IRC log for #qi-hardware on 20180311

01:20.30*** join/#qi-hardware infobot (ibot@rikers.org)
01:20.30*** topic/#qi-hardware is Copyleft hardware - http://qi-hardware.com | hardware hackers join here to discuss Ben NanoNote, atben/atusb 802.15.4 wireless, anelok and other community driven hw projects | public logging at http://en.qi-hardware.com/irclogs and http://irclog.whitequark.org/qi-hardware
01:20.31*** mode/#qi-hardware [+v infobot] by ChanServ
02:44.35*** join/#qi-hardware opbolt (~starcard@host-92-15-35-5.as43234.net)
03:51.13*** join/#qi-hardware _whitelogger (~whitelogg@uruz.whitequark.org)
03:51.13*** mode/#qi-hardware [+v _whitelogger] by ChanServ
05:53.29*** join/#qi-hardware wpwrak (~werner@52-32-17-190.fibertel.com.ar)
05:54.25*** join/#qi-hardware erichvk (~esh@ppp14-2-124-244.lns21.adl6.internode.on.net)
11:07.47*** join/#qi-hardware erichvk (~esh@ppp14-2-124-244.lns21.adl6.internode.on.net)
11:09.04DocScrutinizer05HMMMM
11:09.07DocScrutinizer05blobs
11:10.52DocScrutinizer05how about a phone with modern SoC, premium grade build quality, OpenHardware(!!!) *and* attachable hw kbd genuinely made for that purpose?
11:11.47DocScrutinizer05OpenHardware means you got the *hardware* docs, not some "FOSS" code snippets thrown over the wall
11:35.45*** join/#qi-hardware fdcx (357@ceata/consiliu/FDCX)
11:36.00*** join/#qi-hardware fdcx_ (357@ceata/consiliu/FDCX)
12:08.24opboltDocScrutinizer05, yep sounds more like it :)
12:09.29DocScrutinizer05then i'm working with the right project right now :-)
12:09.37opboltneo900 ant it :)
12:10.05DocScrutinizer05looking into Neo900 STEP2
12:10.17opboltthe qtmoko update... its more than just upzip to sd card isnt it?
12:10.17DocScrutinizer05~step2
12:10.17infobotfrom memory, step2 is http://talk.maemo.org/showthread.php?p=1444602, or http://talk.maemo.org/showthread.php?p=1444625
12:10.42opboltill have a look
12:11.02opboltre qtmoko update, its use makesd on a new, unused sd card right?
12:11.24opboltwas trying to update my current sd card but dont want to over write other partitions.
12:12.09opbolti remmber having makesd over write a drive nearly or something... so im err assuming its stupid and not fool proof :)
12:12.22opboltand use it with caution
12:12.37opboltive got a gta04 ya see
12:14.19opboltoow DocScrutinizer05 i would be interested it hear what you think and feel about librem5 phone?
12:17.17DocScrutinizer05librem5 tries to do the right thing but they don't have their concept/basics sorted. They don't have smart isolation of rogue subsystems and they think throwing FOSS snippets over the wall was the right thing to do
12:18.09DocScrutinizer05and they have no hw kbd ;-D
12:19.43DocScrutinizer05afaik at least
12:19.46opboltmmh. yea i dont have great faith in them. i fear a typical fancy aluminium smartphone with gnu/linux slapped on top but floss and securety wise, disapoiting insides considering all the money and effort.
12:20.17DocScrutinizer05yes
12:20.24DocScrutinizer05exactly my take on it
12:20.26opboltthey dont design the pcbs, from the looks of it, they pay a company to do it
12:20.44DocScrutinizer05they did for their tablets
12:21.42DocScrutinizer05that's not a big issue in itself. However for the tablets they don't even own permission to disclose the schematics since the copyright is held by the OEM
12:21.43opboltdesign company will do what they are most familiar  with. which is proprietary, modem is king,
12:22.14opbolt*modem is king and integrated
12:22.30DocScrutinizer05nah, I *think* they try to design this librem5 from scratch inhouse
12:22.31opboltyea thats sucky
12:23.00DocScrutinizer05but obviously they have no experience in phones, and a weak idea of openness
12:23.01opboltyea but will they do much different than they normaly do?
12:23.25opboltTi soc and?
12:23.33DocScrutinizer05I can't tell, had no more than brief conversations with them
12:23.36opboltyea
12:23.37opbolt<PROTECTED>
12:24.09DocScrutinizer05seems they aim at iMX SoC which isn't bad I guess
12:24.25opboltyea, better than snapdraggon :)
12:24.31DocScrutinizer05hehe yes
12:25.04opboltDocScrutinizer05, you know pyra right? the mobile, version. the extra sensors, would it be possible to have them without the mobile modem?
12:25.43DocScrutinizer05sure I know Pyra, Nikolaus "shared" between Neo900 and Pyra and I even reviewd the schematics
12:25.44opboltdoes any map nav apps, like navit make use of hardware compass? i dont know of any
12:25.51opboltheh
12:26.34DocScrutinizer05you can shut down the modem in pyra
12:26.40opboltor arr the sensors all on same pcb?
12:26.44opboltoh ok
12:26.59opboltarr=all
12:26.59DocScrutinizer05Pyra is pretty sane concept
12:27.07DocScrutinizer05for a ganing console ;-P
12:27.34opboltyea, im thinking of getting one. i just wish gps was separated from modem :(
12:27.39DocScrutinizer05it still doesn't emply Neo900's unique sandboxing
12:27.49opboltyea.
12:27.58*** join/#qi-hardware eintopf (~eintopf@unaffiliated/eintopf)
12:28.02opboltyummy neo900
12:28.06DocScrutinizer05I don't know if Nikolaus added a separate GPS chip, he planned to
12:28.31opboltoh, i tought they where using gps feature of modem?
12:28.45DocScrutinizer05prolly they do, yes
12:28.54opboltthat would be nice if it was isolated
12:29.15opbolti guess they get agps by using the gps built into the modem?
12:29.38DocScrutinizer05the P*S8 modem knows a mode where GPS enabled but modem radio inactive. However Pyra can't guarantee the modem obeys such mode, Neo900 can
12:29.55opboltyea, what i was expecting
12:30.07opboltassisted-gps
12:30.14DocScrutinizer05yes, AGPS is a benefit from GPS-modem integration
12:30.18opboltarr
12:30.22opboltdarn
12:30.33opboltnice to have yet also a snag....
12:30.40opboltaww
12:30.48DocScrutinizer05you *can* provide the assistance data externally to the modem
12:30.54opboltow
12:30.56opbolthow
12:31.02DocScrutinizer05see TRM
12:31.13DocScrutinizer05not sure, I think there's an AT command
12:31.42opboltoh so still using modem?
12:31.45DocScrutinizer05the format of that data however is proprietary
12:32.05opbolti cant have a-gps with my bluetooth gps for example?
12:32.51opboltwell i guess apgs means a modem powered up.. cant have everything
12:33.11DocScrutinizer05no, you switch off "the modem" (GSM) and download the assistance data e.g. via WLAN and send it to the modem like "ATgpsdata "9F33A14E6A..."
12:33.58opboltoh but dont it need to monitor mobile err anterents around the area in order to work?
12:34.03DocScrutinizer05only the AT interpreter and the GPS functions active in modem
12:34.19DocScrutinizer05that's another form of "assisted"
12:34.25opboltoh right
12:34.39opbolti wonder how that works
12:34.43opboltinside it
12:34.51DocScrutinizer05~u-tdoa
12:34.51infobotit has been said that u-tdoa is http://en.wikipedia.org/wiki/U-TDOA
12:35.07DocScrutinizer05~gsm-agps
12:35.08infobotRRLP is the Radio Resource LCS (Location Service) Protocol as specified first in GSM TS 04.31, or http://osmocom.org/projects/security/wiki/RRLP
12:35.44DocScrutinizer05U-TDOA is faster and more accurate than any GPS now, and works indoors
12:35.54DocScrutinizer05but need WWAN aka "GSM"
12:37.56opboltmore than 50M?
12:38.04opboltaccurate
12:38.26DocScrutinizer05typically better than 10m
12:38.51DocScrutinizer05depending on RF-path between mobile and BTS
12:39.49opboltoww so wikipeada is out of date then
12:39.50DocScrutinizer05note how U-TDOA can locate *any* mobile (actually any transmitter, depending on the LMUs)
12:39.58opboltyea
12:40.24opboltexpains why gps only in citys as been a struggle
12:40.29opbolthehe
12:40.42opboltcompared to typical agps smartphone
12:41.34opboltdidnt know they get 10m accuracy just from mobile
12:42.41DocScrutinizer05basically the accuracy is just limited by timing in the LMUs and varaibles in the air RF path
12:43.15DocScrutinizer05they just don't need more than 10m precision
12:43.59DocScrutinizer05and I don't even know if they provide the results of U-TDOA to the mobile equipment with that accuracy
12:44.27opboltsomehow i can imagine the powers that be saying they need more than 10m precision for there mass surveillance wants. 10m is like gps
12:45.37opboltyou would not happen to know what the most floss statnav is ?
12:45.45opboltmy sister wants a satnav
12:45.48DocScrutinizer05in an acient post by me you can see how I evaluated a sort of U-TDOA based on GSM TimeAdvance parameter, which is ~550m precision and predates the official U-TDOA implementation
12:46.06opbolton neo900 forum?
12:46.55opbolttomtom are anti OSM :(
12:48.35DocScrutinizer05http://lists.openmoko.org/pipermail/openmoko-kernel/2008-June/002987.html  sorry takes ages to load, the server is messed up
12:48.55opboltok
12:48.57opboltthanks
12:50.00DocScrutinizer05Load average: 23.95 22.82 26.27   -  it's dead Jim!
12:50.11DocScrutinizer05even frozen now completely
12:51.19DocScrutinizer05http://susepaste.org/88388688
12:54.33DocScrutinizer05I hope the sysop will fix it soon
12:54.55opboltknow anything about gta04 batts?
12:54.58DocScrutinizer05you might find a copy in wayback archive
12:55.11opboltarr yea, thats an idea
12:55.17DocScrutinizer05umm, aren't they the GTA02 smarrtbat still?
12:55.22opboltyes
12:55.31DocScrutinizer05bq27200
12:55.52DocScrutinizer05~batteryfaq
12:55.53infobotextra, extra, read all about it, batteryfaq is http://wiki.openmoko.org/wiki/Battery_Questions_and_Answers
12:55.58DocScrutinizer05also dead server
12:56.04opbolti reused bms from old gta batt to a 18650 cell :) but it didnit work :(. any ideas? gta i dont think was powered from it
12:56.26DocScrutinizer05please rephrase
12:56.58DocScrutinizer05see bq27200 datasheet
12:57.04opboltdead gta04 batt with bq27200 BMS board. i remove BMS board and ....
12:57.18opboltok yea, datasheet of course
12:57.35DocScrutinizer05sorry for the dead server
12:57.57DocScrutinizer05I got a complete docs for the GTA02 smartbat in my homedir on openmoko
12:58.28DocScrutinizer05it consists of the bq27200 plus a protection circuit
12:59.42DocScrutinizer05you could replace the cell by a e.g 18650 but the bq27200 will probably not cover the capacity, register range too small
13:00.14DocScrutinizer05not sure though
13:00.39DocScrutinizer05you could change the value of the shunt resistor
13:00.52DocScrutinizer05it's 20mR now, I think
13:01.07DocScrutinizer05could change that to e.g. 2mR or whatever
13:01.37DocScrutinizer05~bq27k
13:01.46DocScrutinizer05~bq27k-detail
13:01.46infobot[bq27k-detail] http://maemo.cloud-7.de/maemo5/usr/local/sbin/bq27k-detail2, or ~bq27
13:01.57DocScrutinizer05~bq27
13:01.57infobotit has been said that bq27 is http://maemo.cloud-7.de/maemo5/usr/local/sbin/ bq27*, or http://maemo.cloud-7.de/maemo5/patches_n_tools/
13:03.13DocScrutinizer05this server works :-)
13:04.04opboltthanks
13:09.52DocScrutinizer05you see how all register values (except voltages) are based on RS (shunt resistor)
13:10.18kyakDocScrutinizer05: the latest news on Neo900 are really promising! It's great that you will have now more time to spend on the project
13:10.44DocScrutinizer05the actual news are not even out yet :-)
13:11.12DocScrutinizer05but I think STEP2 is the way to go
13:11.37DocScrutinizer05Neo900 (step1) got somewhat dated and obsolete over time
13:12.38DocScrutinizer05though there are not many news publsuhed, i'm *very* busy behind the scenes
13:12.50DocScrutinizer05published even
13:13.27DocScrutinizer05go figure
13:13.33DocScrutinizer05:-D
13:15.45DocScrutinizer05maybe openmoko was just a 10 years ahead of its time ;-)
13:17.37DocScrutinizer05the idea of open hardware still going strong
13:18.55DocScrutinizer05kyak: any comments on RISC-V / lowrisc for phone SoCs?
13:19.46DocScrutinizer05obvious "problem": it's not ARM, so hardly any existing app will work without either an emulator or a recompile.
13:20.13DocScrutinizer05then, they have no decent GPU yet aiui
14:03.24opboltopenmoko was before iphone?
14:05.33opbolti think
14:05.37opbolt:/
14:07.51opboltre bq27200 with 18650 li-ion cell. how does changing the shunt/current sense resistor make the ic able to accepct a larger capacity li-ion cell?
16:01.51kyakDocScrutinizer05: that, and also the linux kernel support is quite new (i think it has recently landed and now get polished): https://www.phoronix.com/scan.php?page=news_item&px=RISC-V-Early-2018
16:05.45DocScrutinizer05kyak: tbh I don't even know if they got ANY real silicon out already
16:06.16DocScrutinizer05all I see are emulations in FPGAs
16:08.07DocScrutinizer05opbolt: all registers are "coun't * CONSTANT" where constant is "Divide by Rs in milliohms to convert µV to mA or µVh to mAh"
16:08.27kyak"There is now a RISC-V public development board in form of the HiFive Unleashed, a 4+1 multi-core dev board using the SiFive FU540 SoC. This board has 8GB of DDR4 memory, Gigabit Ethernet, 32MB SPI flash, and microSD card. But before getting too excited, this first RISC-V development board will set you back $999 USD."
16:08.44DocScrutinizer05see http://maemo.cloud-7.de/maemo5/usr/local/sbin/bq27k-detail2
16:09.15DocScrutinizer05kyak: great, many thanks
16:09.52DocScrutinizer05kyak: link?
16:12.02DocScrutinizer05ooh prolly the one you gave, eh?
16:12.43kyakright :)
16:41.37DocScrutinizer05do they care about idle/zeroclock?
16:42.06DocScrutinizer05I see highperf but no lowpower
20:12.28*** join/#qi-hardware wpwrak (~werner@52-32-17-190.fibertel.com.ar)
20:14.28*** join/#qi-hardware 7GHAABORT (357@ceata/consiliu/FDCX)
20:14.28*** join/#qi-hardware fdcx (357@ceata/consiliu/FDCX)
20:20.19*** join/#qi-hardware eintopf (~eintopf@unaffiliated/eintopf)
21:18.26opboltexciting India risci cpu + in a card: http://libre-riscv.org/shakti/m_class/
21:31.30*** join/#qi-hardware wpwrak (~werner@52-32-17-190.fibertel.com.ar)
21:54.50*** join/#qi-hardware wpwrak (~werner@52-32-17-190.fibertel.com.ar)

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