**** BEGIN LOGGING AT Fri May 11 02:59:56 2007 May 11 03:46:19 re May 11 03:48:19 (script) wiki RecentChanges: [[Buying_Interest_List]] May 11 04:39:22 Sufflope: I think I have seen mentions of funding further development from money they get from Neo1973 May 11 05:49:19 (script) wiki RecentChanges: [[Wishlist:BuiltInScriptingLanguage]] May 11 06:12:41 counter May 11 06:12:41 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) P1: 3 days 17:47:16 (3.741 +-3.5 days) (1143;199) May 11 06:14:38 morning May 11 06:14:49 Morning :) May 11 06:15:38 LaF0rge: Any news about the production run yesterday? :) May 11 06:20:55 Elrond: you wanted something? May 11 06:49:55 hmm I can't make calls with the latest images May 11 06:50:05 the gsm radio won't turn on I guess May 11 06:53:00 bmidgley which phone revision do you have? May 11 06:53:03 Are you going to integrate the javaFX into the phone by deault? May 11 06:53:04 fix__: 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.) May 11 06:53:24 roh: p1 with 1May images May 11 06:54:00 bmidgley which kernel revision? May 11 06:54:37 if its 2.6.21 try updating to the latest builds. there were some bugs on the switch from 2.6.20 to .21 which are afaik resolved in the meantime May 11 06:58:07 roh: yeah, 2.6.21.1 May 11 06:58:48 which -mokoX is it? X=9 is recent May 11 06:59:04 2.6.21.1-moko10 May 11 07:04:02 fwiw, I did not flash uboot since it sounds risky May 11 07:04:19 but the kernel and root are the latest May 11 07:06:06 bmidgley: risk is in flashing wrong image (->brick -> need debug board) May 11 07:06:40 and that has happened I think at least to 3 P0 people, probably more May 11 07:07:08 would any recent uboot be ok on p1? May 11 07:07:25 bmidgley do you have a debugboard? May 11 07:07:29 n May 11 07:07:53 umm.. no, only one that is specifically for v4, v3 u-boot -> brick May 11 07:08:19 yes. because its bv3 and bv4 what the people outside of fic have May 11 07:08:23 bmidgley: I would check with somebody who knows for sure (and I think there is no need to update u-boot) May 11 07:08:49 reflashing kernel and rootfs is safe with DFU method May 11 07:08:56 * aloril nods to hrw May 11 07:08:56 yeah I'm not going to change it unless it's absolutely necessary May 11 07:09:10 reflashing u-boot is safe too but it *NEED* to be correct one May 11 07:09:12 v3 and v4 were before bv3 and bv4, just without bluetooth and some gpio changes May 11 07:09:32 yeah I had to reflash kernel/root right away using dfu (kernel and root didn't match version) May 11 07:09:44 we really should remove the v3 and v4 from the standard builds soon May 11 07:09:58 for P0 devices flashing u-boot is necessary, but not for P1 May 11 07:10:36 combinations happened: v3 u-boot into bv3 device -> brick May 11 07:11:04 bv3 into bv4 device -> brick (or was it bv4 into bv3?) May 11 07:12:15 aloril: any incorrect one will rather req debug board May 11 07:12:57 hrw: guessed so, above is what I have heard has happened in practice May 11 07:13:51 is the gsm icon in the top bar supposed to change? it always has the question mark for me May 11 07:16:33 "power up gsm antenna" and "autoregister with network" don't appear to do anything May 11 07:16:37 Hello; any news about the P1.5? The e-mail Sean Moss-Pultz mentioned will be sent to announce@lists.openmoko.org ? May 11 07:16:55 yes May 11 07:23:44 bmidgley yep. does not do anything afaik May 11 07:26:30 aloril: thanks:) May 11 07:48:21 (script) wiki RecentChanges: [[Bluetooth_Support]] [[Shipping_Notes]] May 11 08:03:25 morning May 11 08:03:41 hi Mac May 11 08:04:19 hi Marcin. Is the new Neo still working or already bricked ;-) ? May 11 08:05:14 discharged May 11 08:06:16 TRIsoft: and my debugboard is few km from me - I gave it to friend which needed to debug his ARM project May 11 08:06:36 hrw: Ouch May 11 08:06:56 TRIsoft: no problem - it is only discharged battery ;) May 11 08:07:01 I have external charger for them May 11 08:10:24 hrw: The Nokia charger ? May 11 08:12:43 no! better one ;) May 11 08:12:53 3rparty totally noname charger May 11 08:13:02 nokia ones refuse to charge !nokia batteries May 11 08:13:25 hrw: nice May 11 08:13:28 Do they id the battery? May 11 08:13:43 (the Nokia ones I mean) May 11 08:13:58 hads: sort of May 11 08:14:09 Annoying. May 11 08:14:15 openmoko: 03laforge * r1947 10/trunk/oe/packages/uboot/uboot-gta01_svn.bb: May 11 08:14:15 openmoko: * make sure lowlevel_foo.bin ends up in DEPLOY_DIR for non-neo1973 May 11 08:14:15 openmoko: * don't build u-boot for gta01v3 / gta01v4 May 11 08:14:35 ~hail LaF0rge for that change May 11 08:15:09 I don't know anything about batteries but I believe lithium has some sort of smarts inside which would make that a possibility. May 11 08:16:11 hads: fic battery is smart too, maximus 3rd party which I have here is smart enough too. May 11 08:16:22 hads: but nokias will not charge them May 11 08:16:32 Yeah, I meant lithium batteries in general. May 11 08:16:54 Just showing my battery ignorance :) May 11 08:19:01 i think liion without smarts would mostly blow up May 11 08:19:11 Yeah :) May 11 08:20:11 I just was guessing that the overcharge smarts may have some way of identifing manufacturer and the Nokia chargers may do that. May 11 08:30:59 hi guys May 11 08:31:39 some can help me? where can i find .pot files? May 11 08:38:46 hrw hehe... i just talked to laf. about the v3/v4 issue, and thats the current result. in the end we also want to remove all the ifdef-hell for v3/v4, since nobody works with these units anymore anyhow May 11 08:41:22 isn't $350 a bit much for "only" a gprs phone? i mean the v3xx with hsdpa is $400.. May 11 08:41:22 plipp: 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.) May 11 08:43:15 plipp on the contrary. its cheap for a smartphone May 11 08:43:29 v3xx is a featurephone afaik May 11 08:44:16 normal phones have no features? :) May 11 08:45:22 tigert thats only a classification. featurephones are normal phones. smartphones have a (usually) seperate application cpu and the gsm/umts stack runs on a second one May 11 08:46:18 ok but what am i paying for? a linux interface and gps? May 11 08:46:40 im just worried the gprs is too slow for my needs... May 11 08:46:53 which boils down to: linux is only possible on smartphones atm since nobody will and can opensource a gsm stack at all nor is there any possibility to get a device certified as needed by fcc and others.. May 11 08:47:51 plipp youre mostly paying for some quite nice hardware and the complete access to it besides the gsm part (where we do everything to make shure all neccessary and usable interfaces are accessible May 11 08:48:10 ok but that definition of smartphone has no meaning to the end user... but any way what is the max theoretical speed of the 2.5g gprs? May 11 08:48:33 touch screen with 4x more pixels at least May 11 08:48:51 i think gprs is aboud 4 KB/s max May 11 08:48:54 *about May 11 08:49:14 plipp i can't give you exact numbers. afaik it was about 6kbyte/sec or so in reality. i havent tried gprs on my moko yet May 11 08:49:36 4 KB/s the practical max i was getting with a pcmcia GPRS modem May 11 08:49:40 yikes.. so any chance of seing edge any time soon? May 11 08:50:11 soon = like within a year May 11 08:50:25 plipp atm its GPRS Class12/CS4 on the moko May 11 08:52:13 hard to tell May 11 08:53:01 pitty, i'd love to get one, but with those speeds.. May 11 08:53:35 plipp: I suspect v2 next year is more than edge (maybe even much more), but that is just 'rumor' May 11 08:54:17 guess i'll have to keep my crappy nokia till then =( May 11 08:54:39 it will also have *much* faster CPU May 11 08:54:47 * aloril suspects it might also be more expensive ;-) May 11 08:55:14 there should be announcement at mass market release time about v2 May 11 08:55:27 P1+ sometime in summer should have wifi May 11 08:56:22 It's surprising to see so many people join here and complain about things rather than being excited about the open software. May 11 08:56:55 wishlisters... May 11 08:56:57 users complain, the sky is blue, news at eleven May 11 08:57:13 heh May 11 08:57:36 im not complainin, was wishing for a marketing pitch =) and wanted to get more facts on the gprs May 11 08:57:44 * aloril thinks that there is no other option than Neo1973 at the moment: openness is the most important feature May 11 08:57:46 * CM spots a Neo: http://www.almaer.com/blog/archives/001485.html May 11 08:59:06 plipp if we would do so much marketing as we do on developing you would all be thrilled and in the end nobody would have a usable device. so we rather do development and don't talk about things which are not there yet or we don't know of for shure ourselves ;) May 11 08:59:21 :) May 11 08:59:32 i can see nothing worse than promising some feature and not keeping it. May 11 09:00:19 especially in an opensource product where it's more about trust than at normal commerial stuff where you know the vendor drops support completely when its out there for 18 month May 11 09:00:31 yes May 11 09:00:43 * hrw looks at nokia770... May 11 09:00:48 hrr May 11 09:01:42 hrw i think nokia still doesn't get what opensource really means. May 11 09:01:54 but thats my personal opinion *remark* May 11 09:02:11 I think they've hired a lot of people lately who do get it May 11 09:02:12 the OSSO people do, the rest of nokia doesn't May 11 09:02:20 including legal and management May 11 09:02:30 so your SOL if you have a nokia tablet May 11 09:02:42 roh, i get you and i think thats super, but thats also why i needed to ask =) May 11 09:02:55 I've avoided all the 'smart'phones until now. I've been waiting for the true linux phone. So I'm happy for each and every feature the Neo has. ^^ May 11 09:03:01 SQL? May 11 09:03:05 SOL May 11 09:03:31 ~sol May 11 09:03:55 apt has been missing all week May 11 09:04:21 Shit Out of Luck? May 11 09:04:53 Still Only Luser? May 11 09:05:33 Somewhat Oldfashioned and Lame? May 11 09:08:03 http://www.urbandictionary.com/define.php?term=S.O.L. (XorA wins) May 11 09:20:39 hrw - Yep. If you get your P1, can you update the "Flashing Openmoko" page for it? May 11 09:21:28 I have my p1 May 11 09:21:42 but today it is offline and I am too busy to play with it May 11 09:23:03 hrw - No problem. It's not top priority, just would be "nice to have" :) May 11 09:28:39 Elrond what needs to be updated? May 11 09:28:56 * lrak has unsuccessfully tried to join openmoko-devel May 11 09:30:01 lrak sorry.. just internal developers there atm. just ask what you wanna know here May 11 09:30:45 lrak: yah not in the CABAL :-) May 11 09:31:40 roh - It's probably only "add entry for phase 1" to the download list. But main point is "make sure, it works for P1 and doesn't break anything". May 11 09:32:11 counter May 11 09:32:11 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) P1: 3 days 14:27:46 (3.603 +-3.5 days) (1142;199) May 11 09:32:13 Elrond the most important part is that users dont flash uboots with the wrong version May 11 09:32:58 roh - Right. But if they want (should), there should be a correct one for them on the page. May 11 09:34:55 yep. May 11 09:35:29 eventually we should make paragraphs for each revision with the corresponding known-to-be-working filenames in the command lines (and urls where to download them) May 11 09:37:10 roh: I was very interested in getting a Neo .. but I am under the impression that the few people who are hacking are facing difficulties with gps and others May 11 09:37:10 roh - Yeah, that would be quite good. Especialy the "known-to-be-working" part. May 11 09:37:31 roh: plus I am interested in the investigations into the wifi device May 11 09:37:59 The phone is a brick if you flash the wrong uboot and ain't got the debug board? You can't reflash a working uboot? May 11 09:38:00 lrak - gps still does not work, right. Most of the rest should work. May 11 09:38:21 Rakshasa - Basicly "yes". May 11 09:38:53 open source difficulties dont scare me .. I fear the last minute "wtf .. this brick is proprietary" May 11 09:38:53 Rakshasa - So don't flash u-boot unless you "know what you're doing". May 11 09:39:03 Rakshasa: with debugboard it is few moments to get it working again May 11 09:39:53 lrak - The gps is the only "proprietary" issue on the phone. May 11 09:40:20 * Rakshasa thinks is going to get the db too. May 11 09:40:50 Elrond: well .. that is a huge thing to me May 11 09:40:52 I've got the bad habit of putting every darn bleeding-edge sw in my devices. May 11 09:41:13 bbl in a minute May 11 09:41:18 Elrond: I wanna know up to what point May 11 09:41:42 lrak - It's currently assumed, that P1 will only be sold if and when the GPS is known to work. May 11 09:42:38 Rakshasa: I have that habbit as well, so I ended up buying a windows phone to stop myself going crazy and actually having one device that works May 11 09:43:29 Elrond: I do not care if the gps does not work .. I only want true -open source- attempts May 11 09:44:52 I'm willing to make minor sacrifices if it means I can talk on an American carrier. =] May 11 09:45:01 lrak - Then better assume "no gps" for now. There are people working on reverse engeenering the gps, but you shouldn't hold your breath on it. May 11 09:45:36 Elrond: thx .. then Ill stick to some other project May 11 09:47:04 lrak - If you can live with a small binary blob in userspace, then the rest is "open source" May 11 09:47:58 lrak: if you find one ;) May 11 09:48:05 dottedmag: :D May 11 09:48:16 The GPS looks likely to work. May 11 09:48:22 The binary blob is present in the images. May 11 09:48:27 It's setup for the neo. May 11 09:48:34 It's just the setup scripts aren't present May 11 09:48:34 XorA: I won't touch a single winblow$ device anymore. I'll rather hack a brick than have a 'working' m$ device. =P May 11 09:48:56 Rakshasa: its nice to have a phone that when wife phones I can actually answer May 11 09:49:33 XorA: Well, sometimes it's convenient to be able to blame the phone for not answering May 11 09:49:37 ;) May 11 09:50:01 XorA: That's why I've got the simple nokia piece of crap... May 11 09:50:29 Elrond: aint there scheduling constraints with the gps stack ? May 11 09:50:35 SpeedEvil - huh, the binary thing is already in the image? So in theory we could do "clean room reverse engeering" if we wanted to? May 11 09:50:47 lrak - What do you mean? May 11 09:51:48 CM: thats why I chose a windows one May 11 09:51:58 Elrond: possiblt May 11 09:52:03 Or the strace thing May 11 09:52:14 strace, with lots of data should be much easier. May 11 09:52:17 Proabbyl. May 11 09:52:33 Elrond: I mean openmoko's gprs stack does not need real time stuff ? May 11 09:53:01 lrak: it's basically ppp over a serial device May 11 09:53:01 lrak, gsm/gprs runs on a separate chip with a proprietary RTOS May 11 09:53:14 from OpenMoko's perspective, it's a GSM modem May 11 09:53:14 you talk to the modem over serial May 11 09:54:39 lrak - Wait... do you mix up gps and gpRs ? May 11 09:55:18 Elrond: Is there a Wiki page about the GPS hardware? May 11 09:56:04 sparq - Hardware:AGPS May 11 09:56:23 sparq - http://wiki.openmoko.org/wiki/Hardware:AGPS May 11 09:57:45 so does anybody know how well the first production run went? May 11 09:57:58 * SpeedEvil looks into his crystal ball. May 11 09:58:10 * SpeedEvil shakes the crystal ball, and the snow falls back on the landscape. May 11 09:58:10 :P May 11 09:58:22 * Elrond looks over SpeedEvil's shoulder, using his monocular. ;) May 11 09:59:28 LuitvD - some people in here might or even should know it. But they're under NDA ;o) May 11 09:59:33 (not me :-) ) May 11 09:59:48 "Outlook good." May 11 10:00:13 <-- runs for food. :) May 11 10:00:34 * LuitvD stops running, realizing he's the food May 11 10:01:29 Hmm... Why does the GPS need non-free software? It looks like it just speaks a GPS serial dialect over at tty. May 11 10:01:57 sparq, licensing and/or US ITAR restrictions May 11 10:01:59 sparq its a softmodem. it does not have its own cpu May 11 10:01:59 we haven't been able to reverse engineer it May 11 10:02:08 and also licesing May 11 10:02:26 mjr nope. its licensing issues from globallocate May 11 10:02:29 isn't there lots of explaining notes on the wiki about this topic? May 11 10:02:31 it's widely expected to be feasible to reverse-engineer in not too long a time May 11 10:02:32 it has to stop working when the velocity is above 8000 knots or something like that :D May 11 10:03:04 Stephmw: The Wiki just seems to say that the plugin for gpsd will be ready sometime in April May 11 10:03:21 guaqua: at a guess it's to stop such hw from being used in missiles, no? May 11 10:03:34 sparq: wonderful, update it with what you've just learnt :) May 11 10:03:36 roh, *nod*, hence mentioning "licensing" first :) May 11 10:03:39 Stephmw: yes May 11 10:03:47 of course still excellent use for guided bomb cars May 11 10:03:51 guaqua read the ITA stuff again. its only us-EXPORT (we produce in china) and also devices which are 'designed to work at $speed/height' which does also not really apply for mobile phones in my eyes May 11 10:03:56 Elrond: "a separate chip with a proprietary RTOS" !! May 11 10:04:09 lrak called gsm. May 11 10:04:24 roh: okay. i'm just repeating other people's words, my bad :) May 11 10:04:31 and we all know that there are more bomb cars than cruise missiles May 11 10:04:41 roh: but, but, surely you could see this big $10million missile setup with Neos glued to the side? May 11 10:04:45 My understanding is that that the speed restriction for retail GPS devices is a hoax. It's just a computational limitation. May 11 10:04:57 * lrak has made up his mind May 11 10:05:13 The model rocketry folks ran into this some years ago. May 11 10:05:27 Stephmw somebody able to build a missile does not have a problem builing a gps completely on its own, so no issue May 11 10:05:31 sparq, satellites go pretty fast, though May 11 10:05:41 so you need to be able to compute with high speeds May 11 10:06:04 roh: , sorry, I should've added it :) May 11 10:06:11 The satilites have a high velocity, but the relative angles are what taxes the CPU. May 11 10:06:14 ;) May 11 10:07:49 mjr: I seem to remember the Iridium satilites moved at about a degree per second in the sky, but that's just eyeballing it. May 11 10:09:50 * aloril2 looks also at crystall ball and imagines seeing testing going May 11 10:10:24 maybe they want to test before they talk about it? May 11 10:13:28 roh: ? May 11 10:15:49 roh: US import too. May 11 10:16:19 GPS satellites are _much_ higher. 12 hour orbits May 11 10:17:31 I suspended my Bv4 yesterday and now it doesn't wake up. the backlight is on, the lcd flickers strangely. May 11 10:17:38 I can ping it but not ssh into it. After about 30s the speakers did a 'pop' and the backlight turned off again. May 11 10:18:17 SpeedEvil: Cool. A new fact to squirl away! May 11 10:19:08 heh, now I get a 'pop' from the speaker every time I press the power button and a short white flash on the lcd every second time I do that. May 11 10:19:26 time to reboot May 11 10:19:30 vive galileo May 11 10:21:05 lrak: "E pur si muove." May 11 10:31:47 Huh. If the p1.5 model has a WiFi chip, the Neo will have seven radios in it. May 11 10:32:04 Eight if it's an a/b/g chip May 11 10:32:19 a is rare though. May 11 10:34:17 * CoreDump|home could live with that =) May 11 10:34:22 I wonder if someday a savvy supplier will stuff all the baseband hardware for bluetooth, GSM, GPS and WiFi into once chip. May 11 10:34:40 You could probably do it with a modern FPGA May 11 10:34:47 Probably. May 11 10:34:54 FPGAs are higher power though. May 11 10:35:10 SpeedEvil: Point. May 11 10:36:06 However... the idea of Universal Software Radio inside a device like the Neo... No, I must turn away from the blinding awesomeness. May 11 10:38:31 I've written a crummy but functional 802.11b MAC in python. The GNU Radio tools really are quite good. May 11 10:41:48 Universal software radio is cool. May 11 10:41:59 Given several more iterations of Moores law. May 11 10:42:11 At the moment - not really ready for primetime. May 11 10:42:56 * sparq shrugs May 11 10:42:57 * zecke|assignment smiles May 11 10:43:46 You also tend to run into problems with universality - it's possible in hardware to implement very sharp filters. May 11 10:43:59 SpeedEvil: Maybe not for mobile computing, but it works well enough for most wideband realtime signal analysis you might want. May 11 10:44:33 Doing it in software tends to be much more problematic, as you have to sample not only the signal at enough bits, but you have to sample all the interfering signals in your input bandwidth at high enough bits to discard them,. May 11 10:44:37 that can get nasty. May 11 10:45:04 SpeedEvil: e.g., I've seen it used to capture and pre-process electron cycletron emissions from a tokamak... May 11 10:45:36 Oh yeah. It's awesome for that sort of instrumentation stuff. May 11 10:45:52 SpeedEvil: True. I imagine it would work better if you designed your protocol to be software-radio friendly. May 11 10:46:06 Especially when it's possible to wideband record large chunks of time. May 11 10:47:01 SpeedEvil: Yep. The Plasma lab here at UCLA has some otherwise ordinary machines with an absolute *embarassment* of RAM. May 11 10:47:23 counter? May 11 10:47:24 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) source: http://wiki.openmoko.org/wiki/MokoMakefile ; P0: http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review; 3 days 13:12:34 (3.550 +-3.5 days) for devices for *ANYBODY* for $350 (targeting developers) (preorders: not yet) (2007-05-13); 4 months, 5 days (128.050 +-15.0 days) for mass market (2007-09-15): see topic for more info (1143;199) May 11 10:47:46 'capture before event' is _such_ a nice thing. May 11 10:50:20 Weird. TI has a dual-core mobile CPU. May 11 10:51:35 wats the point of dual-core in a mobile? May 11 10:52:17 wehttamb_: You can turn one of them off. ^_^ May 11 10:52:32 the same as in every smartphone. one for the application stuff, one for gsm/umts with an rtos on it May 11 10:52:32 oh cool May 11 10:52:52 http://www.taylorriver.com/content/view/140/25/ May 11 10:53:07 That's what's in the Nokia 770 May 11 10:53:31 without a second cpu anything has to run inside the rtos, which means no linux or wince since both arenn't hard realtime May 11 10:55:28 Well - there are patches for linux that do pretty much hard realtime. May 11 10:55:37 It depends on requirements. May 11 10:56:17 I suspect not adequate to push stuff direct to radios and keep in GSM/... timings. May 11 10:56:40 * sparq deletes a halfway typed line May 11 10:57:12 hard means hard in that aspect. and linux is just not possible because of the license. you cannot get anything certified which somebody must be able to change themselves at all May 11 10:57:43 Yeah. - though linux can be certified that way. May 11 10:57:50 But not if the user can change it May 11 10:57:55 which is the interesting case. May 11 10:57:57 This is why, I guess, the Neo talks to it's GPS device over a serial port. The GPS module has it's own little world, sealed away behind the UART. May 11 10:58:12 Not really. May 11 10:58:17 No? May 11 10:58:21 The GPS is designed to do this. May 11 10:58:30 SpeedEvil nope. not at all. because its gpl. which means you cannot make it not updateable May 11 10:58:34 ahgh, I meant GSM May 11 10:58:35 It's designed specifically so that you don't need hard-real-time. May 11 10:58:41 roh: Of course you can. May 11 10:59:01 roh: You have to supply source code. You don't have to make it possible to load the source code into your device. May 11 10:59:08 SpeedEvil lets not talk about this further. neither do i want linux in there because hard realtime is really pain, neither will this happen any time soon May 11 10:59:29 s/soon/as far as i can look into the future May 11 10:59:54 There are many completely locked down linux devices out there that the user can't touch. Even if they have source. May 11 11:00:31 It'd be nice - but AIUI it's completely impractical with current regulators in most places. May 11 11:00:57 SpeedEvil all completely locked linuxes are violating the licence May 11 11:01:28 which means for example in germany every customer can file a compliant and shut down the company from selling it at all till changed. May 11 11:02:06 roh: Possibly, but not necessaraly. The GPL doesn't say anything about *physical* impediments to playing with the software. May 11 11:02:49 yay, devel-image builds agin May 11 11:03:28 roh: This is what Stallman means when he says "TiVoization" May 11 11:03:50 sparq sparq i can think of no way to really lock it. if you mean the bootloader with physical its afaik a violation May 11 11:04:42 sparq just read up on gpl-violationss.org and the links there May 11 11:05:20 i can think of now way GPL can prevent any locking down, the license only talks about the use and availability of the software May 11 11:05:30 roh: Yes, I'm aware. I'm just pointing out that physical impediments are not violations of GPLv2. The vendor doesn't really care if you hack open your devices and solder a DB9 connector to the debigging interface so you can mess around with their Linux widget. May 11 11:06:00 Or if you have to crack a 1024 bit RSA firmware update key. May 11 11:06:34 Yep. It's lame, but probably not a GPL violation. May 11 11:07:19 Hence GPLv3. We can only hope it helps. May 11 11:07:36 sparq leaving away the mechanical parts needed to connect the cable is ok, yes, but actively hindering the user via crypto or parts linked against gpl code is not ok May 11 11:07:53 I agree that it's not OK. May 11 11:09:26 But, ultimately, I hope the market will do away with the practice, and we won't need the GPLv3. That's why I want a Neo! May 11 11:10:16 * CoreDump|home laughs as he finds TRIsoft being already stored in his phonebook May 11 11:10:28 * SpeedEvil agrees with sparq. May 11 11:11:31 sparq i, personally are a fan of gplv2 because it exactly says what you want to do May 11 11:11:43 s/want/can/g May 11 11:12:44 hmm... what about a company desktop with bios password and a protected root password? May 11 11:13:32 However, as a May 11 11:13:32 special exception, the source code distributed need not include May 11 11:13:32 anything that is normally distributed (in either source or binary May 11 11:13:32 form) with the major components (compiler, kernel, and so on) of the May 11 11:13:32 operating system on which the executable runs, unless that component May 11 11:13:33 itself accompanies the executable. May 11 11:13:54 That seems - to me - to completely allow you not to distribute the encrypted boot-loader. May 11 11:13:56 drath: Presumably, the actual owners of the machine have the passwords, and thus the control. One hopes anyway. May 11 11:14:13 sparq: yeah, right.... May 11 11:14:15 Which only allows you to load new encrypted firmware images. May 11 11:14:55 CoreDump|home: Sry, phone busy, will call you back May 11 11:15:03 wehttamb_: without '?' you get shorter version of counter May 11 11:15:18 Hi TRIsoft! May 11 11:15:26 Tri hi. May 11 11:15:35 * sparq smirks at the stacks of laptiops he's seen returned to the vendor with forgotten BIOS passwords May 11 11:16:02 sparq - hd passwords are also nice ;o| May 11 11:16:12 d'oh. May 11 11:17:27 They completely forgot the emergency erase without password feature. Making password protected HDs a brick. May 11 11:18:34 Well. It'd be nice if you could specify that that couldn't be done. May 11 11:18:43 As it makes it harder for thieves to resell. May 11 11:18:52 But leaving it out is a bit broken. May 11 11:19:29 SpeedEvil - Here the opposite happened. laptop stolen, thieve setting HD passwords, police found the laptop. HD==brick. May 11 11:19:50 I have heard that there are trojans that will change your HD password, and then ransom your data back to you. May 11 11:19:59 well, hopefully you can only set a password on blank HDDs May 11 11:20:23 so at least the thief in question can not stop you from accessing an otherwise working HDD May 11 11:20:43 CoreDump|home - No, you can set it on running HDs unless your BIOS has done a "security freeze" (you can also do the security freeze in software. My boxes do that straight after boot) May 11 11:21:10 I would consider that a serious bug May 11 11:21:34 IMHO it is. May 11 11:22:13 We had to throw away the HD (which otherwise was working and had (unneeded) adata on it) and buy a new one. May 11 11:22:29 Ouch May 11 11:23:08 Sigh. PC hardware is filling up with a lot of silly, parinoid security snake oil. It realy would be better if people just stopped buying from software vendors that write shitty code. May 11 11:23:33 Of course you could set the "master password" (that's the one needed for emergency erase) for all new HDs, but who thinks about that when buying a new HD? May 11 11:23:48 BIOS is a software afterall May 11 11:24:06 balrog-kun - Yeah. So what? May 11 11:24:42 picking on your "unless your BIOS has done a "security freeze" (you can also do the security freeze in software." May 11 11:25:10 balrog-kun - Okay. s/software/userspace/ ;-) May 11 11:25:38 balrog-kun - Or more s/userspace/your normal application OS/ ;) May 11 11:25:46 :) May 11 11:26:05 and switch to crypto lvm May 11 11:26:57 hrw - Which is probably even more secure. May 11 11:27:21 it's rare that police manages to find a stolen laptop btw. May 11 11:27:24 * XorA has crypto lvm setup May 11 11:27:28 did they also get the person? May 11 11:27:39 Elrond: and a hell of a lot more portable May 11 11:28:19 balrog-kun - They found the thieve. He was too dump. May 11 11:28:35 XorA: did you do any testing on the CPU burden and / or performance penalties for running fully encrypted disks? May 11 11:28:37 * sparq just doesn't have any data anyone would want (and has insurance) May 11 11:29:06 CoreDump|home: ive not noticed a significant slowdown May 11 11:29:12 CoreDump|home: ever when I do kernel compiles May 11 11:29:35 I would be *delighted* if some nerferious laptop theif would take the trouble to read the draft of my thesis. ^_^ May 11 11:29:46 interesting, that is good to know thanks! May 11 11:29:54 sparq: heheh May 11 11:30:19 sparq - *giggle* May 11 11:30:30 sparq - I don't even have a laptop. ;o) May 11 11:30:31 and [erhaps comment on the places that need more work in the thesis May 11 11:30:43 "We stole yer laptop, and we red yer thesis. Ekwashun 32 has a sign error." May 11 11:30:53 hehe May 11 11:31:03 sparq - *LOL* May 11 11:31:48 bbl May 11 11:31:55 Sometimes, when I'm at a conference, I would have liked a laptop, so I cuold securely ssh into my box without fering keyboard snifers ;) May 11 11:32:42 Elrond: used x40's and x30's are cheap, small and run Linux beautifully May 11 11:32:58 hmmm...used laptops.... May 11 11:33:27 maybe w/ a new keyboard and a new battery =) May 11 11:33:39 I have dell d400 - size of x30 May 11 11:33:45 sparq - What do they cost? (laptops aren't really confortable to use for me. So buying one for using it three times a year is a strange investment) May 11 11:34:04 CoreDump|home: dual core CPU so that probably alleviates the cpu load May 11 11:34:20 Elrond: I paid 300 eur for d400 with pentium M 1.6GHz, 512M ram and 40GB hdd. battery still works May 11 11:34:58 hrw - Hmm, that's nearly acceptable. May 11 11:35:46 Elrond: no cd/dvd drive, 12" May 11 11:36:30 My dad has a used t40, which he later might buy from his company... maybe that could be a nice thingy. May 11 11:36:48 Elrond: the t40 is very nice May 11 11:37:22 Instead of BT it has a normal modem thoguh. May 11 11:38:00 I sprang for an top of the line x40, but I use it for almost all my writing and daily use. It's been worth it. May 11 11:39:54 I want a 13.3" Santa Rosa. Good linux drivers.. May 11 11:39:59 sparq - Well, I have very bad eyes and using my normal glasses, I am 5cm away from the sreen. That's no fun with a laptop. So laptops are not confortable for me usually. May 11 11:41:23 You need a tablet, with an external keyboard and a head-strap. May 11 11:41:25 how's the Neo for people with short sight? May 11 11:41:35 balrog-kun: Just fine. May 11 11:41:48 You can use 80*25 text screens at 10cm easily. May 11 11:41:54 isn't the screen just too small? May 11 11:41:55 :D May 11 11:41:56 hmm, okay May 11 11:42:10 balrog-kun - I'll tell you, when I have a phone for testing. :) May 11 11:42:11 Short sight = you focus easily on close objects. May 11 11:42:14 Elrond: that's rough May 11 11:42:35 SpeedEvil: ah, true May 11 11:42:36 SpeedEvil - head-strap? May 11 11:42:46 Elrond: To keep it 5cm from the screen. May 11 11:43:37 SpeedEvil - Ahh. Well, the 5cm are an estimate. I have special glasses and love to move my head around the screen. ;) May 11 11:44:00 Elrond: One of the things I've been exceedingly annoyed about is the incredibly aweful support (on all OSs) for people with vision problems. May 11 11:44:26 Indeed. May 11 11:44:31 Elrond: I've spent *months* trying to figure out how to set something up for my aunt to use email, and it's just impossible. May 11 11:44:43 I was surprised to learn that you can learn to listen to text at 600wpm. May 11 11:44:51 sparq - Linux is quite okay mostly. I *hate* web designers who all assume people with 130% sight. May 11 11:44:57 At least, tsp over in #freenode-social can. May 11 11:45:26 sparq: elm? May 11 11:45:38 Console apps tend to be better in that respect. May 11 11:45:53 Elrond: True. If I lost my vision today, I could get along OK with a TTY an a screen reader. But I don't think I could have figured it out without having been able to see it first. May 11 11:45:57 Most people can read 80*25 on a 21" monitor. May 11 11:47:22 SpeedEvil: I tried that with her. It worked for a while, but after a few weeks it didn't anymore (she has macular degeneration) May 11 11:47:28 cu May 11 11:48:26 AMD is nasty. May 11 11:48:47 i did some search on screen readers but i don't think i could easily get used to them as an alternative for the monitor May 11 11:49:02 i even had a presentation about them May 11 11:49:39 If I could fork off another me-process, I would write gnome-desktop-nonvisual May 11 11:49:51 heh May 11 11:50:22 interpreting ascii art would be a pita May 11 11:51:17 I bet if there were a good nonvisual desktop for Linux, the userbase would double. May 11 11:52:45 There are a lot of blind or nearly-blind people in the world. May 11 11:53:04 sparq - Yep. May 11 11:54:00 sparq - I use 80x24 with a large font in an xterm (it's something like a 16" xterm) on a 19" screen. May 11 11:54:05 SpeedEvil - - I use 80x24 with a large font in an xterm (it's something like a 16" xterm) on a 19" screen. May 11 11:54:10 * CoreDump|home pokes TRIsoft May 11 11:54:20 * CoreDump|home must leave for work real soon May 11 11:54:42 Elrond: at least it works May 11 11:55:46 * TRIsoft slaps TRIsoft around a bit with a large trout May 11 11:55:55 sparq - Yep. I use a lot of console based apps. My only real uses for X: firefox (with imagezoom installed) and my own graphics tools. May 11 11:56:12 Elrond: Ouch.. I really wonder how the 2.8" Neo screen will work for you May 11 11:56:51 CM - I will let you know, when I get one for testing :) May 11 11:57:02 * CM has 8pt font on 80x24 19" 1600x1200 crt May 11 11:57:07 * arw Hello to all May 11 11:57:16 it's going to be tough by the sound of it May 11 11:57:24 13*4 text mode = 80*25 @ 16" May 11 11:57:42 Elrond: heh. Aside from IRC, that's not a bad way of getting more work done. When I want to get something done, I just run chmod 000 /usr/X11R6/bin/X May 11 11:59:13 sparq - IRC is my major time killer ;o) May 11 12:00:07 counter May 11 12:00:07 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) P1: 3 days 11:59:55 (3.500 +-3.5 days) (1144;199) May 11 12:00:30 aloril2 - Let me guess... now it's running at half speed? ;o) May 11 12:00:33 lower limit is right May 11 12:00:33 again ;-) May 11 12:00:48 Elrond: yes May 11 12:02:05 making uncertainty into +- 1 year would have made counter more likely to be correct, but also would have made it meaningless May 11 12:02:44 sparq: the several blind members of the Freenet I admin are all using MSDOS dialup programs with a screenreader.. (They're with us because we're about the only ones left who still do plain ASCII dialup shell accounts..) May 11 12:03:00 mehhhhhhh May 11 12:03:09 aloril2 - Someone already deemed it meaningless ;o) May 11 12:03:13 each USB disconnect == total freeze on Neo grrr May 11 12:03:26 well, every second disco actually May 11 12:03:31 baird: tsp on #freenode-social is blind - and was active a few mins ago, and uses linux May 11 12:03:40 Maybe worth asking May 11 12:03:42 CoreDump|home - P0 or P1? May 11 12:03:50 p0 May 11 12:03:54 Elrond: I would say there are degrees of meaninglessness May 11 12:03:59 ;-) May 11 12:04:01 but I believe I saw it w/ P1 as well May 11 12:04:32 * CoreDump|home waves goodbye to TRIsoft and #openmoko and leaves for work May 11 12:04:44 CoreDump|home - Have fun at work. :) May 11 12:04:49 there was a bling user on #gentoo, he even made a couple of working ebuilds that are now in the main cvs May 11 12:04:51 one idea would have been/is to have several counters and associate each one some propability May 11 12:04:53 will try =) May 11 12:04:53 *blind May 11 12:04:53 <-- sits at work and distracts himself using irc ;o) May 11 12:06:37 balrog-kun: Was a blind girl in some of my CS courses at the university. Had a linux laptop with some kind of finger braille reader May 11 12:07:05 for example, give current counter 70%, 1-4 weeks 15%, 1-2 months 10%, 2 months-never 5% May 11 12:08:13 CM: oh, that's a brltty (or something) May 11 12:09:19 She was pretty fast reading the console output on it. I just felt like my fingertips were vibrating.. ;) May 11 12:10:42 Imagine a blind person tryig to use the neo... May 11 12:10:43 baird: that's good to know May 11 12:12:31 I had a girlfriend who knew braille--she was a special needs tutor in the Armidale.au area. Her foremost student that she helped into University ended up being a selfish jerk after exposure to the Student Union there and Windows NT. Take your pick at which did the damage. :) May 11 12:13:10 baird: d'oh. May 11 12:14:05 I had a TA for a CS clas who was blind. She was the most astonishing programmer I've ever met. May 11 12:14:56 She used to show off by logging into her account *without* a reader, writing a fairly complicated program, compiling it, and knowing for a fact that it just worked. May 11 12:14:57 :-) May 11 12:15:26 Reader's gotta help memory May 11 12:15:42 sparq: Im evil I would have changed her keymap to dvorak May 11 12:15:43 And then she'd say, "Now go fix your silly little syntax error." May 11 12:16:11 XorA: Tried that. She noticed the console beeps were wrong. May 11 12:16:31 impressive May 11 12:16:35 sparq - Yeah. I use my memorirs for this stuff too. When I'm into a bigger programming project, I now nearly every line of the code. May 11 12:17:21 Elrond: Yep. I'm starting to get that with my textbooks, since I'm preparing for the comp exams. May 11 12:18:05 It's rather terrifying what the human mind can hold onto. May 11 12:18:36 anyway, this human mind had better get some sleep May 11 12:18:54 well i'm drinking and preparing for mine, i'm doing a management module this year, the easiest thing ever. good luck sparq May 11 12:19:32 exams, time to get drunk :-D May 11 12:19:45 Ideally not before sitting down. May 11 12:19:52 revision always goes quicker if you don't remember most of it May 11 12:20:12 XorA: Quantum Electrodynamics only makes sense when drunk. May 11 12:20:22 sparq: as does AI May 11 12:20:30 * sparq shudders May 11 12:20:37 yes, i'd agree with that XorA May 11 12:20:49 webjames: thanks! May 11 12:20:58 * sparq -> sleep May 11 12:21:54 QED makes me want to go back to classical elements. May 11 12:22:49 I think the quantum mecanics courses were ok, but electrodynamics sucked :P May 11 12:32:42 at first week of April it could have been something like 40% for 3-7 days, 40% for 1-5 weeks, 20% for later May 11 12:35:38 then as time goes by without any official info keep counters but adjust propabilities May 11 12:36:18 or change it to 'soon', 100% May 11 12:39:32 ;-) May 11 12:44:12 'soon' already passed ;) May 11 12:46:23 screen locking idea: unlock is gesture (nothing new here yet), new idea is that shape and timing must be pretty accurate May 11 12:48:08 a bit like a signature; if you fail, then there would be less accurate unlocking gesture (like draw circle roughly) May 11 12:48:50 this would activate program that asks now PIN or password May 11 12:50:03 summary: signature gesture would unlock; simple gesture + PIN would also unlock May 11 12:50:23 i would choose to shutdown the phone - wrong gesture = i'm drunk May 11 12:51:56 ynezz, yeah man, drunk phone calls killer! who would have thought so much bad could happen May 11 12:52:13 and trying to enter pin on party after 15 beers isn't considered as a good idea also May 11 12:52:34 3 drunken attempts = phone locked May 11 12:52:44 till morning May 11 12:53:03 rather till tuesday :p May 11 12:54:06 either that or 'drunk mode' is enabled and before each call a automatic disclaimer is automatically read out saying your are drunk and not responsible for your own actions and continue the call at your own risk May 11 12:54:59 yea, key feature missing in every phone :) May 11 12:55:48 maybe also something bluetooth perhaps that undoes the effect of beer-goggles May 11 13:06:14 this is why i want this phone! so many people with good ideas! and you even could implement it if you want! drunk mode.... :D May 11 13:24:52 Drunk mode really needs the alcohol sensor though. May 11 13:28:38 koen: hi? May 11 13:29:21 hey bmidgley_ May 11 13:29:44 it sounds like you are building your own images May 11 13:30:02 Is the Neo1973 available yet? May 11 13:30:14 No. May 11 13:30:17 Maybe week. May 11 13:30:31 Ah darn... May 11 13:30:55 bmidgley_: yes, I got tired of waiting a week before bugs got fixed in the openmoko version of the buildsystem May 11 13:31:17 are you publishing them? or are you just building from svn trunk? May 11 13:31:52 bmidgley_: http://www.angstrom-distribution.org/unstable/images/fic-gta01/ May 11 13:33:16 koen: will it work ok on P1? May 11 13:33:24 it should May 11 13:34:00 (I didn't respond in time to get a P0 so I guess they put me at the head of the line for P1) May 11 13:34:36 koen: why both zImage and uImage? May 11 13:38:49 bmidgley_: for people playing with kexec May 11 13:49:21 (script) wiki RecentChanges: [[Buying_Interest_List]] [[User:Danil]] [[Bluetooth_Support]] [[Wishlist:Expansion_Back]] [[Bootloader]] May 11 14:00:25 I asked this yesterday but had to leave before I saw an answer. With the case closed, is there space for a daughter board plugged into the JTAG socket? May 11 14:01:04 * DukeOfURL looking at the expansion back page May 11 14:03:00 DukeOfURL: not really May 11 14:05:01 DukeOfURL: if you expanded back cover, jtag cable easilly get into battery compartment though May 11 14:05:17 AIUI - you might get a tiny board in there. 5*20*2mm or so. May 11 14:05:39 JTAG is less interesting. May 11 14:05:51 I2C and USB are interesting for that thoguh. May 11 14:06:10 if you stacked then there is a vioid beside battery May 11 14:06:14 just remove the bluetooth baseboard May 11 14:06:22 are the data and io busses accessible on the JTAG? May 11 14:06:23 that gives you usb and power May 11 14:06:39 True. May 11 14:06:55 The nicest solution would have been a ~10 pin connector next to the battery. May 11 14:06:57 But... May 11 14:07:06 So that the case doesn't need opened. May 11 14:07:12 SpeedEvil: its called the microSD slow :-) May 11 14:07:28 SpeedEvil: that is haning on GPIO so you could bitbang SPI or I2C May 11 14:08:16 VCC, GND , USB A, USB B, +3.3V, I2C, I2C May 11 14:08:18 Or something May 11 14:10:09 You can even hang USB ports off SPI/I2C. May 11 14:10:17 There are cheap host chips. May 11 14:11:40 * SpeedEvil hopes the first batch is >300. May 11 14:14:16 well, if they did 200 last time, I'd think this batch would be noticably larger (though probably on the same order of magnitude) May 11 14:15:50 They moved it to china from .tw so we could guess, that the first batch is larger. May 11 14:16:53 koen: are you able to make calls with your image? May 11 14:19:17 * aloril2 *guesses* > 1000 May 11 14:20:28 aloril - Yeah, I had guess for 1000 on the first run, right. May 11 14:23:24 guessed over month ago that actual sales during first 1 (or 2) week(s) would be > 10x byuing interest count May 11 14:23:27 Hi guys, I have some questions about openmoko May 11 14:24:13 I believe it is based on ebian and you can install additional software with apt May 11 14:24:20 well just ask em May 11 14:24:33 N`Kr1pt - No debian. May 11 14:24:49 (also I will probably try debian/arm on it. ;o) ) May 11 14:25:28 N`Kr1pt: it uses ipkg which is similar to apt, this about as far as based on debian goes May 11 14:25:59 And OpenEmbedded for build system May 11 14:26:40 is it possible to write your own applications, and in wat language? May 11 14:26:40 N`Kr1pt: 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.) May 11 14:27:07 N`Kr1pt: yes, and in any language you feel at home in May 11 14:27:12 N`Kr1pt - In any langua that works on Linux/arm. May 11 14:27:27 N`Kr1pt: its an open phone so its no different development wise than your desktop linux machine May 11 14:27:40 about screen unlocking: there could be 3 modes: 1) signature recognized -> unlock; 2) signature recognized as drunken signature -> drunken mode; 3) signature wrong -> ask simple gesture + PIN May 11 14:27:41 very very nice May 11 14:28:13 aloril - *LOL* I like (2) ;o) May 11 14:29:04 aloril - drunken mode should also be triggered for "mic receives user singing horrible songs" ;o) May 11 14:29:12 aloril2: I think being able to SHOUT taxi at the phone in a drunken stupour and have it phone and give directions to a taxi for you would work for 2) May 11 14:29:52 "*censored* *Censored* Neo, I need a Taxi, NOW" ;o) May 11 14:30:41 "Warning girl you are dragging home may not look like Pamela Anderson in the morning" May 11 14:32:04 only one more question: ipkg knwos the ;deb format now, so you can write your software, make a deb package of it en install it on your phone through ipkg; simple as that? May 11 14:32:54 N`Kr1pt: pretty much May 11 14:33:15 kewl, I'm definitely going to but the Neo 1973 as soon as it's available :) May 11 14:34:04 Oh dear... somebody just added this image to their email about an upcoming deadline... http://despair.com/ambition.html May 11 14:35:30 N`Kr1pt - you might have to adapt some paths, as the Neo-FS-layout is a bit different from stock debian. May 11 14:35:45 yea I see, better to make a .ipk package May 11 14:35:50 but ok, the idea is the same May 11 14:35:50 s/Neo/openmoko May 11 14:36:15 N`Kr1pt: see www.openembedded.org for build system May 11 14:36:35 so OpenMoko doesn't use FHS then? May 11 14:36:44 thx for link May 11 14:37:16 N`Kr1pt - Well, I have't yet investigated all of it, but some parts do not look like Debian's intepretation of the FHS. At least to me. May 11 14:37:33 YMMV. May 11 14:38:29 Some thing are in /var on debian, openmoko has them in /usr. May 11 14:39:04 * Stephmw hands N`Kr1pt and Elrond some hefty symlinking May 11 14:39:14 lol May 11 14:39:20 Stephmw - Thanks. :-) May 11 14:41:00 I really hopes it hits the stores on september 11th in europe ( more specific belgium/netherlands), but I read somewhere that operators were being difficult about the openness of the phone May 11 14:41:00 Stephmw - I'll see for sure, when I have a neo :-) May 11 14:41:20 but the GSM/UMTS stuff is closed? are you expecting delays because of that? May 11 14:41:47 I at least do not expect delays due to GSM issues. May 11 14:42:05 N`Kr1pt - Note: No UMTS on the current edition. May 11 14:42:32 hmz too bad, but ok, UMTS is far from standard in Europe anyway May 11 14:43:28 dome operators even seem to skip UMTS and upgrade their sites to HSDPA instead May 11 14:43:38 N`Kr1pt - The GSM system is far away from the CPU (only connected via serial) and pretty standad chipset. So no real problems there. (except for someallish ones which will be fixed soon / are fixed already) May 11 14:44:53 "HSDPA is part of the UMTS standards from release 5 onwards" May 11 14:45:24 yea HSDPA/HSUPA is actually an improvement to UMTS ( some kind of patch if you like) May 11 14:45:32 yes May 11 14:45:39 a tad like EDGE/GPRS May 11 14:45:44 conceptually May 11 14:46:54 <-- heads home. May 11 14:47:11 are there plans to put in a hard disk in future editions? liek the iPhone? May 11 14:47:46 I think that would be silly. Don't have inside plans though. May 11 14:48:15 silly? I think it would be great if you couls store some music:vids on the phone May 11 14:48:55 I plan to store some vids on the phone. Not much at once though admittedly. May 11 14:49:21 anyway, hard disks add malfunctions and slurp power May 11 14:49:25 * koen ponders about putting the bv3 board on a new case and use it as a gps May 11 14:49:28 some family guy episodes for the train rides to work ^^ May 11 14:49:30 true May 11 14:49:40 flash cards are getting bigger all the time May 11 14:49:57 ok some kind of static memory woud be better indeed May 11 14:50:27 neo's been tested with a 2 gig microsd, enough for a couple of eps surely :] May 11 14:51:03 though of course it would be nice if one didn't have to swap them around, but hd, still, I dunno May 11 14:51:32 no flash is better, you're right May 11 14:51:39 shock resistancy and all that May 11 14:51:57 faster... May 11 14:57:12 or get battery powered USB hub + battery powered 160GB harddisk and you have plenty of space May 11 14:58:14 reminds me again of those RSN-coming wifi/bt hard drives, which would be a saner choice May 11 14:59:03 mjr: yeah May 11 15:01:41 koen: my 1gb card works on the openmoko May 1 image but not yours May 11 15:02:08 when trying mount... /dev/mmcblk0p1: Input/output error May 11 15:08:33 I'm going home now May 11 15:08:34 bb May 11 15:11:14 tag May 11 15:13:15 * aloril2 guesses Monday is now earliest date for news May 11 15:13:34 unless the core devs want to spend the weekend working, yes :) May 11 15:13:49 or the core team in general May 11 15:13:53 and sean especially May 11 15:14:40 sean works 28/7 anyway :-) May 11 15:15:02 :) May 11 15:15:40 i guess this whole project is more or less his dream May 11 15:15:47 and most of the devs also May 11 15:16:05 we too *startsdreaming* May 11 15:22:25 * lrak is wondering why openmoko-devel is closed May 11 15:23:36 lrak: because we plot against you May 11 15:25:52 XorA: that is an honour ;) May 11 15:33:28 XorA: beg your understanding .. I have been following this project for a while May 11 15:33:55 lrak: ? May 11 15:35:09 XorA: I am becoming frustrated because of the huge (to me) proprietary stuff .. and I wanted to knock on openmoko-devel to be reassured .. but it seems that the gps stuff is a lost cause May 11 15:35:56 lrak: it is easy to dissamble the GPS module May 11 15:36:05 lrak: I cant comment on GPS stuff, but that was always advertised, what other huge stuff is there? May 11 15:46:14 w transmitter at lirc.org "http://www.lirc.org/transmitters.html" it is the transmitter at the very bottom of the page, that was submitted by Petteri Aimonen. It is small, compact, simple, robust, and the only connections that need be made to the main board are a 5v pwr (I think we can steal that from the CPU power supply??), ground, and an input pin (which I think we can just steal an unused GPIO pin from the CPU). May 11 15:46:21 2. LIRC: The LIRC.org project has already done all the software legwork for interfacing with an IR device, as well as creating customizable configuration files for a whole slew of remote controlled devices, that can be selected on the fly. This will make a perfect software back end! May 11 15:46:26 3. Device driver: Obviously we need a driver module to be able to talk to the device. This is by far the part I need the most help with. I have NO idea how to produce this part. I assume all we would need is a driver that will interface with LIRC, communicate on the appropriate GPIO pin, and be sure that there isnt any voltage (data) on the GPIO pin when the IR is not being used, so we dont end up with power drain. (Shameless call for HELP!) May 11 15:46:46 wonderful, sorry for the flood. didn't mean to do that :< May 11 15:48:21 (script) wiki RecentChanges: [[Wish_List_UserProfile_Employee]] [[Wish_List_UserProfile_Child]] [[Wish_List]] [[Buying_Interest_List]] May 11 15:56:48 gtg, cu May 11 16:08:05 lrak: do you know that kernel is untainted? May 11 16:09:01 aloril2: was not aware .. but I truly had great excpectations regarding the GPS May 11 16:10:05 closed gps part is library: we have complete control over its input/output -> should lead to deciphering of its protocol quite soon May 11 16:10:50 lrak - If you have followed the project for a while, the gps issue was known from the start May 11 16:11:06 and once we know protocol we can s/cloed library/some existing GPL library/ May 11 16:11:21 (and Morgaine already noticed, that it is one of the things making the phone "not fully free".) May 11 16:11:27 s:cloed:closed: May 11 16:11:52 aloril2 - Are there any gpl-gps-libaries out there? May 11 16:13:18 considering OpenMoko Linux distribution, GPS is only non free part (and propably not long is my guess) May 11 16:13:46 Elrond: at least one, maybe more May 11 16:14:28 ie.. they calculate positions from raw satellite data May 11 16:14:39 (doppler, etc..) May 11 16:15:02 what was the DM2 thing? i understood it was one more non-free part? May 11 16:15:14 aloril2 - URLs? May 11 16:16:33 counter May 11 16:16:33 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) P1: 3 days 09:51:42 (3.411 +-3.4 days) (1145;199) May 11 16:16:36 GSM (voice calls, GPRS, etc.): its hardware modem (*not* softmodem) and you communicate with it using standard and documented AT commands May 11 16:22:44 blarog-kun: first guess uncommitted code? May 11 16:25:36 Elrond: using limited shell in mobile phone 16x32 : you propably can google it faster May 11 16:26:07 (32x16 chars) May 11 16:27:12 aloril2: ah, okay May 11 16:31:28 Elrond: http://gps.psas.pdx.edu/ May 11 16:31:56 hm is there yet a programm to read the simcard? May 11 16:34:08 ah okay May 11 16:37:21 (script) openmoko-devel: Werner Almesberger Re: Home-Brew IR port. Was - Please Ignore RS232 message. The May 11 16:40:00 aloril2 - Thanks. May 11 16:40:07 woglinde - What happened to your charger? May 11 16:40:29 elrond let me test it right now May 11 16:40:44 hm there are usbports on computers that dont charge it too May 11 16:40:55 Huh, strange. May 11 16:41:04 yeah May 11 16:41:11 Make sure to use the same procedure for testing it and an usb port. May 11 16:41:50 It's maybe not a bug of the usb port, but of the way the neo notices "Oh, I have power, let's charge" or in some cases not notices it. May 11 16:42:16 *sigh* May 11 16:42:25 i need tab key and strg May 11 16:42:45 Huh? May 11 16:43:08 whats the tab and strg key on the keyboard? May 11 16:43:14 virtual May 11 16:43:23 strg? May 11 16:43:28 ctrl May 11 16:43:30 hehe May 11 16:43:31 yes May 11 16:43:33 sorry May 11 16:43:33 ah May 11 16:43:52 control in german is "steuerung" -> strg May 11 16:44:24 SpeedEvil - This channel is undermined by the "german connection" ;o) May 11 16:45:22 is it really strg on german keyb? :D May 11 16:45:29 yes May 11 16:45:39 guaqua - Yes. May 11 16:45:42 I have here an original german digital keyboard May 11 16:45:53 came with an alphastation May 11 16:45:57 Yeah, even my Model M has "Strg". May 11 16:46:18 hm May 11 16:46:20 okay May 11 16:46:25 than I will use ssh May 11 16:46:59 Use BT and bnep ;o) May 11 16:47:41 hm dont know if I will spend the 60 euros for bt keyboard May 11 16:48:08 woglinde - BNEP is BT-networking, ssh on top of that. :) May 11 16:48:19 hm yes May 11 16:48:30 but usbnet works too May 11 16:48:47 Yes, but you can't test the charger then ;) May 11 16:48:49 hm is the charger getting warm when its connect to power? May 11 16:48:57 elrond yes May 11 16:49:11 but I need to figure out the power states in proc May 11 16:49:19 and with out tabkey May 11 16:49:50 I should buy an multimeter May 11 16:50:21 We have a precision multimeter in the lab. ;o) May 11 16:51:55 we should have Hlnt in finnish keyboards May 11 16:57:28 hm and my gsm istn working May 11 17:03:12 he zecke May 11 17:16:10 zod: and how is 'Alt' called? May 11 17:17:01 loufoque: Alt May 11 17:17:44 loufoque: but there's also second modifier called Alt Gr May 11 17:19:04 Alt Gr is available everywhere AFAIK May 11 17:19:14 loufoque: nope May 11 17:19:25 us english doesn't have it May 11 17:19:52 wikipedia says it has May 11 17:20:48 http://en.wikipedia.org/wiki/Keyboard_layout#US May 11 17:21:00 us-international has it, the refular one doesn't have May 11 17:21:09 I thought US and US-international were the same thing May 11 17:21:11 oh ok May 11 17:21:21 (script) openmoko-devel: Werner Almesberger Re: Home-Brew IR port. Was - Please Ignore RS232 message. The May 11 17:25:05 hey all. May 11 17:25:21 Anybody have a clue why we have portmap running in openmoko? May 11 17:25:22 he phillipe May 11 17:25:29 hehe May 11 17:25:34 hi woglinde May 11 17:25:48 task from oe I guess May 11 17:26:51 my guess is that it should not be there. It is one of the annoying services I always disable on my desktop installs. Together with that exim4/postfix/favourite email server that most distros install by default May 11 17:27:10 hm May 11 17:27:16 maybe it's for nfs May 11 17:27:27 yeah May 11 17:27:41 if you dont want to stress your flash May 11 17:27:49 start applications from nfs May 11 17:27:49 ynezz: just my thought... May 11 17:28:50 however nfs will does not seem to be running. Anyway I guess on production it should not be there... up to bugzilla :-) May 11 17:29:26 well, someone might run nfs over gprs :) May 11 17:29:51 ynezz: nfs is already painful enough on a normal network... May 11 17:30:07 ynezz: oh your are talking about networked SM :-) May 11 17:30:49 i think people do use nfs on openmoko May 11 17:31:46 no-one tried coda yet? May 11 17:31:58 and there's no reason to not have it supported on production too, as a way for sync'ing files to desktop May 11 17:32:19 balrog-kun: for development it makes sense. But no need to have it running by default. Seems it roughly eats up 400K of memory May 11 17:32:36 hmm, that's a lot May 11 17:33:11 Greetings! May 11 17:33:23 So, how is everyone (who has one) finding G4? May 11 17:34:00 balrog-kun: memfree with portmap running 68328 without portmap 68728 May 11 17:34:11 and about 20k change in buffers. May 11 17:34:29 * noidd makes a redundant comment May 11 17:35:09 And for file sync it makes more sense to have a ftp/http/rsync/ssh or dedicated solution May 11 17:35:52 but for testing May 11 17:35:58 ssh via usb is okay May 11 17:36:01 aeh nfs May 11 17:36:03 sorry May 11 17:38:34 woglinde: you don't like nfs also? I personally hate it. Too much file corruption and mysterious failures. May 11 17:39:05 you should never see file corrupption on nfs May 11 17:40:35 noidd: *should* May 11 17:42:33 never had any problems with it May 11 17:43:14 philippe no I like it May 11 17:43:23 especial nfsroot May 11 17:44:32 woglinde: true nfsroot is nice. May 11 17:46:16 philippe - nfs on AIX is quite good. May 11 17:48:21 (script) openmoko-devel: Werner Almesberger Re: Home-Brew IR port. Was - Please Ignore RS232 message. The May 11 17:50:21 (script) wiki RecentChanges: [[Buying_Interest_List]] May 11 18:02:42 counter? May 11 18:02:43 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) source: http://wiki.openmoko.org/wiki/MokoMakefile ; P0: http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review; 3 days 08:58:37 (3.374 +-3.4 days) for devices for *ANYBODY* for $350 (targeting developers) (preorders: not yet) (2007-05-13); 4 months, 4 days (127.748 +-15.0 days) for mass market (2007-09-15): see topic for more info (1146;199) May 11 18:06:30 kiney - You can use "counter" without the "?" to get a shorter form. May 11 18:06:43 I wanted the longer form May 11 18:08:16 You can also counter? in a query May 11 18:12:29 /counter should be made a network wide command :) May 11 18:12:41 *G* May 11 18:12:49 Not really. It's too inaccurate. May 11 18:13:08 it's enough that it spits out a lot of numbers May 11 18:29:59 hey all, so what the latest exciting news? May 11 18:33:34 nothing new, did i hear with monday we would hear some news on P1? May 11 18:34:26 Dunno. May 11 18:34:30 Where did you hear it? May 11 18:34:42 kiney: this works too: /msg aloril counter? May 11 18:35:07 i'm sure someone said it on here May 11 18:35:50 webjanes: I would say that Monday is lower limit for possible news at some point in next week May 11 18:36:09 excellent! May 11 18:36:54 Of course its possible that there are no news next week either .. May 11 18:37:27 hi May 11 18:37:29 counter May 11 18:37:30 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) P1: 3 days 08:41:14 (3.362 +-3.4 days) (1147;200) May 11 18:37:31 lets hope what is possible is not probable May 11 18:38:49 It's possible that the neo1973 is merely a cunningly executed fake. May 11 18:39:10 And that the web store due to pop up next week is really simply to take 500 orders, then run off with the money;. May 11 18:39:19 * aloril2 is just guessing May 11 18:39:28 * SpeedEvil too. May 11 18:39:43 I'd be surprised at no news next week. May 11 18:39:46 SpeedEvil, it is entirely possible May 11 18:40:24 SpeedEvil, here's to hoping for news monday *raises glass to freedom* May 11 18:41:03 current guess is that there will be news once they have tested this run May 11 18:41:22 which run is that? May 11 18:41:38 would that be what i've just compiled into qemu? May 11 18:41:42 SpeedEvil - Well, I'll let you guys fakebuy the first 500. ;o) May 11 18:41:59 webjames: the run from the new factory May 11 18:42:15 Elrond, you're not fooling anyone, we all know you're in for a 20% cut May 11 18:42:15 webjames: my p0 and p1 were produced in taiwain, the new plant is in mainland china May 11 18:42:56 koen, thank you, i did hear about the move. so once they've tested the new P1 from china then we will see some action May 11 18:43:17 webjames - How did you know? ;o) May 11 18:43:44 webjames: that's what I suspect May 11 18:44:11 koen - BTW: On the twin pic, which is the old, which the new? May 11 18:44:23 koen, excellent May 11 18:44:30 the one with the penmarks in the FIC logo is the old one May 11 18:45:26 koen, do they physically look any different? May 11 18:45:40 webjames: after I added the penmarks they did :) May 11 18:45:54 otherwise they are identical on the outside May 11 18:46:06 SpeedEvil: somebody posted link to some web store that has Neo1973 on sale for about $500; image was very old mockup one (green Neo + mapping software) May 11 18:46:15 koen, that's a no then. but we all know what's important is what's in the inside! ;) May 11 18:46:44 elrond, got a cut in that to?! ;P May 11 18:46:53 * koen is still hoping for the UWB demo to happen May 11 18:47:11 koen, what's that? May 11 18:47:56 webjames: swapping the bluetooth board for a UWB board May 11 18:49:02 is UWB: Ultra-wide-band May 11 18:49:28 koen - So the different coloring is really only due to light and viewing angle? May 11 18:49:59 yes, the bottom one is in direct light, the top one isn't May 11 18:50:10 Ahh. May 11 18:50:25 UWB appears in IEEE 802.15.4a draft PAN standard May 11 18:50:41 I've read in bugzilla, that they planned to change the AUX button somewhat to not be so much indented into the case. Did they fix that? May 11 18:50:48 UWB rocks May 11 18:51:02 Elrond: p0 and p1 have identical buttons May 11 18:51:15 dang May 11 18:51:20 it does seem to, however for all those other non rockin' devices that use bluetooth would be useless May 11 18:51:31 I wish I could buy one now and rush-order. The LCD panel on my phone just died ;P May 11 18:51:49 z May 11 18:51:56 bummer :( May 11 18:51:58 'lo koen May 11 18:52:03 hey noidd May 11 18:52:05 paroneayea, on your neo? or something else? May 11 18:52:10 hows G4 working for you? May 11 18:52:23 webjames: on my normal old generic cellphone May 11 18:52:36 but since I'm gonna buy a new cell phone May 11 18:52:45 I wish it had died right when you could order 'em May 11 18:53:01 paroneayea, what i've done is get a contract with a nice expensive phone, sold it and it pays for the contract for the year May 11 18:53:19 leaving me with some cash to get a neo when it comes May 11 18:53:29 funny that you can't make or receive calls with the installed p1 firmware May 11 18:53:51 koen - Still not?! May 11 18:53:51 it's also funny that the gps test app kills the dial and gsmd May 11 18:54:07 Yuck. May 11 18:54:15 Elrond: I am able to make calls with the images I build myself, that don't use the svn overlay May 11 18:54:47 koen - Oh well. And the images from buildhost are also borked? May 11 18:54:55 dunno May 11 18:55:05 using the arm OABI is so last year May 11 18:55:19 Offtopic, but i'm buying a new computer should i get a e6700 cpu or wait for the amd true dual core? May 11 18:57:54 koen - You know by any chance, what debian/arm uses as abi? May 11 18:58:29 Elrond: they are switching to EABI after they stop being utter idiots May 11 18:58:34 koen: I can't get mine to stay powered on for more than 3 mins. May 11 18:58:49 I probably need to hit uboot to increase the charging rate againb May 11 19:30:44 counter? May 11 19:30:45 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) source: http://wiki.openmoko.org/wiki/MokoMakefile ; P0: http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review; 3 days 08:14:36 (3.343 +-3.3 days) for devices for *ANYBODY* for $350 (targeting developers) (preorders: not yet) (2007-05-13); 4 months, 4 days (127.687 +-15.0 days) for mass market (2007-09-15): see topic for more info (1148;200) May 11 20:20:46 anyone using a bluetooth keyboard with the neo1973 already? May 11 20:21:41 if not, are there pieces missing at the moment? which ones? May 11 20:21:54 йниенде Чsomeone uses neo already?! May 11 20:22:37 thresh: There's about 40 ppl having Neo's, but early beta versions May 11 20:22:38 someone does May 11 20:24:09 that was a joke :) May 11 20:24:58 Heh, sorry.. I guess I should go stand in shame in the corner :P May 11 20:25:56 :) May 11 20:33:15 CM - Some already have a P1. May 11 20:33:25 * ossman waves his May 11 20:33:26 :) May 11 20:34:50 speaking of which, is this the proper development channel? or mostly people with a general interest? May 11 20:34:50 ossman: 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.) May 11 20:35:56 ossman: there's a separate development channel May 11 20:36:01 it's invite-only May 11 20:36:28 balrog-kun, oh. who do I poke? May 11 20:36:50 ossman - Just poke around. Many can help you without being "official developers". May 11 20:37:13 ossman: not sure, i guess one of the core people (they're here too) May 11 20:38:14 what's the primary makeup of people here? developers or users? May 11 20:38:32 I'd say half/half. May 11 20:38:35 i'm sure there's nothing interesting going on there ;) May 11 20:39:36 balrog-kun - You mean in #*-devel? My guess is, they're planning "what to do at the next evening". ;o) May 11 20:40:06 as in where to go drinking (the german gang)? May 11 20:40:22 code is best produced hammered May 11 20:41:08 balrog-kun: drinking, where? ;) May 11 20:41:22 * stefan_schmidt is watching :) May 11 20:44:17 balrog-kun - As in "trying to acquire the world dominion" (I'm trying to translate "Pinky and Brain"-animated cartoons from local language to english) ;o) May 11 20:47:06 oh, heh May 11 20:47:54 balrog-kun - Do you know Pinky and Brain? May 11 20:48:24 Elrond: "trying to TAKE OVER THE WORLD!" May 11 20:48:51 relatedly: Animaniacs is now availalbe on DVD, I hear. May 11 20:49:36 pjz - Ahh, good to know the real english version now. May 11 20:49:52 Sounds like a 'must buy' May 11 20:51:34 Elrond: yea, i knew what phrase you meant but i had never watched the english version either May 11 20:52:12 balrog-kun - Ahh. :) May 11 21:25:03 any debug board owners here? May 11 21:44:32 ossman - Why do you need someone with a debug board? May 11 21:44:48 experience and help getting started :) May 11 21:47:33 ossman - You have a debug board?! May 11 21:48:38 indeed. making me do kernel work without one would be very evil ;) May 11 21:48:49 *g* May 11 21:50:03 ossman - Well, what I heard: The FPC seems to only work 20 times or so. So if you have connected the debug board, keep it connected for a while. ;-) May 11 21:50:30 delightful May 11 21:51:21 (script) openmoko-devel: Werner Almesberger Re: Home-Brew IR port. Was - Please Ignore RS232 message. The May 11 21:54:23 ossman - What part of the kernel do you intend to work on? May 11 21:54:32 Elrond, sd reader May 11 21:55:23 Huh, I assumed, that it worked? May 11 21:55:56 but it's slow May 11 21:56:22 Ahh. May 11 21:56:22 and issues might show up as things progress May 11 21:56:51 :-) May 11 21:57:23 Have you been specifically been elected or did you ask to work on this? May 11 21:59:20 well, I wrote most of the code, so a little of both :) May 11 21:59:29 hehe, okay. :-)) May 11 22:00:47 incidentally, IIRC there were issues in writing SD card reader drivers some while back relating to non-disclosure and SD DRM features I guess, that's over with or? May 11 22:01:29 yes and no May 11 22:01:39 we figured out the protocol without signing the nda May 11 22:01:52 we still have no idea how the drm bits work though May 11 22:01:56 righto May 11 22:02:12 well, most here don't probably care about the drm bits May 11 22:02:20 btw, the protocol is available without nda these days May 11 22:02:20 as long as it's a usable block device :) May 11 22:02:34 indeed May 11 22:02:56 From a very high point of view, how is that DRM stuff intended to work? May 11 22:02:58 now I just need to get my hands on one of these 4 GB cards May 11 22:03:43 Elrond, I think it's similar to how TPMs work. you provide them with a hash that identifies the machine. if that hash matches the one that stored the data, the card will allow you to read it May 11 22:04:15 I have seen exactly zero products that support it though May 11 22:04:24 so it's just a waste of bit May 11 22:04:26 +s May 11 22:04:36 Ahh. May 11 22:05:56 I never tried to understand, how the TPM thing works, when you want to update somethint -- say your kernel. But that's offtopic. :) May 11 22:11:29 ossman: Nice to see sd maintainer hacking on the neo. May 11 22:11:37 ossman: So http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=93 is for you? May 11 22:12:03 indeed May 11 22:12:12 ossman: I just dropped a mail to support@sandisk.com for available microSDHC cards May 11 22:12:13 should work, but better safe than sorry :) May 11 22:12:22 exactly May 11 22:12:24 http://shopping.bit-tech.net/UK/product/35029443/SanDisk_SanDisk_microSDHC_4GB_CardPRE_ORDER_ONLY_/ May 11 22:12:47 I have a miniSDHC, but no micro May 11 22:12:54 you should have gone with a bigger slot ;) May 11 22:13:09 ossman: real SD would rock May 11 22:13:20 ossman: Perhaps next model May 11 22:13:32 by then we'll have SDIO ;) May 11 22:13:39 ossman: I still don't get why this card is not available yet May 11 22:14:10 ossman: We should have microSDIO anyway, right? May 11 22:14:13 probably takes time to get good yields May 11 22:14:21 perhaps May 11 22:14:30 I don't think SDIO exist for anything but fullsize May 11 22:15:14 stefan_schmidt, btw, have you frozen kernel version or are you tracking linus? May 11 22:15:28 ossman: tracking releases atm May 11 22:15:34 ossman: .21 May 11 22:15:49 rcs aswell? May 11 22:15:54 ossman: today some alsa stuff got merged and Ben is preparing more for .22 May 11 22:16:06 ossman: No, not atm May 11 22:16:25 there are a whole bunch of suspend fixes in Linus' HEAD right now May 11 22:16:30 for mmc/sd that is May 11 22:16:30 ossman: Sometimes small patch backports. We did it for the SDHC patch in .20 May 11 22:16:59 ossman: hmm, easy to backport? May 11 22:17:19 there are a lot of commits unfortunately May 11 22:17:34 if you don't care about history, then you can just copy over the mmc dirs May 11 22:17:52 ossman: If you think it's worth you could suggest it to openmoko-kernel ml May 11 22:18:13 ossman: nah, we like clean patches May 11 22:18:15 :) May 11 22:19:15 ossman: We could add recipes for the -rcX kernels, but I fear Harald is overloaded with so much stuff that this would not help much May 11 22:19:44 No. Please stick to released stable kernels. Please. May 11 22:20:08 stefan_schmidt, not delegating as much as he perhaps should? May 11 22:20:09 Elrond: Not for the default image May 11 22:20:33 ossman: That is one point. The problem of the perfectionists ;) May 11 22:20:37 btw, how do you update the thing? May 11 22:21:13 ossman: But we have only two reak kernel hacker working on it or money. Harald and Werner May 11 22:21:18 ossman: Update what? May 11 22:21:35 ossman - dfu. May 11 22:21:40 kernel, apps, libs, etc May 11 22:21:52 ossman: apps, whatever. May 11 22:21:56 ossman: ah, as Elrond say. DFU May 11 22:22:02 ssh, or some mechanism not released. May 11 22:22:08 dfu? May 11 22:22:11 Kernel, bootloader - DFU only May 11 22:22:21 USB firmware update standard. May 11 22:22:24 ossman: Device Firmware Update. Hardware independent standard from USB forum May 11 22:22:33 ossman - DFU is a standard for firmware update via usb. u-boot does it. May 11 22:22:34 ossman: Pretty neat but mostly unknown May 11 22:22:37 apps or libs might be done through 'normal' network type. May 11 22:22:38 fancy May 11 22:22:45 ossman - http://wiki.openmoko.org/wiki/Flashing_openmoko May 11 22:23:06 chance of bricking it? May 11 22:23:22 ossman: If you flash the wrong bootloader, yes. May 11 22:23:37 Unless you overwrite u-boot with crap... it's not bricked. May 11 22:23:40 ossman: All other partitions are fine May 11 22:23:53 no plan b in place for that scenario? :) May 11 22:23:54 ossman: And with debug board even that is fine. May 11 22:24:06 ossman - plan b is the debug board. ;o) May 11 22:24:09 Or with cats May 11 22:24:14 ossman: We think about this problem for the mass market launch May 11 22:24:17 lucky I have that then :) May 11 22:24:27 SpeedEvil - cats? May 11 22:24:42 * SpeedEvil was making a silly reference to a current meme. May 11 22:25:20 ossman: btw, what do you think about the sd card holder? May 11 22:25:21 meme? May 11 22:25:31 * stefan_schmidt thinks is purely shit :) May 11 22:25:32 For mass-market, probably going to get more complex. Signed blobs only unless you go through some process to say you have a clue. May 11 22:25:44 SpeedEvil: nah May 11 22:26:07 s/signed/marked-for-this-phone/ May 11 22:26:22 SpeedEvil: In the end it must be something that can not be broken by software. Something really safe. May 11 22:26:36 "I'm a GTA01Bv05 I don't take u-boot for GTA01Bv04" ;) May 11 22:26:43 Yeah May 11 22:26:44 IMO, that doesn't matter. May 11 22:26:50 User gets to break phone. May 11 22:27:03 As long as user can't break phone without actual effort. May 11 22:27:37 stefan_schmidt, no strong opinions about it. full size would of course have been better. as would have an external (from a usage point of view) May 11 22:28:25 ossman: External: Ack, Full size: No opinion. But I fear it is to easy to break May 11 22:28:45 There have already been reports of breakage. May 11 22:28:48 Well - bendage. May 11 22:29:07 Right. May 11 22:29:17 And then there is the whole 'opeerate phone without microSD and it may fry' thing. May 11 22:29:25 come again? May 11 22:29:26 When I get a phone, I need to ask someone with good eyes and good hands to insert the sd :) May 11 22:31:39 stefan_schmidt, how much contact do you have with samsung? May 11 22:31:55 ossman - "come again"? May 11 22:32:37 ossman: hmm, Harald and Sean had many meetings with seem. And the contacts getting more and more senior. Why you ask? May 11 22:32:53 Elrond, as a response to SpeedEvil's claim about frying the phone May 11 22:33:38 ossman: The metal the microSD is made from is very fragile. May 11 22:33:50 stefan_schmidt, I'd prefer it if they submit their drivers to mainline May 11 22:33:54 ossman: if you press on the top of SD slot too hard you short the power on the phone, although I think there is a current limit in there on P1 May 11 22:33:55 There is a reported concern from FIC/or some of their engineers May 11 22:34:05 that it can short out the power supply. May 11 22:34:24 nice May 11 22:34:26 did the vibrator GPIO change place since Bv3? May 11 22:34:38 balrog-kun: yes May 11 22:34:55 stefan_schmidt: thanks, gonna look it up May 11 22:35:22 i noticed the new u-boot doesn't talk to the vibrator in qemu May 11 22:35:30 ossman: We also work with Ben Dooks, so if you have a request I think Harald and Ben should be able to help you a bit perhaps. May 11 22:36:58 ossman - The holder is *censored*. If you want, I can dig up the bug logs about it. May 11 22:37:19 are there other GPIOs that changed place recently? May 11 22:37:33 stefan_schmidt, first I'd just like to know if they have an interest in maintaing a driver in mainline. most of the samsung code in the kernel seems to come from other parties May 11 22:37:51 balrog-kun: Sorry, no idea. Just had vibrator in mind. May 11 22:38:12 Elrond, I get the picture :) May 11 22:38:17 stefan_schmidt: okay, thanks, i will find some ChangeLog May 11 22:38:41 ossman: I work with Samsung to push sound for some stuff to mainline May 11 22:39:22 XorA|gone, will it be you or samsung that maintains it though? May 11 22:39:38 ossman: I think Ben would be the best person for such questions. If we need more pressure some contacts from FIC/OpenMoko could help of course, too. May 11 22:39:47 ossman: me, as they happen to use my companies chips May 11 22:40:42 now I'm confused. the chips are produced by you, sold to samsung, which in turn sells them to fic? May 11 22:41:03 ossman: FIC also use chips from my company May 11 22:41:39 XorA|gone, but how did samsung get into that chain? :) May 11 22:42:49 ossman: sounds runs over I2S or AC97 buses on s3c24XX cpu familly, I maintain the codec drivers and the i2s/ac97 protocol drivers for samsung, Ben Dooks supplied the i2s driver, samsung supplied the ac97 driver to me May 11 22:43:57 ok May 11 22:44:16 stefan_schmidt, I've registered in the bugzilla now in case you want to attach me to some bugs May 11 22:44:46 ossman: ok May 11 22:45:29 ossman: which email? May 11 22:45:46 drzeus-buzilla at drzeus.cx May 11 22:45:51 bugzilla May 11 22:47:58 ossman: Assigned you the SDHC bug May 11 22:48:19 great May 11 22:48:41 btw, what's with the lack of stylus? May 11 22:48:54 ossman: Don't ask :( May 11 22:49:06 :) May 11 22:49:38 ossman: But hey, now you have a 4-feature stylus. ;) May 11 22:50:01 http://wiki.openmoko.org/wiki/Supported_microSD_cards May 11 22:50:09 is this still accurate? May 11 22:50:58 ossman: Should be mostly May 11 22:51:48 ossman: Harald fixed some nasty SD/MMC problem in u-boot I hope that is what the last collum reflects May 11 22:52:02 ossman: Let me check svn which rev this was May 11 22:52:02 balrog-kun - GPIOs: There's a *.h with per hw-revision GPIO naming. May 11 22:52:50 stefan_schmidt, ok. otherwise that might be something I can look at May 11 22:53:14 I can also take a whack at #282 if I get access to a card May 11 22:53:23 ossman: Updates are always welcome for this table. I gues you should have one or two different cards. ;) May 11 22:53:47 only one I'm afraid. I have mostly full size cards May 11 22:54:12 ossman: ah, damn May 11 22:54:54 * stefan_schmidt waits for svn May 11 22:55:43 ossman: I can't make the decision if we could send you a bunch of problematic cards, but I like the idea. May 11 22:56:37 Someone has to make a microSD to SD adaptor don't they? May 11 22:57:16 SpeedEvil: Harald soldered some wire on it once for testing. I would not call this an adapter though ;) May 11 22:57:59 Yeah - i was meaning something with a flex-cable. May 11 22:58:12 hehe. May 11 22:58:30 stefan_schmidt - I might test with the microsd of my dad's phone ;o) May 11 22:59:29 Elrond: Would be good May 11 22:59:40 is battery required to boot? May 11 22:59:45 Yes. May 11 22:59:55 Well - or some low impedance source of 3-4V May 11 23:00:19 but USB is insufficient? May 11 23:00:25 It won't boot off USB May 11 23:00:34 it starts in 'low power' mode - 100mA limit May 11 23:00:37 and that can't power it. May 11 23:01:14 It should be enough to power the CPU, which then could request 500. May 11 23:01:21 True. May 11 23:01:28 But not in the current state of the code. May 11 23:01:38 As it tries to wake the display May 11 23:01:42 ossman: The table seems to be up-to-date after the last u-boot fix May 11 23:02:08 Elrond: thanks May 11 23:02:38 i remembered there were header files with the GPIO wirings but i was looking at u-boot patches instead of kernel now and i was confused because everything was inline May 11 23:02:59 balrog-kun - Ahh. May 11 23:03:07 application-wise, what should be functional? May 11 23:03:14 I haven't understood, why u-boot has code for the vibrator, anyway. May 11 23:03:35 Elrond: it's only for the "neo1973 vibrator (on|off)" command i think May 11 23:03:49 so you can use e.g in scripts etc May 11 23:03:56 Should be able to turn it off on boot. May 11 23:04:01 s/use/\& it/ May 11 23:04:01 balrog-kun meant: so you can \& it e.g in scripts etc May 11 23:04:04 If it can do that, then ... May 11 23:05:08 Ahhh, turning it off at boot is a good idea. May 11 23:05:29 stefan_schmidt, do you know what is needed on the debug boards to get a "normal" serial and jtag? May 11 23:05:34 But wouldn't it just make sense to put all GPIOs to 0 at boot and then the important ones back to 1? May 11 23:05:56 ossman - It's on the wiki about the debug board. May 11 23:06:00 * balrog-kun wonders what the "gta01-led.0" device is that linux probes on boot May 11 23:06:26 ossman - http://wiki.openmoko.org/wiki/Connecting_Neo1973_with_Debug_Board_v2 May 11 23:06:38 Elrond, that isn't entirely the same board they ship out May 11 23:06:44 mine's missing pieces :/ May 11 23:06:53 Huh! May 11 23:07:24 ossman - Then kick someone until the wiki is updated for debug board v*. ;o) May 11 23:07:31 :) May 11 23:07:39 ossman: huh, what is missing? May 11 23:07:48 pin headers at least May 11 23:07:53 that looks like my P0 devboard May 11 23:07:56 ossman: That's normal May 11 23:08:21 I have no doubt about that. I was just wondering if that's all that is missing to turn it into a generic jtag if May 11 23:08:47 ossman: Indeed, for generic JTAG you need the header May 11 23:08:53 ossman: For neo JTAG not May 11 23:09:04 there are some empty pads near that header May 11 23:09:17 * stefan_schmidt fetches his debug board May 11 23:09:27 ossman: Yeah, I think you just need those pads populated. May 11 23:09:53 cjb, argh! I can't hide anywhere! May 11 23:10:30 debug board v2 pic: http://wiki.openmoko.org/wiki/Image:Neo1973_debugboard_v2_top_annotated.jpg May 11 23:10:33 ossman: btw, werner checks the problematic cards he has again with newer software to verify May 11 23:10:42 stefan_schmidt, sounds good May 11 23:11:03 Elrond: the boards we actually received don't have pins on the bottom ports. May 11 23:11:15 ossman: With tracking main perhaps some problems are already gone May 11 23:11:27 cjb, and I wonder what that jumper is doing there May 11 23:12:16 cjb - Ahh, so they sent v2-simple instead of v2.... May 11 23:12:32 ossman: ok, what is missing on your board against the picture? May 11 23:12:46 ossman: Just the two pin header downside? May 11 23:13:26 ossman: If yes, that's totaly fine May 11 23:13:27 stefan_schmidt: the "non-neo1973" ports are bare. May 11 23:13:32 stefan_schmidt, as far as I can tell. but I thought it'd be better to ask someone involved in the process :) May 11 23:13:41 cjb: same here May 11 23:14:12 cjb: They decided that people interested in using it for other devices should be able to solder pin header itself. :) May 11 23:14:29 yeah :) May 11 23:14:30 ossman: So you board is ready May 11 23:14:38 no sweat. I just didn't know if that was it May 11 23:14:59 but the jumper still isn't explained ;) May 11 23:15:01 * stefan_schmidt gets remembered to test this with his EZX phones some day. May 11 23:15:16 ossman: No idea about it, sorry. May 11 23:16:07 stefan_schmidt - Is the FPC as fragile, as some people noted? May 11 23:16:24 Elrond: yes May 11 23:16:38 Elrond: Just be careful :) May 11 23:17:05 I wish FIC had used a somewhat more rigid cable :\ May 11 23:17:12 I had no problems with it yet, but did not used it much on the other hand May 11 23:17:22 CoreDump|home: ACK May 11 23:17:56 stefan_schmidt - I hope not to need it ever :-) May 11 23:18:00 CoreDump|home: It's a question to find something that fit the needed numbers of wires and place on pcb May 11 23:18:06 Elrond: heh May 11 23:18:11 stefan_schmidt: I know =) May 11 23:19:08 I would prefer an external connector for this anyway. As the data connectors from older siemens phones for example. May 11 23:19:24 That would also help people like SpeedEvil to build extensions May 11 23:19:27 So except for the fragile sd-holder (and sim holder), how does the rest of the hardware feel? May 11 23:19:48 Elrond: good May 11 23:20:03 Elrond: bv4 better then bv3 May 11 23:20:12 Good :-) May 11 23:20:44 Connector: Perhaps even something like empty extension cases with preinstalled connector for hardware hacking May 11 23:21:07 With i2c, SPI and GPIO on the connector that would rock May 11 23:21:10 :) May 11 23:21:35 stefan_schmidt - The siemens ones mostly have rs232 on it. May 11 23:21:56 But I think we will see how the neo1973 will sell and how much more phone generation we can produce May 11 23:22:29 [wishlist] a rugged version. ;o) May 11 23:22:34 Elrond: Yeah, I know. Just refer to the type of connector. Something robust on the bottom. May 11 23:22:55 Ahh, yeah. May 11 23:45:30 openmoko: 03andrew * r1948 10/trunk/src/host/qemu-neo1973/ (304 files in 25 dirs): (log message trimmed) May 11 23:45:30 openmoko: Merge changes from cvs.savannah.nongnu.org:/sources/qemu. May 11 23:45:30 openmoko: Merge Paul Brook's I2C/SMBus framework and drop our own I2C fwk. May 11 23:45:30 openmoko: Rewrite interrupt handling to use the new framework. May 11 23:45:30 openmoko: Update MMC host to use the new APIs. Update CPU selection to the new API. May 11 23:45:34 openmoko: Rewrite GPIO handling to use only the IRQ fwk. May 11 23:45:36 openmoko: Convert memory mapping to the Fabrice Bellard's new API. Convert I2C slaves and S3C2410 I2C master to new API. May 11 23:46:57 counter May 11 23:46:57 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a optimistic guess) P1: 3 days 06:06:30 (3.255 +-3.3 days) (1149;201) May 11 23:47:03 bah May 11 23:47:38 * balrog-kun didn't realise that was 304 files... May 11 23:48:00 P1 is coming Real Soon May 11 23:49:21 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Wish_List_UserProfile_Employee]] [[Wish_List_UserProfile_Child]] May 12 00:16:24 BTW: I remember, that the P0 had problems when the battery was completely exhausted. It could not even recharge it. Is that a hardware problem? Or has it been fixed in u-boot? May 12 00:16:45 the latter May 12 00:18:55 So I can use a completely empty battery and charge it using the neo? May 12 00:19:08 s/neo/neo P0 May 12 00:24:56 yeah May 12 00:25:17 assuming you've updated u-boot May 12 00:28:44 Good! May 12 00:44:55 Any news on gsm firmware updates? May 12 00:45:05 (for p0) **** ENDING LOGGING AT Sat May 12 02:59:56 2007