#qi-hardware IRC log for Thursday, 2011-04-21

kristianpaulxiangfu: ha, i catch you disasabling electronics  ;)03:38
kristianpaulthis lady is very smart03:39
kristianpaul"Information (including all of my work) is not scarce. Attention (time) is."03:39
xiangfukristianpaul: sorry. what you mean?04:03
xiangfukristianpaul: the case of that adapter. is not open by me, it's open by exploded, there are some black things on that screw04:35
xiangfuexploded dust  :)04:36
wpwrakrejon: (slowpan) of course, there's the lingering question: why call it SLoWPAN if 6LoWPAN already has a perfectly good name ? after all, we don't, for example, measure current in Torvalds to honor our benevolent dictator10:52
larschm... 5owWPAN10:53
wpwrakrejon: the trademark issue because of which we had to rename "MicroSD" to "8:10" doesn't exist there10:53
larscas a compromise10:54
wpwraklarsc: Sow-pan is what you call it when it doesn't work or is too slow ;-)10:54
dvdkwho's done the mplayer port for ben?11:12
dvdkkyak, was that yoyu?11:12
dvdkneed somebody to complain to :)11:13
wpwrakdvdk: just imagine it was kyak while you hurl your abuse at him ;-)11:19
dvdkwpwrak :)11:20
dvdkbtw have hardware-scaling support working in mplayer.11:24
dvdkso full-screen playback possible at 24 FPS (using less-than-fullscreen video)11:24
wpwrakdvdk: kewl. when will we have full HDTV ? :)11:24
dvdkunfortunately the ogg demuxer of our mplayer version has some problems11:25
dvdkwpwrak: as soon as you added the dvi port to ben :)11:25
Action: wpwrak wonders if one could bit-bang DVI ...11:25
kyakdvdk: hi there11:26
dvdkkyak: mahlzeit11:27
wpwrakdvdk: in what range is the maximum original resolution that still yields an acceptable frame rate ?11:27
dvdkwpwrak: depends on the codec.  with current theora (compiled with -Os), I get 256x144 without problems (maybe 80% CPU use)11:27
dvdkin theory we should be able to decode 320x240 if the codec were properly optimized and compiled11:28
dvdk(using the hardware for yuv->rgb step only)11:28
dvdkquality is quite fine when looked at from 0.5m :)11:28
dvdkbtw kyak, is that mplayer port maintained by you?11:29
kyakdvdk: kind of. :)11:29
dvdkkyak: i think we need to upgrade.  the ogg/theora demuxer seems to have a bug.11:29
dvdkmy test video drops video after 60s (even without hw accel -vo fbdev).11:30
dvdksome strange message about "frame duration too long)11:30
dvdkPC mplayer (svn) does not have the problem11:30
dvdk(however mplayer svn has yet another ogg/theore bug)11:30
kyakdvdk: you don't need to ask for my permission to upgrade ;)11:30
wpwrakdvdk: not bad :)11:30
dvdkkyak: yeah, but I'm a little short on time here :)11:31
kyakdvdk: i somehow have a feeling that increasing the version in makefile would do it11:31
kyakthere are no patches11:31
dvdkkyak: is there even a newer version released?11:31
dvdkok, ok, going to have a loog at it.11:31
kyakdunno, you just told me like it did :)11:31
dvdkkyak: maybe we have to use svn?11:32
kyaki think you already proposed it once. Are you sure it's a better options than using release?11:32
dvdkwait, going to upload my demo video (big buck bunny) plus the new vidix driver, to help motivate you :)11:33
kyakall right :)11:33
kyakhave to go now. btw, mplayer is missing libggi dependency11:34
kyakjust letting you know11:34
dvdkok. cu.11:34
kyakthis have been in my TODO list for some time now, i was waiting for a godo change to add it in Makefile :)11:35
wpwrakkyak: nice ... the transition from to-do to go-do11:38
dvdkkyak: looks like mplayer does not release tarballs any more.  use of svn is recommended.  just looking into patching our Makefile accordingly.12:55
vladkorotnevhello everyone :P16:12
rjeffries__wpwrask there is a thin line between "realist" and "defeatist" //grin17:38
kristianpaulrjeffries__: what side you already cross?17:39
wpwrakrjeffries__: indeed, and there is a line between "can't do" and "can do". very often, it exists only inside someone's mind :)17:41
rjeffries__may 1,000 flowers bloom.17:41
rjeffries__beleive it or not, I am not at all risk averse.17:42
rjeffries__new topic?17:43
rjeffries__smiles again. anyway, why do we think Ben sales are so sluggish?17:44
kristianpaulmay be more people thinking like some body we know...17:45
kristianpaulyou got your ipad finally, right?17:45
wpwrakrjeffries__: ben sales, perhaps lack of exposure. nanonote sales on general, i'd blame lack of a follow-on product17:48
kristianpaulyou mean a missing/strong marketing strategy?17:49
urandom__what about lack of usb-host? :P17:50
wpwrakkristianpaul: (exposure) visibility17:50
rjeffries__yes without a strong follow-on BEN is a VERY VOOL niche inside a niche inside an imperceptibly small target market17:50
wpwrakurandom__: "see also \"follow-on product\""17:50
wpwrakrjeffries__: indeed. continuity also creates long-term credibility creates access to additional markets17:51
rjeffries__the business case for creating a new Ben unfortunately does not exist.17:52
rjeffries__I can understand why wolfspraul is focusing Sharism resources in the direction og Milylisy, a totally different marejt17:53
wpwrakrjeffries__: oh, i think the biz case for a better ben does exist.17:53
rjeffries__wpwrak including a new case to allow a decent keyboard and a larger touch sensitive screen?17:54
wpwrakrjeffries__: it's just difficult to convey to potential investors. at least difficult enough that, combined with the general lack of investor-attracting skills in the qi-hw community, it's insufficient (so far)17:55
wpwrakrjeffries__: i would consider a new case a requirement. i wouldn't change the form factor, though17:55
rjeffries__why would an investor risk capital funding something that by definition can be easily reproduced anywhere anytime? it does not compute17:55
wpwrakrjeffries__: if you want a netbook, there are dozens of choices. it would be insane to go there.17:56
wpwrakrjeffries__: likewise for a "tablet"17:56
rjeffries__no much smaller than netbook but enough larger to allow more comfortable use17:56
rjeffries__there are many existence proofs (read smart phone in various packages) where one gets a larger screen, and a better keybaord17:57
rjeffries__actually I am not positive physical keyboard is required17:57
wpwrakrjeffries__: the easy reproduction anytime only happens if the reproducer is convinced there's a market share worth stealing and that there won't be severe enough punishment17:57
urandom__form factor is good as is but marketing must focus more on showing how small it is and that the keyboard really is awesome, though we need bigger screen so that it doesnt look like a toy anymore17:58
rjeffries__what are the use cases that make a Ben compelling...17:58
rjeffries__other than "I want a totally open device"17:58
rjeffries__That's it, end of story. it does not approach a smart phone form two years ago in usability17:59
urandom__i would like to see an nanonote with fpga, so we get the fpga hacker also and are more flexibel17:59
kristianpaulmay be fpga with built-in arm chip.. :-)18:00
wpwrakrjeffries__: well, it all depends on what you're trying to sell in the end, and to whom. if it's devices to regular consumers, it's difficult. if it's devices to vars, openness is an enabler and a very strong differentiator as well.18:00
mththe biggest NN limitation for the average consumer is no wireless internet, imo18:00
kristianpaulis ebook market still young?18:01
wpwrakurandom__: i think an fpha nanonote would only make sense after a ben successor. there are still many important things we don't "own" yet.18:01
rjeffries__mth you are correct and that is off the table because of philosophical reason18:01
wpwrakurandom__: such as mechanical design18:01
kristianpaulbut. nah devel time sloq down those toughts..18:01
kristianpaulwpwrak: how long take freerunner to be developed btw?18:02
rjeffries__I don't want to own this mechanical design, but a new fresh approach would be interesting18:02
urandom__the nanonote makes a good translator/e-book, those stuff isnt cheap, we just need to make user friendly software18:02
wpwrakmth: my list of troublemakers has: usb host, wireless, display resolution, memory. not sure about the ranking.18:03
kristianpaulindeed with urandom__18:03
wpwrakmth: (troublemakers) for first sales18:03
mthwpwrak: bluetooth is also encumbered? because sharing a network connection with a phone would also be an option18:03
rjeffries__wpwrak add some more i/o in particulat I2C, also a second 8:10 port18:03
kristianpaulbetter fint the another vesion of the color digital dictionary and free it?18:04
kristianpaulthat take time a cost, but surelly is cheap than develop own hardware18:04
wpwrakmth: BT is also a problem, yes. it may not even be so much the standards per se, but access to chips/modules. if you can't get chips with useful documentation, you lose.18:05
kristianpaulanyone with all the interes and time, can help on the support of touch screen displays, so a mechincal side could be develop faster i gues..18:05
wpwrakrjeffries__: oh, i definitely want two 8:10 ports :) one inside, one outside18:05
rjeffries__keyboard needs backlight18:05
wpwrakrjeffries__: yup18:05
kristianpaulbtw you can use a UBB for a DIY led light ;)18:06
wpwrakkristianpaul: true ;-))18:06
wpwraki have a set of mean high-power leds. should adapt one ;-) unfortunately, they're all soldered to some board that's not very useful in hindsight18:07
kristianpauli think thats something rjeffries__ can achieve at home easilly :-)18:07
kristianpaulhigh-power leds = how may watts? :o18:07
wpwrakkristianpaul: maybe around one. don't quite remember.18:09
kristianpaulthats fair good18:09
wpwraki think it's these: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=475-2617-1-ND18:12
wpwrakmaximum power 1075 mW18:12
kristianpaulhe, for a moment i tought re-use the freerunner touch-display with milkymist  ..18:16
mthwpwrak: what about an internal USB bus? for example the wifi in the Wii is accessed via USB. or is that not open enough?18:28
kristianpaulhttp://openwear.org/lookmap, we add that to the woldgang freedom pyramid18:43
kristianpaulwe should*18:44
kristianpaulmoment, where is the source..18:44
wpwrakmth: you mean to have a bus in general ? or specifically for wlan ?18:48
mthwpwrak: the bluetooth is also connected to it18:48
mthmaybe even the SD reader, but it's been a while since I was messing around with that kernel18:49
mthin any case, it has 2 busses, one which is connected to a USB port and another for built-in peripherals18:49
rjeffries__USB hub inside next nanonote is a Good Idea woukd allow Much Freedom (to add new goodies)18:56
Jay7usb over slowpan :)18:57
rjeffries__wpwrak you did not mention this on your list but any Ingenic-based next Nanonote would use a new SOC such as 4760 for many valid reasons18:58
rjeffries__for a new product are people willing to stay with 320x240 resolution?19:00
mthwpwrak: wiibrew.org says the wifi is actually accessed via an internal SDIO bus, but the bluetooth is via an internal USB bus19:02
urandom__rjeffries__ i would but of course higher screen resolution would help with many apps19:02
urandom__actuelly being limited to 320x240 is also interesting. forces you to focus on what is important19:04
rjeffries__the advantages of increased screen real estate are significant for usability19:06
wpwrakrjeffries__: (4760) yup. it may just not be an easily externally visible shortcoming.19:06
kristianpaulwpwrak: you thiks this schm is okay for a 3v3 to 5V level convertion for Ben with a UBB https://sites.google.com/site/pinguinotutorial/wrt54gl/conv533.png?attredirects=019:22
kristianpaulI was thiking an opto isolated addon, but i need to source if the local electronic shop sells "optoacopladores"19:24
kristianpaulhmm LVT-xx19:31
wpwrakkristianpaul: the local shop will almost certainly have them. they're very very common.19:34
kristianpauli hope here..19:34
kristianpaulbtw to save me some reading do you rememnber mac current i can driver with the 8:10 port?19:35
wpwrakkristianpaul: your diodes probably have better counterparts inside the chips. so they may not do much. also, if one side is entirely 3.3 V and the other entirely 5 V, you don't need to clamp the 3.3 V -> 5 V direction19:36
wpwrakkristianpaul: on the I/Os ... i think about 4 mA19:36
kristianpaultunner need SDA and SCL,19:37
kristianpaulalso a CE pin19:37
kristianpauland a LOCK signal to be read19:37
kristianpauli dont get the part about ( you don't need to clamp the 3.3 V -> 5 V direction)19:37
wpwrakkristianpaul: WRT54GL.TX -> PINGUINO.RX, why the diode ?19:40
kristianpaulbad sch19:40
kristianpaulpinguino is 5V ttl, not 3v3 compatible19:41
kristianpaulbut in the case of the nanonote i need 3v3 to 5V in both directions19:41
wpwrakalso,  the chip on the WRT54GL.RX side probably has its own diode. and it probably has a lower voltage drop than the 1N4148. so you may still send about (5-3.3-0.3)/330 A through it, and very little through the 1N414819:41
kristianpauli see19:42
wpwrakyou may want to check how the chip on the WRT54GL side is specified19:43
kristianpauli think i'll just end using a transitors as on/off gate to driver the 5V signal19:44
wpwrakyou may need two, so that you don't invert it19:44
kristianpauland for the coming LOCK signal from tunner just  adiode or voltage divider to drop it to near/bellow 3v319:45
kristianpaulah, you mena darlington bridge?19:45
wpwrakkristianpaul: hmm, no, darlington also inverts19:48
kristianpaulneed to leav, read you later19:48
kristianpaul"Interoperable products can be recognized by the logo Qi"19:49
kristianpaulmy ben have a Qi logo ;)19:49
mthoh great... in a few years we'll have people joining this channel when they have problems charging their phone...19:52
kristianpaullol ;)19:52
kristianpaulwpwrak: http://ur1.ca/3x2ab20:25
kristianpaulyeah, i got now what you mean with invert it :-)20:26
wpwrakyeah ;-)21:07
--- Fri Apr 22 201100:00

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