#qi-hardware IRC log for Thursday, 2012-08-02

wolfspraulif anyone can handle heavy reads (and for the archives), here is a good TSMC earnings conf call transcript... http://www.tsmc.com/uploadfile/ir/quarterly/2012/2vyik/E/12Q2%20Conf%20Call-Transcript.pdf02:16
wolfspraul25 pages, but the usual solid quality from TSMC...02:17
larscwpwrak: or the economical crisis has hit people from the qi-hardware community so hard that they can't afford internet access anymore.08:26
wpwraklarsc: i thought of that, but then we would have heard them first complain about lack of food. i mean, you can go a few days or even weeks without eating, but without internet ? quite impossible. so they'd give up food first.08:34
larsc:)08:35
Action: larsc complains about the lack of coffee and gets up to get some08:36
kyakyeah, i liked werner's twist in that e-mail :)08:48
whitequarksigh08:53
whitequarksomeone on a russian mailing list complains that Xburst1 CPU is slower than a P6 one08:53
whitequarkyeah, like, really. Xburst is a single-issue MIPS32 CPU without any L2 cache whatsoever, and P6 is superscalar, out-of-order microarch with speculative execution and huge fucking L2 cache08:54
whitequarkof course it is way faster.08:54
wpwrakwhitequark: congratulate him on his research and encourage him to implement his conclusion, i.e., use a P6 ;-)09:02
whitequarkwpwrak: nope, it was MY research09:03
whitequarkhe is supposedly too stupid to open up Wikipedia, much less to, like, think with his brain.09:03
whitequarkand yes, already did09:03
wpwrakwell, his discovery then :)09:04
Action: whitequark sometimes hates everyone09:04
wpwrakwho would have thought ;-)09:05
kyakoverreacting is almost as bad as stupid e-mails :)09:27
wpwrakan interesting low-cost ISM spectrum analyzer, in the 2.4 GHz band also with signal generator: http://micro.arocholl.com/index.php?option=com_content&view=article&id=48&Itemid=6211:11
wpwrakthey're working on a wideband extension now: https://groups.google.com/forum/?fromgroups#!topic/rf-explorer/oxfBLNvfNGI11:12
wpwrakunfortunately not open source or open hardware. but at least most of the serial protocol is documented.11:14
rohwpwrak: well.. there is osmosdr...11:21
rohand umtrx11:21
larscand <shameless-plug>FMCOMMS1</shameless-plug>11:37
larscalthough that's not so low-cost anyway11:37
wpwrakroh: osmosdr doesn't get to 2.5 GHz, does it ?13:04
wpwrakroh: umtrx seems a bit bulky. no mention of the frequency range.13:08
rohdunno (frq).. i thought 2.713:10
rohah. no.. 1.413:10
wpwrakhmm, http://www.hack4fun.eu/2012/03/umtrx-open-source-hardware-transceiver-for-gsm/13:12
wpwraksays that umtrx does 375 MHz - 4 GHz. if that's really what they mean. (and not just the specs of one part of many in the chain)13:12
wpwrakseems bulky, though13:13
wpwraklarsc: and yours needs a fair bit of extra hw, too13:13
larscone bit to be exactly ;)13:15
larscIt is a modular solution, you have to have a FMC host board, but you can choose the host board according to your requirements13:33
kristianpaulwpwrak: rtl-sdr can still go/hack to 2.2Ghz.. but i think you lost gain a lot13:40
wpwraklarsc: yeah, but in the end you're probably in the USRP price range (~USD 2k for box plus RF board)13:42
wpwrakkristianpaul: for the "2.4 GHz" band, you still need 350 MHz more ...13:44
wpwrakwell, or at least 300 MHZ, depending on which bits you're after13:45
kristianpauldamn, i'm the only receiving a excessive bounces accoutn disabled from qi mailman?..13:45
kristianpaulwpwrak: i'm after my rc plane remote control :)13:45
wpwraknever had that :)13:45
kristianpaulhmm i had lost 3 weeks mails..13:49
kristianpauldamn13:49
larscwpwrak: yea, as I said not really low-cost. But I think there will be a bundle of FMCOMMS + ZYNQ baseboard for about 1k USD soon13:51
wpwraklarsc: USD 1k sounds nice. you'd beat the USRP+SBX at that price point and you have more bandwidth.14:09
larscand we even have video out on the board14:11
wpwrakgrmbl. and now i need to figure out how to mill a piece of plastic from both sides. capacitative touch sensors are finicky. the price to pay for low BOM cost and no mechanical wear ...14:12
larscwpwrak: http://wiki.analog.com/resources/tools-software/linux-software/iio_oscilloscope14:12
larscthat's running on the target system14:13
wpwraknow we know what you've been doing lately :)14:13
wpwrakdoes the critter work with gnuradio ?14:13
larscnot yet14:13
larscit's on the nice to have feature list14:14
larscit should be a matter of implementing a IIO backend for gnuradio14:15
wpwrakand, and the scope comes with source. very nice !14:15
larscwpwrak: everything comes with source14:15
wpwrakyeah, gnuradio is a bit the lingua franca for everything SDR14:15
wpwrakwith that attitude, analog should also make SoC (with GPU), FPGAs, and WLAN chips ;-)14:16
wpwrakthe world would be considerably easier :)14:16
larscwe'll see ;)14:17
larscthe chips on the board are obviously not opensource14:17
wpwrakbut with available data sheets ?14:21
larscyes14:22
larscand linux drivers14:22
wpwrakyou're seriously trying to kill the job market for reverse engineers, aren't you ? :)14:23
larscwell market demands kind of drives this14:25
larscbeing 'open' is a competitive advantage14:26
viricwhat board?14:27
larschttp://wiki.analog.com/resources/fpga/xilinx/fmc/ad-fmcomms1-ebz14:27
viricwe have a bitscope oscilloscope here... it works enough for our needs, and it felt like helping those who provide schematics.14:28
wpwraklarsc: i wish more companies would see it like this. especially in the aforementioned sectors.14:28
wpwrakbtw, speaking of socs, i hear that allwinner is very liked in certain open source circles. yes they don't seem to have any public data sheets for their chips. does anyone know more ?14:29
wpwrake.g., was i just too stupid to find them ? are there there but in chinese ? have they been "leaked" and everyone has a copy but not officially ? or did i just get all this wrong ?14:30
kristianpaulgnuradio blocks should be sinthesized on the fpga as well14:38
wolfspraulahh :-)22:51
wolfspraulis that real? they force lars to add a small but significant extra line into the middle of an otherwise unsuspicious 'bsd style' license?22:52
wolfspraul"Use of the software either in source or binary form, must be run on or directly connected to an Analog Devices Inc. component"22:52
wolfspraulgood morning everybody22:52
wolfsprauland that's for a 1500 line sample app22:53
pabs3weird requirement22:54
wolfspraulnot too unusual, it's just sad to see the entire circumstances with this being a small sample app, certainly this extra line was not lars idea, analog does not gain anything from it, etc.22:55
wolfspraulfree software obviously would have gone nowhere if everybody who managed to hack up a 1500 line sample app would have also spend the time to think about extra clauses like this one to make 'their' thing special and keep it connected to 'their' stuff22:56
wolfspraulso then :-) it's probably still best for lars to not waste his time in turn to 'convinve' anybody inside analog, better just ignore and keep working22:57
wolfspraulconvince22:57
wolfspraulbut naturally this piece of coding is trapware22:57
Action: pabs3 misses some context22:57
wolfspraulpiece of code, arg22:57
wolfspraulthe oscilloscope stuff lars published22:57
wolfspraulgreat to see him publish on analog.com now :-)22:57
wolfspraulhttp://wiki.analog.com/resources/tools-software/linux-software/iio_oscilloscope22:58
wolfspraulalso it's not nice imho to hide this kind of restriction in the middle of a bsd license22:58
wolfspraulnot many people like me read the fine print to the last tiniest line (where in real life all the good stuff is - OF COURSE :-))22:59
DocScrutinizer05hey, anybody over at TPE? All well with all our friends who live there?23:32
wolfspraulI think so :-)23:47
wolfspraulthe flooding tends to get worse in the news as well23:47
wolfspraulwhat's the future of Qt without Nokia?23:51
mthnot sure, but the future of Qt inside Nokia seems to be non-existent, so going on outside is probably better23:55
--- Fri Aug 3 201200:00

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