#qi-hardware IRC log for Wednesday, 2011-06-15

wpwrakhmm, can i rename a page in the qi-hw wiki ? e.g., Ben_WPAN -> Ben_WPAN_Background (there's way too much philosophy before this comes even near any technical details)01:19
wpwrakand while the philosophy can be useful, it's not well written. (e.g., jumps from 3rd person to 1st person and so on)01:27
DocScrutinizermediawiki page rename is a pita, I heard. Basically you copy and delete I guess01:37
wpwrakheh ;-)01:37
wpwrakwell, what i'll do is make a static page in HTML, like i did for ubb-vga01:37
DocScrutinizer(business model) too late, dude, way too late ;-D01:38
wpwraklife is too short to wrestle with mediawiki :)01:38
wpwrakDocScrutinizer: (shit blog) ah, i was afraid that someone may have had that brilliant idea already :)01:38
wpwrakDocScrutinizer: maybe my plan B will work, though. post a daily video clip of the killing of a kitten. or has that frontier been crossed too ?01:39
DocScrutinizer3 years lated they invented arpanet01:40
DocScrutinizersexual connotation aside: http://www.google.de/search?q=killing+kitten+video01:42
wpwrakrejon: here's your opportunity for a moment of greatness: clean up the language of http://en.qi-hardware.com/wiki/Ben_WPAN01:47
wpwrakrejon: ... such that it meets basic courtesy when using the written word :)01:47
wpwrakrejon: meanwhile, I'll do a shorter version of the technical side with only the minimum of underlying philosophy etc.01:48
DocScrutinizerawesome kitten-hands-up pic in that youtube vid01:56
DocScrutinizerhttp://share.ovi.com/media/joerg900.pennystore/joerg900.1012902:03
rejongreat wpwrak 02:25
rejonhow is FISL plans coming?02:25
tuxbrain_awayplease spread retweet, ident, facebook it, linkeding it, send to other related mailing lists, sumarizing spam it as far as your ethic can hold02:25
tuxbrain_awayhttp://www.tuxbrain.com/en/content/free-and-open-replacement-wireless-lan02:25
tuxbrain_awayin spanish02:27
tuxbrain_awayhttp://www.tuxbrain.com/content/una-alternativa-libre-y-abierta-las-comunicaciones-wi-fi-wireless-lan02:27
rejonok, doing02:29
rejonwpwrak, tuxbrain_away i have a question, what are the next steps beyond the sales side?02:29
wpwrakrejon: (fisl) haven't heard from them for a while. not sure if they silently dropped me.02:30
rejonno they haven't02:30
rejonthey have hotel for you02:30
rejonthey are super brazilian02:30
rejonin handling these things02:30
rejoni forgot about it a bit and then looked at the calendar today02:30
wpwrakrejon: oh, they even have organized a hotel. wow !02:31
wpwraktuxbrain_away: great page !02:38
rejonjust email them02:43
kristianpaulwpwrak: i can be rename, or technically moved should be called02:45
kristianpaulDocScrutinizer: no copy and delete as you will lost editing history02:45
kristianpauls/i/it02:46
kristianpaulwpwrak: if is okay for you can move it to Ben_WPAN_Background02:46
DocScrutinizerkristianpaul: I'm well aware of that02:47
DocScrutinizertuxbrain_away: already spread the word :-)02:47
kristianpaulDocScrutinizer: he :-)02:48
DocScrutinizerkristianpaul: 3 thoughts in there are with my contribution. ok, 1.502:48
wpwrakkristianpaul: naw, let's keep the name for now. my page will be html-only anyway. with mediawiki, doing this would take forever02:49
DocScrutinizerI can't do different than just publish it :-)02:49
kristianpaulwpwrak: np, 02:49
kristianpaulwpwrak:  (html-only) thats cool02:49
kristianpauleven more if is tracked with git :-D02:50
DocScrutinizeraahyeah, bisect html02:50
DocScrutinizerwhy not02:51
DocScrutinizer:-)02:54
wpwrakkristianpaul: (git) of course02:54
DocScrutinizerwpwrak: what's that curing thing? you drowned the atusb in resin?02:54
DocScrutinizeraah, and what's the roadmap for future porcessors/SoCs?02:56
DocScrutinizer(superbrazilian) sounds like the next now term in my vocbulary after "to tepco"02:58
DocScrutinizernew*02:58
ds3DocScrutinizer: what was factor pushing for an atmel chip?03:02
DocScrutinizerask wolfgang, when he joins again. OR anybody else here, I'm the least involved person :-)03:03
DocScrutinizer"ask the cook, I'm just the waiter"03:03
ds3donno who's who... guess I'll lurk a bit03:04
Action: kristianpaul thinks that was because gcc and hobbist comuminty awareness 03:04
DocScrutinizerwpwrak: werner almesberger03:04
DocScrutinizerwolfgang (spraul)03:04
kristianpaulalso i heard that usb stack was well supported?03:04
DocScrutinizertuxbrain_away: himself03:04
kristianpaulanyway, ask wpwrak indeed03:05
kristianpaulgn803:05
ds3I see.03:05
DocScrutinizersourcability another main factor for small projects03:06
DocScrutinizerincredibly hard to get <100k of some of the good big SoCs03:06
ds3I see. 03:07
DocScrutinizerwhen you order 10^7 you probably can get your custom made SoC, all included. For 1000 you have to take what you can get03:08
ds3I'd think for the SoC, ARM is a more natural choice then a MIPS03:09
DocScrutinizerdepends03:09
ds3I like my mobile devices to have a long battery life but that might just be me03:10
DocScrutinizerbut honestly I don't know what's been the design rationale for Ben03:10
DocScrutinizeroh, MIPS are rather great on power conservation, some of them03:10
ds3there has to be a reason why they don't show up often in low power devices...think the Alchemy was the only one that was sold for that03:11
ds3OTH, I could be wrong/skewed due to the large representation on the router SoC's03:12
kristianpaulah,almost forgot i think woflgang suguested a SiGE chip for wpan, but that means a SDR implementation in the nanonote03:15
kristianpaulwell,i think werner said this atmega rf chip also support sdr, so i think thats a plus too03:15
wpwraktuxbrain_away: s/atben y atusb to enchance/atben and atusb to enhance/ ;-)03:30
wpwraktuxbrain_away: btw, you should really indicate that your prices contain EU taxes. US visitors will have kittens when they see your prices.03:33
kristianpaulwpwrak: (taxes) but when simulating a buy the're droped03:34
Action: DocScrutinizer frowns at 4.99 for http://service4handys.de/index.php/de/_/f-a002462 and the fact you can't do that at home. Got one, mod is hard03:34
kristianpaulthey*03:35
DocScrutinizernobody's going to simulate a buy of sth that's already way beyond03:37
Action: kristianpaul hides03:38
Action: DocScrutinizer disassembles that huge yellow plastic blob03:39
DocScrutinizerto reveal the 2 resistor secret inside, that makes the modem enter service mode when sensing the magic on BSI of battery03:40
wpwrakfirst draft: http://downloads.qi-hardware.com/people/werner/wpan/web/03:42
kristianpaulfirt question for a coming FAQ, can i increase transmit power?03:45
DocScrutinizerMOAR POWER!03:45
DocScrutinizeranybody a good source for pogopins?03:46
wpwrakkristianpaul: yes, but you'd have to add 1-2 chips03:46
wpwrakDocScrutinizer: digi-key :) search for spring-loaded03:46
DocScrutinizerta03:46
kristianpaul"While not an immediate project goal, it may be possible to write firmware for atusb that turns it into a transparent gateway for legacy operating systems and protocols."03:47
wpwrakkristianpaul: (taxes) yes, once you've logged into the shop it knows where you live and gives you the non-EU price. nice idea, but confusing.03:47
DocScrutinizerwhile I can't do the yellow blob for 4.99, I still may drill additional holes in that one03:47
wpwrakkristianpaul: do you like the "legacy" ? :)03:48
kristianpaulwell. is a bit confusing, but surelly is my crazy english03:48
DocScrutinizeranybody a clue abut what e.g logitech is using for their proprietary 2.4GHz wireless mouse stuff?03:49
kristianpaulbut i get that include bill? :-)03:49
wpwrakDocScrutinizer: (curing) tuxbrain put sugru around an atusb: http://en.qi-hardware.com/wiki/File:Sugrued_atusb_01.jpg03:49
DocScrutinizer~sict sugru03:49
DocScrutinizer~dict sugru03:49
DocScrutinizermeh03:49
kristianpauls/get/guess03:49
DocScrutinizer/invite infobot03:49
wpwrakDocScrutinizer: but it dampened RF a lot. now we need to see if that is because of the sugru per se or because of the solvent in the sugru.03:49
DocScrutinizersugru sounds like sago here03:50
wpwrakds3: (atmel chip) open documentation, availability, low price03:50
kristianpaulhum, Sige was bought by skyworks inc03:51
DocScrutinizeris that sugru some playdoo stuff?03:51
DocScrutinizernever heard about it03:51
wpwrakds3: (that's for the transceiver) for the MCU: mainly popularity in the community. i had a perfectly working design based on silab's 8051 chips before. but 8051 isn't quite as popular as avr.03:52
DocScrutinizerWUT?03:52
DocScrutinizer8051 is BASIC03:52
Action: kristianpaul still like PICs ;-)03:52
DocScrutinizeruaergh03:52
kristianpauli never had a 80xx btw, so :-)03:53
DocScrutinizerget an atmel 89c1051 or sth like that03:54
DocScrutinizerawesome03:54
wpwrakkristianpaul: (sdr) no, the at86rf23x don't do sdr. afaik, sdr is still big and high-power (well, compared to ben-wpan. you don't need to get a whole USRP, but still ;-)03:55
DocScrutinizer20pin DIL, like 3 external components, GPIO can drive 16mA+ each03:55
DocScrutinizerfor instance03:55
DocScrutinizerof course they have other formfactors and chip flavours as well03:55
kristianpaulatmel still not very widespread on my country, so, well i can run navre on the milkymist at least ;)03:57
kristianpauloh right, namuru core adaptation for milkymist CSR here we go :-)03:59
DocScrutinizerthere are exactly 2 things I hate til hell freezes over; assembling a device and then noticing you forgot battery, and soldering a plug to a cable and "duh the friggin hood"03:59
kristianpaulfor now just one correlator, as seem with 12 it will eat 20k LE03:59
wpwrakah, firmware ...04:00
kristianpaulha, looks this http://www.skyworksinc.com/Products.aspx?CategoryID=6204:01
kristianpaulnah,datasheets looks poor04:02
kristianpaulany way, gn804:02
Action: DocScrutinizer wonders why speedevil isn't here - also a GPS hacker04:03
wpwrakany comments on my overview page ?04:04
DocScrutinizerurlrlrlr04:04
wpwrakDocScrutinizer: still at http://downloads.qi-hardware.com/people/werner/wpan/web/04:04
DocScrutinizerwpwrak: unify orthography: seems Wolfgang calls then AtUSB and AtBEN04:07
wpwraki usually call them atben, sometimes ATBEN, but i'm trying to shake that latter habit04:08
wpwraknot sure where wolfgang got the AtBEN from :)04:08
DocScrutinizerplease agree on one writing _now_04:10
DocScrutinizer~spell homebrewn04:11
wpwraki agree with "atben" and "atusb" :) feel free to capitalize it or not at the beginning of a sentence.04:11
DocScrutinizernah, first of all I like to get topic right04:12
wpwrakDocScrutinizer: "homebrewn" should be fine. it's semi-jargon.04:13
DocScrutinizeryou might want to explain if it's possible to use two atusb together, without Ben04:13
DocScrutinizereven have 'networks' with multiple atusb04:15
DocScrutinizerit's not entirely clear if the hw supports it04:16
wpwrakyeah, that seems to be a bit of a FAQ04:18
wpwraki'm also adding a warning that it's not SDIO04:18
DocScrutinizererrr?04:19
wpwrakit's bit-banged SPI04:19
wpwrakplus few control signals04:19
DocScrutinizerah, whatever that might explain to unbiased readers04:20
wpwrakno MMC/SD/SDIO protocol04:20
SpeedEvilIs there a compelling reason you can't bitbang SDIO?04:20
SpeedEvil(other than the obvious performance and lack of drivers issue)04:20
DocScrutinizerhi SpeedEvil :-D04:20
SpeedEvilhi04:20
wpwrakSpeedEvil: bit-bang -> SDIO peripheral works just fine (been there, done that ;-). but SDIO host -> on-SDIO peripheral won't work04:21
SpeedEvilI'm pondering doing the reverse - making a not-SD card to goin a (micro)SD slot04:21
SpeedEvilAnd be driven by bit-bang04:21
SpeedEvil(compass for n900)04:21
wpwrakSpeedEvil: that is, unless you can convince the host to give you access to plain GPIOs04:21
SpeedEvilTrue. At least on the n900, you can.04:21
DocScrutinizerSpeedEvil: there's been a mentioning of correlators some minutes up in backscroll04:23
SpeedEvilhmm04:23
DocScrutinizerthought this might be of some interest for you04:23
SpeedEvilMy client lacks backscroll like that. :)04:23
DocScrutinizerwe got chanlog04:23
DocScrutinizer[2011-06-15 05:59:25] <kristianpaul> for now just one correlator, as seem with 12 it will eat 20k LE04:24
SpeedEvilLE?04:25
DocScrutinizer(milymist with FPGA, aiui)04:25
DocScrutinizermilkymist even04:26
SpeedEvilIdeally you want a couple of thousand correlators per sat.04:26
SpeedEvilBut...04:26
DocScrutinizer[2011-06-15 05:59:03] <kristianpaul> oh right, namuru core adaptation for milkymist CSR here we go :-)04:26
DocScrutinizerbtw that's been about all04:26
SpeedEvilI had some GPS proto-code-diagrams that looked sort-of-plausible to do on a STM32/70MHz04:27
DocScrutinizerping him if you're interested04:27
SpeedEvilDecoding bitstream from a GPS chip like maxim's http://www.maxim-ic.com/app-notes/index.mvp/id/344704:27
SpeedEvilNasty hand-rolled assembler code.04:27
SpeedEvilAnd given you can now get GPS modules at ~$5 that output NMEA...04:28
DocScrutinizersorry, reconnect04:29
DocScrutinizerassembler was last04:30
SpeedEvil And given you can now get GPS modules at ~$5 that output NMEA...04:30
SpeedEvilAt least you need something that can take a 18mbits/s bitstream and DMA it04:30
SpeedEvilAnd then you process it.04:30
ds3who has GPS modules for $5?04:30
DocScrutinizerindeed, just interesting for special development04:30
SpeedEvilBitsliced DES is interesting04:31
SpeedEvilhttp://www.aliexpress.com/product-gs/423767224-gps-module-SKG11B-12-12-4mm-the-smallest-dimension-wholesalers.html04:31
wpwrakadded a compatibility section: http://downloads.qi-hardware.com/people/werner/wpan/web/04:31
SpeedEvilThat was one I happened to have on bookmarks, it's not the cheapest out there.04:32
ds3that isn't exactly useable04:34
ds3it lacks an antenna04:34
SpeedEvilWell - yes.04:34
SpeedEvilI don't know the parameters that were wanted.04:35
wpwrakSpeedEvil: i think kristianpaul's chip is a bit slower. 8 Mbps or such.hmm, can't find the data sheets :-(04:35
SpeedEvilThere are various sorts - you can trade bit-depth for datarate04:35
SpeedEvilIf you've got an 18MHz sample of the IF, or a 8MHz at 3 bits, it's more or less similar.04:36
SpeedEvil2 bits04:36
wpwrakrejon: how's the style cleanup going ?04:47
rejonwpwrak, great to get a big picture up top on that article04:49
rejonwpwrak, looking at now04:50
qi-bot[commit] Werner Almesberger: web/: Ben-WPAN overview page http://qi-hw.com/p/ben-wpan/ea7103e04:55
wpwrakrejon: (progress) great !05:00
wpwrakany last changes before i hit [Submit] on /. ? :)05:01
rejonyeah, i'm scanning05:01
rejonalmost done05:01
DocScrutinizerhaha05:01
qi-bot[commit] Werner Almesberger: web/Makefile: also upload the ~800 pixels images http://qi-hw.com/p/ben-wpan/394ca5605:02
wpwrakhuh. irc reset. how odd.05:08
SpeedEvil(06:07:27 AM) wpwrak left the room (quit: Read error: Operation timed out).05:09
DocScrutinizerirc / freenode is sick today. See mohammad* in #maemo. Suddenly an alter ego popped up from "nowhere"05:10
DocScrutinizernetsplits as well, which might have made that ZNC reconnect and thus re-register mohammad-05:11
wpwrakSpeedEvil: that's a lie ! i didn't even leave my chair in the last few minutes :)05:12
whitequarkwpwrak: maybe that was an office chair, one of these on small wheels?05:12
wpwrakwhitequark: hmm, it has wheels. kinda tricky to get though the door with it, though. and the path there contains some obstacles as well :)05:14
rejontuxbrain_away, 05:19
rejonoi05:19
rejonaren't these available?05:19
rejonwpwrak, this needs an update i believe: http://en.qi-hardware.com/wiki/Ben_WPAN#Progress05:20
wpwrak(available) what ?05:21
rejonnm05:21
rejoni'm fixing05:21
tuxbrain_awayrejon : ?05:21
rejontuxbrain_away, "Also, a lot of software work still needs to be done. The driver level code for 802.15.4 has been demo'd. A production test program is working which verifies that a board is not severly compromised.05:22
rejonThere are a small number of programs to send and receive data. However, there's still a lot to do in order to have a useable 6LoWpan stack."05:22
rejonis that accurate still?05:22
wpwrakyes05:23
wpwrakwhat we have at the moment is a very basic IEEE 802.15.4 stack in the kernel. it lacks things like CSMA and ACKs, let alone encryption. so it's really *very* basic.05:24
wpwrak6LoWPAN code has been ported over on one or two occasions, but that's out of date now and was considered defective back then. so this needs another effort.05:25
rejonwpwrak, how are these tasks tracked?05:25
rejonhow can people get involved?05:25
rejon(after buying one of course :)05:25
wpwrakthe place to go should (still) be the linux-zigbee list05:25
rejonno i mean on this project05:25
wpwrakit's very quiet there, though05:25
wpwrakah, our mailing list ?05:26
rejonso all the tasks are in your brain you are saying?05:26
wpwraksee the bottom of http://downloads.qi-hardware.com/people/werner/wpan/web/05:26
wpwraki actually spelled them out there :) well, not the details05:27
wpwrakthe 6lowpan stack isn't qi-hw specific. so people who want to play with that should go to linux-zigbee05:27
wpwrakor, better, find some sponsoring for me, and i'll be glad to solve that little problem ;-)05:27
rejonyeah, that is what i'm getting at05:28
rejonwpwrak, i want to know the problems left and the next steps05:28
rejondo you want to tell me now, or add to the page?05:28
wpwrakmy plan for now is to first clean up the infrastructure that's already in the kernel05:29
wpwrakstarting with the at86rf230/231 driver (which doesn't support atusb yet, only atben)05:29
wpwrakthen tackle the IEEE 802.15.4 protocol stack until it works properly. that should give us decent dirtpan performance05:30
wpwrak6lowpan is less certain. by the time i'm done with dirtpan working well, i should find me some real work. so i don't know it i'll have much time to work with 6lowpan.05:31
wpwrakwhat i'll do is contact atmel and see if they're interested in sponsoring such an effort, but i don't really expect this to lead anywhere05:33
rejonyeah, that atmel thing will have to be handled well so they don't get their hooks into it05:34
rejonwpwrak, i have another source that is interested too05:34
rejoni will follow up with05:34
rejonwpwrak, tuxbrain_away what about targets?05:34
rejonjust send and see?05:34
wpwraktargets ?05:35
rejoni mean target market?05:35
rejonjust hackers for now?05:35
rejonalso, is there a case for the exposed circuitry?05:35
wpwrak(market) i think at the moment it's mainly of interest to the ben-owning community. i don't know about the ieee 802.15.4-centric market. it may be huge, but i don't know if these products are of interest to them.05:36
wpwrakno case yet. that's still under investigation.05:37
rejonok great05:37
rejonhow many did you guys make?05:37
wpwraktuxbrain put sugru on an atusb but it dampened the signal considerably. then someone got his gadgets for a photo session and we have to wait for them to return before we can see if curing improved the sugru's RF transparency05:39
wpwraki think something like 120 atusb and 135 atben. 3 atusb didn't work. all the atbens passed testing.05:40
wpwraktuxbrain_away could confirm the exact numbers.05:40
rejonwpwrak, what about hardware improvements?05:42
rejonanything off the top of your head?05:43
rejonbeyond the case05:43
tuxbrain_awayrejon let me elaborate later on, but until we have a case defined and the software is not fully plug and play and atusb is not supported our main target is hackers and NN owners, then well let me elaboorate and alse will add to the study the reation of this "for hacker" releases05:43
wpwraknothing that needs fixing immediately. one thing that's completely missing is certification. that may or may need hw changes.05:44
rejongreat05:44
wpwrakthe boards are a little tweakable: 1) you can make a one-way modification to add an U.FL connector (not for the faint of heart, but doable) 2) there is an unpopulated capacitor and a 0R resistor that can be used/changed for RF tuning.05:45
wpwrakbesides that, i don't see any urgent need for hw changes. one could think of adding an RF amplifier, to extend the range. but that's significant new work (~2 more chips, and so on)05:46
wpwrakalso, the transceiver supports antenna diversity, which could be interesting05:47
wpwrakmeanwhile, there's also one competing chip that could be interesting. that one an internal balun, reducing the component count. but of course, totally incompatible in terms of layout and low-level driver.05:48
rejonok, great05:49
wpwrakthat's the sort of things to consider if tuxbrain gets a order for 100k+ units :)05:49
rejonexcellent05:49
rejonadded http://en.qi-hardware.com/wiki/Ben_WPAN#Future 05:49
wpwraktuxbrain_away: wouldn't you like to sand, flash, and test 100k+ units ? ;-))05:49
rejonhttp://en.qi-hardware.com/wiki/Ben_WPAN#HOWTO_Get_Involved05:49
rejonand http://en.qi-hardware.com/wiki/Ben_WPAN#Outcomes05:49
rejonand cleaned up the whole page, and added TOC05:50
wpwrakoh dear, direct irc quotes ;-)05:50
rejoni cleaned them up05:51
rejonwpwrak, tuxbrain_away this is some badass shit!05:52
rejonlike totally perked me up05:52
rejongreat work dudez05:52
rejontotally inspiring!05:52
wpwrakthanks :)05:53
rejoni gotta cop me some in brazil05:53
tuxbrain_awayI have modified this due NN doesn't have USB host :P Until we have a case defined, the software is fully plug-and-play and atusb is supported on kernel main line, the market for the Ben WPAN units is hackers and Nanonote owners. 05:54
rejoncool tuxbrain_away i already fixed05:54
tuxbrain_awayregarding certification, when teh moment comes I have two colaborator on universities that can do so05:57
wpwrakrejon: page looks a lot better now. thanks !05:57
tuxbrain_awayyeah thanks rejon :)05:57
rejonyou guys did the hard work!05:57
wpwrakwe can merge the remaining bits later05:57
rejontuxbrain_away, wpwrak i started to rev up the press engine, but thinking might be better for batch 205:58
wpwraktuxbrain_away: (university resources) always good to have :)05:58
rejonbut come on, companies come out with vacuous announcements without products EVERYDAY05:58
rejonnerd channels better imo05:58
wpwrakrejon: yeah, we should release a roadmap with Ya and 2-3 successors firmly announced right now. wolfspraul ? ;-)05:58
tuxbrain_awayto consider this a real sucess I have to get rid of this first batch by the end of the month :)05:59
wpwraktuxbrain_away: two weeks ? wow :)05:59
wpwraktuxbrain_away: btw, do you still have the sugru device or is it among the ones you lent to that other guy ?06:01
wpwraktuxbrain_away: because, if you still have it, you can just test with another atusb. the tests don't have to be atben-atusb06:01
rejonwpwrak, you know the drill ;)06:02
rejonok, added pictures: http://en.qi-hardware.com/wiki/Ben_WPAN06:04
tuxbrain_awaywpwrak: ok I will try this later, just point me what will be the order, it can be done in same computer?06:04
wpwraktuxbrain_away: yes. you need to specify the path to the USB device. digging out the posting ...06:07
wpwraktuxbrain_away: can you make sense of what i wrote here ? http://lists.en.qi-hardware.com/pipermail/discussion/2011-April/007715.html06:10
wpwrakthe command would simply be: atrf-path -g usb:... usb:... 1006:11
wpwrakyou can debug the usb:... with  atrf-id -d usb:... try until it works ;-)06:12
wpwrakrejon: (pics) nice !06:13
wpwraktuxbrain_away: btw, i really like http://en.qi-hardware.com/w/images/thumb/8/8b/Atben_atusb_prod_10.jpg/800px-Atben_atusb_prod_10.jpg06:13
wolfspraultuxbrain_away: if you have a slashdot account, vote up David's submission at http://slashdot.org/submission/1648862/A-Free-and-Open-Replacement-for-Wireless-LAN06:15
wolfspraulmaybe the voting helps, maybe not :-)06:15
rejonis there some policy against asking people to help vote it up publicly?06:17
qi-bot[commit] Werner Almesberger: web/: added zoomed-in view of atusb in action http://qi-hw.com/p/ben-wpan/7e1b47b06:23
qi-bot[commit] Werner Almesberger: web/index.html: removed duplicate IMG tag http://qi-hw.com/p/ben-wpan/cce781006:23
wpwrakrejon: don't ask, don't tell ;-)06:25
wolfspraulrejon: why should there be such a policy?06:33
wolfspraulthe opposite is right, if you want your friends to vote it up you should do so publicly06:34
qi-botThe build has FAILED, see log here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.trunk-full_system-06142011-0345/06:51
rejonok wolfspraul tuxbrain_away wpwrak posted up on fab blog and on qihardware twitter and identica07:54
wolfspraulrejon: nice!08:03
rejonbrings me out of microblogging retirement...i setup ping.fm account for qi too08:06
rejonok, night dudez08:06
kyaktuxbrain_away: any update about shipment cost to Russia?10:32
dvdkok, the /. article failed11:06
dvdknext try: ycombinator.com.  you can upvote the article by pressing the uparrow sign: http://news.ycombinator.com/submitted?id=dvdkhlng11:07
lekerneldvdk, sometimes it takes a few days11:08
wolfsprauldvdk: how do you know that it failed?11:22
dvdklekernel: (/.) but it's already invisible on slashdot.org/recent even when going far, far  back. thought /. editors would only look at /recent. don't know for sure, though11:23
wolfspraulwe wait a week then repost? :-)11:24
dvdkwolfspraul: but different article next time.  already building one up in my head.11:27
dvdkmaybe wait 2 weeks.  what's the half like of memory in the average /. editor?11:27
wolfsprauldvdk: thinking about the /. editors, I cannot imagine that they just pick articles out of the incoming flow of posts11:45
wolfspraulI would do it in a multi-layer system where I first mark good posts, then later review them again, possible throwing out some more, and finally selecting based for inclusion on the frontpage so that the flow of articles on the frontpage is interesting11:46
wolfspraulthat would mean - it could very well still pop up after a few days, like lekernel said. complete guess of course :-)11:46
wolfspraulthey have to worry about the readability of the frontpage, a good mix of articles in different categories, etc.11:47
wolfspraulwe will know soon :-)11:47
wpwrakmy guess would be a ~24 h-ish timeout on /. after that, they may consider it not bleeding edge enough. maybe a few hours more.11:51
jow_laptopslashdots quality is declining rapidly11:53
jow_laptopalso many dupes lately11:53
jow_laptopand articles pointing to crappy blog articles11:53
jow_laptopwhich seem to be solely made to get ad revenue11:53
jow_laptoptypos, wrong summaries11:53
jow_laptopso I don't think the editors actually do editing anymore11:54
lekernelwpwrak, my milkymist article took 2 days to get through11:55
lekerneldvdk, "underdog" ...hahahaha12:05
wpwraklekernel: yeah, right. so there's still hope for when they hit a slow spell ;-)12:08
lekernelmaybe they're underdogs but:12:09
lekernelthey have better technology (mmu, asic, etc.)12:09
lekernelthey have better compilers and linux port, contributed by open source people (hahahahaha) compared to milkymist linux12:10
lekerneland they have much larger market share, publicity, etc.12:10
lekernelnot bad for a underdog, he12:10
wpwrakdvdk: don't you want to use the new summary page at link target ?12:11
wpwrakdvdk: http://downloads.qi-hardware.com/people/werner/wpan/web/12:11
wpwrakdvdk: also, it would be good to mention that tuxbrain's prices are with EU taxes, which do no apply to non-EU shipments. since you're not the merchant, you can in fact just give the price without tax and mention that it is without taxes.12:13
wpwrakdvdk: (tuxbrain, being in the EU, may be legally obliged to show his prices with tax)12:13
tuxbrain_awaywpwrak connecting atben to arduio pins directly can work in some way?12:14
wpwraktuxbrain_away: if you connect every pin, yes.12:14
wpwraktuxbrain_away: make a little PCB with an uSD holder that connects to a cable for 100 mil headers, and there you are12:15
wpwraktuxbrain_away: just make sure you really have power on VDD and that you can switch it on and off. the board should draw about 15-20 mA. so you want this proper power, not just a GPIO12:16
wpwraktuxbrain_away: if you don't have switchable power, add a transistor. BJT or FET, either ought to work fine12:17
tuxbrain_awayarduino pins can give 40mA per pin so I think there will be no problem12:21
wpwraktuxbrain_away: no idea how that behaves. don't forget that the voltage can drop quite a bit if you really load these pins. so you'd get voltage variations that are synchronized with the transceiver activity.12:22
tuxbrain_awayalso instead of pcb we can adapt an uSD to SD adapter12:23
wpwraktuxbrain_away: (voltage) e.g, if you load a pin with 15 mA at room temperature, the voltage will drop by about one third. you're still above the transceiver's minimum supply voltage (1.8 V), but you're putting quite a bit of stress on the regulators12:25
tuxbrain_awaywill atben work at 5V?12:25
wpwraktuxbrain_away: of course not. nothing built since humans came down from the trees works with 5 V ;-)12:27
wpwraktuxbrain_away: (uSD to SD) if you have an SD slot or an SD-to-GPIOs adapter, why not12:27
wpwraktuxbrain_away: (SD) the question is whether you have a full SD slot :) the two schematics you showed me didn't12:28
kristianpaulSpeedEvil: hi14:05
SpeedEvilhi14:05
kristianpaulsorry i was sleeping at the time you write my nick 14:05
kristianpaulSpeedEvil: about the correlators, sure i'll implement all 12 as soon as i managed to use at least one14:05
kristianpauluse and sinthesize :-)14:06
kristianpaulthe maxim chip looks interesting too14:06
SpeedEvilin FPGA?14:06
kristianpaulyes14:06
SpeedEvilIn principle, you can do it on nanonote class hw14:06
SpeedEvilin software14:06
kristianpaulsure14:06
SpeedEvilbut it's a lot of assembler14:06
kristianpauli know, and thats a goal, for later14:06
SpeedEvilWhat's your general correlator architecture you're thinking of?14:07
kristianpaulgps receiver and some logic glue in a cpld, but may be not needed as seems dma capabillities of the SD-like slot are not so bad for sampling beyond 10Mhz..14:07
kristianpaulSpeedEvil: is called namuru14:08
SpeedEvil'GPS reciever' ?14:08
kristianpaulsorry14:08
kristianpaulgps frontend*14:08
SpeedEvilWhat bits of architecture?14:08
SpeedEvilI mean in the CPLD14:08
SpeedEvilOr are you just meaning to go serial->parallel-SD-like 14:08
SpeedEvilJitter is annoying - you absolutely can't lose a sample without knowing about it14:09
kristianpaulnamuru is implementation of a zarlink gps correlator, gp2021 think14:10
kristianpaul(Jitter) hum well,the signal showed in the logic analizer looks sane it think14:13
SpeedEvilAh - right14:13
kristianpaulserial to parallel aprouch in the cpld, yes, as the sampling data is synced in a sort of nibble14:13
SpeedEvilAh - I misread.14:13
kristianpaulbut that for when thinking in another board for making nanonote a gps-receiver14:14
kristianpaulnow i'm focused on making all this work in the fpga14:14
SpeedEvilIf you're doiung it with a FPGA - it's proably pointless, compared to a GPS module for consumers.14:14
kristianpaulwell, we like take some risks ;)14:14
SpeedEvilIt's interesting from other aspects though.14:14
kristianpaulalso i dont see too much performance in the nanonote for correlating (using FFT i guess) 12channels..14:15
kristianpauland i dont want mess with asembler again..14:15
kristianpaulmay be implement the correlator on a small cpld?14:16
kristianpaulor a spartan3?14:16
kristianpaulwho knows..14:16
SpeedEvilI was basically assuming a fairly nasty approach borrowing from bit-sliced DES ideas.14:16
kristianpaulas i said at least 20k LE seems to be needed, well that was for an altera fpga14:16
kristianpaulDES?14:16
SpeedEvilhttp://www.cs.cmu.edu/~dkindred/des/bitslice.html14:18
kristianpaulhad you experience with maxim chip?14:18
SpeedEvilI have a sample - other than that no.14:18
kristianpaulas it is more like an "ADC" 14:18
SpeedEvilThis was so far a paper excersize.14:18
SpeedEvilyes.14:18
kristianpaulsample  = EVB or just a chip?14:19
SpeedEvilIt relies on you doing all the heavy lifting.14:19
SpeedEvilchip14:19
kristianpaulcause EVB costs ~500usd14:19
kristianpaul:/14:19
kristianpauli already asked ;)14:19
SpeedEvillaurenceb over on #highaltitude has some logs of raw GPS signals from a similar device14:19
kristianpauloh14:19
SpeedEvilAnd there are other poeple in that channel who have worked/stidued SDR GPS14:20
kristianpaulwich sdr?14:20
kristianpaulthe one from CTAE?14:20
kristianpaulusing scilab? fastgps?14:20
SpeedEvilI don't recall - it was something similar to the maxim chip.14:22
SpeedEvilA bit or two at a few MHz sampling hte IF14:22
kristianpaulone or two, not much difference it seems14:22
kristianpaulmore if you are in a high altitude globe :D14:23
kristianpaulballon*14:23
SpeedEvilPart of the interest is due to GPSs not dealing well with stuff. 14:23
kristianpaulhe :)14:24
SpeedEvilFor example - if you have an accellerometer on a rocket - you really want to feed that accellerometer into the core of teh position filter.14:24
kristianpauli was told gps receiver stop working at some altitude, how that is true?14:29
kristianpauli also was told, once reason was to avoid comercial closed-source gps receiver be put on rockers ;)14:30
kristianpaulrockets*14:30
dvdkwpwrak: no way to give an article to ycombinator, so i went for the announce list post, which is concise.  would be nice if the announce list post  linked to your excellent wpan page, but unfortunately we cannot edit it :)14:41
dvdkwolfspraul: i think slashdot.org/recent already implements the multi-layer system you describes.  this is why it's number of articles gets lesser and lesser the further you move into the past.  and benwpan-article was already dropped.14:42
wpwrakdvdk: (edit it) what, the announce list post or my page ?14:44
dvdkwpwrak: the next SD post is going to link to your new page.  maybe editors evaluate submissions for slashdottability and outright deny articles that link to mediawiki pages? d)14:44
dvdk:)14:44
wpwrakthat would actually make a lot of sense ;-)14:45
dvdkwpwrak: the announce list post archive version14:45
dvdkdid you hear: jane thinks she can just quit and walk out of the community. ;)14:46
dvdklet's just ignore any attempts at unsubscribing from the list :)14:46
wpwrakyeah. she didn't hear the clicking of the lock on those chains, did she :)14:47
dvdklol14:47
Action: kristianpaul really happy at #highaltitude14:51
kristianpaulthaknks for the info SpeedEvil !14:51
lekernelwhat's that?14:52
wpwrakdrugs ? :)14:53
wolfspraulkristianpaul: that's another channel?14:56
kristianpaulwolfspraul: yup :-)14:57
kristianpaulvery interesting one14:57
Action: kristianpaul like hihg altitude baloon projects as well :-)14:58
kristianpaulballoon*14:58
kristianpaulwpwrak: better :-)14:58
SpeedEvilkristianpaul: It's not quite true. The COCOM/ITAR limit is it should not produce positions at over 60000 feet AND 1000 knots.15:09
SpeedEvilSome malfunction, and do OR15:09
SpeedEvilAnd some basically will take any excuse - slight signal fading - to not believe that they are not really 30Km in the air, and produce a solution on the ground.15:10
lekerneldo we give a shit about the COCOM/ITAR limit?15:10
SpeedEvilThe penalties are _really_ severe if you get found violating it.15:11
lekernelthen this should be illegal? http://lea.hamradio.si/~s53mv/index.html15:12
lekerneland who enforces that?15:12
SpeedEvilThe next section in ITAR deals with satellite kick motors with enough impulse to drop reentry bodies on targets, the prior one deals with high speed cameras used for nuclear weapons development.15:12
lekernelsorry http://lea.hamradio.si/~s53mv/navsats/theory.html15:12
SpeedEvilIt's specifically hardware + software as a package15:13
kristianpaulhow they enforce that :-)15:14
SpeedEvilEnforcement is a seperate question.15:15
lekernelit seems to be a US-only thing anyway, no?15:15
SpeedEvilNo - http://en.wikipedia.org/wiki/Wassenaar_Arrangement15:15
kristianpaulyay i'm out !15:17
kristianpaulwee :-)15:17
SpeedEvilReading the details are fascinating.15:17
lekernelthat's again bureaucracy retardation/security theatre as its best, as if building a GPS was a difficult part in developing an ICBM15:17
SpeedEvilFor example - why are 80GHz spectrum analysers controlled?15:17
SpeedEvilBack in 1980, a GPS was actually really quite hard to build.15:17
lekernelsee link above, that (admittedly skilled) ham did it in 199115:18
lekernelwhat is it with 80GHz spectrum analyzers?15:18
SpeedEvilThey are controlled goods. 15:19
lekernelwhy?15:19
SpeedEvilI'm unsure why, but they are.15:19
kristianpaulhum?15:19
lekernelmaybe the reason is just that the US government are control freaks. an amazing list of things are "controlled goods" to varying degrees; even the FPGAs that we use in MM15:20
SpeedEvilTo some degree, yes. Though wasamar is _much_ more widespread than the US15:21
lekernelhttp://www2.rohde-schwarz.com/en/products/test_and_measurement/spectrum_analysis/FS-Z60_-Z75_-Z90_-Z110.html15:24
lekerneldoesn't seem they are particularly hard to order (just probably very expensive)15:24
lekernelyou can also try to hack a luxury car's cruise control radar15:28
wpwrakbe afraid. be afraid. obey. be afraid. consume. be afraid. be afraid. don't see the fnords. be afraid. ... the concept sounds familiar ;-)15:28
lekernelactually I'll probably get that last option a try, when I get time (difficult) and money (those things are about 900¬)15:29
lekernelan USRP module above 70GHz from a car cruise control radar sounds exciting :)15:31
wpwrakhmm :)15:34
SpeedEvilI may have misremembered the exact frequency - it was 15:35
SpeedEvil            1."Signal analysers" having a 3 dB resolution bandwidth (RBW) exceeding 10 MHz anywhere within the frequency range exceeding 31.8 GHz but not exceeding 37.5 GHz;  15:35
SpeedEvilhttp://www.wassenaar.org/controllists/index.html - category 315:35
lekernelluxury car cruise control radars have 76GHz up/down baseband converter chips on them15:36
lekernelhacking them shouldn't even be particularly hard - they have two PCB processes, one with the RF part with a high quality substrate and microsolders for the flip-chip 76GHz silicon15:37
SpeedEvilSome car stuff is controlled, and specifuically must stop functioning if you remove it drom the car15:37
lekerneland the other is a normal PCB with ADC/DAC, DSP and logic15:37
SpeedEvilI think that's IR though15:38
lekernelif you throw away the logic board and put an USRP instead, you should get something interesting :)15:38
SpeedEvilSpeaking practically - there are many realms where increasing commodity hardware has obsoleted this.15:38
SpeedEvilFor example - several of the best webcams pointed at the sky from disparate locations will get you accurate positions on those aircraft in near real-time.15:39
lekernelthe major difficulty will be to figure out the pinouts of the connector between the two PCBs and the control commands for the RF chip15:39
SpeedEvilNot really.15:39
SpeedEvilThe RF chip will be dumb15:40
SpeedEvilLogic in high frequency RF products is expensive.15:40
lekernelhttp://www.infineon.com/dgdl/RXN774x.pdf?folderId=db3a30431689f4420116a096e1db033e&fileId=db3a30431a5c32f2011aa5431f0f4d9915:40
SpeedEvilI guess it's not going to work very well on veroboard.15:41
lekernelthe car cruise control radars' RF board just have that chip, a 12GHz reference oscillator for the PLL, and stripline antennas15:41
lekernelperiod :)15:41
lekernelsure, just take the complete RF board apart from a radar15:41
lekernelI wouldn't even be afraid of mounting the flip chip (there are companies who offer the service at an affordable price) but using the pre-made board is just easier and cheaper15:43
lekernelit would be nice to know what the real frequency range of those critters is15:46
lekernelthe datasheet only says "for radar applications in the 76-77GHz range", but...15:46
SpeedEvilGet a friend in a scrapyard that deal with high end cat-d cars.15:58
rejonwpwrak, tuxbrain_away any uptake on the news?17:45
kyakhttp://old.nabble.com/Dropping-DirectFB-backend-from-GTK-3-td29430982.html17:47
kyakseems that we are stuck at gtk+-2.1717:48
kyaki wish i saw that message before i started some effort porting 2.24 only to see that it is (indeed) broken :)17:49
rohkyak: well.. if you maintain it they maybe dont drop it ;)17:55
kyaki wonder if it only the matter of a time that qt drops directfb support?17:56
kyakroh: maintain it? That's too much for me :) I barely made it build17:56
kristianpaulbut gtk3 is not too bloated anyway for the ben?17:57
kristianpauli tought the rule was stick on gtk1 btw ;)17:57
whitequarkhm, another jz47xx device: http://translate.google.com/translate?hl=en&sl=auto&tl=en&u=http%3A%2F%2Fwww.3dnews.ru%2Fnews%2F612638%2F17:57
kyakgtk1 didn't have directfb support at all17:58
kristianpaulah, true17:58
kristianpauljlime uses it with X, sorry :-)17:58
rohkristianpaul: gtk1 is dead and thats good17:58
kristianpaulif the code is there well :-)17:59
mthwhitequark: looks cool, but how do you know it's jz47xx?18:03
kristianpaullol http://xkcd.com/912/18:12
whitequarkmth: the screenshot18:50
whitequarkit's the Ingenic's minios18:50
whitequarksomeone here on our russian list has asked chinese developers and they sent us sources, asking not to redistribute them18:51
GNUtookristianpaul, what if the info pages are not packaged?19:13
GNUtoofor instance in embedded systems mans are oftens not in the main package or not packaged19:13
kristianpaulindeed GNUtoo 19:35
GNUtooimagine you're in that plane and someone comes out of the pilot cabin and ask if someone know the unix command line19:36
GNUtooor the command line(not precising)19:36
kristianpaul:-D19:37
rejonGNUtoo, hahahha19:49
mthwhitequark: do you know what resolution it has?19:56
mthif I could run OpenDingux on it, I might want to buy one19:56
rejonfunny, that open hardware logo is a broken gear19:57
rejonhahaha19:57
mthah, the ChinaVision page lists resolution as 480x32020:08
wpwrakrejon: /. hasn't taken it yet20:11
rejonwpwrak, i think the title is not very good20:11
dvdkkyak: hacked ASE to give it its own integrated mouse emulation.  want to give it a try?  key mapping described in the wiki http://en.qi-hardware.com/wiki/Applications#Allegro_Sprite_Editor20:11
rejonneeds more bitcoin20:11
rejonhahaha20:11
rejontitle needs to be more explosive20:11
dvdk(wrt bitcoin) i'm sure slashdot editor insider trading is going on over there20:12
rejonwpwrak, another thing is to resubmit20:12
rejonyou didn't syubmit that one right?20:12
dvdkwpwrak: /. article?  were can we read it?20:14
wpwrakrejon: (title) the whole announcement/PR thing was a bit rushed. no proper welcome page, confusing pricing, etc.20:15
wpwrakrejon: i'm not sure if tuxbrain has fixed the price information yet, or if he even considers doing so. if he does, then i'd want to wait for him to get this done before pushing again20:16
rejonwpwrak, more important to get hackers on and plan out for something that is more public facing20:17
wpwrakrejon: i.e., i don't think it helps if the boards appear 20% more expensive than they really are20:17
rejonwpwrak, yeah, that sucks20:17
dvdkwpwrak: let's not resubmit, just write up a new article with a different spin20:18
wpwrakdvdk: (atricle) i mean your submissions, and what they link to20:18
dvdkyes, just saw your comment attached to the earlier article20:18
wpwrakyes, might as well use the 2nd try to improve things20:18
dvdkgoing to edit one here: http://en.qi-hardware.com/wiki/Slashdot-20110615-benwpan20:20
rejonyeah20:20
dvdkmaybe submit again on a saturday around noon (utc), that's what was successfull last time20:21
dvdkunfortunately mediawiki is no multi-user parallel editor :/20:21
dvdk:)20:21
rejoni think fine now20:21
rejonor whenever tuxbrain_away gets those numbers up20:21
rejonthe main thing is the title imo20:21
rejonsomething a bit more outrageous20:21
rejonhow did EFF get all that press on freeing wireless though???20:22
dvdkmaybe we should ping the EFF about our efforts?20:23
rejondvdk, that is a good idea20:24
dvdkalso maybe FSF?20:25
rejonsure dvdk 20:25
rejonasking who is there now at eff20:25
rejonhttp://mobile.slashdot.org/story/11/04/28/150241/EFF-Advocates-Leaving-Wireless-Routers-Open20:26
wpwrakwould also be nice if we could sneak in a word somewhere (doesn't have to be the article per se - could be the overview page) that sonsorship would be rather welcome20:26
rejonhttps://www.eff.org/about/staff/peter-eckersley20:27
rejonwpwrak, totally20:27
rejongoal of this is to get money20:27
rejonmake sales20:27
wpwrak(sponsorship) IEEE 802.15.4 may strike a chord where the Ben didn't. that "internet of things" is more en vogue these days :)20:27
rejonoh, i know that guy who wrote that EFF article20:28
rejoni'm emailing him20:28
wpwrakrejon: i don't quite subscribe to wolfgang's approach of financing the project through sales. i think the ramp-up is too slow for that. we should emphasize that we're creating a lot of value outside the traditional product-to-market chain, namely an open process, free tools, etc.20:28
rejonwpwrak, i hear you20:30
rejonanyway we can get money to support20:30
rejondevelopment20:30
wpwrakrejon: (eff getting coverage) ah, but it's a transgression. everybody *loves* to read about transgressions ;-)20:30
rejonok emailing you wpwrak tuxbrain_away to that eff guy...please don't over hammer him20:30
rejonwe'll see what he says20:30
rejonwolfspraul doesn't hammer people i intro as much now like he used too20:31
rejonhahaha20:31
Action: wpwrak brings the hammer collection from the workshop ;-)20:31
kristianpaulrejon: what is that symbol in the network ring from the freedom stack in the home page of qi-hw.com ?20:32
wpwrakrejon: yeah, i think wolfgang has pretty much withdrawn from seeking external support20:32
rejonthe worst one was an email thread i was having with eben moglen that ended in a huge slamfest to him20:32
kristianpauli'm guessing guess you updated the pic :-)20:32
rejonbut, to his credit, i had meeting with moglen at CC a while back and he said to me: "say hi to wolfgang for me.."20:32
rejonthat guy has amazing memory20:32
rejoni showed him the mm1 board i just got that day20:33
rejonkristianpaul, which symbol?20:33
rejonthe qi logo i want to replace?20:33
rejonor the new http://autonomo.us logo?20:33
rejonwpwrak, yeah, i'm all ears...i'll followup on any and all money/grant leads20:33
rejonpart of the reason for that Ben-WPAN page is to help that20:33
rejonneed more pages like that to detail how its innovation, and how people can get involved20:34
kristianpaulrejon: http://en.qi-hardware.com/wiki/File:Freedom-stack-sphere.svg20:34
kristianpaulrejon: the thing with the "A"20:34
kristianpaulah, you just said it :-)20:35
rejonyeah, that is autonomous20:35
rejonyeah, that effort is not that far along20:35
kristianpauli dint knew that autonomous logo20:35
rejoni think web projects and community are the wimpiest20:35
rejonkristianpaul, yeah, i had that created a long time ago, but didn't get it applied until we wrestled that server from mako20:36
rejonhahaha20:36
rejonautonomous is the least known project unfortunately20:36
rejonanyway, ok, emailing eff dude20:36
kristianpaulmay be eff can sponsor wpwrak to develop the sdr implementation  of "Wifi" ;-)20:37
rejonlets see!20:39
rejoni know john perry barlow and mitch kapor, but those guys are in the clouds often20:40
rejoni'll email them20:40
rejonfuck it20:40
rejoni like to email people in separate threads 20:40
wpwrakrejon: at least it's not the a-in-a-circle logo ;-)20:40
rejonfrom same company20:40
rejonpit them against each other20:40
wpwrakkristianpaul: sdr-wifi has been done with USRPs. seems to consume quite a bit of resources. not a likely candidate for a "make a product within the next 6 months" project. even if you accept generous overruns20:42
kristianpauloh i dint knew it20:42
SpeedEvilI looked at the wifi standard for a bit.20:42
SpeedEvilMy head melted.20:42
wpwrakrejon: "pretty, wouldn't you like ... . mitch, i think we all hate ... ."20:42
SpeedEvilI was trying to work out how the powersave logic worked.20:42
SpeedEvilThere are _so_ many layers.20:42
kristianpaulwpwrak: i guess is same history for the wpan?20:43
SpeedEvilIf you 'just' want to do a raw packet interface, and are willing to ditch everything but making it a 'raw' link that you rely on tcp to do retransmits - it'd be a fair bit easier.20:43
wpwrakSpeedEvil: did they add some new ones lately ? didn't look much more horrible than USB, last time i looked20:43
SpeedEvilMaybe it was just the bits I was looking at.20:44
kristianpaulwpwrak: so the goal is reduce USRPs size, or jump in to FPAA..20:44
wpwrakSpeedEvil: (rely on tcp) funny. that sounds like dirtpan ;-)20:44
wpwrakkristianpaul: wpan would be much easier. less complexity, slower, etc.20:44
rejoni'm emailing moglen and bdale too20:44
kristianpaulrms? :-)20:44
wpwrakkristianpaul: SDR wpan may be doable as a "half year" project20:44
kristianpaulwpwrak: yay :-)20:45
wpwrakrejon: hmm, maybe we should sync on what exactly you're mailing ? :)20:45
rejonhahaha20:45
rejonjust keep them flowing20:45
wpwrakrejon: would hate to have the whole world propagate some misunderstandings :)20:45
rejoncheck that last email20:45
rejoni have an update one20:46
rejonthese are mostly informational20:46
rejonnot expecting a lot20:46
rejonfrom them20:46
kristianpauli think tuxbrain_away should be prepared to run out wpan stock soon :-)20:47
rejonthat usually happens because its in something they need, we are in the same room, or just some random coincidence20:47
wpwrakrejon: hmm, "parent-free" may be an exaggeration. i'm sure there are patents covering the transceiver, probably also the balun, maybe even the antenna20:47
rejonwpwrak, of course, just trying to get their attention20:47
rejonsuggest better language20:47
rejonthen why is this different?20:47
wpwrakrejon: the absence of worrying patents applies more to this being an open standard with now known patent trolls20:47
rejonok i'm going to stick with patent-free20:48
rejonand put a clarifying comment20:48
rejonthe real thing is to be able to see inside the tech20:48
wpwrakrejon: also, the difference to wlan is more on the level that you can get transceiver chips for wpan, with usable documentation and such, while the situation for wlan (or also BT) is decidedly bleaker20:48
rejonjust have to get them ith keywords...emails over 1 paragraph already too much for these dudes20:49
rejonlets get them to comment first20:49
wpwrakrejon: i think also 802.11 is pretty much "open" in terms of specs, at least the basic ones (11, 11b), and there are USRP implementations to support this claim.20:50
wpwrakrejon: (keeping it short) yeah, terse is good. but let's not make vastly exaggerated claims. that only backfires.20:50
wpwrakrejon: so let's work out where exactly the freedom is20:51
wpwrakrejon: the main restriction is see with wlan is in the closed nature of the chips20:51
wpwrakrejon: you basically can't buy chips that let you implement wlan without entering some deep NDAs. what you can buy are modules that are black boxes.20:52
rejoni'm bcc'ing you guys on barlow, kapor and moglen and bdale...lets get them to comment first20:52
rejonwpwrak, can you add blurbs here: http://en.qi-hardware.com/w/index.php?title=Ben_WPAN&action=edit&section=1420:52
rejonand20:52
wpwrakrejon: some of them with closed source firmware, some with even the APIs/protocols (above trivial things like SPI) under NDA20:52
rejonhttp://en.qi-hardware.com/w/index.php?title=Ben_WPAN&action=edit&section=1520:52
rejonhttp://en.qi-hardware.com/w/index.php?title=Ben_WPAN&action=edit&section=1620:52
wpwrakrejon: wait !20:52
wpwrakrejon: let's fix this BEFORE you mail the world20:53
rejon(ok wpwrak )20:53
wpwrakrejon: if you don't have time now, we can also do this later. one day more won't make a big difference.20:53
rejonnah, do while energy20:54
wpwrakrejon: hehe, good :)20:54
wpwrakrejon: so, i'm not sure about the patent situation. there are probably patents on implementation details and such, but nobody outside the few chip-making companies has really penetrated this far20:54
wpwrakrejon: the main problem with wifi is that you can only get black boxes. wpan is much more friendly. there, you get chips with complete documentation where all the registers are, what they do, etc.20:55
rejonright20:56
rejonso can say20:56
rejonwe are attempting to make a patent-free wireless solution20:56
wpwrakrejon: also, because wpan is simpler, there often isn't even any firmware in the chips. some have firmware. some of that is closed, but there's now also open source firmware for at least one.20:56
kristianpaulnot patent-free hardware fabbing 20:56
wpwrakrejon: the chip we use has no firmware. everything hard-wired. rms should like that ;-)20:57
kristianpaulwich is diffrent from manufacture i think20:57
wpwrakrejon: i'm not quite sure on the patent-free side. that word may not describe the aspect properly20:57
wpwrakrejon: it's more making a solution that doesn't require you to be part of the inner circle that has sworn a vow of silence20:58
rejonwpwrak, that is cool20:59
wpwrakhehe ;-)20:59
wpwrakthe mechanisms of that inner circle themselves are even obscure. can be patents, trade secrets, copyright, you name it20:59
wpwrakif you're member of the inner circle, you probably know some of those mechanisms (and may employ them yourself)21:00
wpwrakif you're not a member of the inner circle, they can't tell you. fight club reloaded.21:00
rejonok wpwrak sent you the draft21:01
rejonplease edit and/or tweak then get back to me21:01
rejonwould like to get this b4 end of the day (1 hr)21:02
wpwrakalso, the wlan situation clears up when you go to bigger sizes. i think there are some relatively open chips for USB-based wlan (would have to check, though. "open" is used in so many ways these days ...)21:02
rejonsure21:02
rejonjust trying to hook into these guys21:02
rejonnot hook into the details so much21:03
wpwrakbut with the small chips, for cellphone-sized mobile devices, there's all that veil of secrecy21:03
rejontotally21:03
rejonthat is good21:03
rejonwpwrak, can you channel these ideas onto that wiki page21:03
rejoni can't transcribe right now21:03
wpwrak(details) yeah, i just want to avoid making statements they know are wrong21:03
rejonthose high up dudes have filter on things that don't increase their value or reputation21:03
wpwrakpoliticians :)21:04
rejoni say paid or laid aka p(l)aid21:04
kristianpaulhahah21:04
wpwrak;-))21:04
wpwrakwe need a better word for "patent-free". maybe just "Free" ?21:05
wpwrak(in the sense of "compatible with the principles we uphold")21:06
kristianpaulfreedom  is overrated21:06
rejonyeah, probably best to ignore patent word overall21:06
rejonfreedom is def. overrated21:06
rejonits not really competitive either to real world21:06
rejonits like, does that shit work?21:06
rejoncan i connet my shit with it?21:06
wpwraki'd take a benevolent dictatorship any day ;-)21:06
rejonyeah, patent and free are trigger words21:06
rejonbetter to avoid mostly21:07
kristianpaulpatent club free 21:07
kristianpaulah, ok21:07
rejonif someone has enough moula, then can take it21:07
rejonsadly21:07
dvdknew slashdot article.  first draft: http://en.qi-hardware.com/wiki/Slashdot-20110615-benwpan21:07
dvdkthis time much longer 21:08
dvdkneeds lots of polishing.  too tired already.  has time anyways21:09
kristianpaulwifi-dongles, humm21:09
kristianpaulthats a bit confusing, so wifi is wireless nor wlan?21:10
wpwrak(... processing rejon's mail ...)21:12
SpeedEvildvdk: On first read. Does the dongle plug in or what. 21:14
kristianpauldvdk: vSpeedEvil: i think usb-dongles is a right word to mention21:16
Action: SpeedEvil is annoyed zigbee is closed.21:17
wpwrakSpeedEvil: don't be annoyed with it. ignore it :)21:20
wpwrakSpeedEvil: it's dead anyway. everybody is moving on21:21
wpwrakdvdk: i'd change "all WLAN" to "all suitable WLAN" (or "all otherwise suitable"). e.g., there may be some large chips that would have better documentation. but the board would be as big as the ben itself.21:22
wpwrakwhitequark: what was your ghettoblaster's chip again ? and is there a non-NDAed data sheet for it ?21:23
dvdkwpwrak: it's a wiki, change what you want :)21:23
rohwell.. there are wifi chips with free drivers... but they are usually not intended for _ultra_ mobiles21:23
dvdkwhen should we post it?  use wpwrak's account this time to disguise the redundancy?21:23
wpwrakdvdk: sigh. wiki, the downfall of the culture of authorship and editing ...21:23
dvdkwait, don't edit just yet.21:24
wpwrakdvdk: wait .. still reading ...21:24
rjeffriesbeing anti-WiFi is a charming notion. Reminds me of being opposed to the sun rising each day. 21:24
wpwrakroh: yeah. plus, you probably still can't get the chip per se.21:24
wpwrakrjeffries: you'd be surprised by how many people here are rather successful at avoiding the evil day star ;)21:25
rejonwpwrak, you have to dm me if you want me21:25
rejonworking on other things21:25
dvdkwpwrak: ok done.  :-P21:25
rjeffriesWiFi gives one a single-ended soution for internet connectivity, at say your fav coffee shop. the wonderful 802.15.4 stuff is two ended, so will work at you home just fine, but not at a public wfifi hot spot.21:25
dvdkwpwrak: just ctrl+r and add your changes21:26
rohwpwrak: you can.. but usually just in some real numbers ( 5 digits at least... the rest is 'samples')21:27
dvdkuh, already changed, does it merge mulitple edits? not bad.21:27
dvdks/but be energy efficient/but at energy efficiency/ ?21:27
rejonwpwrak, what is this font on here again? http://downloads.qi-hardware.com/people/werner/m1/tmp/back.png21:27
wpwrakdvdk: i didn't try any editing so far. first list the issues21:28
rohrjeffries: find a 'mobile access point/bridge' hardware with an usb port and see if openwrt is useable for it21:28
rejoncxadams, needs to know21:28
Action: dvdk is startled and looks at the history of the page21:28
wpwrakroh: and usually NDA :)21:28
rohrjeffries: i know there are devices like that, even battery operated ones. usually with also a ethernet connector. to use in a hotel room for example21:28
dvdkah, evil tuxbrain is adding comments.  Yuck /me hates parens21:29
SpeedEvil'not hurt battery life'21:29
rohwpwrak: some dont.. i think ralink for example gave out some datasheets.21:29
wpwrakdvdk: (at ... efficiency) yes21:29
rohwpwrak: not the 'big ones' . sure.21:29
dvdk(tuxbrain) that's why the broken english :-P21:29
rjeffriestroh you are correct, and so am I. WiFi access is virtually ubiquitous. I can't go to my fave coffee shop and plug wpwrak 's dongle in and connect. ;)21:29
dvdkwpwrak: saw my comment about authorship :)21:29
dvdk?21:29
wpwrakdvdk: also, 2 Mbps is a non-standard extension by atmel. the standard only goes to 250 kbps.21:29
Action: dvdk knows that21:30
wpwrakmaybe i'll take a shower and come back when the edit war is over ? then i only have to fight the winner :)21:31
Action: tuxbrain_away will stay quite for a while :P21:32
kristianpaulthen use a piratepad instead 21:33
Action: dvdk is falling asleep21:33
kristianpauland no more wars :-)21:33
kristianpaulok, bteer sleep :-)21:34
dvdkguys, keep the story simple and concise and fun to read.  don't know whether throwing around with numbers is going to attract readers?21:34
rejonwpwrak, font21:34
wpwrakrejon: oh. dunno. that's from the altium-generated gerbers21:35
rejonok cxadams wpwrak doesn't know...its probably some standard font from that program21:36
wpwrakdvdk, tuxbrain_away: yeah, i'd drop the bandwidth figure. we have all that on my overview page anyway21:36
wpwrakdvdk: (authorship) ;-))21:38
wpwrak"more open 6LoWPAN" maybe "IEEE 802.15.4 and eventually 6LoWPAN". we don't have the latter yet21:39
wpwrakdvdk: the first sentence doens't quite ring true. those chips/modules with persistent storage usually come pre-programmed.21:41
wpwrakdvdk: e.g., in the case of openmoko, where we used such a module, they wuoldn't even give us a reflash utility. instead, the modules had to be returned (from a team scattered all over the world, ha ha) to the company in taipei for flashing a critical update21:42
wpwrak(the original "consumer grade" version the shipped would die after a few seconds of SSH)21:42
wpwraktuxbrain_away: also, don't put (ugly) in parentheses :)21:43
dvdkwpwrak: no i think it's ok.  production can be "produciton of the chip"21:43
dvdkwpwrak: that was me :)21:43
dvdk(paraanthesis)^21:43
wpwrakdvdk: (production) phew ;) okay21:44
whitequarkregarding the wifi chipsets, I heard that Atheros has open-sourced driver and firmware21:44
dvdkalso we shouldn't go for 100% accuracy.  Conciseness means information loss means approximations (like 6LowPAN)21:44
xMffthey're actually firmware-less21:44
wpwrakdvdk: (ugly) hah ! with parentheses it means that the project is an ugly mess. without parentheses it says in an ironic way that the details are non-trivial.21:44
dvdkwpwrak: thought the opposite.  parenthises for the irony. hmm21:45
xMffonly eeprom with a bit of callibration data21:45
dvdks/Low/LoW/21:45
wpwrakwhitequark: firmware too ? i know of drivers. for all but the 6k series, which are the ones targetted at small mobile systems21:45
wpwrakwhitequark: that's why i'm saying we should qualify that statement21:46
whitequarkthis page http://linuxwireless.org/en/users/Drivers/Atheros says that firmware is opensourced under a good license, and has a link to repository21:46
whitequarknot sure which devices use that, through21:46
wpwrakdvdk: "ugly details" is a fixed expression. doesn't mean that everyone would find them ugly.21:46
dvdkwpwrak: also the article does not need to reflect the exact state of work where we are.  it's about the wpan _project_ and if that goes for 6LoWPAN, then these are 6LoWPAN dongles (a few slashdotters can hack the drivers over the weekend, won't they? :)21:47
wpwrakwhitequark: that would be the 9k series21:47
dvdkthat's why the irony it's {"ugly details", "details"} encoded via parens21:48
wpwrakdvdk: (6lowpan) well, i wouldn't want to advertize a feature we haven't even started to implement, and where we're not certain whether it will be implemented from our side21:48
whitequarkwpwrak: ah yes, I see now. the 6k embedded series is still encumbered21:48
wpwrakdvdk: naw, the parentheses don't work this way. you could use "ugly" details. but keeping it simple conveys the same meaning.21:49
dvdkwhitequark wpwrak: that's why it reads '_nearly_ all WLAN chips' .  buy one may want to add a suitable21:49
dvdkwpwrak: that's the nice thing about communication: what it means only depends on the receiver :)  21:50
dvdkbut change it as you wish.  too tired to see any nuances currently21:50
dvdks/only/entirely21:50
wpwraki'd say "requite proprietary firmeware or are otherwise encumbered." (e.g., if you can't use the chip without a nasty NDA, it's unsuitable for copyleft hardware)21:50
wpwrak"use" = circuit design21:50
dvdkwpwrak: (encumbered) sounds neat21:51
dvdkand it gets longer and longer.  should have kept the first version even shorter.  at this growth rate it's not going to fit in the submit box :)21:51
whitequarkdvdk: (encumbered) hey, that was my idea! :)21:51
dvdk:)21:52
whitequarkI should have registered a patent on it...21:52
wpwrakrjeffries: we need your instinct as a native english speaker: "the ugly details", "the 'ugly' details", or "the (ugly) details" ? when the intended meaning is "intricate", with a dose of irony21:52
dvdkwhitequark: hurry, before we submit it21:52
whitequarkdvdk: isn't that prohibited by your religion? ;)21:52
wpwrakdvdk: whitequark's patent lawyers should be asleep now. maybe we can sneak ours in via rejon or rjeffries in the us ;-)21:53
dvdkwhitequark: yes i'm already feeling the bad karma from it21:53
whitequarkdvdk: if you do, you may spend your next life as a tiny and horribly closed wifi chip...21:55
dvdkwhitequark: or as a tiny gate, having to work at 4 GHz21:56
whitequarkdvdk: or, in the worst case, as an x86 core21:56
dvdkit's a little difficuult to fit 'encumbered' into the long sentence with proprietary firmare, have to drop the last part of the sentence21:56
dvdkwhitequark: as an A20 gate ? :)21:57
whitequarkdvdk: definitely!21:57
whitequarkdvdk: (fitting) that sentence just becames too encumbered, isn't it?21:57
dvdkok, time for bed21:59
dvdkcu guys21:59
whitequarkbye21:59
wpwrakso .. nobody else editing ? then i shall have my way with it :)22:12
wpwrakdone with my edits. http://en.qi-hardware.com/wiki/Slashdot-20110615-benwpan22:25
wpwraksee history for details: http://en.qi-hardware.com/w/index.php?title=Slashdot-20110615-benwpan&action=history22:25
SpeedEvilI'd use wifi, not wlan22:27
tuxbrain_awaywifi == wlan22:28
tuxbrain_awaywireless !=wifi22:28
SpeedEvilTrue - the former acronym is much more common though22:28
tuxbrain_awaywireless!=wlan22:28
SpeedEvilAnd in context - wlan is used as a synonym for wifi.22:28
SpeedEvilI've hacked on it - revert if you hate it. :)22:34
wpwrakSpeedEvil: WIFI or WiFi ? :)23:10
SpeedEvilyeah - probably23:10
SpeedEvilI forget how they capitalise it.23:10
wpwrakSpeedEvil: WiFi is more widely understood than WLAN. alas, it's a forbidden word (trademarked). but i guess i this context we can get away with using it.23:11
wpwraktuxbrain_away: have you seen the numerous comments on the prices ? regarding indicating taxes and not ? your current system is very confusing23:13
tuxbrain_awaywpwrak: ok I will try to fix it ASAP, with an "IVA includo" is enough?23:14
wpwraktuxbrain_away: do you know the legal requirements for what you can write there and what not ? e.g,, i think you're obliged to include taxes in prices that are potentially destined at EU customers, but i don't know if you can mention that this includes EU taxes, and whether you're allowed to also show non-EU prices or have a link/button for non-EU prices23:15
wpwraktuxbrain_away: are you in good terms with pulster ? if yes, you may want to ask him. or maybe Nikolaus Schaller.23:16
wpwraktuxbrain_away: germans tend to be very well-informed when it comes to laws ;-)23:16
tuxbrain_awaythere is no rule but meanwhile it's clear in the final confirmation order sumary23:16
tuxbrain_awaybut mostly 90% of my cosrtumers are from EU so if I show prices without Vat in the final order sumary before confirmation they feel cheated and don't buy.23:17
wpwraktuxbrain_away: heh, pulster is making fun of your shipping costs ;-) http://www.pulster.de/nanonote.htm23:18
wpwraktuxbrain_away: (tuxbrain: EUR 99 + 23 shipping = EUR 122. ergo my price is better ;-)23:18
tuxbrain_awayI can try to make more clear that is with vat included23:19
wpwraktuxbrain_away: i think it's okay to show prices with VAT by default. but you should also mention that these are in-EU prices. so, say, US customers, who are used to prices without taxes won't think it's very expensive.23:19
tuxbrain_away(pulster) yeah I already know that , he asked me  permision for that :)23:19
wpwraktuxbrain_away: (permission) nice :)23:20
wpwraktuxbrain_away: (mainly EU customers) also don't forget that you're the only global distributor for atben/atusb for now :)23:20
tuxbrain_awayit's not true anyway but I don't care, the NN they sell I have selled to him first :P23:21
SpeedEvilWhat was the price?23:21
wpwrakso he adds EUR 9 and optimizes the shipping. fair enough23:21
SpeedEvilAnd can the microSD thingy fit in arbitrary sockets?23:21
DocScrutinizerwpwrak: (trademarked) it's legal to say "we are NOT <trademark>(R)"23:22
wpwrakDocScrutinizer: that's what i hope :)23:23
wpwrakSpeedEvil: no. they must have the right physical clearance, you must be able to bit-bang the I/O pins, and you must be able to switch power on/off.23:24
tuxbrain_awaywpwrak: (WWD of atben and atusb) Ok I will try to make it clearer, somthing with little lettes beside the price telling (vat included, are you not from EU? click here)23:24
SpeedEvilwpwrak: I guess not the n900 - physical clearance would kill it.23:24
wpwrakSpeedEvil: ah, and you must have all the I/O pins. no cheating with 1-bit bus devices. (e.g., various things in the arduinoverse)23:24
SpeedEvilAnd on reflection - it's under teh battery)23:24
DocScrutinizerwpwrak: (VAT) in EU you need to show prices incl VAT for products intended at end users. You are free to also show the netto price w/o VAT, as long as you have the all-incl price too, and you clearly state what is what23:24
wpwrakSpeedEvil: you should get a ben. it fits perfectly there ;-))23:25
SpeedEvilwpwrak: REgrettably.23:25
wpwraktuxbrain_away: sounds great23:25
DocScrutinizertuxbrain_away: ^^^23:25
SpeedEvilI have _WAAAAAY_ too much random hardware I haven't touched again.23:25
wpwrakSpeedEvil: lock away the rest. focus on what's important, namely the ben ;-)23:26
SpeedEvilAs I'm now on a sharply limited budget, I'm trying to limit my purchases to sutff I'm actually going to be doing stuff on next week.23:26
SpeedEvilThough I am looking at solar panel bits on ebay, which is probably silly.23:26
wpwrakSpeedEvil: don't worry. none of us can really afford our addiction. don't let this stop you ;-)23:26
DocScrutinizerhah23:27
SpeedEvilI had no heating on last winter, to save for insulation work on the house - so I have zero spare cash at the moment.23:27
SpeedEvilhttp://cgi.ebay.co.uk/2-KW-3x6-Short-Tabbed-Solar-Cells-Diy-Panel-/120717367254?pt=LH_DefaultDomain_0&hash=item1c1b50d7d623:28
SpeedEvilSilly solar.23:28
wpwrakSpeedEvil: so you solar-paneled everything ? the cost of is would probably for a century of inefficient heating ;-)23:29
wpwraks/for/pay for/23:29
SpeedEvilwpwrak: naah. Though 12m^2 of solar cells for $800 is silly cheap.23:29
SpeedEvilI've bought conventional 10cm foam insulation, and am installing it.23:30
wpwrakhmm. the kWh costs here about 2.5 US-cents. so USD 900+ USD 149 shipping, taxes (30%?)... that's 68 MWh. if you panel has an overall yield of 10% per year, that's 1.75 MWh per year. you'll break even in about 40 years. i hope you're young and in good health ;-)23:41
SpeedEvilHere it's $.2023:41
--- Thu Jun 16 201100:00

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