stupid about that Mar 25 03:25:26 As soon as I get it isolated I will certainly add it to the wiki Mar 25 03:27:14 Ocibu: as you may see there's also a workaround mentioned on the MokoMakefile discussion page. Mar 25 03:27:51 aevin: I hadn't see that, Im looking now Mar 25 03:28:10 Ocibu: (but not for your problem, i mean) Mar 25 03:28:23 i was thinking of it could be smart to collect them in one place. Mar 25 03:29:01 ah, I was wondering if I was missing something. Mar 25 03:29:31 yes, I'll stick it on the bottom of that page; when I know something more specific. Mar 25 03:29:34 i don't know what's better; either on discussion page (where people might not find it) or perhaps make a Troubleshooting or Workaround section below the Tips, or something... Mar 25 03:30:29 aevin: is there a pointer to the discussion page in the Tips? Mar 25 03:30:53 dotmef: no. but i think that would be a good idea. Mar 25 03:31:12 So where is the /discussion/ page, I have just been looking at the main page Mar 25 03:31:16 and then put all workarounds in the discussion, where they can be discussed... Mar 25 03:31:38 never mind. Mar 25 03:31:44 Ocibu: at the top of each wiki page there is a tab pane: article , discussion ... etc Mar 25 03:31:58 thanks, I see it now. Mar 25 03:32:16 I have been living in tWiki 90% of the time and didn't see thte tab Mar 25 03:32:35 gotta love all the different wikis... Mar 25 03:32:37 aevin: right - I guess the best or most frequent/persistent ones can be put into the main page for easy access Mar 25 03:33:00 Ocibu: yep - first time I've seen it too Mar 25 03:33:16 I guess I see those tabs as Wiki administration things Mar 25 03:35:17 I added a link in the main page to make that more obvious. Mar 25 03:35:57 rwhitby: Doh, I was just doing that. Thanks. I'll throw my chage away Mar 25 03:36:13 dotmef: my build just completed. Mar 25 03:37:40 Now maybe I can get on with getting the agps daemon going. Mar 25 03:37:46 Anyone have any tips there? Mar 25 03:38:06 Ocibu: you have P0 hardware? Mar 25 03:38:16 rwhitby: cool. Mar 25 03:38:57 Yup Mar 25 03:40:09 My focus is getting the proprietary gps stack opened up Mar 25 03:40:28 Are you aware of the current status? Mar 25 03:40:32 I'm a gps nerd in my day job. Mar 25 03:41:15 The status of the GlobalLocate daemon? Mar 25 03:41:16 No Mar 25 03:41:22 <[g2]> Obitus_ what qualifies as a gps nerd ? Mar 25 03:41:46 It being my professional focus for 12+ years? Mar 25 03:42:30 <[g2]> neat do you build stuff like this ? http://www.sparkfun.com/commerce/product_info.php?products_id=8238 Mar 25 03:44:10 No, my focus is more on the control segment support. That means dual frequency. Mar 25 03:44:48 But yes, that is a cool toy. I am working on a device similiar to that in concept except for wideband Mar 25 03:44:48 Ocibu: last I heard OM was still waiting for the daemon from the vendor. And last I saw here, people were trying to talk to the chip via serial but couldn't get a response from it. Mar 25 03:45:04 groan Mar 25 03:45:15 rwhibty: thanks for the update. Mar 25 03:45:45 perhaps you can be the first :-) Mar 25 03:45:46 it's cool to know we have gps pros on the project :) Mar 25 03:45:50 rwhitby: is there any way for me to get updates on the status of that part of the phone Mar 25 03:46:24 Ocibu: I'd saying talking to LaF0rge and mickey|working here is the best info you'll get publicly. Mar 25 03:47:27 There are several of us at work who are hoping to contribute to this effort. There are lots of neat things that can be done with GPS... Mar 25 03:47:32 speaking of the MokoMakefile wiki page, the "Bugs" sections at the top should be removed now, right? Mar 25 03:47:46 aevin: it should be gone Mar 25 03:48:10 i can do it. Mar 25 03:48:19 ah Mar 25 03:48:21 distributed DGPS could be very interesting. Mar 25 03:48:29 i had a cached site. Mar 25 03:49:02 hi SpeeDevil ! Mar 25 03:49:21 Upload from charging static neos their relative delays to satelites. Compute from that to an ionosphere/satelltite clock mode. Mar 25 03:49:22 l Mar 25 03:49:25 :) Mar 25 03:50:02 Yea, but the IGS has servers that are of much better quality and the data is available... Mar 25 03:50:14 Isn't it US only? Mar 25 03:50:14 how accurate is the time resolution resolved with GPS? Mar 25 03:50:23 It depends what you mean. Mar 25 03:50:23 IGS, no, it is worldwide Mar 25 03:50:47 Ocibu: do you know what that service costs? Mar 25 03:50:49 It measures the time to within hundreds of picoseconds skew between satellites. Mar 25 03:51:02 Most of the data is free. Mar 25 03:51:18 okay. Mar 25 03:51:27 The products are often what the groups want to charge for Mar 25 03:51:51 products being the things that are computed from they raw gps observations. Mar 25 03:53:04 How variable is the ionosphere - IOW, how fresh do you need a correction to be to be of optimal use, and when does it become useless. Mar 25 03:53:06 SpeedEvil: Heh, someone who knows just how difficult a question as to time accuracy is Mar 25 03:53:12 Ocibu: is gpstk something which we should get into OpenEmbedded so it can be used on the Neo? Mar 25 03:53:19 Not yet. Mar 25 03:53:41 Much of that is for precise positioning. Not simple nav solutions. Mar 25 03:53:57 Well... Mar 25 03:54:06 The stock GPS gets 2m. Mar 25 03:54:39 I would want to hand pick the parts that would be usefull and make a separate package. Mar 25 03:54:39 It would be rather nice on occasion to be able to click a 'get precise fix' button, and it gets the current state of the ionosphere/... and figures it out to within 10-20cm Mar 25 03:54:53 Ocibu: not even for the library functions, like time conversions? Mar 25 03:55:22 Ocibu: ah, cool. Are you familiar with OpenEmbedded - there are a number of packages which get split up into useful chunks ... Mar 25 03:55:32 10-20 cm with a single frequency device is not really achieavable. Mar 25 03:55:37 (script) wiki RecentChanges: [[MokoMakefile]] Mar 25 03:55:44 Well - I assume better than 2m is. Mar 25 03:55:50 Even with differential Mar 25 03:56:43 To get a reliable 2m you still need dual frequency or some form of DGPS to get better iono/orbit corrections Mar 25 03:57:03 yeah - I was talking about the spec sheet - which says 2m. Mar 25 03:57:07 I have doubts. Mar 25 03:57:38 How fresh does iono correction need to be? Mar 25 03:57:42 Heh, the spec that the air force drives the satellites to is only 15 m dual frequency Mar 25 03:57:56 But they regularly do in the 2 m range Mar 25 03:58:47 There are some really, really smart people spending their lifes squeezing every last bit of accuracy out of the GPS system Mar 25 03:59:30 Its one of the few really nice things that the US military has done for the world Mar 25 03:59:36 I wasn't meaning 10-20cm as a specific number - just better than what it'll get for the rest. Mar 25 04:00:20 I haven't been able to find what a good GLONASS reciever can do. Now that the more high resolution version has been opened and more birds launched. Mar 25 04:02:22 If you want to spend $10-20k (US) you can get a survey grade box that will do GLONASS+GPS and give you a static position that is well under a meter. Mar 25 04:02:50 You can get it in the 10-20 cm in post processing with a precise orbit Mar 25 04:02:51 Is that using the declassified code though. Mar 25 04:02:54 Yes Mar 25 04:03:31 The US turned off all the selective availability back in 2001 Mar 25 04:03:36 Err - no. Mar 25 04:03:53 They still encrypt the P code, however Mar 25 04:04:05 That is called AS, not SA, however Mar 25 04:04:08 GLONASS made available for civilian use the more precise code that I forget the name of in January Mar 25 04:04:38 Or at least they announced they had. Mar 25 04:04:56 gpstk passed configure fine in OE, let's see if it cross-compiles ... Mar 25 04:05:15 The survey grade receivers would even do this with the classified codes. They use a tracking techniques often referred to as 'codeless' Mar 25 04:05:35 It allows them to track a PRN bitstream without knowing the actuall PRN Mar 25 04:05:53 Wacky. Mar 25 04:06:02 Damn smart. Mar 25 04:06:09 I came up with a silly method for using P code that way. Mar 25 04:06:18 It wasn't very portable though. Mar 25 04:06:29 As it needed a 35m dish or so. Mar 25 04:06:32 Just shows how hard it is to really protect information that needs to be widely disseminated Mar 25 04:06:36 heh Mar 25 04:07:05 SpeedEvil: You an RF or comm type? Mar 25 04:07:12 Umm... Mar 25 04:07:17 Ocibu: what would be a one-line description of gpstk? Mar 25 04:07:23 I'm a 'play with everything' type. Mar 25 04:08:31 Ocibu: would you know anyone able to comment on opensource GPS legalities? http://wiki.openmoko.org/wiki/Hardware:AGPS#Q:_Is_an_open-source_GPS_daemon_able_to_be_distributed_by_FIC.3F Mar 25 04:08:42 rwhitby: Open source libraries and applications to facilitate working with GPS data for research and high accuracy uses. Mar 25 04:08:44 Which isn't yet relevant. Mar 25 04:09:00 But could be. Mar 25 04:09:42 I have disccussed this with several people at work who have to deal with the ITAR; the US laws that cause the problems Mar 25 04:10:20 ITAR is such a bundle of fun... Mar 25 04:10:32 Fortunately I haven't had to touch it. Mar 25 04:10:37 We believe that we can write code based upon publicly available data, make the source available and meet the regulations Mar 25 04:10:54 As I read it, source is one thing. Mar 25 04:11:05 A complete GPS reciever with that source may be something quite different. Mar 25 04:11:20 Ocibu: LGPL - nice :-) Mar 25 04:11:35 This is important to me because I would rather not loose my job over this Mar 25 04:11:55 rwhitby: we like to play nice Mar 25 04:12:14 Ocibu: I expect that's quite important for commercial collaboration and such. Mar 25 04:13:31 Ocibu: gpstk seems to be compiling fine under OpenEmbedded. Is there a categorisation of it into smaller packages which would be useful to you? Mar 25 04:13:32 rwhitby: Actually, my lab is not allowed to collaborate with commercial ventures. Mar 25 04:14:25 Ocibu: why is that (if you don't mind me asking publicly) ? Mar 25 04:14:30 If the src dir compiles, that would be a good starting point. Mar 25 04:15:01 Ocibu: it's up to WxObsMap Mar 25 04:15:19 rwhitby: We fall under a piece of legislation that requires us to do this so we can be a 'trusted advisor' to the US Government. Mar 25 04:15:51 Yea! Mar 25 04:15:51 that seems to make good sense. Mar 25 04:16:31 Ocibu: it should be built well before your MokoMakefile build finishes :-) Mar 25 04:16:40 * SpeedEvil is planning on using his neo to play with openstreetmap further. Mar 25 04:16:41 groan.... Mar 25 04:17:07 I want it so I can look up maps when I am on long bike rides. Mar 25 04:17:16 I am a roadie and like to explore Mar 25 04:17:32 There is unfortunately no nice map solution. Mar 25 04:17:41 google-maps may work if you have flat-rate GPRS Mar 25 04:17:53 or openstreetmap if you've mapped the area :) Mar 25 04:17:58 Ocibu: src finished, onto apps now Mar 25 04:17:58 That is what I figured would be a good start Mar 25 04:18:29 Ocibu: will this lib give us an easy way to get precise time from the GPS stream? Mar 25 04:18:39 rwhitby: excelent. if you see any build errors, pop me an email Mar 25 04:18:41 openmoko-libs-0.0.1+svnnow openmoko-libs-0.0.1+svnnow task do_configure: failed Mar 25 04:18:50 rwhitby: yes, it will Mar 25 04:19:12 We should be able to get that to better than 10-3 away from UTC Mar 25 04:19:16 LWT001: full pastebin of the output of following http://wiki.openmoko.org/wiki/MokoMakefile#Reporting_Problems please. Mar 25 04:19:58 My email is on the openmoko wiki for ocibu Mar 25 04:20:28 Umm - precise time depends on the gpsd plugin. Mar 25 04:20:35 if it screws that up, ... Mar 25 04:21:29 Ocibu: would you like me to check this gpstk recipe into OpenEmbedded? then when your MokoMakefile build is finished you will be able to "bitbake gpstk" and ipkg install it. Mar 25 04:22:05 I would expect we will just send NMEA strings to the gpsd for a starting point. A custom daemon would be better, however Mar 25 04:22:23 there are no NMEA strings Mar 25 04:22:28 rwhitby: Yes, that would be wonderful Mar 25 04:22:44 the hardware is a dumb correlator that you talk a custom unreleased serial protocol to. Mar 25 04:22:50 dumbish. Mar 25 04:23:06 the gpsd plugin then cooks this into a format that gpsd understands. Mar 25 04:23:19 That is the point of what I want to do, get the position from the raw correlation delays Mar 25 04:23:47 First - reverse engineer the plugin :( Mar 25 04:24:08 Unless it releases the correlation delays. Mar 25 04:24:57 Exactly what I want to do. Mar 25 04:25:27 that's why *I'm* making sure Ocibu has everything he needs to get started :-) Mar 25 04:25:28 Start with figuring out what the Hammerhead needs, get it tracking, and figurre out what it outputs Mar 25 04:25:44 Ah - I was going to have a stab at that. Mar 25 04:25:49 be nice to have someone with more clue. Mar 25 04:26:23 It'd be lovely if the hardware could be kicked with an initial doppler, and get an even faster than spec lock. (it specs a 5m position after 1s). Mar 25 04:26:32 (for power saving) Mar 25 04:27:10 If you do, let me know what you find. I will likely start a wiki page when get the agpsd going. Mar 25 04:27:34 Ocibu: I've just done the stable release. Is that fine for initial work, or will you need to track SVN? Mar 25 04:27:34 Well, it's all kind of contingent on hardware :) Mar 25 04:27:36 Actually, time is more important than doppler. Mar 25 04:28:00 Stable release will be good for now. Mar 25 04:28:35 You mean if you know which 1/50th of a second you are in it helps you determine where in the bitstream you are faster? Mar 25 04:28:40 The routines we need for this are quite mature; I hope they won't be chaning a lot Mar 25 04:29:20 Actually, getting sub ms into the device is when you can really crank Mar 25 04:29:37 The 1/50 doesn't help that much for a C/A acquisition Mar 25 04:30:03 I am assuming that you have a current ephemeris, however Mar 25 04:30:08 Yeah. Mar 25 04:30:14 --not just the almanac Mar 25 04:30:30 Not usually much of a hastle - especially as you may be able to get a good signal while charging. Mar 25 04:30:40 Hopefully Mar 25 04:30:43 As the GPS is allegedly quite sensitive. Mar 25 04:31:00 [http://gpstk.sourceforge.net/papers/ion-gnss-2004/ION-GNSS-2004-paper.pdf <- Introduction to gpstk, Ocibu is one of the authors] Mar 25 04:31:06 The gpstk has all the routines for parsing the ephemeris data as it is stored in the IGS archives Mar 25 04:31:16 don't remember the number - it's 10db more than my GPS-12 though. Mar 25 04:31:39 That is needed for the E911 to work indoors Mar 25 04:32:14 We have a newer paper that also talks about some of the apps that we presented at the ION last fall Mar 25 04:32:36 AIUI there is no GSM standard for E911 though. Mar 25 04:32:59 http://www.gpstk.org/pub/Documentation/GPSTkPublications/gpstk-ion-gnss-2005.pdf ? Mar 25 04:34:12 I hope that the GSM standards get something there. I would hate to have this phone not support E911 just because the carriers all used a private protocol Mar 25 04:34:39 I believe that the carriers all just use triangulation. Mar 25 04:34:45 which is generally pretty accurate. Mar 25 04:34:53 I may be wrong. Mar 25 04:35:14 Some of the newer handsets do use GPS. I just got a new LG that does Mar 25 04:35:52 The byzantine world of the networks is wierd though. Who knows if it's a generic protocol or a custom-rolled one. Mar 25 04:35:56 They make you pay $20 / month just to use the gps on your phone for anything else Mar 25 04:36:22 that is another reason for this effort Mar 25 04:36:35 Indeed - locked down hardware is insane. Mar 25 04:39:17 Ocibu: apps/multipath/Makefile didn't get created by configure ... Mar 25 04:40:44 rwhitby: We use Jam for the standard build tool. There is a guy in spain that contributes the makefiles so I don't stay up-to-date on the status of that. Mar 25 04:41:04 ok, Jam isn't packaged for OE, so I'll just fix the configure.ac script. Mar 25 04:41:18 Tell it not to build that? Mar 25 04:41:37 it's a one line fix in top-level configure.ac so far ... Mar 25 04:41:41 I will still try to get it fixed in subversion Mar 25 04:43:01 Many of the apps in there are for scientific analysis of gps data and won't be of interest to 99% of even the openmoko developers Mar 25 04:44:33 * Ocibu happy! Mar 25 04:44:34 ok, build continues ... Mar 25 04:44:43 My build is up to the kernel!!! Mar 25 04:45:41 Ocibu: here's the patch (against the 1.2 stable release): http://pastebin.ca/408858 Mar 25 04:46:18 (the patch will be in OE, and will be automatically applied to the stable release before building) Mar 25 04:47:32 rwhitby: You are fast. Thanks 1e6 Mar 25 04:48:51 Ocibu: if there's one thing I've learnt in my 40 years on this planet, it's to make life easy for the people who know magnitudes more than you in subject areas in which you're interested :-) Mar 25 04:48:59 What would be a good way to contact LaF0rge and/or mickey Mar 25 04:49:22 <[g2]> 40 years -- You geezer ! Mar 25 04:49:25 mickey is usually here in EU hours. Mar 25 04:49:32 [g2]: heh - you can't talk! Mar 25 04:49:45 <[g2]> what do you mean old man Mar 25 04:49:54 I just hope I can do something good with this. Mar 25 04:50:11 [g2]: I thought you're kids were almost as old as me ... Mar 25 04:50:20 s/you're/your/ Mar 25 04:50:21 rwhitby meant: [g2]: I thought your kids were almost as old as me ... Mar 25 04:50:27 <[g2]> that's grand kids :) Mar 25 04:51:03 <[g2]> it's ashame the Fortran guy died Mar 25 04:51:18 <[g2]> this year is the 50th anniverary Mar 25 04:51:26 Ocibu: I'd suggest you announce your presence on the oe-devel mailing list - Harald and Mickey definitely read that. Mar 25 04:51:39 <[g2]> well night all Mar 25 04:51:56 (I'm sure they would be very keen to talk with you) Mar 25 04:52:01 night [g2] Mar 25 04:52:16 NOTE: package gpstk-1.2-r0: task do_compile: completed Mar 25 04:53:06 rwhitby: Once again, thanks Mar 25 04:55:37 (script) wiki RecentChanges: [[Talk:MokoMakefile]] [[Hardware:AGPS]] Mar 25 04:57:53 Hmm - no "install" target ... Mar 25 05:00:29 Ocibu: it compiles - I Mar 25 05:00:53 will add packaging later. I'll check in what's there so you can at least compile it with OE. Mar 25 05:02:16 rwhitby org.oe.dev * r921c450e... / (5 files in 4 dirs): gpstk: Open source libraries and applications to facilitate working with GPS data for research and high accuracy uses. Mar 25 05:02:23 rwhitby: I just sent out a quick query on the devel list. Maybe they can get me a pre-release (but I won't hole my breath) Mar 25 05:03:55 Ocibu: heh - you're too modest - you should have mentioned your backround :-) Mar 25 05:04:09 I'm sure mickey|working will read it here anyway. Mar 25 05:06:25 Well, I gotta get some sleep. Thanks for the help/support/chatting. I'll try to stay on here while I am working this. Mar 25 05:06:36 Goodnight! Mar 25 05:06:43 Ocibu: welcome aboard! Great to have your experience here. Mar 25 05:07:20 (I'll look at the Makefile.am files to add install targets so gpstk creates packages for installation) Mar 25 05:23:21 Ocibu: should any of the libraries from apps be installed? Mar 25 06:10:38 rwhitby: build completed - thanks again. Mar 25 06:12:16 rwhitby: btw - I see an error with the build trying to access files in oa/packages. is this normal? Mar 25 06:26:32 Hi, I am writing up an application for openMoko SOC. Mar 25 06:27:30 Can anyone tell me if there are any specifics I must address. Mar 25 06:27:45 I am making an app for Ad-hoc bluetooth communication Mar 25 06:30:09 hello... Mar 25 06:34:38 what is the status of development on bluetooth in openMoko til now Mar 25 06:39:04 dotmef: oa/packages? Mar 25 06:40:10 can nebody hear me ? Mar 25 06:40:10 technomaniac: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 25 06:43:34 Has Bluetooth-alsa been ported to openMoko Mar 25 06:51:26 rwhitby: yep - here's the output in question http://pastebin.ca/408912 Mar 25 07:00:10 What is the status of Bluetooth ALSA ... on openMoko Mar 25 07:35:21 dotmef: you don't have a typo in one of your config files do you? (local.conf, setup-env)? Mar 25 07:37:45 technomaniac: http://buildhost.openmoko.org/tmp/deploy/ipk/armv4t/bluez-utils_3.9-r2_armv4t.ipk and http://buildhost.openmoko.org/tmp/deploy/ipk/armv4t/alsa-*.ipk are the current state. Dunno what functionality that gives. Mar 25 07:40:56 rwhitby: thanks for adding gpstk, I had some problem with 'make install' not working Mar 25 07:45:27 koen: yeah, the base package does not have install targets in the Makefile.am's. Mar 25 07:46:16 and when I added some, it no longer compiled for some reason. Need to go back and look at it. Compilation is enough for the Ocibu who wants to use it (he's part of the team that wrote it) Mar 25 07:49:04 Ocibu: the -Qy problem is known, but we haven't been able to figure out what triggers it (i.e. unable to reproduce locally) Mar 25 07:52:44 * rwhitby changes just noinst_LTLIBRARIES to lib_LTLIBRARIES in gpstk/src/Makefile.am and sees if compilation still works. Mar 25 07:55:38 (script) wiki RecentChanges: [[User:Openmarko]] Mar 25 07:57:27 yep, compilation fails from that single change in the Makefile.am Mar 25 07:57:30 rwhitby: looks like a typo in openmoko/trunk/oe/conf/site.conf Mar 25 07:57:50 :q Mar 25 07:58:42 dotmef: interesting - looks like that's intentional - OM core devs must have an 'oa' symlink linking to something else that we haven't seen yet ... Mar 25 07:58:52 BBFILES := "${OMDIR}/openembedded/packages/*/*.bb ${OMDIR}/oa/packages/*/*.bb ${OMDIR}/oe/package\ Mar 25 07:58:53 s/*/*.bb" Mar 25 07:59:12 * rwhitby wonders what 'oa' stands for ... Mar 25 07:59:30 OpenAppliance? Mar 25 07:59:44 openalchemy? Mar 25 07:59:49 OpenAccess (old PalmOS?) OpenAngstrom? Mar 25 08:00:08 OpenAllHours? Mar 25 08:00:11 koen: know of any other package repos forked from OpenEmbedded? Mar 25 08:00:37 rwhitby: harald is involved with http://openalchemy.org/index.php/Main_Page Mar 25 08:00:39 rwhitby: I know to ignore it now - ta Mar 25 08:01:05 ahh. that makes sense then Mar 25 08:01:06 rwhitby: http://openalchemy.org/pipermail/openalchemy-general/2007-March/000009.html Mar 25 08:01:32 that explains it then. Mar 25 08:01:43 one to add to the Wiki? Mar 25 08:05:31 Hi everyone :) Mar 25 08:05:32 counter Mar 25 08:05:33 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 5 days 03:54:27 (12.163 +-7.0 days) (621) Mar 25 08:07:21 someone with better autofoo than I will need to look at this gpstk problem. Mar 25 08:08:00 rwhitby: iirc someone promised to send a request to the gpstk list to fix the install targets Mar 25 08:08:57 rwhitby: http://purl.rikers.org/%23openmoko/20070110.html.gz Mar 25 08:09:14 ~seen pben Mar 25 08:09:51 pben was last seen on IRC in channel #openmoko, 73d 11h 50m 37s ago, saying: 'koen: I can't reproduce your problem with gpstk's make system. problem with my system. i'll post a question to the developer list to see if anyone has actually installed using make and not with jam'. Mar 25 08:10:53 koen: ah, you've been through all this before I see. Mar 25 08:11:18 yes Mar 25 08:11:38 I didn't check in gpstk since make install didn't work Mar 25 08:12:12 should I remove it? Mar 25 08:12:23 nah Mar 25 08:12:38 let's see if we can fix is with some cp, grep and awk :) Mar 25 08:12:46 heh Mar 25 08:12:52 back later ... Mar 25 09:21:42 rwhitby: do_install is a bit better now Mar 25 10:23:34 good morning Mar 25 10:25:48 hi Mar 25 10:29:04 mornin' Mar 25 10:47:32 counter Mar 25 10:47:32 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 5 days 01:12:27 (12.050 +-7.0 days) (622) Mar 25 10:54:58 evanm89mob: in 5 days counter will start to reduce uncerainty (and counter will also move in half speed) Mar 25 11:46:41 hi Mar 25 11:59:47 will the mass market price of the Neo be lower than the developer price? Mar 25 11:59:47 amayera: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 25 12:08:23 counter Mar 25 12:08:23 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 4 days 23:51:36 (11.994 +-7.0 days) (623) Mar 25 12:26:27 counter Mar 25 12:26:27 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 4 days 23:33:33 (11.982 +-7.0 days) (624) Mar 25 12:29:20 while(1) counter; Mar 25 12:33:49 Hopscotch - while(1) sleep(10), counter; Mar 25 12:35:37 ah, with power-management now? Mar 25 12:36:53 Yep. ;) Mar 25 12:37:21 10 seconds should be enough to recharge ;) Mar 25 12:54:35 * root slaps RedDog around with a small 50lb Unix Manual Mar 25 12:58:01 Has there been any interest expressed in porting other OSes to the Neo1973? Mar 25 12:58:01 dho: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 25 12:59:01 dho: what OSes? Mar 25 12:59:18 knowing NetBSD folks they are probably working on it Mar 25 12:59:19 netbsd probably could be doable Mar 25 12:59:20 balrog-kun: WinCE ;) Mar 25 12:59:29 considering that Neo1972 is not a toaster Mar 25 12:59:36 *1973 Mar 25 12:59:37 I was personally thinking Plan 9/Inferno Mar 25 13:00:09 Just as long as its binary compatible with the GSM modules ;) Mar 25 13:00:09 and from NetBSD support immediately gets ported to OpenBSD Mar 25 13:00:39 Oh, ok. I guess not then. I was under the impression that was done via serial only. Mar 25 13:01:23 It is. Mar 25 13:01:26 GPS is binary Mar 25 13:01:37 Ah Mar 25 13:02:00 Are there details of the ABI or just ``Linux glibc xyz''\ Mar 25 13:02:58 hmm, did it switch yet to EABI? Mar 25 13:03:16 It's not been stated. Mar 25 13:03:27 (Not that Plan 9 would be compatible with it anyway) Mar 25 13:03:42 actually since gps is userspace and the BSDs have linux userspace emulation gps might work, too Mar 25 13:03:44 It is just 'the GPS will be enabled by a binary plugin to GPSd" with no more details. Mar 25 13:04:05 heck i have run binary java on FreeBSD Mar 25 13:04:27 if that works, a small GPSd probably does so, too Mar 25 13:05:10 prolly. Mar 25 13:05:26 Depends on the syscalls it uses. Mar 25 13:05:32 gpsd should be rather portable, Mar 25 13:05:36 IIRC it just talks over a serial port. Mar 25 13:05:37 it's a userspace code Mar 25 13:06:00 plugin to gpsd - not gpsd... Mar 25 13:06:03 gpsd is portable Mar 25 13:06:11 i think someone even ran oracle linux on freebsd Mar 25 13:06:19 just to prove it's possibl Mar 25 13:06:19 e Mar 25 13:06:22 the plugin to gpsd works in userspace as well Mar 25 13:06:41 make world on neo, muahahaha Mar 25 13:06:47 the OS should make no difference Mar 25 13:06:53 Well, it does :) Mar 25 13:07:12 Plan 9 will never be binary compatible. Its stack doesn't even go in the same direction ;) Mar 25 13:07:21 i say gpsd would run on the bsds Mar 25 13:07:28 I'm sure it would Mar 25 13:07:31 there's no need for any binary compatibility Mar 25 13:07:44 probably not Mar 25 13:07:50 tho it depends just how the plugin works Mar 25 13:07:54 gpsd is supposed to manage the loading of the binary plugin and calling it Mar 25 13:08:12 (Plan 9 doesn't have an rtld either) Mar 25 13:08:15 :) Mar 25 13:08:46 the same plugin is used on all system that use this gpd chip Mar 25 13:08:49 Inferno does though. Anyway, the GPS is a small portion of functionality of the phone. Mar 25 13:08:51 *systems Mar 25 13:09:21 *gps Mar 25 13:09:48 But the general concensus is that based on specifications and what's currently running, it should be possible? Mar 25 13:10:01 Sure. Mar 25 13:10:07 Sounds fun. Mar 25 13:10:22 There are completely open hardware specs - well - documented in the code. Mar 25 13:10:33 so it's simply a matter of porting. Mar 25 13:11:19 the bootloader also does a significant part of hardware setup, so you get this free Mar 25 13:11:24 Right Mar 25 13:12:53 Well. My 8125 broke so I may as well get one of these and play around with it. Mar 25 13:13:57 you should een be able to boot from an sd card Mar 25 13:14:08 so you can actually keep the installed stack for general use Mar 25 13:14:25 Hm, does it virtualize that at all? I'd probably have to prot the SD stack too. Mar 25 13:14:25 ~/een/even/ Mar 25 13:14:41 s,prot,port, Mar 25 13:14:43 dho: the bootloader itself treats it as mmc card i think Mar 25 13:15:24 as for the os itself, you will have to port an SD stack i guess Mar 25 13:15:51 openbsd understands SD Mar 25 13:15:54 Yeah, I can figure out details of what needs what later, I suppose. Mar 25 13:16:02 FreeBSD too, so it shouldn't be too big a problem Mar 25 13:16:03 NetBSD does too Mar 25 13:16:09 moko/openbsd Mar 25 13:16:12 for the truly paranoid Mar 25 13:16:29 I was mostly just wondering if there had been explicit express interest Mar 25 13:16:40 i think youre the first to ask Mar 25 13:16:41 IIRC you're the first. Mar 25 13:16:52 fingerprint, retina scanning and a password to unlock, and the flash is off course encrypted Mar 25 13:16:53 At least you seem saner than the iSCSI over GPRS guy. Mar 25 13:16:56 but i had wondered when people would start with that ;-) Mar 25 13:17:09 * buz still wonders if that one is trolling or serious Mar 25 13:17:17 iscsi over gprs :D Mar 25 13:17:32 * balrog-kun 's guess would be that NetBSD people have probably already started Mar 25 13:17:51 well they dont have hardware Mar 25 13:18:00 but yeah they might have ;) Mar 25 13:18:04 true.. Mar 25 13:18:16 wouldnt put it beyond them to release a distro before the hw even ships Mar 25 13:18:31 I would Mar 25 13:18:37 They're largely busy with other stuff right now Mar 25 13:18:44 Especially since 4.x is late Mar 25 13:18:52 personally i want to see kde ;) Mar 25 13:19:49 is there kde in openembedded? Mar 25 13:19:57 there's OPIE Mar 25 13:20:11 thats like Qt 2 based right? Mar 25 13:20:29 i'm not sure Mar 25 13:20:37 Oh yeah Mar 25 13:20:40 then again, to kde neo should mostly be a somewhat limited pc Mar 25 13:20:42 does the S3C2410 have an MMU? Mar 25 13:20:48 dho: yes Mar 25 13:20:50 ok Mar 25 13:20:56 otherwise they wouldn't use Linux on it :p Mar 25 13:21:01 ucLinux Mar 25 13:21:23 can run MMU-less Mar 25 13:21:26 i doubt they would go with uclinux then Mar 25 13:21:34 Yeah, was just making sure. Mar 25 13:21:49 Though I guess Inferno can run MMUless Mar 25 13:24:59 kde on the neo wouldn't be fun. Mar 25 13:25:13 opie doesn't look good at all Mar 25 13:26:29 Elrond: about the same as gnome ;) Mar 25 13:27:00 which is to say largely not very useful without a keyboard Mar 25 13:27:08 then again, amarok... Mar 25 13:27:18 Anyway, it might be a fun SoC project for someone too, and I'd be glad to mentor that over in that camp. Mar 25 13:28:10 i've seen kde run (maybe walk is better) on p-100 Mar 25 13:28:14 so its not that out of touch Mar 25 13:28:41 the neo packs about as much punch as the pcs we had 10 years ago Mar 25 13:28:46 and double the ram Mar 25 13:28:46 Is inferno used anywhere notable? Mar 25 13:29:04 buz - Yeah. :) Mar 25 13:29:07 Elrond: various government and academia projects, I don't know details. Mar 25 13:29:28 kde2 should be almost good on p-100 with reasonable amount of ram Mar 25 13:29:29 both at this point are really for research for general purposes. Mar 25 13:30:00 dho - So it's more a research os then a real useful thing? Mar 25 13:30:09 * thresh had p-120 with 96 ram and it was enough Mar 25 13:30:17 Depends on what you're defining as useful Mar 25 13:30:39 For most people, probably not. Mar 25 13:31:08 (Of course, most of the Plan 9 / Inferno nuts would probably go berzerk trying to tell you how much better it is than whatever you're using) Mar 25 13:31:19 i keep thinking that my pentium m 1.6ghz kde laptop is slow ;) Mar 25 13:31:21 the believe that older software is more suitable for older hardware is often wrong Mar 25 13:31:39 i'm pretty sure current kde would run better than kde2 Mar 25 13:31:39 balrog-kun: true enough, kde got faster with every release in 3.X Mar 25 13:31:56 It has a lot of good ideas that could make it better for practical purposes, but right now, it's not really useful for most people, and the learning curve is probably steeper than learning any give UNIX or UNIX-like OS Mar 25 13:32:00 and kde4 will make another leap Mar 25 13:32:29 some benchmarks i've seen put qt4 apps at about 20% faster while using 20% less ram Mar 25 13:32:35 buz - But kde tends to get larger, no? Mar 25 13:32:49 They at least try to compartmentalize it Mar 25 13:32:51 Ohh, less ram. I'm surprised! Mar 25 13:32:56 Not as much as the gnome people, I guess. Mar 25 13:32:58 (same goes for osx, btw, it's just MS land where each subsequent release needs more of everything) Mar 25 13:33:23 same goes with MS sometimes, too, take the windows 2003 Mar 25 13:33:34 never used that Mar 25 13:33:46 but it might be the one exception in the whole family Mar 25 13:33:58 balrog-kun - Is 2003 better then 2000? Really? It just feels better thought out, but really better to resources?! Mar 25 13:34:05 it was extremely stable and a lot faster than windows xp, also required less ram Mar 25 13:34:28 Well, xp is "for the dumb users at home/desktop" ;) Mar 25 13:34:41 and it supported both windows 2000 and windows xp drivers, themes etc Mar 25 13:34:42 ... those want to buy new hardware every half year. ;o) Mar 25 13:35:06 Er, didn't mean to start a UI/OS flame war in case it goes in that direction, heh Mar 25 13:35:18 seems quite civilized so far Mar 25 13:35:22 dho - Don't yet see the flamewar :) Mar 25 13:35:42 its rare but you actually CAN discuss os objectively ;) Mar 25 13:35:44 I'm pre-empting it ;) Mar 25 13:36:03 dho - Okay. ;) Mar 25 13:36:18 So what's about hurd? ;o)) Mar 25 13:36:22 in that case you cant discuss Win9X objectively Mar 25 13:36:22 * dho runs Mar 25 13:41:13 hurd has no arm support whatsoever at the moment :p Mar 25 13:41:40 Shouldn't be too hard? Mar 25 13:41:57 i don't know, but that's more work than for linux or BSD Mar 25 13:42:23 ARM's not that difficult at all Mar 25 13:42:24 but Mar 25 13:42:31 It's always really application specific Mar 25 13:43:37 porting an os to a whole new architecture is hard i think, even if it's a relatively simple arch Mar 25 13:43:43 all the caches stuff Mar 25 13:44:12 Hmm, yeah. Mar 25 13:44:21 you see even the ARM Inc. people on linux-arm-kernel mailing list sometimes get lost in this Mar 25 13:44:44 (there are 5,000,000,000,000,000 different ARM processors in 50 times as many applications) Mar 25 14:26:37 hi people Mar 25 14:27:29 any GoSoC mentor here :) Mar 25 14:30:31 Is there anyway for a non-programmer to help in the OpenMoko Project? Mar 25 14:30:32 Kablamo: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 25 14:51:04 SpeedEvil: I think mickey mentionet that they hope to move to EABI between P1 and P2 Mar 25 14:55:38 (script) wiki RecentChanges: [[Wishlist:Bootable_USB_device_emulation]] [[Summer_of_code]] Mar 25 15:02:29 If it was a bootable device it would be handy. You could implement a good few features that way without having to worry about cross-OS support. Mar 25 15:05:24 Bootable like when booting you're pc from a usb pendrive? Mar 25 15:05:40 that shouldn't be exactly difficult assuming usb block device support Mar 25 15:13:34 MortenLysgaard: Yup Mar 25 15:23:23 Is it possible to develop openmoko applications with an emulator? Mar 25 15:23:48 Yeah, I think so, 2 sec. Mar 25 15:25:31 Think this says it: http://wiki.openmoko.org/wiki/FAQ#Q:_Is_there_an_emulator_available_for_OpenMoko.3F gotta love wiki =) Mar 25 15:26:37 MortenLysgaard: I just read it, thanks. So I would be able to see the openmoko ui, too? Mar 25 15:27:24 I don't think i'm the right person to ask, sorry, I'm just a student who got to hear of this project Mar 25 15:28:17 * dho sighs. Mar 25 15:28:36 But, of general knowledge i can say this. I's wery hard to emulate the neo on a pc. Because it has so many functions, your pc doesen't. Eg. touch screen. Mar 25 15:28:46 Not really. Mar 25 15:28:53 the touchscreen is about it. Mar 25 15:28:57 you can fake the rest. Mar 25 15:29:06 gps and gsm? Mar 25 15:30:16 plain text files Mar 25 15:30:20 300DPI? Mar 25 15:30:23 calling really is another thing Mar 25 15:30:35 285DPI and the touchscreen is the biggy. Mar 25 15:30:47 gpsd already comes with a 'fake' module Mar 25 15:30:52 or you could just plug in a GPS. Mar 25 15:30:57 what's the problem with touchscreen? Mar 25 15:31:01 True Mar 25 15:31:06 most PCs don't have them. Mar 25 15:31:12 :P Mar 25 15:31:16 which makes developing a little harder. Mar 25 15:31:37 i think it doesn't matter for the emulator whether there's a touchscreen or a mouse or trackabll Mar 25 15:31:47 Strictly, no. Mar 25 15:31:56 if you don't care about touchscreen and dpi, why do you need an emulator? Mar 25 15:31:57 For usability, it may help a fair bit. Mar 25 15:32:01 just run your app under X Mar 25 15:32:11 qemu already emulates a couple of touchscreens, including the neo one (this is not public code though, but will be public in not too long) Mar 25 15:32:51 For example, you simply can't get an idea with a mouse about what gestures on the screen 'feel' like in real life. Mar 25 15:33:05 And how practical drawing glyphs on the screen is. Mar 25 15:33:38 you can do less things with a touchscreen than with a mouse, Mar 25 15:33:52 You can do different things with a touchscreen than a mouse. Mar 25 15:33:55 emulating the other way would be more problematic, but not mch more Mar 25 15:34:12 One could test is with those wacom tablets. Mar 25 15:34:17 'finger painting' with a touchscreen, and using a mouse are _very_ different actions. Mar 25 15:34:24 SpeedEvil: well, yes, but thinking in terms of the set of possible signals that X receives Mar 25 15:34:40 the touch event is equivalent with mouse left button press Mar 25 15:34:52 In terms of that, hundreds of buttons on the screen at once would make sense. Mar 25 15:35:16 I have wondered about a mouse wanderer. Mar 25 15:35:26 it makes mouse clicks displaced by a given radius. Mar 25 15:35:33 up to a given radius. Mar 25 15:35:35 no, X receives through evdev only three kinds of data: single x coordinate, y coordinate and touch (up/down, no pressure) Mar 25 15:35:39 and tracks a bit blobby. Mar 25 15:35:51 no - I mean to simulate finger inaccuracy. Mar 25 15:36:40 wind it up for finger, down for pointer. Mar 25 15:37:38 touchscreen doesn't know what is touching the screen Mar 25 15:37:46 No - for the emulator. Mar 25 15:37:59 to make the mouse as inaccurate as a finger. Mar 25 15:38:13 ah Mar 25 15:38:14 For practical purposes, how accurate is qemu as far as developing things for OpenMoko Mar 25 15:38:21 from an OS standpoint Mar 25 15:38:51 actually qemu is inserting slight noise in touchscreen measures that it reports, because tslib depends on this (which i think is a tslib bug) Mar 25 15:39:36 It seems like it emulates two different platforms, and it's not clear whether it serves as good enough of a hardware reference to implement support for the phone in another OS Mar 25 15:51:59 I cannot say about UI but afa Network connectivity and GPS is concerned, ther have been emulators for mobile devics which do a pretty decent job Mar 25 15:52:23 and me too think Mouse Pointer is nearly as good as styles Mar 25 15:53:56 i read in the FAQ, if your first question is about an emulator, chances are you would never develop. =P Mar 25 15:55:38 (script) wiki RecentChanges: [[Wishlist:Bootable_USB_device_emulation]] Mar 25 15:56:55 Prayag: excuse me? Mar 25 15:58:54 yeah, that jab is due to the fact that for most stuff that you'd usually use an emulator for in developing for a cell, there is no _need_ with openmoko Mar 25 15:59:30 mjr: Only if you're planning on using Linux on it. Mar 25 15:59:54 If you're planning on porting an OS to it, you need either the device or some sort of reference Mar 25 16:00:13 Openmoko is a 'distribution' for linux. Mar 25 16:00:16 I'm trying to figure out how easy it would be to entice an SoC student to do it, and whether I'm going to have to spend $550 on it. Mar 25 16:00:22 it's not inherently linux. Mar 25 16:00:24 "for _most_ stuff that you'd usually use an emulator for in developing for a cell" :] Mar 25 16:00:41 Pardon the terminology Mar 25 16:00:42 Is it me, or is the wiki slowish? Mar 25 16:00:55 * SpeedEvil thinks the wiki needs a graph. Mar 25 16:01:30 np: (at 11) The proclaimers - I would walk 500 miles. Mar 25 16:01:44 Elrond: the wiki Mar 25 16:01:53 it seems to be down Mar 25 16:02:03 Okay. Mar 25 16:02:12 loaded recentchanges in 8s or so for me. Mar 25 16:02:24 quickdev: working fine for me Mar 25 16:03:10 might be ;) Mar 25 16:04:36 hey whats the security on an OpenMoko device ? Mar 25 16:04:58 in what way? Mar 25 16:04:59 It has a poultergeist. Mar 25 16:05:04 it's a linux system. Mar 25 16:05:04 haha Mar 25 16:05:09 It throws things at potential thieves. Mar 25 16:05:13 it probably doesn't have many servers on by default Mar 25 16:05:22 I mean servers listening on network sockets Mar 25 16:05:23 More seriously - it's a linux box. Mar 25 16:05:29 with some prototype phone stuff. Mar 25 16:05:47 no i mean like other phones have a 4 digit "unlocking code" Mar 25 16:05:52 questions about features - rather than desires for features is mostly misplaced at the moment. Mar 25 16:05:56 not implemented. Mar 25 16:06:12 Prayag, that's actually the SIM that has the code Mar 25 16:06:34 YEs the SIM does have PIN code (Personal IDentification doe) Mar 25 16:06:49 but do does most of the phones.... Mar 25 16:07:45 *so do most of the phones Mar 25 16:08:05 so if you steal the phone, you need to know the code before taking out the battery and sim Mar 25 16:08:20 yeah well, that Mar 25 16:08:35 the battery power won't be the critical factor ;) Mar 25 16:08:38 there have been a lot of ideas for "stolen mode" etc. but probably not much implemented so far :] Mar 25 16:09:03 call 911 and play a wav that says i'm stolen Mar 25 16:09:11 rhrr :) Mar 25 16:09:21 with gps position reported ;) Mar 25 16:09:21 and giving the location with the gps Mar 25 16:09:25 hahahaha Mar 25 16:09:27 * dho ^5 quickdev Mar 25 16:09:51 PIN protection has a problem that because its a smartcard if u enter the code wrong three times.. its locked Mar 25 16:10:32 Who is comparing the code? the card internally? Mar 25 16:10:38 yes Mar 25 16:10:48 quickdev - for sim-pin: Yes, the card. Mar 25 16:10:50 not exactly... because the card is the client Mar 25 16:10:58 Who is it at the moment? Is no pin code required? Mar 25 16:11:02 the card has minimal processor and RAM on it. Mar 25 16:11:18 Prayag: doesn't have to be Mar 25 16:11:26 dho: asolutely, but it does have a PIN Mar 25 16:11:49 dho: din't get u :/ Mar 25 16:11:51 Prayag: you can choose to not boot without a certain SIM without a certain PIN Mar 25 16:12:22 djo: still din't get u Mar 25 16:12:56 IMHO, this security is not for the phone, its for the card Mar 25 16:13:10 ....Right, I'm not arguing that. Mar 25 16:13:24 to access certain files on the card, you need to have certain previledges... Mar 25 16:13:53 Than I don't understand the base of your question. Why ask about security if you already know about smart card security implementations? Mar 25 16:13:53 therecan be upto 16 key sets, as they are called Mar 25 16:14:10 there is inherent problems with SIM security Mar 25 16:14:22 For who? The card or the phone? Mar 25 16:14:32 for the user Mar 25 16:14:49 In what sense? Mar 25 16:15:12 1. The problem of locking as i already said Mar 25 16:15:15 dos? A malicious software could disable the sim. ;o) Mar 25 16:15:39 Elrond: require escalated privileges to write to that memory Mar 25 16:15:41 2.Since its a 4 digit code... brute force doesn't take much time Mar 25 16:15:41 If the user is too stupid to remember a 4 - 6 digit code, they _should_ be locked out. Mar 25 16:15:46 Yes it does Mar 25 16:15:48 :) Mar 25 16:15:59 It gets locked after 3 tries Mar 25 16:16:02 yes.. Mar 25 16:16:06 And then you need a much longer PUK code. Mar 25 16:16:12 and u need a PIN2 to unlock that Mar 25 16:16:22 10 tries to that Mar 25 16:16:31 and after that .. puff!! its gone Mar 25 16:16:38 unusable Mar 25 16:16:38 So Mar 25 16:16:53 Why is this not a problem for any other phone company? Mar 25 16:17:06 I don't see the issue here. Mar 25 16:17:24 :) its not an issue because they have their security mesures Mar 25 16:17:28 Brute forcing a 10000 digit space with 13 tries Mar 25 16:17:52 is 1/13th of 1 percent chance of unlocking the device Mar 25 16:18:10 Which is sufficiently small IMO Mar 25 16:18:24 I'd be more concerned about someone stealing the phone and using a different SIM card in it Mar 25 16:18:27 Prayag - I still don't get the issue. Mar 25 16:18:40 In that case, I would have it trash the SIM card Mar 25 16:18:46 You can buy a gsm module for about $50, which lets you do the same. Mar 25 16:18:55 The issue is phone security thats all Mar 25 16:19:07 u might have sesitive data on phone, how do u secure that Mar 25 16:19:15 Ahh. Mar 25 16:19:17 How I said. Mar 25 16:19:18 encryption. Mar 25 16:19:25 yes, encryption Mar 25 16:19:29 Require a valid SIM code to boot Mar 25 16:19:34 possibly encrypt data Mar 25 16:19:52 ``don't store something really private on your phone that you leave at the restaurant?'' Mar 25 16:19:53 and how do u authorize the user then ??? Mar 25 16:20:04 Prayag, several methods Mar 25 16:20:16 mjr:for example Mar 25 16:20:19 you could store the encryption key on the sim card, in which case it'd be behind the PIN Mar 25 16:20:21 passprhase to decrypt. Mar 25 16:20:27 let's go back 45 years when the idea of user authentication first came up Mar 25 16:20:28 or you could just have a separate passphrase, yes Mar 25 16:20:53 (or one of the keying methods customized for a touchscreen) Mar 25 16:21:11 draw a complex squiqqle ;) Mar 25 16:21:12 mjr: exactly, biometrics... Mar 25 16:21:48 most secure methods... keystrokes or voice recognition Mar 25 16:21:51 well, _you_ brought up biometrics Mar 25 16:21:52 the touchscreen is nowhere near sensitive enough to read biometric data Mar 25 16:22:07 dho - Okay, we're 45 years back in time. Now what? Mar 25 16:22:11 Except a custom squiggle Mar 25 16:22:13 Elrond: read :) Mar 25 16:22:20 Did we get the time machine out for nothing? Mar 25 16:23:00 rbs-tito: I'm hoping Prayag will hop in and research the myriad successful, secure, proven user authentication schemes that have been devised in that timespace. Mar 25 16:23:04 Elrond, buy IBM Mar 25 16:23:09 hahaha Mar 25 16:23:09 i know that a user can recodnised by the way he presses the keys.. can we do it using touch screen too Mar 25 16:23:41 What about a dental check, you impy bite the touch screen Mar 25 16:23:44 *simply Mar 25 16:23:48 :) Mar 25 16:23:49 haha Mar 25 16:23:53 rbs-tito: better Mar 25 16:23:55 rbs-tito: DNA Mar 25 16:23:56 :D Mar 25 16:23:57 sneeze on it Mar 25 16:24:14 it will identify the level of bacteria in your snot Mar 25 16:24:29 Possible DoS methods: Mar 25 16:24:33 Getting sick Mar 25 16:24:33 btw dental marks is a biometric method u see... but the problem is.. how do u ask a user to bite his own phone ??? Mar 25 16:24:35 taking antibiotics Mar 25 16:24:58 Prayag: By attaching the DentalPro 1000 dongle Mar 25 16:25:04 dho - *LOL* Mar 25 16:25:09 :D Mar 25 16:25:38 (script) openmoko-devel: Harald Welte Re: agpsd status and 1 other mails Mar 25 16:26:38 (script) openmoko-devel: Harald Welte Re: USB on-the-go (OTG)- how to use the Neo1973 with USB host and Mar 25 16:26:39 (script) openmoko-neo1973: Harald Welte Re: increasing vibrator strength Mar 25 16:27:25 * mjr needs to read -neo1973 Mar 25 16:27:25 The Openmoko doubles as a vibrator? O_o Mar 25 16:27:49 * mwester checks to see that he's on the right IRC channel... Mar 25 16:28:22 HAHA Mar 25 16:28:38 rbs-tito: That's the other dongle.... Mar 25 16:29:48 dho: Ooh, USB or Bluetooth? Mar 25 16:30:01 Wait a minute, GPRS sex. Mar 25 16:30:40 * dho heads to USPTO Mar 25 16:30:51 www.thetoy.co.uk got there first. Mar 25 16:31:24 Dammit. Mar 25 16:31:31 I had that idea at least 5 years ago :( Mar 25 16:31:45 lol Mar 25 16:32:05 * SpeedEvil started writing scripts for an online auction site, before deciding it could never work, and leaving it. Mar 25 16:33:27 Will the Openmoko decode and play OGGs? Mar 25 16:33:49 yes Mar 25 16:33:53 it can do that Mar 25 16:34:21 Good :) This means I might not need to find a seperate MP3 player Mar 25 16:35:08 rbs-tito: You'll what a micro SD card for storage space Mar 25 16:35:40 Yeah. Maybe a standalone MP3 player would be needed still Mar 25 16:36:04 Maybe that could be the next FIC project, an MP3 player running rockbox Mar 25 16:37:17 I do intend to use my Neo also for music Mar 25 16:37:33 a 2GB microSD has space for a couple of songs too.. Mar 25 16:39:32 i'm more than happy with my normal collection of 128 MB Mar 25 16:44:08 youl get a good bit of vorbis encoded music on a 2gb card Mar 25 16:44:18 Indeed. Mar 25 16:44:42 as for a player. you can always just install mplayer on it Mar 25 16:45:48 rbs-tito, ogg is a container not a encoding. just btw Mar 25 16:45:49 :P Mar 25 16:48:01 theholyduck: or then again, actually make up a gui Mar 25 16:48:18 vanilla mplayer might not be too handy on touchscreen Mar 25 16:48:22 true Mar 25 16:48:54 but console playing is just so cool Mar 25 16:49:08 generic media player with playlist support. not too complicated. might also be handy to have good podcast integration for seamless podcast updates Mar 25 16:49:52 well for the playlist etc. you could always just use some sort of modifyed mplayer Mar 25 16:50:06 with a gui that works with a touchscreen Mar 25 16:50:23 yes Mar 25 16:50:59 mplayer is a good place to start anyways. since it runs on ARM Mar 25 16:52:51 i'd use madplay on a non-fpu system like the arm920t Mar 25 16:53:10 mad uses only fixed-point arithmetic Mar 25 16:53:34 mpg123 runs too and does only fixed-point Mar 25 16:54:48 mplayer -ac mad Mar 25 16:55:38 (script) wiki RecentChanges: [[Wishlist:Bootable_USB_device_emulation]] [[Wish_List_-_Hardware]] Mar 25 16:56:35 Does gstreamer have any fixed point decoders? Mar 25 16:56:57 They said they were working on a lot of embedded and phone stuff now Mar 25 17:01:35 theholyduck: there's xmms-embedded, too Mar 25 17:04:59 xmms-embedded is a farce Mar 25 17:05:30 CM: gstreamer was targetted at arm years ago Mar 25 17:05:39 google for ridgerun inc. Mar 25 17:06:06 mplayer works for ogg IIRC at a little more CPU than mp3 Mar 25 17:06:13 33% IIRC, compared with 25. Mar 25 17:06:42 I suspect every decoder is using the tremor codec Mar 25 17:07:17 As a datapoint - hardware little DSPs that are used for decoding ogg/mp3 uses more power for ogg. Mar 25 17:09:34 koen: Was just thinking of the last lugradio show where Christian Schaller said they were doing a lot of embedded things right now. :) Mar 25 17:17:36 Has the display a solid surface or is it rather 'soft' and deformable? Mar 25 17:18:07 rather solid, glass-like Mar 25 17:18:20 soft plastic Mar 25 17:18:24 with a glass underlay Mar 25 17:18:35 the travel is very small till you hit the glass Mar 25 17:18:44 softness depends on the force you apply :p Mar 25 17:19:19 so you can't damage the display very easily? ;) Mar 25 17:19:20 I mean not very hard. think slightly thinner than a plastic coke bottle, and of much the same stuff. Mar 25 17:19:37 It'll shatter as easily as any touchscreen. Mar 25 17:19:44 Probably. Mar 25 17:20:13 One thing I want to find out - when P0 people are upgraded to P1 - or even if only the internal ones are - is drop testing. Mar 25 17:20:15 touchscreens usually survive and LCDs break from my experience Mar 25 17:20:55 drop from 10cm onto concrete from all orientations, test functionality, multiply distance by 1.5, repeat. Mar 25 17:21:01 SpeedEvil: but it won't damage in your pocket? :) Mar 25 17:21:16 It will be instantly scratched by any metal objects. Mar 25 17:21:29 repeat with another phone onto carpet. Mar 25 17:21:39 * koen hands SpeedEvil a metal ballbearing Mar 25 17:21:45 provoke a collision of a Neo1973 with a OLPC moving with the same velocity in opposite directions :p Mar 25 17:22:13 I suspect the lid is the first thing to come off Mar 25 17:22:37 True. I want to know when actual irreversable non-cosmetic damage happens though. Mar 25 17:23:01 I.E. LCD breaks, or GSM stops working. Mar 25 17:24:55 buy another one! Mar 25 17:26:37 "buying" is currently a bit of a problem ;) Mar 25 17:26:49 lid - lcd - gps - gsm antenna Mar 25 17:26:57 I suspect it will fail in that order Mar 25 17:27:03 counter Mar 25 17:27:03 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 4 days 18:32:56 (11.773 +-7.0 days) (625) Mar 25 17:27:56 I think the first 500 will go away in days ;) Mar 25 17:28:38 But if /window goto 3 Mar 25 17:28:39 But if /window goto 2 Mar 25 17:28:40 But if /window goto 3 Mar 25 17:28:42 gah Mar 25 17:28:58 Your irssi didn't understand conditionals? Mar 25 17:29:02 * dho gives SpeedEvil a cookie Mar 25 17:29:35 * CM uses alt + [0-9] or Esc + [0-9] Mar 25 17:29:53 * dho uses meta 1 to l Mar 25 17:30:02 I do to - but my alt key has a bit of rice underneath it. Mar 25 17:30:04 I'm in a lot of channels :P Mar 25 17:30:05 SpeedEvil: the display is a lot hardier than it looks like. Mar 25 17:30:11 SpeedEvil: lol Mar 25 17:30:17 * balrog-kun uses Esc + [0-9] and Esc + arrows Mar 25 17:30:33 jeez Mar 25 17:30:39 am I the only one that uses letter keys as well? Mar 25 17:30:39 SpeedEvil: its quite bendable, too. i think that the biggest danger to it are hard, pointy things. pressure shouldnt be much of a problem. Mar 25 17:30:40 balrog-kun: Thanks, I didn't know about the arrows :) Mar 25 17:30:45 irssi works with q to o by default also Mar 25 17:30:55 for 11 to 19 Mar 25 17:31:02 dho: Nice :-) Mar 25 17:31:13 I just happen to need p to l Mar 25 17:31:28 Start a second irssi. ;) Mar 25 17:31:30 * balrog-kun /whoises dho Mar 25 17:31:38 and put both in screen ;) Mar 25 17:31:49 I use ctrl+p and ctrl+n instead of arrows Mar 25 17:32:39 inz: I didn't know that one either.. Been to lazy to read all man files ;) Mar 25 17:32:40 SpeedEvil - I wonder, if the FIC people got my screen protctor thread. ;o) Mar 25 17:33:02 CM, man files be damned, just blindly press all the keys at random and see what happends... ;) Mar 25 17:33:03 balrog-kun: I'm on other networks too heh Mar 25 17:33:14 * CM ;) Mar 25 17:33:28 CM - Meta-A switches to another window with activity. I leave the excercise to find out, which of many it chooses to the reader ;) Mar 25 17:33:39 * SpeedEvil likes his windows to be visible, so only has 8 open. Mar 25 17:34:09 80*90 terms ftw. Mar 25 17:34:25 Elrond, I have actually a script that prioritised some windows when pressing meta-a Mar 25 17:34:36 -typo Mar 25 17:35:04 to play guitar hero or not to play guitar hero Mar 25 17:35:07 that is the question Mar 25 17:35:18 The priority ones are lower numbered, so I can just do Meta-3 Mar 25 17:36:42 Speed, 80x50 ftw! Mar 25 17:37:03 What exactly qualifies as a deliverable ?? Mar 25 17:37:04 * Tv 's going to present a talk about openmoko in an hour Mar 25 17:37:24 Can anyone please tell me... :) Mar 25 17:37:59 Prayag: "deliverable" in what context? Mar 25 17:38:18 A deliverable is something deliverable in terms of a contract. Mar 25 17:38:22 GSoC? Mar 25 17:38:26 in the context of present Projects Mar 25 17:38:26 It may - for example be source code Mar 25 17:38:28 oh Mar 25 17:38:41 oh google Mar 25 17:38:58 it means just goals, i think Mar 25 17:39:11 deliverables are what you are contracted to deliver. Mar 25 17:39:13 basically. Mar 25 17:39:30 in some cases it's just an end product - a binary. Mar 25 17:39:43 in others it'll be all the working notes, source, binary, documentation, ... Mar 25 17:39:57 ok.. do modules of the project count as deliverable?? Mar 25 17:40:27 i mean... (4 neatly defined module) * integration = project Mar 25 17:41:06 now, can these modules be quoted as deliverables ? Mar 25 17:41:30 * koen ipkg install gpstk-bin Mar 25 17:42:09 i m sorry i din't realise this is not GoSoC specific channel, I mean in refernce to GoSoC only Mar 25 17:45:12 gpstk works on arm Mar 25 17:45:47 SpeedEvil, if you would please :) Mar 25 17:46:08 I dunno. Mar 25 17:46:16 Go and read the GSoC terms and conditions. Mar 25 17:48:28 ok.. thanks.. Mar 25 17:48:46 there was nothing given there, so i just asked Mar 25 17:51:51 maybe think of deliverables as what users get from your project Mar 25 17:53:03 I'm running mokoMakefile 'make setup' - Is it ordinary, that it runs hours? Mar 25 17:53:24 the whole thing takes about 7 hours on my computer Mar 25 17:53:51 argh - mainly depends on the internet connection? Mar 25 17:55:38 (script) wiki RecentChanges: [[Buying_Interest_List]] Mar 25 17:58:42 quickdev: i have 100/100, so no Mar 25 17:58:47 balrog-kun: thanks :) Mar 25 17:58:50 compile times are notorious Mar 25 18:00:39 I hope my dual core will proceed fast :P Mar 25 18:10:35 is GTA02 p1.5? Mar 25 18:11:25 are there references to GTA02 somewhere? Mar 25 18:11:26 I'd guess on p2 Mar 25 18:12:12 balrog-kun: harald mentions on the list that there MIGHT be SOME usb power in GTA02 Mar 25 18:12:22 ah ok Mar 25 18:12:35 that would be nice Mar 25 18:12:46 gimme enough to run an USB stick on it Mar 25 18:13:00 that shouldnt take much Mar 25 18:13:03 I did some tests. Mar 25 18:13:12 And found a USB stick took 50-60mA Mar 25 18:13:15 less at USB1.1 Mar 25 18:13:18 oh really? Mar 25 18:13:20 they'll give you half of the voltage :p so you can keep on dreaming Mar 25 18:13:35 Logically. Mar 25 18:13:42 You'd want 100mA@5V Mar 25 18:13:52 as that should work for all standards compliant stuff. Mar 25 18:15:04 only self powered one really Mar 25 18:15:10 or very low power devices Mar 25 18:15:20 printers Mar 25 18:15:27 keyboards, ... Mar 25 18:15:31 i dont think i care for a printer on a phone Mar 25 18:15:46 besides, drivers are binary only for my samsung laser Mar 25 18:15:49 People will want to print faxes. Mar 25 18:15:53 For example. Mar 25 18:16:10 re Mar 25 18:16:18 100mA is guaranteed to be enough power for standards compliant stuff to read the actual device info. Mar 25 18:16:28 And then to set it up in low power mode if it supports it. Mar 25 18:17:03 i only care for keyboard and usb memory ;) Mar 25 18:17:16 You can tell that a wifi stick has been plugged in, for example, and read the descriptor that says it uses 300mA and not enable it for example. Mar 25 18:17:17 and for keyboard i can always go bluetooth reall Mar 25 18:17:40 well i really hope for internal wifi for 1.5 ;) Mar 25 18:17:50 I mean that as an example. Mar 25 18:17:53 (atheros ar6001 seems like it could work ;) Mar 25 18:17:54 Scanner for example. Mar 25 18:18:03 Or joystick with force feedback. Mar 25 18:18:16 you can easily use a powered hub with a scanner Mar 25 18:18:21 True. Mar 25 18:18:25 its not like a scanner fits into your pocket Mar 25 18:18:29 (not a normal one anywaY) Mar 25 18:18:42 (same goes for the keyboard, anyway) Mar 25 18:18:46 The point is that you can plug any USB thing into it, and it may function somewhat. Mar 25 18:19:03 webcam = Mar 25 18:19:04 ? Mar 25 18:19:06 For example, a camera might let you read the memory card, though it won't charge or take pictures. Mar 25 18:19:18 they tend to be moderately high powered IME. Mar 25 18:19:26 never had one Mar 25 18:19:37 for webcams I mean. Mar 25 18:19:43 yeah i know ;) Mar 25 18:19:48 moderately useless devices Mar 25 18:20:15 i wonder if there would be a camera small enough to fit into the frame of glasses Mar 25 18:20:26 to get like 24/7 filming of your life Mar 25 18:21:09 vga would be quite enough for that purpose Mar 25 18:21:18 SpeedEvil: are you developer? Mar 25 18:22:41 buz: depends on the purpose :p Mar 25 18:23:04 well with more than vga you quickly need a lot of cpu power for encoding of the video Mar 25 18:23:12 mpeg4@vga is feasible Mar 25 18:23:25 right, and glasses don't have that much cpu power :P Mar 25 18:23:36 at least the usual type Mar 25 18:23:45 space for the camera is more of an issue i gues Mar 25 18:23:54 you could always make bigger frame in the back to host a cpu Mar 25 18:24:04 but ideally it should be stealth Mar 25 18:24:44 you could have a second camera somewhere under the tongue, recording everything you're eating Mar 25 18:25:08 mhh that movies might even make money Mar 25 18:40:13 I just have to say it: Gimme wifi! Mar 25 18:41:09 * koen hands MortenLysgaard some 2.4GHz EM waves Mar 25 18:41:27 Thanks :P Mar 25 18:41:43 *laughing* Mar 25 18:44:42 But, I'm wondering. Where I live, in Norway, everybody have wifi in their houses, and many doesn't have encryption. So if Neo had a wifichip I could get internet almost evetywhere where there's people. How is to you? Mar 25 18:45:25 * koen has flat-rate EDGE Mar 25 18:48:49 How much do you pay koen, and neo doesn't have edge? Mar 25 18:49:11 I suspect the plan covers non-edge gprs as well Mar 25 18:50:24 I could switch to t-mobile Mar 25 18:50:38 flatrate umts/edge/gprs for €9.50/month Mar 25 18:51:16 here that costs 10 euros Mar 25 18:54:49 in norway you have to pay 450 nok for flatrate gprs, that about 50eur Mar 25 18:55:39 (script) wiki RecentChanges: [[Talk:Summer_of_code]] [[Wishlist:Bootable_USB_device_emulation]] Mar 25 19:25:40 hey mentors any project in OpenMoko (GoSoC) not applied for yet ?? :) Mar 25 19:56:39 (script) wiki RecentChanges: [[Talk:Summer_of_code]] Mar 25 20:09:11 i want a neo1973 Mar 25 20:09:19 counter Mar 25 20:09:20 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 4 days 15:50:40 (11.660 +-7.0 days) (626) Mar 25 20:09:32 r u excited Mar 25 20:09:50 im gonna try and get my work to buy me one Mar 25 20:10:03 Heh, that would be nice Mar 25 20:11:51 ya Mar 25 20:12:02 i think i gotta switch providers tho Mar 25 20:12:09 Hi ppl :) Mar 25 20:47:59 I'm looking at MokoMakefile, and when I do make setup, I get an error indicating that "branch org.openembedded.dev has multiple heads." is there a suggested resolution to that issue? Mar 25 20:48:59 BrianHV: I got that too, and had to do a mtn merge or something like that Mar 25 20:51:59 Heya, everyone. Mar 25 20:52:00 * BrianHV has no experience with monotone Mar 25 20:52:16 BrianHV: Me neither, sadly ;) Mar 25 20:52:43 mickeyl blogged that he has working dialer and basic PIM applications... Mar 25 20:52:50 * pavelm would love to test that ;-) Mar 25 20:53:02 * CM waits patiently for P1 Mar 25 20:53:35 ...ouch and word of warning: Neo charger is crap that puts ~60V AC (relative to ground) on its output. You can feel that. Mar 25 20:53:53 Hehe Mar 25 20:54:09 relative to which ground? Mar 25 20:55:26 psi: relative to earth Mar 25 20:55:32 hehe Mar 25 20:55:49 are its outputs floating? Mar 25 20:56:11 ...fortunately openmoko does not seem to have any metal parts on the outside :-) Mar 25 20:56:16 psi: probably. Mar 25 20:56:39 (script) wiki RecentChanges: [[Buying_Interest_List]] Mar 25 20:57:13 Ok, I have to get some sleep. Bye bye. Mar 25 20:57:35 Night Mar 25 20:57:43 dream of your neo ;) Mar 25 22:51:01 greetins Mar 25 22:51:36 when would openmoko be released? Mar 25 22:51:36 darx: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 25 22:54:37 darx - What do you mean? Mar 25 22:56:18 Elrond: i'd like to buy one.. so i'd like to know when it'll be released and when i'll be able to buy one Mar 25 22:56:21 darx, it has been released already, though a rather early version Mar 25 22:57:36 mjr: cool.. and i suppose that would not be available for purchase? Mar 25 23:00:04 not yet Mar 25 23:00:49 and it'll be overly expensive at first Mar 25 23:01:48 i wonder if they are going to make more money with the developer preview version or the final mass sale version Mar 25 23:02:50 hmm last i heard they were always both going to be US$350 Mar 25 23:03:27 oh? i thought that was only the preview version Mar 25 23:10:03 balrog-kun - expensive? Mar 25 23:11:11 Elrond: as for a phone Mar 25 23:11:32 Yeah, okay. But for a linux pda? Mar 25 23:11:39 you can also find cheaper arm based pdas with higher clocking (but probably less ram) Mar 25 23:12:39 The rumours tend to say, that the mass market version would have a faster cpu. Mar 25 23:13:31 Tv - How was your talk? Mar 25 23:13:58 balrog-kun - What is a reasonable pricing for a useful linux pda with gps? Mar 25 23:14:21 Elrond: Faster CPU is for sure. Other stuff is still in progress. Mar 25 23:15:08 Still.. an HTC P3600 is like $600, and it's QVGA. Mar 25 23:15:22 Elrond: there probably aren't any with gps Mar 25 23:15:39 you could count N800 with a bluetooth gpd Mar 25 23:15:41 *gps Mar 25 23:16:03 So what'a n800 + bt-gps in $? Mar 25 23:16:17 balrog-kun: Which has the same price and no GSM. :) Mar 25 23:16:18 developer versions of N800 are EUR 90 Mar 25 23:16:23 :P Mar 25 23:16:46 balrog-kun: If you count dev versions, my Neo waa free. Mar 25 23:17:12 Well, what's an n800 + gps for _me_ ? Mar 25 23:17:14 balrog-kun: The interested numbers to compare are normal prices. Mar 25 23:17:21 true.. Mar 25 23:17:25 Elrond: ~370$ Mar 25 23:17:25 Elrond: i'm not sure Mar 25 23:17:47 Elrond: 350 for N800 and 20 or more for GPS Mar 25 23:18:16 And now add the cheap GSM phone and the space is your pocket for carry this all. ;) Mar 25 23:18:32 Okay. So in the end, if I want a linux-pda with gps, the neo seems less-expensive. Mar 25 23:18:54 The n800 has no gsm? Mar 25 23:19:02 The price is ok imho. We just need to get newer chips in it. :) Mar 25 23:19:10 Elrond: no Mar 25 23:19:24 Elrond: Not designed as a phone Mar 25 23:19:44 Ahh. Mar 25 23:19:58 Elrond: nope, you would need a modem for this i think Mar 25 23:20:04 pcmcia or something Mar 25 23:20:38 That' Mar 25 23:20:50 s not going to last through the day though/ Mar 25 23:21:43 balrog-kun: N800 has no PCMCIA, right? Would be braindead in this small case Mar 25 23:22:28 * stefan_schmidt tries to remenber the N800s he has seen Mar 25 23:22:54 hmm, i thought it had, but i may be wrong Mar 25 23:23:35 balrog-kun: I can't say for sure, only N770 here, but I really doubt it. Mar 25 23:24:56 stefan_schmidt: appears you're right, looking at wikipedia Mar 25 23:25:06 there are two SD slots instead Mar 25 23:25:20 yeah Mar 25 23:26:26 Elrond: you want N770 GPS? Mar 25 23:26:26 Moo^^: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 25 23:27:06 Moo^^ - Huh? Mar 25 23:27:32 I just read some random lines above - Navicore has N770 based GPS product: http://www.navicoretech.com/ Mar 25 23:28:11 I guess, the battery life wont be fun. ;) Mar 25 23:28:22 On the n800 webpage I read soemthing like 3,5 hours browsing time. ;) Mar 25 23:29:40 I think display sucks the most of it Mar 25 23:29:46 800x600 is not cheap, you know :) Mar 25 23:29:59 Yep. Mar 25 23:30:08 800x480 I read. Mar 25 23:30:18 ops, sorry Mar 25 23:30:25 I yet have to find the size of the screen. Mar 25 23:30:26 must be 800x480 since it's not 4:3 Mar 25 23:31:25 And let me guess, the n800/770 have no free (as in speach) OS? Mar 25 23:32:03 it's debian based Mar 25 23:32:12 there are some binary blobs Mar 25 23:32:22 but kernel is open AFAIK Mar 25 23:32:31 Elrond: maemo.org Mar 25 23:32:37 Ahh. :-) Mar 25 23:32:47 I thought, maemo.org made only userspace trickery or so. Mar 25 23:34:04 Moo^^ - The binary blobs are for wifi or so? Mar 25 23:34:14 their kernel changes are slowly pushed to linux-omap-2.6 tree (and from the changes go to mainline, even slower) Mar 25 23:34:16 and for handwrite recognition Mar 25 23:34:39 and several other things... Mar 25 23:34:41 also, N770 team has done a lot of Gnome tweaks to make it run better Mar 25 23:35:06 good evening mjr, my owlery friend :) Mar 25 23:35:26 they hired o-hand.com for a the Gnome changes and some of the maemo parts Mar 25 23:36:13 Hmmm. Mar 25 23:37:08 If I want a linux thingy, I want it to be without binary blobs. The gpsd(-plugin) already annoys me. If I wanted a binary thingy, I could buy a normal phone, or a garmin gps. Mar 25 23:37:58 * Moo^^ is happy in the world where he can hack and build applications Mar 25 23:38:31 Elrond yes, but as far as i know the question was small binary blob in userspace with only serial to the hardware or no gps at all.. so i'm kinda glad we have at least that Mar 25 23:40:01 roh - Yeah. Well, a sirf3 would have eaten little more power and would have used open protocols. Mar 25 23:40:36 and really shitty reception AND lock time if the specs of the hammerhead are right Mar 25 23:41:01 i know sirf3 and it sucks big time to wait 30seconds till it knows where it is when under open sky Mar 25 23:41:21 lets not think about locking indoors ;) Mar 25 23:41:31 Moo^^ - I was dependent on a binary kernel blob for my pc at home. Guess what? Right. The vendor dropped supporr for my hardware I stuck at 2.0.x, when 2.4.20 was out there. I'm pissed by that. Mar 25 23:42:11 and yeah there are other not-softmodem-gps chips for mobile use.. but they are in a completely other cost-domain so are afaik out of the question Mar 25 23:43:01 Elrond its NOT a kerneel-binary-blob (which is really important) and it does not touch any registers of the hardware (which is important to contain and debug with it) Mar 25 23:43:27 roh - Yes, I know. That's, why I'm mostly happy with it. Mar 25 23:44:22 Although it might become annoying in 10 years.... imagine the userspace thingy would be a.out ;o) Mar 25 23:44:45 and hey.. we can easyly swap the hammerhead out for another, better solution on the way to gtaY where Y is > 1 when we find some chip which is similar in price and fits out licence-problems Mar 25 23:45:37 I'd be happy with TTFF in the 30sec order. :) Mar 25 23:45:37 Elrond it might. but lets complete the function first and get it working so we have the time to make it nice before the hardware isn't buyable anymore Mar 25 23:46:29 Yeah. I anyway guess, that some geeks will decode the protocol sooner or later and we'll have a nice open hammerhead library ;) Mar 25 23:46:46 yep.. have fun with it :) Mar 25 23:48:29 Elrond but hey.. youre in germany... the land of the (mostly/best known) reverse-engineering laws *g* Mar 25 23:48:47 Europe FTW! Mar 25 23:50:49 the userspace thingy is not in any popular format Mar 25 23:51:03 ..or rather, even if it is, the gpsd needs to implement a loader for it anyway Mar 25 23:54:57 good night Mar 25 23:56:11 roh - What do you mean? Mar 25 23:56:39 (script) wiki RecentChanges: [[Talk:Neo1973_Power_Management]] Mar 25 23:57:34 Elrond that reverse-engineering is relatively harmless and relaxed here, compared to e.g. the US where you can go to jail for voiding some stiupid eula Mar 25 23:58:09 here you can even disassemble sth. if you only use what youve learned to generate some kind of 'compatibility' Mar 25 23:58:54 but be careful. neither am i a lawyer nor can i give (or are allowed to give by law) any law-wise advise ;) Mar 26 00:02:24 Yeah, law-stuff via irc is always "no warranty". I know. Mar 26 00:03:10 Well, I'm in the black box reverse engeenering business for 7 years now. I don't need to disassemble. :) Mar 26 00:04:18 morning Mar 26 00:04:33 Elrond: so you're a hacker :) Mar 26 00:08:48 Moo^^ - developer/maintainer/... ;) Mar 26 00:17:48 Elrond ;) nice to hear Mar 26 01:07:05 *sigh* ... why does it build cups now? Mar 26 01:39:35 \ Mar 26 01:39:42 back up? Mar 26 01:40:19 press [AUX] to select, [POWER] to execute. ;) Mar 26 01:42:44 counter Mar 26 01:42:45 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 4 days 10:17:15 (11.429 +-7.0 days) (627) Mar 26 01:56:39 (script) wiki RecentChanges: [[Talk:Messages]] **** ENDING LOGGING AT Mon Mar 26 02:59:57 2007