IRC log for #openmoko-cdevel on 20131015

01:10.02*** join/#openmoko-cdevel cmair__ (~chris@ppp-93-104-56-0.dynamic.mnet-online.de)
01:46.34*** join/#openmoko-cdevel prahal (~prahal@ARouen-553-1-153-29.w92-152.abo.wanadoo.fr)
03:00.57*** join/#openmoko-cdevel jonwil (~jonwil@27-33-137-199.static.tpgi.com.au)
04:54.50DocScrutinizer05sounds pretty recent
04:56.27DocScrutinizer05whether it's "stable" or not is another question
05:01.20*** join/#openmoko-cdevel daniele_athome (~daniele_a@net-2-39-114-47.cust.dsl.vodafone.it)
05:32.43pabs3hmm at http://shr-project.org/trac/wiki/Rock-solid%20Farewell%20Release%20%28GTA02%29
05:34.59DocScrutinizer05>>pretty much lost its initial fraction<< hmm indeed
07:35.56*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
07:55.06*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
08:37.07*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
08:43.10*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
08:44.47*** join/#openmoko-cdevel pespin (~pespin@51.pool62-37-71.dynamic.orange.es)
08:45.49*** join/#openmoko-cdevel ao2 (~ao2@2001:1418:117::1)
09:06.39*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
09:16.16*** join/#openmoko-cdevel dos11 (~dos@unaffiliated/dos1)
09:23.20PaulFertserdos11: hi
09:23.40PaulFertserdos11: regarding wifi, do you remember my mail with "tips and tricks" for ar6001?
09:25.12PaulFertserdos11: regarding fso-gpsd, hm, why? Upstream gpsd doesn't save ephemeris, that's true. But in general it seems like a saner solution.
09:26.19dos11PaulFertser: hello, but sorry, have to go
09:26.29dos11I don't rememeber, but I'll check it later
09:27.03dos11fso-gpsd - well, looks like gps doesn't work oob now (at least here), so read that as just "fix gps", either with gpsd or fso-gpsd :)
09:27.21PaulFertserdos11: please ping me if you have any questions.
09:27.35PaulFertserdos11: i have issues with gps myself but i suspect hardware failure
09:27.45PaulFertserdos11: it feel like my antenna has reduced sensitivity :/
09:27.49dos11sure, thanks, cya :)
09:31.41*** join/#openmoko-cdevel jonwil_ (~jonwil@27-33-137-199.static.tpgi.com.au)
09:31.44*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
09:44.01DocScrutinizer05agps broken?
09:44.40DocScrutinizer05we recently seen quite a few messups with agps servers
09:45.36DocScrutinizer05and I'd wonder if and how gta02 is using agps data, and which. I guess ublox no longer supports their supl server
09:45.49DocScrutinizer05s/agps server/SUPL/
09:46.05PaulFertserDocScrutinizer05: not agps, but we do not even have flash to save current ephemeris and almanac on gps module itself, so it's recommended to store them on disk before powering i off.
09:46.28PaulFertserI'm not sure about the supl server, it used to work, but I have no clue what's going on with it now.
09:46.34DocScrutinizer05sure, that's how many gps work
09:47.50PaulFertserThere're dirt-chip mtk3329 devices on the market that have embedded flash, integrated antenna, that are smaller and way more sensitive than our ublox.
09:47.57DocScrutinizer05n900 agps recently borked. not even supl.google.com did work anymore (supl.nokia.com been broken since quite some time), but it turned out it been a expired or messed up cert in N900 cert store
09:48.26PaulFertserDocScrutinizer05: have you seen http://www.u-blox.com/en/positioning-antennas/gnss-antenna-modules/uc530.html ?
09:48.45DocScrutinizer05no, what's exciting about it?
09:49.41PaulFertserVery small and very sensitive at the same time
09:49.49DocScrutinizer05nice
09:50.24PaulFertserTracking: –165 dBm Cold start: –148 dBm
09:50.41DocScrutinizer05I'd be already happy when the Option modules came with embedded LNA
09:50.56PaulFertserHeh
09:51.16DocScrutinizer05building a LNA is kinda PITA
09:51.59PaulFertserI very much hope you'll finally find the way to fit some module in n900 case.
09:52.00DocScrutinizer05a LNA module would actually be sth nice to find
09:52.21DocScrutinizer05ooh, the fit is already solved
09:52.48DocScrutinizer05we go completely bonkers and extend the case ;-P
09:53.41DocScrutinizer05or - maybe - we just move the SIM tray a bit and cut a hole for it
09:54.01DocScrutinizer05and dig a hole into PCB for the Option
09:55.51DocScrutinizer05but we need a LNA to use the intergrated GPS which otherwise is really fine
09:56.13DocScrutinizer05GPS&GLONASS combined
09:56.31DocScrutinizer05just like the uBlox
09:59.29PaulFertserPCB holes are cheap
09:59.45DocScrutinizer05actually I don't get it if uBlox uc530 is glonass and (us)gps combined. And it seems to provide no assisted mode, cold and warm TTFF 31s is a tad long
10:00.16PaulFertserTTFF of 31s is excellent when you have no supl data
10:00.29DocScrutinizer05yes, that's what I say
10:00.42DocScrutinizer05for a phone you want supl
10:01.24DocScrutinizer05when you start an app that's location aware, user hardly wants to wait half a minute for it to warm up
10:02.00DocScrutinizer05here: heat up
10:02.22PaulFertserhttp://www.gtop-tech.com/en/product/AGPS-Assisted-GPS/Software_Services_07.html they say they support mt3339
10:02.38PaulFertserI've never used agps on my gta02 :/
10:03.32DocScrutinizer05ooh, and it doesn't support GLONASS
10:05.47DocScrutinizer05then OTOH 45mW isn't too much
10:06.20PaulFertserGlonass meh, does it work anywhere?
10:06.36PaulFertserI'm not proposing that gps module for neo-n900 anyway :)
10:07.35DocScrutinizer05well, it seems glonass is a nice supplementary to gps
10:08.00DocScrutinizer05and 10mA is quite nice but too much to keep it always on
10:08.01PaulFertserI was just shocked when I saw such a tiny module with integrated antenna.
10:08.12DocScrutinizer05hehe
10:08.31PaulFertserThis one has some autonomous logging facilities.
10:09.27PaulFertserIt's funny how ublox sells chinese MTK devices now :)
10:09.29DocScrutinizer05yep
10:09.42DocScrutinizer05hm, chinese devices?
10:10.05DocScrutinizer05ooh I see
10:10.21DocScrutinizer05you say it's based on mtk3339?
10:11.36PaulFertserYes
10:16.22DocScrutinizer05I don't get it why ublox doesn't support agps
10:16.45PaulFertserIt should support agps, why do you think it doesn't?
10:17.06DocScrutinizer05because of the 31s cold and warm TTFF
10:18.32PaulFertserThat's unassisted case
10:18.35PaulFertserI think
10:19.10DocScrutinizer05yes, that's why I think they don't offer / support A
10:19.34DocScrutinizer05NB they state best case values
10:19.57*** join/#openmoko-cdevel jonwil (~jonwil@27-33-137-199.static.tpgi.com.au)
10:20.27DocScrutinizer05though the gtop page says 41s is minimum for unassisted
10:20.34DocScrutinizer05is puzzled
10:21.15DocScrutinizer05anyway I just tested N900 and it got a fix in 15s thru the closed window
10:22.09PaulFertserGreat stuff, agps works.
10:23.54DocScrutinizer05yeah, and on second test it took 1s
10:24.09DocScrutinizer05to get 5 sats
10:24.23DocScrutinizer05stored data
10:25.11DocScrutinizer05the u530 seems incapable of either. And it's not what this module is made for. It's made to run 16h continuously
10:29.38PaulFertserI'm not so sure. uc530 stores data, so for 2 hours it should be able to get a fix really fast.
10:30.07PaulFertserAnd it's most probable that it supports agps too.
10:40.35*** join/#openmoko-cdevel ggoZ (~zogg@78-60-10-96.static.zebra.lt)
10:41.12DocScrutinizer05datasheets says "no"
10:42.07DocScrutinizer05err, it says nothing
10:43.11DocScrutinizer05Receiver type MT3339 engine
10:43.12DocScrutinizer05<PROTECTED>
10:43.14DocScrutinizer05<PROTECTED>
10:43.34DocScrutinizer05Acquisition Cold starts: 31  s Warm starts: 31  s Hot starts: 1  s
10:44.32DocScrutinizer05it has backup voltage input, so it probably stores some data for "hot"
10:44.50DocScrutinizer05dunno what's their definition of cold, warm, hot though
10:46.48DocScrutinizer05protocols: NMEA, RTCM (whatever that means, does that include AGPS data upload?)
10:47.38DocScrutinizer05well, this pamphlet doesn't give away a lot
10:48.10PaulFertserMediatek, sigh :/
10:49.07PaulFertserhttp://www.gizchina.com/2013/07/23/how-to-fix-poor-gps-on-mediatek-phones/ they seem to be able to use some assisted data
10:50.15DocScrutinizer05how is that related?
10:50.18PaulFertserftp://ftp.leadtek.com.tw/gps/9023/G-monitor%203.26/User%20Manual/EPO%20Format%20and%20Protocol/LEADTEK%20EPO%20Format%20and%20Protocol%20Reference%20Manual%20V1.0.pdf
10:50.29PaulFertserI'd guess all mediatek gpses to use the same protocol
10:53.22PaulFertserDocScrutinizer05: http://www.adafruit.com/datasheets/GlobalTop-FGPMMOPA6C-Datasheet-V0A-Preliminary.pdf here's data for the same MTK3339 chipset afaict.
10:57.38DocScrutinizer05I don't expect a module to provide all capabilities of the chipset used
10:59.19DocScrutinizer05e.g. only UC530M supports GLONASS
11:00.19DocScrutinizer05http://www.u-blox.com/images/downloads/Product_Docs/UC530M_ProductSummary_%28FTX-HW-12002%29.pdf
11:02.03DocScrutinizer05UC530M supports autonomous A-GPS, allowing a reductio nof the warm start TTFF by as much as 90%.
11:03.00DocScrutinizer05Acquisition
11:03.02DocScrutinizer05<PROTECTED>
11:03.03DocScrutinizer05<PROTECTED>
11:03.05DocScrutinizer05<PROTECTED>
11:06.37DocScrutinizer05while IT530 comes without antenna, without GLONASS but with AGPS and some anit-jamming
11:07.14DocScrutinizer05oops, withOUT AGPS
11:07.35DocScrutinizer05err, what now?
11:07.35PaulFertserHm, probably that's a firmware-only restriction?
11:07.58DocScrutinizer05it says warm:31 but states it supports A-GPS???
11:08.40DocScrutinizer05weird
11:08.45DocScrutinizer05ds-err
11:08.55DocScrutinizer05http://www.u-blox.com/en/download/documents-a-resources/fastrax-gpsgnss-modules.html
11:12.53DocScrutinizer05AHA and MAX-M5Q announces *concurrent* GPS and GLONASS
11:13.13DocScrutinizer05and QZSS (whatever that is)
11:18.22DocScrutinizer05funny enough it yields 3m CEP, while the 530 does 2(.5)m
11:18.33DocScrutinizer05MT3333
11:24.00DocScrutinizer05actually if we hadn't the Option already, or when this one came with LNA built in...
11:24.32DocScrutinizer05though 2.5mm is quite a statement
11:25.09DocScrutinizer05not exactly trivial to integrate into Neo900
13:21.31*** join/#openmoko-cdevel ao2 (~ao2@2001:1418:117::1)
13:40.51*** join/#openmoko-cdevel patinux (~patrick@unaffiliated/patinux)
14:33.56*** join/#openmoko-cdevel ao2 (~ao2@2001:1418:117::1)
14:34.00*** join/#openmoko-cdevel ThibG (~ThibG@faye.sitedethib.com)
14:49.54*** join/#openmoko-cdevel patinux (~patrick@unaffiliated/patinux)
15:27.52*** join/#openmoko-cdevel paulk-collins (~paulk@gagarine.paulk.fr)
16:09.31*** join/#openmoko-cdevel ayaka (~ayaka@2001:250:6801:1450:a60:6eff:fe70:368a)
16:26.20*** join/#openmoko-cdevel ggoZ (~zogg@78-60-10-96.static.zebra.lt)
17:01.04*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
17:36.12*** join/#openmoko-cdevel ao2 (~ao2@2001:1418:117::1)
18:02.13*** join/#openmoko-cdevel dos11 (~dos@unaffiliated/dos1)
18:12.18*** join/#openmoko-cdevel aloniv (~alonivtsa@85.64.17.154.dynamic.barak-online.net)
18:13.08ayakaI am doing some work with exynos 4412's hardware encoder, but I can't find many sample code about it, I hear that xbmc support it but I can't find it in its git
18:35.55*** join/#openmoko-cdevel plotr (~name@unaffiliated/plotr)
19:07.31dos1PaulFertser: funny stuff with ESSID
20:17.06*** join/#openmoko-cdevel ao2 (~ao2@2001:1418:117::1)
21:13.46*** join/#openmoko-cdevel cmair__ (~chris@ppp-93-104-56-0.dynamic.mnet-online.de)
21:17.53*** join/#openmoko-cdevel radekp (~radek@63.120.broadband10.iol.cz)
21:27.41*** join/#openmoko-cdevel patinux (~patrick@unaffiliated/patinux)
21:46.10*** join/#openmoko-cdevel JaMa (~martin@ip-62-24-80-145.net.upcbroadband.cz)
23:26.47*** join/#openmoko-cdevel patinux (~patrick@unaffiliated/patinux)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.