IRC log for #neo900 on 20140501

02:37.04*** join/#neo900 wicket64 (~wicket@gateway/tor-sasl/wicket64)
03:39.51*** join/#neo900 thrakcattack (~pi@193-83-180-135.adsl.highway.telekom.at)
04:59.49*** join/#neo900 AndrewX192 (~andrew@unaffiliated/andrewx192)
05:20.54*** join/#neo900 SAiF (73761a7d@gateway/web/freenode/ip.115.118.26.125)
07:20.41*** join/#neo900 remarc (~sequantz@194.11.24.44)
08:39.25*** join/#neo900 ddark (~ddark@5.102.198.248)
09:00.51*** join/#neo900 kolp (~quassel@55d44388.access.ecotel.net)
09:55.01*** join/#neo900 edgar2 (~edgar2@85-76-139-48-nat.elisa-mobile.fi)
11:39.06*** join/#neo900 dos1 (~dos@unaffiliated/dos1)
12:07.28*** join/#neo900 thrakcattack (~pi@193-83-180-135.adsl.highway.telekom.at)
12:47.52*** mode/#neo900 [+o DocScrutinizer05] by ChanServ
12:49.05*** topic/#neo900 by DocScrutinizer05 -> http://neo900.org | conversations are logged to http://infobot.rikers.org/%23neo900/ | 2013-11-04 - the day our fundraiser reached its goal, 25k EUR | 2013-12-02 - 200 devices reached! | 12-14 50035EUR, 232 units | 01-17 60kEUR, 300 units | 02-28 333 units, 70k€ | 03-28 350 units, 400 donations, 73555€ | 0501 360 410 75k
12:49.25*** mode/#neo900 [-o DocScrutinizer05] by ChanServ
13:38.34*** join/#neo900 nicksydney (~quassel@43.11-134-203.dynamic.dsl.syd.iprimus.net.au)
13:53.35*** join/#neo900 Private_User (~Private_U@unaffiliated/private-user/x-9143494)
13:54.19*** part/#neo900 Private_User (~Private_U@unaffiliated/private-user/x-9143494)
14:33.21DocScrutinizer05is it just me or doesn't tmo offer any reasonable way to view *original size* of attached images like in http://talk.maemo.org/showpost.php?p=1423190&postcount=7
14:35.24dos1yep, noticed that as well :/
14:44.57dos1DocScrutinizer05: just included link to ODS source for my table
14:45.13DocScrutinizer05fine :-)
14:46.41ddarkgoing to do the same
14:47.56DocScrutinizer05funny how Walter's drawing resembles mine, eh?
14:48.21dos1:)
14:48.35DocScrutinizer05I'm sure he'll adore what we're about to bash up
14:53.09DocScrutinizer05tmo is a PITA to embedd/attach images. You either have to link to a usually external URL and get "click here to view full size image" and eventually the external reference goes broken thanks to source server deceased or whatever. Or you attach file of picture, and get that thumbnail view only. I one came up with a 2-step scheme to first attach the photos, then in 2nd step edit msg text to use tmo attachment download URL as "external URL
14:53.11DocScrutinizer05reference" for embedding images into msg text
14:54.01DocScrutinizer05once*
14:54.47DocScrutinizer05it's extremely clumsy
14:55.22DocScrutinizer05and I dunno if it still works, since it seems .jpg etc attachments can't get downloaded anymore??
14:56.32DocScrutinizer05other attachments you usually can click to download them. Either the embedded link for doing that for .jpg/.png etc recently vanished. Or the complete functionality vanished
15:01.54DocScrutinizer05coool¡ dropbox link just made my konqueror segfault
15:02.06DocScrutinizer05luckily not reproducable
15:07.28DocScrutinizer05the P*S8 section of ddark's table seems not all correct, I'll fix that
15:07.52ddarkwhere ?
15:12.05ddarkfound it
15:17.29ddarkfixed
15:17.52DocScrutinizer05don't worry, I'm preparing a template for all spreadsheets to have *all* bands and proper attribution to P*S8* on what they support of them
15:20.15ddarkmaybe some public wiki page, like you said on TMO ?
15:27.15*** join/#neo900 derka (~derka@41.82.162.23)
15:27.28derkaHi
15:27.33derkaHi
15:27.36derkaHi
15:28.56*** join/#neo900 remarc (~sequantz@194.11.24.44)
15:31.33derkaGuys
15:31.36derkaanyone here
15:32.37dos1hello
15:33.43derkaHello DoS1
15:34.05derkaHow soon can i get a neo 900  to dev stuf fon it
15:34.29derkaand then how quickly can i buy lot of them like 50, 100, 500
15:36.58dos1if you would want to get 500 of them, you could basically set your own rules (to some reasonable extend of course :D), as you would be buying the majority of produced devices ;)
15:37.44derkalol
15:37.47derkabut we are not here yet
15:37.57derkain the meanwhile, how quickly can i get some
15:38.00derkato do dev on it
15:38.14dos1how quickly - we're aiming at Q3 this year for the shipment of finished devices, but it might shift a bit towards Q4
15:38.24derkaand up to now are every components used on the phone open and controlled ?
15:38.35DocScrutinizer05we hope to come up with very first prototypes in maybe 2..3 months
15:38.39derkaso basically, july or august ?
15:38.48dos1but there's also a "special offer" for people who donate >= 1500 EUR - they also get first prototypes
15:38.56derkaok
15:39.21derkawhere are you based ?
15:39.30DocScrutinizer05Germany
15:39.38derkaok
15:39.56derkaand the phone will be assembled where
15:40.05dos1also Germany
15:40.07derkaok
15:40.22dos1Bavaria, to be more exact
15:40.25derkaif i was to doate 1500 € for instance when would i get the first prototypes
15:40.59derkaIch kann nicht mehr deutsch sprechen
15:41.12derkaich habe ganz vergessen
15:41.13DocScrutinizer05we got 9 proto-"orders" right now. We are offering protos to those as they get built, first come first serve
15:41.13derkalol
15:41.34derkaok so those 9 orders were not served yet
15:41.50DocScrutinizer05no, we have no proper complete prototypes yet
15:41.51dos1that's right
15:42.20derkais there a way to develop for the neo900 right now even if there is no HW
15:42.26DocScrutinizer05first prototypes that are "useful2 by any metrics expected to show up in 2 months maybe
15:42.31DocScrutinizer05raher 3
15:42.46DocScrutinizer05sure, it's compatible to N900
15:42.53derkaok
15:42.54DocScrutinizer05and to beagleboard
15:43.03DocScrutinizer05~fptf
15:43.03infobotwell, fptf is the Fremantle Porting Task Force, see http://talk.maemo.org/showthread.php?t=91308
15:43.04derkanice
15:43.37DocScrutinizer05note please we don't suggest or offer any particular OS for the hw we ship
15:43.48DocScrutinizer05up to user to decide what they wanna use
15:43.57DocScrutinizer05or for what to develop
15:44.13DocScrutinizer05it basically like buying a bare bones PC
15:44.37derkaok
15:44.44DocScrutinizer05we make sure there are linux drivers available for every subsystem in Neo900
15:44.48derkaok
15:45.19DocScrutinizer05if for some chip there are not, we still provide the datasheets links
15:45.50DocScrutinizer05so you can write the drivers
15:45.59DocScrutinizer05but I don't expect that to happen
15:46.14derkaok
15:46.22DocScrutinizer05I think we can find chips with existing driver support for everything
15:48.47DocScrutinizer05when somebody "preorders" 50 devices (aka 5k donation), we for sure will listen to any special requirements s/he mentions to us
15:51.28DocScrutinizer05we have quite some flexibility in some segments, less in others. E.G we *can* offer separate independent GPS chip, and we'd be willing to see if we can inplement whatever you like when you want 50 of them. For modem we are less flexible, you can pick from the range of compatible Cinterion P*S8 and we'll get you any of those. WLAN is another possible domain to negotiate what can be done
15:52.07DocScrutinizer05just ask
16:06.46derkaok
16:06.54derkaso basically you can provide a bare linux os on it ?
16:07.07derkaor there is not yet any os available
16:23.07x29amaemo/mer
16:23.17x29athere was even an ubuntu ported on the n900 if i recall correctly
16:24.03bencohdepends on what you expect to be working, but debian runs on it
16:24.22bencohwith most peripherals
16:24.27x29aderka: http://talk.maemo.org/showthread.php?t=79027
16:25.09derkathx
16:25.30derkagreat i would certainly stick to debian
17:28.09*** join/#neo900 derka_ (~derka@41.82.157.59)
17:29.15DocScrutinizer05derka: we will ship it with a bare bones debian with some tools to check basic functionality of the hardware
17:29.31DocScrutinizer05a so caled board support package
17:30.14derka_ok cool
17:30.17DocScrutinizer05aybe not even X11
17:30.42DocScrutinizer05bit for sure some tools to read out numerical values from sensors like compass etc
17:30.50DocScrutinizer05but*
17:31.18DocScrutinizer05and from touchscreen
17:31.28derkanice
17:31.33DocScrutinizer05and all other stuff the device comes with
17:33.36DocScrutinizer05we need that for QA after factory anyway
17:34.22DocScrutinizer05so what you get is basically the factory test software
17:36.18DocScrutinizer05we might provide (an URL to) a script to download and install other distros
17:36.50DocScrutinizer05but those other distros are neither developed nor provided nor maintained by Neo900 group
17:37.04DocScrutinizer05rather see
17:37.07DocScrutinizer05~fptf
17:37.07infobotfptf is probably the Fremantle Porting Task Force, see http://talk.maemo.org/showthread.php?t=91308
17:44.04*** join/#neo900 sixwheeledbeast (~paul@2a01:348:6:860a:6da1:fe26:630:c0fe)
17:45.49*** join/#neo900 derka_ (~derka@41.82.137.80)
18:13.27*** join/#neo900 derka (~derka@41.82.137.80)
18:34.26freemangordon~pali
18:34.27infobotsomebody said pali was http://atrey.karlin.mff.cuni.cz/~pali/
19:05.55*** join/#neo900 edgar2 (~edgar2@85-76-55-110-nat.elisa-mobile.fi)
20:37.45*** join/#neo900 xes (~xes@unaffiliated/xes)
20:52.27derkajust a quick idea, would it be cool to allow some interface to plug the raspberry pi compute on it
20:52.41derkaor even run it with it
20:53.25derkahttp://www.raspberrypi.org/raspberry-pi-compute-module-new-product/
21:51.49bencoh?
21:52.46derkayes Bencoh ?
21:52.51derkayou dont like my idea
22:08.46*** join/#neo900 wicket64 (~wicket@gateway/tor-sasl/wicket64)
22:08.46DocScrutinizer05we probably don't understand your idea
22:13.00DocScrutinizer05I don't think the CPU on this thing is significantly better than the OMAP3, and I don't understand the usecase of a Asymmetric Multi Processor desogn like this anyway
22:13.22DocScrutinizer05design*
22:14.38DocScrutinizer05anyway on Neo900 there's hardly any high bandwidth interface to integrate the two systems, like "share RAM" or whatever
22:16.17DocScrutinizer05so all you can do is hook that thing up to the Neo900 via USB2 (since I don't think you want to link the two systems via I2C) - and then the R-Pi becomes a blackbox for Neo900 and it makes no difference if the USB-link goes to R-Pi or to a i7-laptop
22:18.55DocScrutinizer05btw not even the R-Pi-compute offers a high bandwidth interface. What it offers are lots of generic IO
23:03.58derkayeah the idea is to plug many stuff on it thats it
23:04.26derkai have zero knowledge of eletronics, i just know software
23:06.32*** join/#neo900 b1101 (~b@209.222.7.236)
23:10.29ShadowJKNeo900 cpu is omap3 @ 1Ghz or something like that, right?
23:10.50ShadowJKrpi compute is roughly equivalent to omap3 @ 0.35 MHz or so afaik
23:12.28derkawoew
23:12.29derkawow
23:12.33derka3000 slower???
23:21.35ds2rpi compute is more like the N800
23:21.37ds2if even that
23:34.58derkak
23:39.55ShadowJKI meant GHz, not Mhz, sorry :)
23:40.00ShadowJKso factor 3 slower, roughly
23:40.49ShadowJKShould be noticeably faster than N800/N810 though. Omap2 of N8x0 and rpi cpu are similar in performance per clock
23:41.46DocScrutinizer05anyway I don't see the usecase
23:42.37ShadowJKOff topic, but, the idea of using the rpi soc for "IO" amuses me a bit. I've deployed projects using rpi, and because the I/O is so crippled on it, I sandwiched it with an AVR handling all the I/O, and rPi handling writing logfiles to SD, data processing, internet, etc
23:42.52DocScrutinizer05even when it was 10 times as fast, it still wpuldn't benefit Neo900 to have a SODIMM protruding from its back
23:43.34DocScrutinizer05particularly when said SODIM has ony a USB connection to Neo900
23:44.53DocScrutinizer05ShadowJK: sounds about real. I never used rPi so I can't join in commenting
23:44.58ShadowJKthat soc is ok for spi, i2c, serial etc.. but it's pretty tragic at anything that needs bitbanging, and it only has a handful of GPIOs anyway..
23:45.15DocScrutinizer05haha
23:46.01DocScrutinizer05the rPi-compute is clearly ment to beef up a stitching machine display
23:46.07DocScrutinizer05meant*
23:46.31ShadowJKIt's a general trend really. imx233 can bitbang at 10-20MHz or so, iirc.. But the CPU runs at "only" 233MHz, iirc.. The faster CPU you have in a SoC though, the slower its GPIOs tend to be :)
23:48.21DocScrutinizer05yeah, on 1MHz Z80 we created complex 500kHz signals by dynamically writing the opcode sequence to bitbang the stuff
23:49.09DocScrutinizer05actually I think the clock been 4MHz, four clocks per cycle
23:49.19ShadowJKYes
23:49.26ShadowJKfour clocks per instruction
23:49.27DocScrutinizer05except for extended complex opcodes
23:49.29ShadowJKiirc?
23:49.55DocScrutinizer05yup
23:50.01ShadowJK6510/6502 was "only" 1/2 MHz, but most/"best" instructions ran in 2 cycles
23:51.24DocScrutinizer05counting clock cycles for your "program" and inserting a NOP every now and then, this been real fun
23:51.29ShadowJKOn AVR, it's more like 1-2 per instruction, and you can clock it up to 20Mhz, GPIOs are a store instruction more or less, so you can bitbang it quickly enough to produce a modulated AM signal that can be picked up by a normal radio receiver :)
23:51.51DocScrutinizer05yeah
23:52.38DocScrutinizer05on soviet intel-cpu the clock modulates you
23:53.13DocScrutinizer05have fun councting that for timing purposes
23:53.17ShadowJKThen again, fabrice (autohor of qemu, and many other software) wrote a program, that could make a VGA card produce a valid DVB-T signal with content, at specific resolutions of the VGA out :)
23:53.44DocScrutinizer05yeah I know :-d
23:53.50DocScrutinizer05:-D
23:54.05DocScrutinizer05he didn't publish this stunt though
23:54.11ShadowJKoh?
23:54.19DocScrutinizer05afak he didn't
23:54.34DocScrutinizer05he liked to keep it a IP
23:54.54DocScrutinizer05PoC been a blob iirc
23:55.30DocScrutinizer05or he didn't include the part to encode the payload, or whatever
23:56.04DocScrutinizer05so the "transmitter2 been open, but the payload been preprocessed and nobody could make own content
23:56.16DocScrutinizer05can't recall the details
23:57.36ShadowJKmight've been kqemu
23:57.44DocScrutinizer05ShadowJK: totally OT: why does a `mount -o remount,acl /` fail ?
23:57.49ShadowJKthe virtualization without hw supported virtualization
23:57.59DocScrutinizer05raher, why doesn't it show up in `mount`
23:58.11ShadowJKdocscrutinizerpresumably acl cant be changed at remount?
23:58.12ShadowJKoh
23:58.24ShadowJK"mount" is pretty confused these days, I look at cat /proc/mounts
23:58.42ShadowJKthat is also somewhat confused but better..
23:58.43DocScrutinizer052014-05-02T01:10:20.644259+02:00 neo900 kernel: [1988772.136810] EXT4-fs (vda1): re-mounted. Opts: acl
23:59.17DocScrutinizer05<PROTECTED>
23:59.44DocScrutinizer05w. t. f. ?

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