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

rjeffriesCES tidbit in "cheap tablets" category: Ainol Novo 7 Paladin $110, runs Ice Cream Sandwich, Ingenic JZ4770 Xburst, 1GHz; GPU: GC860, 512MB and 8GB, up to 32GB via microSD, 7 inch LCD display with resolution of 800×48001:51
zrafa_kristianpaul: you should make something like this : http://excamera.com/sphinx/gameduino/02:15
zrafa_for the nanonte.. via ubb or something :)02:15
kristianpaulah, yes i saw it but too lazy having a M1 next to me..02:16
zrafa_so we can play some games or use some monitor with nn. Of course. THere is the werner vga output, but maybe it could be more generic02:22
kristianpaulzrafa_: yes werner's software defined vga still have some potencial tought02:40
kristianpauland even the ubb port it self, i still want to have time to see what sdr related tasks can it do02:40
wpwrakhmm, low frequency tasks. maybe you can make an AM radio transmitter ;-)02:41
kristianpaulyes i was thiking something like that !02:43
kristianpaulwhy not? :)02:43
zrafa_can we open orcad files with linux?03:31
kristianpaulsounds unlikely at least it have a floss replacement03:36
kristianpaulor a orcad binary that just run on Linux ;)03:36
wpwrakopen .. sure :)03:46
DocScrutinizerooh, dental floss06:55
cladamwaDocScrutinizer, hi thanks a lot for your great check and idea on else. ;-)09:00
cladamwaDocScrutinizer, wpwrak have really given much thought to the matter (J23 to add AUDIO_AGND), thanks !09:09
wpwrakyeah, it sounds like a good change to me09:11
cladamwawpwrak, okay.09:11
DocScrutinizer51YW, glad to help :-D09:30
DocScrutinizer51cladamwa: you might want to add some of your nice varistors to line-out jack as well09:32
cladamwaDocScrutinizer, yeah... i forgot to mention this too. thanks for this reminds. ;-)09:33
cladamwaDocScrutinizer, one topic for us now will be: we're also planning to take consideration about adding insert/detection idea in LINE-IN/LINE-OUT connector. (i.e. trying to let s/w can have reaction on insert/plug-out), so maybe in the end we'll have changes on both later. Any idea from you? ;-)09:37
cladamwaDocScrutinizer51 we'll consider if add that feature soon. i may ask for your help again. ;-D09:45
wolfspraulwow there are so many things in motion now10:24
wolfspraulI acquired the entire remaining stock of atben and atusb from tuxbrain!10:24
wolfspraul45 atben, 33 atusb10:24
wolfspraulhave to get some fire under this :-)10:24
kyakinteresting - does it mean a higher demand for atusb than for atben?10:33
kyakdoesn't make sense :)10:33
mstevenswolfspraul: what happpened? did tuxbrain give up on the project?10:33
kyakit;s a pity that the lot wasn't sold anyway -\10:34
wolfspraulnot sure, I have a hard time reaching tuxbrain, I think the shop became a part-time project10:34
wolfspraulso he is focusing on the Ben NanoNote10:34
wolfspraulI gladly buy those units because I was unsuccessful in buying some recently ;-)10:34
wolfsprauland I pay him in NanoNotes10:34
kyakwolfspraul: why buy them, when you can just produce them? :)10:34
wolfspraulkyak: not sold yet10:34
wolfspraulthat's a run with another whole bunch of overhead10:35
wolfspraulbut i am working really 100% on and over capacity everywhere10:35
pabs3wolfspraul: many things in motion, sounds like someone should write an article about Qi happenings for LWN?10:35
wolfspraulI gladly gladly take those boards10:35
wolfspraulpabs3: has lwn ever reported about Milkymist?10:35
wolfspraulthey should10:35
wolfspraulit's a bona fide attempt to create a true collaborative effort below the Linux kernel, in IC design10:36
pabs3wolfspraul: yeah: https://lwn.net/Articles/443415/10:36
pabs3no articles though10:36
Action: pabs3 bbl10:36
wolfspraulare you an editor/reporter at lwn? or you know one?10:37
wolfspraulI think we need a quality story, and an editor there needs to be seriously interested10:37
wolfspraulit's not a hush-hush marketing thing, where we have to sell some new thingie before everybody realizes the hype is over10:37
wolfspraulif an editor at lwn would be interested, I would be more than happy to provide him or her with a review unit, and unlimited time to answer questions :-)10:38
whitequarkdoes anyone understand how this actually works? not much info is available10:38
pabs3wolfspraul: neither11:33
pabs3wolfspraul: I've suggested an interview to them before, hopefully they think about it11:33
zrafa_wpwrak: open: we can use "cat file.orcad" :)12:09
wpwrakif its's text. else, hexdump -C :)12:13
GNUtooisn't orcad a program for blind people?12:15
GNUtooah maybe it's orca12:15
GNUtoosorry 12:16
pabs3wolfspraul: they do accept guest writers too: https://lwn.net/op/AuthorGuide.lwn12:16
kristianpaulwolfspraul, atusb/ben !!14:45
kristianpaulI can buy a combo and send it by cheap postal mail :)14:45
wpwrak(atusb/atben) great indeed !14:58
wpwrakkyak: there were fewer atusb than atben to start with. about 20 fewer units produced (due to different pcb panelization) and he had three atusb rejects while all atben tested okay15:00
wpwrakkyak: so they sold roughly at the same pace, though not everyone bought them in an 1:1 ratio15:01
kristianpaulwolfspraul: can i buy atusb&atben with bicoin? how much it will? :)15:15
Action: kristianpaul have some bitcoins saved15:16
wolfspraulsure you can, but I don't have them yet, so please give it some time15:16
kristianpaulah, sure np 15:17
kyakbelieve it or not, when i plug in the jtag cable which came with m1 into my laptop, it powers off momentarily; i have to remove the battery to be able to turn it on again15:40
kyakwolfspraul: any idea? :)15:40
kyaki assume that the cable is shortcut and it triggers some protection on laptop side15:40
kyakcan i use just any mini USB cable with m1 jtag port?15:41
wpwrakbasically yes15:43
kyaknow, when linux starts to swap and i can't do much, i will just insert this cable instead of holding the poweroff button for some seconds :)15:43
kyakthis cable can also be used a DoS exploit15:44
kyaki will just give it to some manager at work15:45
whitequarkkyak: not all notebooks will detect that gracefully15:47
kyakyeah, i hope so!15:48
kyakburn, burn!!15:48
wpwrakjust stay out of the blast radius :)15:48
kyaki think i will get an extension cable :)15:48
kyakok, i know. i will lable it as "cable of death" and leave in the eyesight. And observe!15:49
kyaklabel even15:50
kyakBus 001 Device 004: ID 20b7:0713 Qi Hardware Milkymist JTAG/serial15:52
kyakhm, this is neat.. 15:52
kyakagain, rolling releases win (usb database up to date)15:53
wpwraki just wish the database was updated more frequently15:54
wpwrak(at the source)15:54
kyakwpwrak: where i can look for jtag commands?15:55
wpwrakyou mean  jtag<Enter> help<Enter> ?15:57
kyakhe :) no, i mean, what commands i can issue to the board? like, find the revision etc?16:00
wpwrakmmh. dunno about revision. should be possible if you do a boundary scan. should be interesting to figure out how to do that :)16:02
kyakok, i might not understand completely how jtag can be used for m1.16:05
wpwrakwe use it mainly for flashing16:07
wpwrakoccasionally for booting16:08
wpwrakdebugging and such go through other channels16:08
wpwrakand we don't do boundary scans as part of production testing (yet ?)16:08
whitequarkkyak: a lot of them will just turn off16:19
whitequarkand will still work even without re-inserting the battery16:19
whitequarkyou have some chances of frying USB ports, through16:19
whitequarkI wonder if there are any notebooks crappy enough to make battery explode in such a condition16:20
whitequarkI think no, but you can never know for sure..16:20
wpwraktry tablets. that's where the race to the bottom is at these days16:20
whitequarktablets are useless16:20
whitequark... by the way. latest news from russian sonar company.16:21
whitequarkmy friend has agreed to write a firmware for them16:21
whitequarkso they assembled a prototype and delivered it to him16:21
whitequarkhe turned it on and it did not work, because the main LDO was not present, because they accidentally placed the LDO and its tantalum cap on the same place on board16:22
whitequarkok, he returned it and they wire-wrapped the LDO and the cap16:22
whitequarkhe turned it on and the cap has exploded, because it was reversed16:22
wpwrakbut it was cheap ! ;-)16:23
whitequarkhe returned it and they soldered it the right way16:23
whitequarkthe CPU had none of its bypass caps soldered16:23
whitequarkso it was pretty unstable16:23
whitequarkthe quartz has not had any load caps, too16:23
whitequarkok, he wire-wrapped some all over the board16:24
whitequarkthe LCD did not work afterwards16:24
whitequarkthe CPU has 4 VDD pins16:24
whitequarkthe board author has connected one to the power source16:24
whitequarkand other to the power pin of LCD16:24
whitequarkothers were left floating16:25
whitequarkok, he wire-wrapped some more wires16:25
whitequark... the tantalum cap near the lcd has exploded.16:25
whitequarkguess why.16:25
wpwraki hope they paid him through all this16:25
whitequarkhe soldered another one correctly16:25
wpwrakbecause it's pretty clear that this contraption will never actually run16:25
whitequarkand the switching supply has begun to heat. instead of projected ~250mA it could only deliver 25mA or such16:26
whitequarkbecause the author did not bother to look the correct value for inductor16:26
whitequark... ok, I think he wire-wrapped more stuff16:26
whitequarkthe LCD still did not work16:26
whitequarkit had 5 passive components in a row which should have been mounted near it16:27
wpwrakthat;s what you get when you outsource your engineering to the last tribe of homo habilis16:27
whitequarkthere were only three16:27
whitequarkand, er16:27
whitequarkI do not know what was on his mind16:27
wpwraknothing ? :)16:27
whitequarkbut it has been looking like he just threw them on the pcb and poured some solder from the top16:27
whitequarkthey weren't even connected to the corresponding pads16:28
whitequarkthat is16:28
whitequarka cap C1 was connected between a R1 and R2 pad16:28
wpwrakcargo cult engineering at its best16:28
whitequarkI think there were some more exploding tantalum caps16:28
whitequarkcan he ever mount them correctly at all?!16:29
wpwraki suoopse he did all the research for sport. because it's already obvious after the first 2-3 errors that this can't work.16:29
whitequarkor maybe test the fucking board. just for lulz, you know16:29
wpwrakwhy test it if you know it doesn't work ?16:29
whitequarkwell. they somehow are able to deliver working product with their way16:29
whitequarkI dunno how16:29
wpwraki doubt they do :)16:29
whitequarkatmegas can actually withstand the VDD rape16:30
whitequarkhe did it for last like 5 years for so and "it worked" (quote)16:30
whitequarkthey really do. I've seen that myself16:30
whitequarkthey sell it16:30
whitequarkit fails16:30
whitequarkthey sell some more of "it"16:30
whitequarkit sometimes works.16:31
whitequarkaccidentally, I presume16:31
whitequarkand they pay him, ahem16:31
wpwrakyeah, one million monkeys ...16:31
whitequark$300 a month16:31
whitequarkthat's low even for Russia16:31
wpwrakhe must be desperate :)16:31
wpwrakor he likes slumming16:31
whitequarkhe has his own sofrware business16:32
whitequarkwhich is quite profitable (I know as I made half of its infrastructure)16:32
wpwrakso he doesn't care. fine.16:32
whitequarkit's just fun to watch16:33
whitequarklike Dilbert in real life16:33
wpwrakanyway, i was talking about tablets that actually work at a basic level. yet may still harbour surprises.16:33
whitequarkbut a little bit more stupid16:33
whitequark(tablets) yeah, I know how they're done16:33
whitequark(homo habilis) roflmao, yes. /me has added that to a list of his quotes16:37
whitequarkhave I mentioned he considers 1206 components "quite small" ?16:41
whitequark80% of their "production" device just does not work from the very beginning16:42
whitequarkwhich is expected at least16:42
whitequarkso they throw it out16:43
whitequarkthey only use commercial atmegas. for the device which very purpose is working on below zero temperatures16:44
whitequarkthey just check if they work in a reactor with controllerd temp16:44
whitequark*controlled. sigh. this stuff has quite a negative impact on my English skills.16:45
whitequarkI just can't type correctly while laughing out loug.16:46
whitequark... oh crap. I'm sure you will appreciate the following engineering solution16:51
whitequarkhe neeeded to make a peak detector16:51
whitequarkso he took an amp (not an op-amp, but an old Soviet plain fixed-coefficient amp)16:51
whitequarkthen he adjusted the amplification so it would saturate at any signal > 0 (the ultrasound sensor sends voltage spikes in range of mVs)16:52
whitequarkand that kind of works as a detection16:52
whitequarknow, try to guess how he wants to compute the amplitude of the input signal16:53
whitequarkas a quiz.16:53
whitequark*measure, not compute16:53
wpwrakwhy measure ? put something. resistor, inductor, chewing gum, whatever. if it works, great. if it fails, throw it away and try again.16:55
whitequarkthe amp is connected to the power rail via a resistor16:56
wpwraklet me guess - that's a government-funded company with an inexhaustible money supply ?16:56
whitequarkand he suggests to use ADC in uC to measure the voltage drop, as the current is somewhat proportional to the height of the spike16:56
whitequarknope. that's a private-owned company16:57
kyakwhitequark: what is the name of this company? :)16:58
whitequarkjust read it17:02
kyakwhitequark: well, at least your friend is lucky enough to get the "hardware" before he started working on firmware17:05
whitequarkkyak: I'd suggest the name "failware"17:07
kyakha! :)17:07
Artyomkristianpaul: hi17:35
kristianpaulArtyom: hey17:39
ArtyomI've noticed that you were looking for me ;)17:39
kristianpaulyup time ago..17:39
kristianpauli cant make namuru core run on my soc, since i changed to bram like design..17:40
whitequarkwpwrak: wanna look at their flagship product which allows them to stay alive? it's a $35k water something analyzer, which they make and sell a few ones a year17:40
Artyomwhat does it mean bram-like design?17:41
kristianpaulArtyom: did you never ran m1 bios in your board?17:41
kristianpaulArtyom: your simplified version of the baseband17:41
ArtyomI ran bios...17:41
whitequarkand that's a "production" variant17:41
kristianpauli dont know what  broke, system got hang..17:41
Artyomyou want to put bios + my_code to bram? Not in DRAM?17:42
kristianpaulno no17:42
kristianpauli was sjut migrating hdl part17:42
kristianpaulwith a minor modification i added this wire assign ram_we = wb_cyc_i & wb_stb_i & wb_we_i;17:43
kristianpaulso i can write 8'hF1: if (ram_we) instead of 8'hF1: if (ram_we = wb_cyc_i & wb_stb_i & wb_we_i)17:43
kristianpaulArtyom: i still had not looked at debug code, already made the menus still not tested the algoryth for fiding  correlation peak17:44
kristianpaulwhen i mean bram i meant for the namuru core17:44
kristianpaulfor accessing the reigisters and writing to it too17:45
kristianpaulthis is the code http://pastebin.com/AQ2FaED317:46
ArtyomI'm studing your code... 17:50
ArtyomAnd you have problems with it now?17:50
kristianpaulmy soc hang and bios trought garbage..17:50
kristianpaulthis was working https://github.com/kristianpaul/milkymist/blob/gps-sdr-testing/cores/namuru/rtl/namuru_baseband.v17:52
kristianpauli'll sinthesize old version again and see17:55
ArtyomWhen I started to play with MM SoC + my_cores I faced similar problems. So at first I added simple bram core - in order to test BIOS writing/reading commands on the correlator-address... When I fixed all bugs - I started to experiment with 17:55
Artyomcorrelator core17:55
kristianpaulyeah i wanted to skip that part :)17:56
kristianpaulbut i also created a very simple core derivated from csrbrg when i started that one..17:56
kristianpaulArtyom: had you able to write to the wiki of qi-hardware?17:56
kristianpaulI was thinking in a roadmap with tasks and etc..17:57
ArtyomI haven't tried it yet...  I will experiment this week with it...17:57
kristianpaulBte had you tried all PRN codes from namuru datasheet work?17:59
kristianpaulconsidering u have a gps simulator :D17:59
Artyomno... I think it's easier to do with verilog-simulator ;)18:01
ArtyomIf you would ask this question half a year ago - I would agree to test with simulator. But now I think that if hdl-simulator can be used then it must be used...18:07
kristianpaulArtyom: what about a nother namuru +milkymist combo + dac that generatate the simulated gpa signal?18:08
kristianpaulwhat you think about it*18:08
kristianpaultoo elaborate ?18:10
ArtyomDo you think to make something like GPS-simulator? 18:16
kristianpauljust intelually curios for now yes18:17
ArtyomSomething simple? Like generating only one or two satellites signals?18:18
ArtyomOr more intellectual? To generate signal that can be used to calculate position18:20
ArtyomWhat is the aim of such a project? To test gps equipment? Or some other reasons?18:23
kristianpaulbasically test equipment18:25
kristianpaulgps simulators are prety expesinve18:25
kristianpaulbuy anyway just curious18:25
ArtyomIf you keep in mind testing of GPS equipment then it's easier to make device that can record and play back GPS signal18:25
kristianpaulhad you seens something un-unual in the pastebin?18:25
ArtyomSuch a toys also exist and cost much cheaper then simulators18:26
kristianpaulrecord and playbak sounds good to18:26
kristianpaulhow are they called?18:26
kristianpaulthe toys yop said18:26
kristianpaulbtw do you plan add glonass support to osgps?18:32
LunaVoraxGood evening everyone!18:32
kristianpaulis okay fot you work with scilab for now?18:32
kristianpaulevening LunaVorax 18:32
LunaVoraxMy girlfriend got her kryoflux a few days ago, she loves it already!18:33
kristianpaulgotta go, greak is off back to work..18:38
ArtyomI didn't find errors in your code... 18:42
ArtyomHere are some links about GPS record-play_back systems: extremly expensive: http://sine.ni.com/nips/cds/view/p/lang/en/nid/206806 ; cheaper alternatives: http://www.spirent.com/Solutions-Directory/GSS6400.aspx and this one: http://www.labsat.co.uk/ 19:10
Artyomyes, I definitly want to add GLONASS support to osgps...19:10
Artyomsorry, what did you mean about scilab?19:11
qwebirc61666Is It Possible To Connect A Camera To The Nano Note?19:14
qwebirc61666E.G a webcam, a stand alone camera (DSLR, Compact, etc) (If not a USB one what about the very old ones that do not use USB?), or some other sort of camera?  If so could it do video as well as stills?  Thanks19:14
Artyomkristianpaul: may you could test my core in your design...19:16
kristianpaulArtyom: yes i could19:17
viricqwebirc61666: it should be a SDIO camera19:19
Artyomkristianpaul: Honestly speaking I think from time to time about possibility to develop gnss-simulator. But the task seems rather difficult. There is number mater's thesis available in internet that can be a good starting point.19:28
kristianpaulthe record and playback looks usefull19:35
qwebirc61666So how could one connect a SDIO camera?19:37
qwebirc61666viric: look up19:37
viricthere is a microsd port.19:38
qwebirc61666viric: so you connect the camera to the nanonote by the cameras sd card port and the nano notes sd card port? I am a neewbie btw.19:40
viricfind a camera that can be connected to a sd card port.19:41
viricI don't know if they exist.19:41
qwebirc61666dad has a old cam that does not use usb it works by a sireal with so many pins i forget19:43
virica serial line camera19:44
whitequarkthere are some SDIO cameras around19:44
whitequarkthey're proprietary, old, unsupported and crappy19:44
whitequarksame for SDIO WiFi19:45
whitequarkyou may have success with serial cameras, through19:45
whitequarkyou can get something like 1-3fps on 640x48019:45
whitequarkon a nice baud like 3mbit which ben should have no problems supporting19:45
qwebirc61666do very compact small ones exsist?19:45
whitequarkafaik yes19:46
whitequarkdefine "compact"?19:46
qwebirc61666ours is not pocket size19:46
qwebirc61666little bit thicker than a smart phone19:47
viricI just found http://www.saelig.com/VI/BOVI001.htm19:47
whitequarkthe ones we have in russia are exactly like that19:48
whitequarkI think it's the very same camera I was talking about19:48
qwebirc61666I think that size is ok19:49
qwebirc61666shame its jpeg and not webp19:50
viricis webp any standard already?19:55
viricor only marketing?19:55
viricroh: ok19:55
viricbut webm is not only marketing, right?19:55
rohwebm is used, but poorly supported19:55
qwebirc61666opera,google chrome19:56
viricbut as for standard?19:56
qwebirc61666hopfuly there will be more support in the future.19:56
rohviric: standards are something everbody uses as default. neither webm nor webp is that19:57
qwebirc61666the stanard is set i thank and with containter. might be the odd bit left to do19:57
viricI wanted to reencode some videos... and I had to decide the encoder19:58
viricI decided for vpx on matroska19:58
viricI hope I'll be able to decode them for long19:58
qwebirc61666roh: what rubbish.19:58
viricdo you think similar?19:58
virichm vp819:58
viricfrom libvpx19:58
qwebirc61666libvpx is codec19:58
rohhm. well.. i dont think thats universal yet. you will have trouble with most devices playing that19:58
qwebirc61666ffmpeg project had even better codec in dev19:59
viricah, devices19:59
viricI don't care much on the devices19:59
rohmatroska as container is ok. we will see about codecs19:59
viricI'm doubting on vp8 or xvid19:59
viric(the only device I could be interested in using plays xvid)19:59
rohxvid in avi is the de-facto standard btw19:59
qwebirc61666not mpeg4 plese19:59
qwebirc61666avi rubbish and mpeg420:00
rohxvid IS mpeg4.220:00
qwebirc61666so STOP USEING RUBISH ARRRR20:00
rohand actually looks better than most h264 codecs20:00
qwebirc61666what a waste of band width20:00
rohqwebirc61666: stop annoying people.20:00
viricthe vpx people give even verilog decoders for what I remember20:01
viricfor free20:01
larscviric: they say they do20:01
rohviric: well.. i know what the hw decoder chipsets support and thats basically mpeg4.2, mpeg4.10 (h264) in sd and hd resolutions, sometimes even all res... sometimes resolutions which are divisible by 16 or so20:02
viriconly say? :)20:02
rohin addition to some wmv9 and similar stuff. but nothing 'new' from the last 2-3 years.20:02
qwebirc61666support is separate from standard20:03
rohit takes time till stuff is supported in asics. on the other hand its much more energy preserving in asic.20:03
rohqwebirc61666: there can not be one standard20:03
qwebirc61666I never said there must be one standard20:04
rohusecases are too different in requirements. (latency/quality/features/support/compatibility). every usecase develops its own combination and thus different codec/container combinations get chosen.20:04
qwebirc61666vp8 h264 and if you must must must then other20:05
rohhw vendors usually adopt the 'support everything major used codec' route and do containers in sw.20:05
qwebirc61666genic hw enc/denc suport vp8. see webm website20:06
rohwebm is not foss useable in hw20:06
qwebirc61666you mean hw not foss?20:07
rohthere is no opensource for webm support in open hardware20:07
rohthey got a design but refuse to release it.20:07
rohunder a license useable for open hw20:07
rohclosed you can make a deal. but not for opensource hw20:08
qwebirc61666from google, this is?20:08
qwebirc61666oh did not know that. thanks20:08
rohsomebody tried getting it from google for the milkymist but they refused/ignored it20:09
rohso the stance of google on opensource stays 'a bit for marketing, but not by heart'20:09
qwebirc61666remmber open source is free software murdered for evil bizness20:11
rohi think you are confused20:11
rohits broken businesscases which conflict with opensource.20:11
rohcannot help there.20:11
qwebirc61666open source was started for "bizness"20:12
qwebirc61666in OS it is ok to have non-free parts in FS it is not20:13
rohfrom my pov opensource is more than just 'free sw'20:13
rohqwebirc61666: thats nitpicking from trolls like stallman20:13
rohwho, moderatly speaking, has obviously no clue about hw and its details20:13
Aylawell, he's not totally wrong, FS implies OS, the contrary is not true20:14
rohopen source and open hw allows one to have all details and documentation on and for a device and its sw. and still have a businesscase.20:14
qwebirc61666can't coment on HW too much neebie me20:14
rohthe service to develop and manufacture.20:15
qwebirc61666roh: thats FS20:15
rohthats what you pay for.20:15
rohAyla: still nitpicking. the details are washed out massivly so some define things this and some the other way.20:15
larscfor some people free software is the stuff you'd get on the pirate bay20:16
rohqwebirc61666: on hw its 'just' more difficult to get all factors into alignment due to contraints like interfacing stuff not completely open20:16
rohlarsc: exactly. so terms are very mangled and confused. we always can just try defining what things mean for us20:17
Aylalarsc, hey, there is legal stuff too on TPB :)20:17
rohtpb has a section on 'physical stuff' now too :)20:17
rohcad/cam files of real stuff20:17
qwebirc61666RMS a troll. narrr.20:18
qwebirc61666when I say Free sofware (FS) i mean what is definded by fsf.org20:20
rohthat definition is very us-mindset centric. not complete for helping open hw20:20
lindi-indeed it does not cover hardware at all20:21
rohe.g. i miss a openhw license which is similar to gpl20:21
qwebirc61666I aggree the term Free Software is not ideal. Posonley I think some thing like freedom(s) software would be better. thought perhapes a bit cheasy20:22
roheven licensing is a compatibility game20:22
qwebirc61666us-mindset centric?20:22
rohin the us everything has to do with business (from py pov)20:22
qwebirc61666FS all about biz? what!20:23
rohthere are nearly no not-for-profit organisational structures and if they tend to be big to compensate the buerocratic overhead.20:23
rohin europe there are many clubs/associations20:24
rohlegal entities NOT a company per default.20:24
qwebirc61666the fsf.org is non-profit20:25
qwebirc61666FS is strict unlike OS. I do see that this is harder in HW to applie20:26
qwebirc61666thought the nano note and arduino to my knowage are more in line with FS than OS20:27
qwebirc61666I would say the the terms Copyleft HW is the FS equenv and Open HW is the OS equiv20:28
qwebirc61666If you are thinking I am in the usa or think they are angle then that is wrong. I think the usa is evil20:31
rohcountries are not evil. governments in power can be sometimes. single people can be for sure.20:31
qwebirc61666sorry my bad wording. good point20:32
lindi-qwebirc61666: arduino schematics are only readable with non-free software20:33
qwebirc61666lindi: oh poo20:33
lindi-qwebirc61666: eagle20:33
lindi-qwebirc61666: probably the most common electronics designed tool used by amateurs20:34
lindi-electronics design20:35
qwebirc61666roh: "xvid IS mpeg4.2" I was not being prosice. thanks for the full version num20:39
qwebirc61666right better go and have diner I will be back with more...20:39
whitequarkroh: webm is in ff, isn't it20:55
rohno clue. doesnt matter. ff is sw not hw.20:56
whitequarkby the way, google not releasing an ip core is fucked up. imho, as much evil as google is/became, it still does more good by net result. are you sure there aren't a patent case or something like that which legally prevents them from releasing the source?21:00
qwebirc61666ffmpeg uses libvpx don't know the status of the better webm codec.21:17
qwebirc61666right me now go and help with dish washer. more to come...21:17
viricwebm isn't a matroska container with vp8+vorbis?21:24
whitequarkviric: afaik it is. and libvpx is vp8 codec in this case then21:24
qwebirc61666libvpx is a codec of the vp8 format21:52
qwebirc61666the webm container is based on matroska21:52
viricI'd call libvpx an implementation21:52
viricof vp821:52
qwebirc61666yes 21:52
qwebirc61666codec = software implementation21:53
larsccodec = compressor-decompressor21:55
viricwhen I used 'codec', I meant the definition in higher level than an implementation21:55
larscdoesn't matter whether it is done in hardware, software or by hand on paper21:55
viricI subscribe.21:56
qwebirc61666viric: webp = lossy and lossless image21:56
viricqwebirc61666: I'm trying it now21:56
viricqwebirc61666: I can't find how to make a lossless image21:56
qwebirc61666viric: webm supported in : firefox, opera, google chrome, IE vir plugin, apple never but stuff them they are evil. M$ & apple make money from h264,mpeg*,mp3,mp4,acc21:58
qwebirc61666I hate avi becuase it is used on so many videos, websites, torrents just so some lazey ass can burn it to a cd for there dvd player. grrrr. avi is nasty for streaming due to not having decent seek.22:03
viricneither allowing subtitles or multiple audio tracks22:03
viricor even some codecs.22:03
qwebirc61666I hate mpeg4.2 + mp3, acc because they are bandwidth hogs. h264 for the least bandwidth vp8 for a bit more bandwidth but less decod power needed. as for mp3 & acc. Vorbis is superior!22:07
qwebirc61666instead of your avage 700MB mpeg4.2 under SD res. if you used h264 you can have 720p!22:08
Aylayou hate AVI. But remember it's super-old22:08
Aylaback in the day, streaming it wasn't really possible22:08
viricalso jpeg has its problems. and it's there :)22:09
Aylaon a 33kbps dialup connection22:09
viricsuch a surviving means quite a success22:09
Aylaactually, is there a good alternative to jpeg?22:09
viricpov? :)22:14
viricneither webp or jpeg2k show a big improvement I think22:15
viricAnd jpeg is still not at its limits :)22:15
viricAyla: http://viric.name/soft/qjpegrest/22:15
viricI also think that people became used to the jpeg noise.22:16
qwebirc61666I hate jpeg too22:16
viricthe brain can learn the jpeg noise enough to reduce the annoyance :)22:16
Aylaheheh, I see you know the topic22:16
qwebirc61666"such a surviving means quite a success" pah! it means humans are lazy22:17
qwebirc61666webp is a huge improvemnt22:17
viricimprovements over jpeg, if any, are not significant enough22:18
viricotherwise there would be a change.22:18
qwebirc61666webp is from 2010 or 2011 I forget. give it a chance!22:19
qwebirc6166630% improment if I rember22:19
virichow did you measure that?22:21
qwebirc61666webp website22:21
Aylaand what '30% improvement' means?22:21
Aylayour eye see a 30% better image?22:21
qwebirc61666size and quarlity22:21
viricReplacing an image interchange convention by another is an oportunity of business. We are going to see that from time to time.22:22
qwebirc61666interchange convention?22:22
qwebirc61666no loss in looks but smaller file size22:22
viricBut the improvement has to overcome the inertia. That's why I meant the improvement has to be that significant enough.22:22
viricPeople working in imaging, loose any ability to evaluate image quality22:23
qwebirc61666think very slow internet22:23
mthMP3 is still the most popular lossy audio format even though there are better algorithms out there22:23
viricThey start to give opinions on images as a PSNR measure would give. :)22:23
qwebirc61666how i mena22:23
viricAnd as I stated in that website...22:24
viricthere is a big field of improvement at jpeg *decoding*22:24
wpwrakviric: (qjpegrest) pretty impressive22:24
viricalthough explored, almost not applied anywhere.22:24
qwebirc61666re mp3 because they don't know aby better/don't care/are lazy22:24
wpwrakmaybe everyone is just waiting for the patents to expire ? :)22:25
viricwpwrak: thank you! :)22:25
viricwpwrak: I never got anyone to even download that web page. :)22:25
viricsince years posted.22:25
qwebirc61666why borth though when vorbis is patent free and better22:26
viricwho knows if it is patent free?22:26
wpwrakviric: it needs to be integrated into libraries. that way, people will use it22:26
Aylaviric: why don't you add those improvements to libjpeg?22:26
viricwpwrak: I started it as a libjpeg replacement, with envvars affecting it22:26
viric(the vcs show that)22:27
viricbut I saw that the results looked far better (to my distorted taste, of course) with a little interaction per image22:27
qwebirc61666re qjpegrest but your starting with a rubbish format to begin with22:27
wpwrakviric: yeah, it should be just part of libjpeh and ffmpeg. then everyone will use it.22:27
wpwrakviric: otherwise, few if any will bother with the extra effort22:27
viricBut it's not trivial to find a formula that always improves22:28
viricit may either blur too much, or work too slow...22:28
viricI thought of gimp integration... but I imagined that'd be hard to maintain, with too much extra work unrelated to the topic22:28
qwebirc61666opera very quickly added suport for webp because it meant better looking images but with out the file size22:29
viricand the libjpeg integration worked fine. It's just in anothe branch.22:29
wpwrakviric: ah, i see. i thought it worked on algorithmic issues, so it would always produce a more accurate image. if it needs interaction, that's bad22:29
qwebirc61666opera use webp in there tourbo service22:29
viricqwebirc61666: a tool to *decode better* has better value the more images there are in that format ;)22:29
viricwpwrak: the interaction also allows to use some almost parameter-less decoders :)22:30
wpwrakviric: few people are determined enough to tune nakedgirl047.jpg if nakedgirl048.jpg promises even more nudity :)22:30
viricwpwrak: but with a little hand help that improves22:30
viricWell, I imagined it could be used as in sound restoration, with human evaluation and interaction22:31
viricIf I had found anyone with a minimal interest on that, maybe I would have developed further.22:31
wpwrakwell, maybe if someone comes across the original hand-coded monalisa.jpg ...22:31
viriceh? :)22:32
wpwrakgimp sounds like a reasonable target. that's a place where people might spend some time tuning things22:32
qwebirc61666the decode better is only for quick previews? before if one waited you could still get the same results right?22:33
wpwrak(monalisa) for making the effort of tuning :)22:33
viricI looked at gimp code back then22:33
viricand it made me jump back22:34
wpwrakmake a lib, show it to them, make them like it, then let them worry about the rest ;-)22:34
viricI tried22:34
viric:) bye22:34
qwebirc61666are there any higher res cameras that one can connect to the nano note then?22:36
zrafa_qwebirc61666: nn does not have usb host22:40
viriczrafa_: he accepts serial port cameras22:41
qwebirc61666the one that was linked too earer had been super seeded by one that could do 30fps video. is that posable over serial? hardware newbie here22:44
qwebirc61666here is the new model http://www.saelig.com/product/BOVI008.htm22:44
qwebirc61666roh: I know ffmpeg is software not hardware22:47
rohthere are different mechanisms at work defining what becomes the de-facto standard and most are not really easy to influence/ are not just based on knowledge now22:50
rohfor example.. mp3 is still used not only because its easy and lots of files are in that format, but because it doesnt matter to the user22:50
rohif the user is pleased with the result, he has no reason to 'change' anything or switch to something better22:52
rohthe switch from something to something else is more likely to happen if the user gains something important to him.22:52
rohwhen your problem is neither bandwith or cpu, and quality isnt your issue, you will most likely use the most easy to use for example.22:53
qwebirc61666roh: and then they don't bother to provide OGG + Vobis or even better FLAC arrrrrrr grrrrrr. selfish numteys!23:02
rohsome do.23:02
qwebirc61666most don't23:02
qwebirc61666and won't :(23:03
rohbut to be fair i am pleased if its in any defined format i have proper decoders instead of flash streams23:03
qwebirc61666I try emailing/posting and get no repliay :(((((((((((((((((((((23:04
qwebirc61666I hate flash too23:04
qwebirc61666I no long have it installed23:04
qwebirc61666jdownloader,savevideo.me and youtube webm23:04
qwebirc61666Magnautne.com is great. I have a member ship so artists get paid and I get great music and FLACs :)23:06
qwebirc61666but theres so much more great music that is only availbe in mp3 128 cbr :*(23:07
GNUtoojamendo can be downloaded in ogg vorbis too23:10
Action: mth is a big Magnatune fan as well23:13
qwebirc61666I know about jamendo and OGG. I only donwload the OGGs. I should of said. as someone was bound to say it. oh well23:18
AylaI use jamendo on my android phone23:18
qwebirc61666right bye dad will be calling...23:19
--- Thu Feb 9 201200:00

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