#qi-hardware IRC log for Friday, 2011-06-17

awwpwrak, hi if my *.brd under folder /A, do i cd into /A then directly use fab/prettygerbv? how's syntax for it?01:14
awwpwrak, /fab/prettygerbv -d whatever "project-name" all allgerber.png ?  "project-name" here is supposed my folder /A ?01:18
wpwraklemem check :)01:19
wpwrakfirst, you need to generate the gerbers01:20
wpwrakthen you go to the directory where your gerbers are and01:20
wpwrakprettygerbv project-name all overview.png01:20
wpwrakit needs the project name because kicad prefixes all the gerber files with the project name. so you get atusb-Front.gtl, atusb-Back.gbl, etc.01:21
wpwrakinstead of "all" you can also use "front", "back", or "conn"01:22
wpwrakthe -d is for the drill file (.drl). if you want to include that too, then the invocation is01:22
wpwrakwell, no. first the drill file needs to be converted from .drl to geber. that would be01:23
wpwrakdrl2gerber drill-file.drl drill-file.gbr01:23
wpwrakand then01:23
wpwrakprettygerbv -d drill-file.gbr project-name all overview.png01:24
awi see, so the "project-name" is now for me is "usb_jtag-xxxxx" >>> "usb_jtag"?01:24
wpwrakso yes, you cd into /A and do things from there01:24
awokay..try now. :-)01:27
awwpwrak, do i need chmod for that script first?01:32
wpwrakit should already be executable (?)01:32
wpwrakif it isn't, yes :)01:33
awadam@adam-laptop:~/extras-m1/jtag-serial$ prettygerbv usb_jtag front overview.png01:33
awwhat words before 'prettygerbv' i need to enter?01:34
awcurrent the prettygerbv is under my /home/adam/eda-tools/fab  01:34
wpwrakjust the path to it, in case it's not in PATH01:34
DocScrutinizerhi aw01:35
wpwrakthen it would be: /home/adam/eda-tools/fab/prettygerbv usb_jtag front overview.png01:35
awman~ i don't know command line type well... i know nothing. :-)01:35
awDocScrutinizer, hi 01:35
wpwrakaw: by the way, the invocation of all these things is  http://projects.qi-hardware.com/index.php/p/ben-wpan/source/tree/master/makefiles/Makefile.kicad01:36
wpwrakaw: it's not very easy to read, but it's a handy reference because everything is in one place01:36
DocScrutinizerot question: what's Ben's audiosystem?01:36
wpwrakDocScrutinizer: you mean the codec ? some ac97 built into the SoC01:37
awwpwrak, do u meant that i need to make your tools firstly then i can just type >> prettygerbv ****** ?01:37
DocScrutinizernah, I mean the system, like PukeAudio vs ALSA, vs dunno OSS4 whatever01:38
wpwrakaw: no no, prettygerbv is a shell script01:38
wpwrakaw: the makefile i mentioned calls prettygerbv (and a number of other scripts). so you can see there how these things are used01:39
wpwrakaw: e.g., when i want to make the overview in atben or atusb, i simply do "make overview". it will then generate the gerbers, the drill file, convert the drill file to gerber, and finally run prettygerbv01:40
DocScrutinizerI'd guess ylu didn't want PissAudio on such a relatively modest CPU01:40
awwpwrak, wow..okay01:40
wpwrakaw: if you look at http://projects.qi-hardware.com/index.php/p/ben-wpan/source/tree/master/atben/Makefile01:40
wpwrakaw: (which is the project-specific makefile) you'll see that all the magic is in Makefile.kicad01:41
awwpwrak, hmm..okay..those two scripts are.01:42
wpwrakDocScrutinizer: duh. i guess that may depend on the distribution. i haven't used audio on the ben so far ;-)01:42
awwpwrak, Loading project _gvp...01:42
awmain_open_project_from_filename():  Opening project = /home/adam/extras-m1/jtag-serial/_gvp01:42
aw** Message: could not read /home/adam/extras-m1/jtag-serial/_gvp[-1]01:42
DocScrutinizersure it's a distro thing, also question of what's supported for the hw01:42
awwpwrak, what does this mean?01:42
wpwrakaw: i mention these makefiles just for reference, in case you want have a closer look at the details and/or automate the process01:43
wpwrakaw: hmm. that means that a temporary file wasn't generated01:43
awwpwrak, okay..if i have more time to have a closer look. but sorry that i have to see those images firstly. :-)01:44
wpwrakaw: can you please paste the exact command you used ?01:44
awwpwrak, adam@adam-laptop:~/extras-m1/jtag-serial$ /home/adam/eda-tools/fab/prettygerbv usb_jtag front overview.png01:45
wpwrakaw: that should work01:45
wpwrakaw: were there any messages before  Loading project _gvp...  ?01:46
awwpwrak, no, then shows that msg though01:46
wpwrakaw: what does   gerbv --version   say ?01:46
awdo i need to use drl2gerber drill-file.drl drill-file.gbr firstly?01:47
awwpwrak, gerbv version 2.4.001:47
awCopyright (C) 2001 -- 2008 by Stefan Petersen01:47
awand the respective original authors listed in the source files.01:47
wpwrakaw: i have the same version. good so far.01:47
wpwrakaq; no, you don't need drl2gerber yet (only with the -d option)01:48
awwpwrak, okay01:48
wpwrakwhat does   ls usb_jtag*  say ?01:48
awusb_jtag.000  usb_jtag.brd  usb_jtag.pro  usb_jtag.rpt  usb_jtag.sch01:49
wpwrakno gerbers ?01:49
wpwrakor did you type  ls usb_jtag.*   ? :)01:49
awmy gerbers are all under /A/gerber/01:49
DocScrutinizerusb_jtag? what's that?01:50
wpwrakurgh. that explains it. okay, cd /A/gerber01:50
wpwrakthen try prettygerbv again01:50
awDocScrutinizer, hi https://github.com/milkymist/extras-m101:51
wpwrakDocScrutinizer: the jtag dongle for M101:51
DocScrutinizerwhat's a jtag dongle?01:51
wpwrakDocScrutinizer: an usb to jtag interface. the name kinda hints at its function ;-)01:52
DocScrutinizernot necessarily01:52
DocScrutinizerso it's M1 debubo?01:53
wpwrakaw: how did it go ? success or sucks ? :)01:53
wpwrakDocScrutinizer: sort of, yes. i think it also has a UART. uses one of the ftdi chips, like our good old friend, the openmoko debug board.01:54
awwpwrak, http://downloads.qi-hardware.com/people/adam/m1/jtagserial/jtag_rc2_overview.png01:54
awwpwrak, GREAT, THANK YOU!01:54
wpwrakaw: now bad :-)01:55
wpwrakaw: now, you should also add the holes01:55
wpwrakaw: do you have a drill file ? (usb_jtag.drl)01:55
awwpwrak, you meant drill gerber?01:55
awwpwrak, let me use your another tool. :-)01:55
wpwrakaw: first the .drl01:56
wpwrakaw: if you don't have it, pcbnew --drill --origin=aux /your/path/usb_jtag.brd01:56
awwpwrak, i generated it already from gui pcbnew. :-)01:57
wpwrakaw: let's hope the coordinates are right ;-)01:57
awwpwrak, second . let me try. don't know. :-)01:57
DocScrutinizer(jtag) just idly wondered if somebody finally put JTAG to purpose as meant by those who invented it01:57
wpwrakaw: okay, then you can convert it with drl2gerber01:57
DocScrutinizeron some qi device01:57
wpwrakaw: if the coordinates are wrong, you'll get a BIG image with the board in one corner and all the holes in another. easy to notice ;-)01:58
wpwrakDocScrutinizer: where would be the fun in that ? ;-)01:58
DocScrutinizerwhere <purpose> where originally automatic QA in manufacturing01:59
DocScrutinizertotal control over each pin of the main chips, sending H on one chio:pin and see if trace and soldering is OK so H arrives at chip2:pin202:00
wpwrakDocScrutinizer: you'll find something quite similar in http://projects.qi-hardware.com/index.php/p/ben-wpan/source/tree/master/prod/atusb :)02:02
wpwrakDocScrutinizer: only controlling one chip, though. the other isn't that sophisticated.02:02
awwpwrak, one question, pcbnew can generate a drill gerber. why you need to write drl2gerber ?02:03
DocScrutinizeraah, you're using a JTAG set of TP to bitbang atmel's IOs and check for soldering defects?02:04
awwpwrak, http://pastebin.com/vw5gcHqR02:04
wpwrakDocScrutinizer: i just toggle the GPIOs via USB. don't have JTAG connectivity.02:05
wpwrakaw: (drill gerber) hmm, a real gerber ? or the excellon (.drl) ?02:06
DocScrutinizerso what's the similar thing then?02:06
wpwrakDocScrutinizer: same kind of test02:07
awwpwrak, it generates a 'usb_jtag-drl.pho'02:07
wpwrakaw: ah, sorry. drl2gerber usb_jtag.drl >usb_jtag.gbr02:07
awwpwrak, oah...type ">" ?02:08
wpwrakaw: yup, just the   >   was missing02:09
wpwrakaw: (*-drl.pho) hmm, i think i've seen this. looks a bit different from the actual holes, though. prettygerbv tries to make "photo-realistic" output.02:10
awwpwrak, http://pastebin.com/uBSD5wpp02:19
wpwrakhmm ... 02:23
awwpwrak, let me try to generate drill by setting "Zeros format" not "Suppress leading zeros" .:-)02:24
wpwrakaw: maybe just use the command-line variant ? i know that one works :)02:25
awi picked "Suppress leading zeros" for close gerber viewer tool. maybe this caused that. 02:25
awwpwrak, sorry , how to generate drill by commands?02:26
wpwrak pcbnew --drill --origin=aux /your/path/usb_jtag.brd02:26
awoah~this one. :-)02:27
wpwrakand then drl2gerber02:28
DocScrutinizerhehe, c&p comes in handy02:29
awwpwrak, pcbnew opened and shows "File <--drill.brd> not existing02:32
awThis is normal for a new project"02:32
awis it supposed to be that?02:32
wpwrakhmm no. seems that your kicad doesn't have our patches for --drill and such02:34
wpwrakaw: your boss will not be happy - all his hard work ignored you :)02:36
wpwrakwolfspraul: btw, how's the upstreaming of the kicad patches going ? (-:C02:36
kristianpauloh, LWM now02:37
wpwrakLWM ?02:37
awwpwrak, hmm...okay..although this way i can't get there now, but i think i still trust those gerbers i generated. :-)02:38
kristianpaulwpwrak: http://lwn.net/Articles/448048/rss02:38
wpwrakwolfspraul: let's try it manually then. drill units: inches, zeros format: decimal, drill origin: auxiliary, all other options off/none02:38
awwpwrak, with your tools, it can let people to see the whole gerber pictures though. :-)02:38
wpwrakwhee, lwn. nice ;-)02:39
kristianpauli feel atusb soon sold out :-)02:41
wpwrakaw: (options) wait .. that doesn't work ...02:41
awwpwrak, hup? what's that doesn't work?02:43
wpwrakaw: drill origin absolute. then it'll work.02:43
wpwrakaw: in the drill file dialog02:43
wpwrakaw: then drl2gerber, etc.02:43
wpwrakrejon: so you have a good channel to LWN ;-)02:44
wolfspraulwpwrak: upstreaming - bah! That's one of the worst things I did this year02:44
wolfspraulfirst make a big plan, ask for feedback, get very nice feedback, then do nothing02:44
wolfspraulreally really bad02:44
wpwrakwolfspraul: this or last year ? :-)02:44
wolfspraulI sent my mails there in January or so, I think02:45
wolfsprauland I got very encouraging feedback02:45
wolfspraulthen silence from me. argh.02:45
wolfspraulnot good, I know02:45
wolfspraulmy worst job this year, by far02:45
wolfspraulthe thing is that I need to push on this freaking m1 rc3 release, it's delayed by several months now02:45
wolfspraulall for good reasons, but I have to get this thing done and into the shop02:46
wolfspraulI'm sold out on m1 right now02:46
wpwrakhmm, don't see them right now. they must be hiding somewhere in the wrong mail folder.02:46
wolfspraulwas it in late December? I don't think so02:46
wolfspraulI hacked up the command line stuff around christmas02:46
wolfspraulso must have been January02:46
wolfspraulbut it doesn't matter, I dropped the ball there02:46
wpwrakprojects that are on time are rushed ;-)02:46
wolfspraulI have to continue 100% on m1 now02:47
wolfspraulimagine tuxbrain wouldn't come to help with ben-wpan!?02:47
DocScrutinizermoo wolfspraul 02:48
kristianpaulxiangfu: hi !02:48
xiangfukristianpaul: hi02:48
kristianpaulxiangfu: in flicernoise, will be nice implement a ramdisk in a proper way, that will help a lot for benchmarks/debug i think, and also me in order to be able to take bigger raw samples from the gps frontend02:49
kristianpaulxiangfu: i can point you again to some sample code and chat log from #rtems about what is needed to do it02:50
xiangfukristianpaul: ok. thanks. 02:50
wpwrakaw: did it go better this time ?02:50
wpwrakwolfspraul: (tuxbrain) yeah, he saved us there02:51
awwpwrak, http://downloads.qi-hardware.com/people/adam/m1/jtagserial/overview.png02:52
rejonwpwrak, yes02:52
wpwrakaw: that's more like it ;-)02:52
wpwrakaw: congratulations !02:52
awwpwrak, guess what? just only one difference on zeros format: decimal compared to my close gerber viewer settings. :-)02:53
wpwrakaw: heh :) yes, the number format in excellon is extremely weird02:54
awwpwrak, congratulations on you too. I am the first one user to use your smart tool though, am i? :-P02:54
wpwrakaw: btw, did you notice ? you have silk screen on your pads :) "FB2", "R14", "FB3"02:54
wpwrak(or very close to them)02:55
kristianpaulxiangfu: http://paste.debian.net/120121/02:55
wpwrakaw: (1st user) yup :)02:55
kristianpaulxiangfu: i think is not so hard, i just dint had time to look at it,02:55
awwpwrak, yah..i can move them a little though. :-)02:55
xiangfukristianpaul: got it.02:56
wpwrakaw: i found these overviews great for optimizing the silk screen. i kicad, you often overlook some things.02:56
awwpwrak, man!..i need to record all those steps, otherwise i'll forget them. :-)02:57
kristianpaulaw: make a script :-)02:57
DocScrutinizerwolfspraul: are you any way interested or even involved in gta02 anymore?02:57
wpwrakkristianpaul: there already is a makefile :) but ... he needs a patched kicad for it02:58
awwpwrak, yeah from pcbnew it's not easy to make all silk screen text to be well-done though. your three-images can look well deeply. :-)02:58
wolfspraulDocScrutinizer: phew, that's hard to say02:58
DocScrutinizerwolfspraul: I recently found Mentorgfx has a free (as in beer) PADS viewer now02:58
wolfspraulthat's like a history question :-)02:58
wolfspraulWerner did a lot of good things with gta02-core02:58
wpwrakaw: like this: http://projects.qi-hardware.com/index.php/p/eda-tools/source/tree/master/kicad-patches/README02:58
DocScrutinizerso we *could* "disclose" the PCB02:58
kristianpaulwpwrak: ah, thats much better02:58
awkristianpaul, man..you know that i was poor on s/w side now. but indeed the script truely great... i like it.02:59
wolfspraulbut I don't know whether/how/when we can make this go anywhere now, or should even try02:59
wolfspraulI am very interested in wireless connectivity02:59
wolfspraulwe have ben-wpan now, we need a lot of luck, and then in 1 year we have a great software stack for it :-)02:59
kristianpaulaw: you will not believe but i never had done a script before.. 02:59
wolfspraulkristianpaul works on gps-sdr, also fine02:59
kristianpauli need improve on that path some day, for now memory still good :)03:00
wolfspraulwhat's missing is a longer range wireless protocol03:00
kristianpaulhow longer?03:00
DocScrutinizerwolfspraul: just pondering about contributing as much as possible to community hive wisdom about that ancient abandoned project03:00
wolfspraulwe looked at the ISM stuff (hoperf modules), but maybe not so interesting03:00
wolfspraulthen we hvae the gsm/3g/edge/lte world, well, I am following now03:00
wolfspraulI support osmocom a little, mostly just watching though03:00
wolfspraulI know where I can source any mtk chip, if needed03:00
DocScrutinizerwolfspraul: I'd care about it, if we could get 'OK' from whoever owns the rights of the gta02 PADS project files now03:01
wolfspraulbut I have no clear feelings on that side now, anything longer range. So for the time being I focus on the 802.15.4 stuff Werner started, short-range stuff03:01
wpwrakwolfspraul: get some chinese company to add a 10 W RF amplifier and we can go ~700 m ;-)03:01
wolfspraulDocScrutinizer: I'm definitely the wrong person to ask about that.03:01
awwpwrak, this http://projects.qi-hardware.com/index.php/p/eda-tools/source/tree/master/kicad-patches/README  you meant i need to git pull and rebuild KiCad?03:02
DocScrutinizerthat's why I asked if you are involved anymore03:02
wolfspraulI'm not sure but I think I have a good chance to still be #1 on Billy's persona-non-grata list03:02
DocScrutinizerI gaher you're also not inclined to talk to <owner> about it?03:02
wolfspraulyou or me have the same powers about sending mails03:02
DocScrutinizeraah ok03:03
wolfspraulmost likely they will go answered03:03
wolfspraulwhen you are high up there (ahem), the problems of us lowly workers are not worth too much attention...03:03
wpwrakDocScrutinizer: well, maybe you should ask about the gerbers. that may cause some uneasiness on the receiving end. like "damn, are they STILL not dead ?" :)03:03
DocScrutinizerwolfspraul: so you got any clue who's the right persn to ask now? Still SMP?03:04
wpwrakDocScrutinizer: err, i mean design files03:04
wolfspraulDocScrutinizer: on the long range stuff, in addition to following osmocom, pondering about lte chips, mtk chips, I am also following some IETF white-space developments03:04
wolfspraulor 802.22 (I think that's what it was)03:04
wolfspraulbut that's all just for crazy thinking once in a while, none of that is real03:05
wolfspraulas werner pointed out I'm not even living up to small upstreaming promises right now...03:05
wolfspraulDocScrutinizer: seriously, you have absolutely none, zero, chance to get anything out of there.03:06
wolfspraulI've tried to tell this politely to Werner years ago when he believed in some support, well, Werner found out later :-)03:06
wolfspraulyou may ask well email ceo@nokia.com to release the N900 gerbers03:07
wolfspraulooh, typo above, I meant "most likely they will go unanswered"03:08
wolfsprauljust try :-)03:09
wolfspraulsend another one about n900 gerbers to ceo@nokia.com, see which one gives you a more meaningful answer :-)03:09
wolfspraul10W RF amplifier...03:10
wolfspraulmakes me thinking here :-)03:10
wolfspraulhow come a good old GSM phone can send signals 5+ km? and ben-wpan only 10m ?03:11
wolfspraulwhy is that so?03:11
DocScrutinizeralso frequency/band makes a difference03:12
wolfspraulif a wpan would just scream out loud with 500mW too, what would happen to other wpans nearby?03:12
wpwraksqrt(500mW/2mW)*10m = 158 m. GSM still beats us :-(03:12
wolfspraulthe 500mW wpan would drown out everybody else within a long range, no?03:12
DocScrutinizerwpwrak: 900MHz vs 2400MHz03:13
wpwrakwolfspraul: everybody, including wifi, bluetooth, ...03:13
wpwrakDocScrutinizer: some f^2 law ?03:13
DocScrutinizerwpwrak: wolfspraul: then you mustn't miss the pretty decent antenna of BTS 03:13
DocScrutinizerwpwrak: kinda03:14
Action: wpwrak calculates ...03:14
DocScrutinizerwpwrak: the higher the freq, the more you need line of sight03:14
wpwrak158m*(2.4GHz/0.9GHz)^2 = 1124 m. we're getting there. add massive antenna gain and the numbers work.03:15
DocScrutinizerI guess with a tower and antennae like on a BTS for one end of wpan, you'd get up to 500m with 2mW03:15
wpwrakoh, and the 10 m range for wpan may be a bit on the low side. tuxbrain got ~12 m and we still don't know how good/bad the boards really are. someone with real RF knowledge and a lab will have to find out.03:16
wolfspraul"real RF knowledge", that must be me03:17
wolfspraulI know how to turn on my phone!03:17
wpwrakwolfspraul: hehe ;-)03:17
wpwrakDocScrutinizer: ah, and GSM is a much lower data rate. that helps too.03:17
wpwrakwolfspraul: btw, how do you like my rants on the ben-wpan wiki page ?03:18
DocScrutinizertaking another nap, got an appointment in 3h03:20
wolfspraulDocScrutinizer: maybe when you get up then - do you know anything about the TV white-space stuff? 802.22? are there any RF ICs we could hookup to get some longer-range (hopefully) communication going?03:23
wolfspraulor is this just a bad direction overall...03:23
wolfspraulthere has been so much wiki editing, not sure what you mean03:25
wolfspraulsomehow Jon got a snippet up on lwn.net, that is good03:25
wolfspraulwhere does that point to? lemme check03:25
wolfspraulgod, points to 10 different places03:25
wpwrakthe wiki page is still a mess. inconsistencies, etc. i just put a pointer to my page at the beginning. i hope people use it ;)03:29
wolfspraulshort and up-to-date is good03:30
wolfspraulif David manages to get something on slashdot, it should point to your page right away, and only to that03:30
wpwrakwolfspraul: (wiki editing) the three sections i mentioned in the mail on which i copied. particularly "How is this patent free wireless?"03:30
wolfspraulin fact that may have been one reason for the rejection last time - I think they hate little articles that point in all sorts of directions03:30
wolfspraulme too, btw03:30
wolfspraulimagine you find out about something new, you only saw some snippets/words that caught your interest03:31
wolfspraulnow you want to learn more, but you have 5 links pointing in all crazy directions on the Internet03:31
wolfspraulnot nice :-)03:31
wpwrakyeah, that's a bit wikipedia style. every word a link :)03:31
wolfspraulfor someone reading an encyclopedia that's fine03:32
wpwraknot that i'd necessarily mind, but there should be at least a consistent story, and if there's one link that's more important than the rest, it should clearly stick out03:32
wolfspraulbut when I read a news post, and I want to read more, I want to be sent to one, and exactly one, place03:32
wpwrakkinda those "(more)" buttons03:33
wolfspraulI think they do that [more] to find out reliably how far people are actually reading03:35
wpwrakyes, that must be the real motivation03:36
wpwrakso, removed some more misconceptions. the 2nd sentence still gives me a bellyache. it's not using UBB, and yes, there is new testing software as well, it's not the main point. brr. confusing.03:42
wpwrakphew. cleaned up the worst. i hope :) let's see what nightmares will crawl out of it :)04:04
kyaktuxbrain_HxxHhzo: is it possible to order now from tuxbrain.com? (you mentioned that you wanted to update the shipment cost to Russia)05:16
DocScrutinizerwolfspraul: I have no idea about 802.22 yet - will dig into it a bit. If I find interesting stuff or remarkable factoids I'll let you know06:02
wolfspraulnice, thanks06:02
wolfspraulI think it's all quite speculative right now06:02
wolfspraulI'm looking for something that can be as free as ben-wpan, but longer range06:03
DocScrutinizergenerally longer range means more trouble about cert06:26
rejonyeah wpwrak wolfspraul that wiki page for ben-wpan was an attempt to get something down, and to understand/explain to others06:33
rejoni completely agree important to have a to-the-point page06:33
rejonpreviously stuff locked in brains and in git repos ;)06:33
rejonideally the project would even have a good name so don't need a page at all06:34
rejonanyway, i don't know how that got onto lwn06:34
rejonwho know who reads my shit06:34
qi-bot[commit] kyak: Store/restore alsa settings correctly. http://qi-hw.com/p/openwrt-packages/ec8d8f808:05
kyakxiangfu: hi! is it time now to stop building backfire images and rename trunk->master? :)08:09
xiangfukyak: I am waiting your order :)08:11
xiangfulet's stop the backfire building first. 08:11
kyakxiangfu: :))08:11
xiangfutrunk->master, I can do that weekend, not today. (I am not ready for that :)08:12
kyaksure, ok08:13
xiangfukyak: backfire build stopped, also I remove all backfire images.08:19
kyakxiangfu: ok!08:19
kyakxiangfu: btw, do you have control over dega-1.14-NanoNote-Edition.tar.gz?08:20
kyaki mean, the tarbal at SF08:20
kyakthere is some problem with it08:20
xiangfukyak: no.08:21
xiangfuI remember someone in mailing list have that right. let me find08:21
kyak1) the tarball seems to have changed on server. You have downloded it before to http://fidelio.qi-hardware.com/~xiangfu/openwrt-xburst.trunk-full_system/dl/, and the checksum doesn't match the current one08:21
kyak2) there is some problem with permissions, we have to do nasty things in Makefile 08:22
xiangfu2) yes. 08:22
xiangfu1) didn't notice this one. 08:22
kyaki think i'll start by adding the md5sum to Makefile. .so that your version would be re-downloaded08:22
kyakcurrently it fails on buildhost because patches can't be applied08:23
xiangfukyak: here: http://en.qi-hardware.com/wiki/Talk:Main_Page#Degap_on_Ben_9108:23
xiangfudega* in dl/ removed 08:24
xiangfuthe dega is very slow in Ben nanonote, but it's a good app.08:26
kyakyeah, it's pretty slow, but i played Mortal Combat a little bit :)08:26
kyakand Sonic the Hedgehog08:26
qi-bot[commit] kyak: dega: add md5sum for tarball http://qi-hw.com/p/openwrt-packages/d8162a708:27
xiangfu(patches can't be applied)?  I just tried, patch fine.08:29
kyakxiangfu: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.trunk-full_system-06102011-0026/BUILD_LOG.bz208:31
kyakcheck this BUILD log08:31
kyakchmod -R 755 /home/xiangfu/openwrt-xburst.trunk-full_system/build_dir/target-mipsel_uClibc-0.9.32/dega-1.1408:31
kyakApplying ./patches/001-fix-compile-in-openwrt.patch using plaintext:08:31
kyakpatching file Makefile08:31
kyakHunk #1 FAILED at 8.08:31
kyakof course it won't fail now if you've re-downloaded the tarball08:32
kyak(it only failed with the previous version of dega* in dl/)08:33
kyaki wonder what takes it so long to accept your patch for sshfs: http://patchwork.openwrt.org/patch/605/08:33
kyakit has been failing to build for months now08:34
kyakthe patch is trivial and works08:34
kyakxiangfu: how do you think, would it be usefull to have a file listing packages that failed to build ?08:40
kyakthat would save some work downloading and grep'ing that BUILD_LOG :)08:40
xiangfukyak: yes. it will be usefull for sure. 08:40
kyaki can have a look into doing that08:41
DocScrutinizer51wolfspraul: seems 802.22 is in fight with 802.11af08:41
DocScrutinizer51wolfspraul: last I heard  about the so called 'digitale dividende' from switching to DVB-T and thus freeing channels (white space) been that it shall get used by carriers to bring proper internet to rural areas here in D. Which would conflict wit the 'licence free' of 802.2208:44
xiangfukyak: great. thanks.08:46
kyakxiangfu: done :)08:48
DocScrutinizerwolfspraul: >> Grundsätzlich lässt sich bezüglich einer Einführung von CR-Geräten sagen, dass Sensing alleine nicht zu einer ausreichenden Interferenzminimierung führt, so dass nur bei Kleinleistungsgeräten, bei denen eine weitestgehende Interferenzminimierung nicht erforderlich ist, diese Technik alleine genutzt werden kann. Für andere Anwendungen, die CR-Technik in hierarchischer Nutzung einsetzen, ist Geo- lokalisierung mit 09:39
DocScrutinizerDatenbanken erforderlich.<< http://www.bmwi.de/BMWi/Redaktion/PDF/Publikationen/Technologie-und-Innovation/frequenzregulierung-fuer-software-defined-radio-cognitive-radio,property=pdf,bereich=bmwi,sprache=de,rwb=true.pdf09:39
DocScrutinizerwolfspraul: so it's basically lost case for long range non-cert trouble-free connections09:40
wolfspraulnot really, because this geo-locations with database backend is what people are working on09:41
DocScrutinizeras you will need a database to make sure you're not causing trouble for nearby receivers trying to catch weak signal of a prioritized broadcast station09:41
DocScrutinizerindeed, that's the prerequisite to use higher TX power, but that would need some sort of network connection AIUI, so you'd have access to that database09:42
DocScrutinizergiven the publication dates of related documents, I'd guess it'll take another few years til this tech is commonly adopted, except of course for 802.11af aka WiFi(R)09:44
DocScrutinizeronce WiFi 802.11af gets established it becomes a prioritized technology on the bands it uses, in a sense that other concurrent technology must guarantee interference free operation, which renders 802.22 basically incompatible and thus useless for the same freq ranges - my shoot from da hip take on it09:47
DocScrutinizerplus, as mentioned, I seem to recall statements of regulatory and gvmt speakers to the effect they want to give TV white space to carriers on a commercial basis, so they can roll out "wimax-II" on it, for DSL alike service in rural areas09:50
DocScrutinizerat least here in germany09:50
DocScrutinizerJFYI, will loosely follow up on the topic and occasionally toss over another remark, link, quotation... whatever09:51
DocScrutinizerone last comment: I guess chip market is sparse atm, as there's no real volume customer base09:53
DocScrutinizerfirst to come: probably 802.11af chipsets, which quite likely are "downscalable" to 802.2209:54
wolfspraulDocScrutinizer: yes correct, ICs is what I am mostly interested in09:58
wolfsprauland I think to a large part ICs will drive regulation too09:58
wolfspraulin that sense I don't care so much what a regulator in Germany has to say about this right now09:58
wolfspraulthey may say something else next year09:58
wolfsprauland in the end, once you have functioning products that you sell to end users directly, that's something that is very hard to stop in any case09:59
wolfspraulnot to say bust the regulator, but my point is - if the FCC moves forward with something, and creates a big market, and some IC makers invest a lot of money to make cheap and powerful chips, that creates a lot of momentum (and pressure) for regulators elsewhere...09:59
DocScrutinizeryou'll see this happen... for WiFi10:00
wolfspraulI'll read up on 802.11af10:00
wolfspraulGerman regulators don't have enough critical mass anymore to drive large-scale IC investments10:01
DocScrutinizeronce the WiFi association sells 802.11af in large numbers, you're screwed on CR 802.2210:01
wolfspraulFCC does, but even they need to speed up, otherwise something is developed in China and sold throughout the entire 3rd world, creating a momentum that will then force the FCC to just slap their stuff around it or become irrelevant10:01
DocScrutinizerIt's about WiFi association mafia, not any German regulator10:02
wolfspraulok fine - if you see interesting chips please holler10:02
wolfspraulthe world is a big place, many things are happening :-)10:02
DocScrutinizerwolfspraul: have you ever seen FCC or USA at large care about rest of the world? At best in a sense like "they use GSM900/1800, so let's do it differently here to protect our markets"10:03
DocScrutinizer(holler) sure thing10:04
wolfspraulDocScrutinizer: I don't know, I don't follow regulation closely10:05
wolfspraulI think there is also competition between regulatory bodies10:05
wolfsprauleverybody wants to have the 'important' standards10:05
DocScrutinizerthen USA FCC is performing really poor on that :-P10:05
wolfspraulso sometimes a standard from a competing body is just renumbered and reissued. quite a bit of heat between ITU and IETF for example.10:06
wolfspraulI'm just looking for great chips10:06
wolfspraulin the end really great chips require a lot of investment10:06
wolfspraulso whoever invests that must have a strong idea how to sell the volume required to recoup the investment10:06
DocScrutinizerevery now and then I get to know about a new chip. I'll drop you a note instantly10:06
wolfspraulonce that has happened, I don't care how or why it happened, I just want to source that chip for a few USD10:06
wolfspraulthe next problem is regulatory oversight in particular markets10:07
wolfspraulwhich I have no problem with, just need to understand and see how to comply10:07
wolfspraulbut in that order, for me :-)10:07
wolfspraulotherwise I get lost debating all sorts of hypothetical regulation10:07
wolfspraulthe white-space stuff looks like it has some momentum10:07
DocScrutinizersure, just saying WiFi is the only interest group with enough power to push chips10:08
wolfspraulhow much? don't know10:08
wolfspraulI'm sure about that10:08
wolfspraulalso I don't think potential multi-billion USD investments are driven by regulatory decisions, I think it's the other way round10:08
wolfspraulso you are right, Wi-Fi has huge capital, they can make chips, and drive regulation10:08
DocScrutinizerthat's the reason why I think 802.22 won't fly in the end10:09
wolfspraultuxbrain_HxxHhzo: how are atben/atusb sales so far?11:17
wolfspraulunless you want to hide your riches and don't tell us :-)11:18
qi-botThe build was successfull, see images here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.trunk-full_system-06162011-0817/11:49
wpwrakwolfspraul: he's probably busy training the new shift of 100 factory workers in the flashing and testing of the boards ;-)12:12
whitequarkwpwrak: if they'd be spanish, that would be quite costly. but I think he could just order some chinese ones wholesale...12:52
wpwrakwhitequark: yeah, a container of 'em :)12:53
methril_worknice, Qi-HW is on LWN14:48
wolfspraulmethril_work: sorry I lost track. Didn't you get a milkymist one too?14:53
methril_workwolfspraul, i get14:54
methril_workwolfspraul, i'm not having time t owork on ... right now14:54
wpwraktuxbrain_HxxHhzo: btw, are my ben-wpan boards already on the way ? if yes, can you please send me the tracking number so that i know when to pay attention ?14:56
qi-bot[commit] Stefan Schmidt: usb/dfu.h: Include usb.h to avoid gcc warning http://qi-hw.com/p/ben-wpan/a48c6dc15:02
wolfspraulmethril_work: ah cool15:13
wolfspraulI just forgot15:13
wolfspraulno worries about no time, the product gets better every day :-)15:13
wolfspraulby the time you are ready, it will be really cool :-)15:13
methril_workwolfspraul, i hope to get ready any time soon (TM) :)15:14
wolfspraulwerner just helped fixed another nasty small hardware bug today :-)15:15
methril_workwerner is amazing :)15:16
methril_workis DocScrutinizer or wpwrak ?15:16
wolfspraulWerner = wpwrak; Joerg = DocScrutinizer 15:18
methril_workUps!! it's right!! 15:18
DocScrutinizeris it spoken slowpan, or blowpan, or glowpan?16:11
DocScrutinizer~spell 6lowpan16:11
Action: whitequark wonders how to spell "Joerg" correctly16:14
DocScrutinizerJörg ;-D16:15
Action: kyak wonders how to spell "kyak" correctly16:17
DocScrutinizermy question not been about spelling  though. See scasy err SCSI16:28
wpwrakDocScrutinizer: some seem to want to pronounce it "slowpan"16:31
wpwrakDocScrutinizer: the official name for IEEE 802.15.4 would be Low-Rate Wireless Personal Area Network. so "slowpan" isn't too bad.16:31
wpwrakDocScrutinizer: (and sounds a lot catchier ;-)16:32
kristianpaulhum, seems gpredict TLE file is not well update, or my closed source gps-receiver is laying?18:44
kristianpaulmissing in gpredict prn 16, 23 and 19..18:53
kristianpaullets find other tle better18:53
rejonyou guys see we got on engadget18:57
rejonthere's no freaking link to buy them or the info about them!!!!18:59
kristianpaulbut the source is tuxbrain19:00
kristianpaulalso prices include taxes it seems19:01
kristianpaulah, okay for some reason gpredict mismatch the name for this satellites19:04
rejonread the comments19:06
wpwrakgood comments so far. it was clear that a lot of people wouldn't like the speed. some will also discover the range and find more to bitch about :)19:08
zearwpwrak, not like bt was ever useful for anybody despite every second handheld device having it ;)19:24
zearpeople will simply bitch about anything, i for one love the idea of an openhardware wireless standard19:25
rejonthe register article is pretty good wpwrak 19:39
kristianpaul"Slow, short range, incompatible ... but it's pure"19:41
wpwrakzear: (bt) oh, don't say that. when my internet-over-catv goes down, bt to my gprs phone is the only way out. it saved me a few times ;-)19:44
wpwrakrejon: hmm, they got the 6lowpan a little wrong. perhaps also a bit confused about zigbee. but it worked - they took it ;-)19:46
wpwraknext: heise.de ;-)19:46
wpwraknow they can't possibly ignore it :)19:46
zearwpwrak, well, i guess it can be used that way, yes19:47
zearbut it's not like this couldn't be done with wpan19:47
zearthat are just rare situations when you need to use bt, but it's not like there are bt hotspots offering free internet everywhere ;)19:48
wpwrakmore balanced reaction on average on the reg. very large std deviation ;-)19:49
wpwrakzear: well, like everything WPAN, BT is for the short range. don't expect to find IEEE 802.15.4 anytime soon at your airport ;-)19:50
dvdksaw jon's mail about the engadget article?20:19
dvdkmaybe some people around here want to add something informative in between the ranting comments?20:20
wpwrakoh cool. i can write but i can't post :-(20:34
dvdkwpwrak: need to log in (openid)20:35
wpwrakif anyone wants to post this, please feel free: http://pastebin.com/QdY46RuJ20:35
dvdktry a different browser :)20:36
wpwrakweb-based interfaces just suck20:36
dvdkno, java-script based interfaces suck20:36
dvdkshall i post it under my name?20:36
dvdkor "forwarded from the designer of wpan?"20:36
wpwrakdvdk: please feel free to use it in whichever way you feel appropriate :)20:37
dvdkok, then i'll post the characters in reverse order.20:37
dvdkbetter, bit-reverse :)20:37
wpwrakdvdk: i don't think it really matters who claims authorship20:37
wpwrakdvdk: don't forget the double-rot13 !20:37
dvdkbut, it matters, whose opinion it is, doesn't it?  crazy engineer, vs. troll etc.20:38
kristianpaulcurious, lwn article havent got comments yet20:39
wpwrakdvdk: naw, opinions are opinions. you believe the ones you like, you find excuses for disregarding those you don't20:39
wpwrakkristianpaul: maybe their troll deterrence is better ;-)20:40
kristianpauloh, yes :_)20:41
dvdkwhat about the /. article.  still need that now that the cat is out of the bag? :)20:49
wpwrakhmm, i guess it can't hurt to post something that points straight to my overview page.20:54
wpwrakotherwise, they'll copy from the others before too long.20:54
wpwrakwith uncertain result.20:54
dvdkwpwrak: talking about /. or engadget?20:54
dvdkyes.  i think conciseness means low count of links with concentrated information.  maybe only your page + link to nanonote wiki page20:57
wpwrakthat sounds quite reasonable, yes20:58
Action: dvdk is looking for the /. draft link20:59
dvdkfound it http://en.qi-hardware.com/wiki/Slashdot-20110615-benwpan20:59
wpwrakthe wiki page adds an air of mystery, which may have been good in the case of the reg, but in general, it may be more confusing that helpful20:59
wpwrak(the Ben_WPAN wiki page)20:59
wpwrakthem: http://www.theregister.co.uk/2011/06/17/qi_hardware/21:00
dvdkahh, yes i know them21:00
dvdk"potentialy suitabley" too long21:02
wpwrakoh, not is has grown sections ;-)21:02
wpwrak"likely" ?21:03
wpwrak"Nokia Windows Phones will launch in six European nations first, Finland not among them" i think they're really set on doing *everything* right ;-)21:05
dvdkwpwrak: (sections) still missing a TOC though :)21:06
wpwrakand maybe an appendix with formulas ? ;-)21:06
dvdkok, rewrote the article.  now much more concise.  people will have to click through werner's page to get to tuxbrain or see details.21:40
dvdk(  http://en.qi-hardware.com/wiki/Slashdot-20110615-benwpan )21:40
dvdk(it's the v2 at the bottom)21:40
Action: dvdk is fixing typos21:43
dvdkgoing to be offline until saturday around 10:00 utc 22:01
rejonwpwrak, did you get your FISL talk in?22:53
--- Sat Jun 18 201100:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!