#qi-hardware IRC log for Thursday, 2011-01-20

rohwhat scope are you talking about?00:28
rjeffries19:52 <wolfspraul> rjeffries: is it the Real Ron? //yes. jave been reading some irc logs, thought I'd watch in real tiem01:00
wpwrak(catching up ... slowly)01:00
wpwrakya/4760: yes, that's th  one that seems to have left the radar.01:01
wolfspraulrjeffries: welcome here!01:05
rjeffriesThanks. I finally realized this must be where the action is, since mail list is so quiet.01:06
wolfspraulyou asked about fpga power consumption once01:06
wpwrak(touch screen) i agree on pragmatism, which shouldn't exclude anyone having interesting ideas, but in the absence of tangible results, take whatever works. i have no strong opinion wrt resistive/capacitative01:07
wolfspraulto give you a rough ballpark starting point, at the same process technology you can assume 10* the price, 10* power consumption and 1/10th of performance01:07
wolfspraulbut that comparison won't get you to understand why it's an interesting chip01:07
wpwrakalternatives could also include a touch pad or an external mouse. but a touch screen would be nice in any case.01:07
wolfspraulthen you asked about more memory once, the answer is that it won't help current Ben NanoNote owners :-)01:08
wolfspraulwpwrak: 4760 hasn't left the radar01:09
wolfspraulbut it's upon Ingenic to make the next move01:09
wolfsprauland my estimate for them making that move in 2011 is...01:09
rjeffrieswolfspraul: I pretty much assumed those factors. the point I was hinting at is MM (fpga based) is very(!) cool, but a very ;omg way from being a pocekatble cheap batter device a la Nanonote. that's all01:09
wolfspraulhmm. let's say 25%01:10
wolfspraulfpga makes a lot of sense together with the cartdrige idea01:10
rjeffrieswolf, the more memory question was "what if you spin a low risk improvment to Ben that can be done at modest cost. it would be ready when the current inventory is depleted.01:11
wolfspraulit's not radical enough, and Ben is maturing nicely so no rush01:12
rjeffriesMM is fabulous stuff, but attackes very different use cases than nanonote, at a much higher cost01:12
wolfspraulfor example ben-wpan may become an excellent 'next level' type of thing01:12
wolfspraullet's first get ben-wpan to work well with Ben01:12
wolfspraulout of the box, including Linux kernel, upstreaming, apps, etc.01:13
wolfspraulit will take time01:13
rjeffriesI assume werener is aiming to have the wpan inside a future nanonote update01:13
wolfspraulthat's something that could be integrated into Ya, and still not exlude Ben owners because they can get it as an add-on to talk to their new Ya01:13
rjeffriesmy pragmatic side would predict that you are very smart to convinve Ingenic you may walk away01:14
wolfsprauldon't understand01:14
wolfspraulI would love them to move forward, we see.01:14
wolfspraulthey are just a chip supplier to me, my model is not to make money because I write free drivers for their proprietary technology01:15
rjeffriesbut, the practiacl Ya Nanonote will have newer Ingenic chip, wpan stuff, more memory. may not get USB host, or lighted eyboard, or larger screen01:15
rjeffrieswhat drivers are you missing now?01:16
wolfspraulmaybe you are so eager about Ya because you regret to never having bought a Ben :-)01:16
rjeffriesno regrets, Ben is not designed for me, but is a delightful product. ;)01:16
rjeffriesI now I sold one for you, a friend in Santa Barbara'01:17
rjeffriesI am afraid it is a drawer computer: stays in his desk drawer01:17
rjeffriesanyway you guys are doing interesting wor.01:17
wolfspraulthe drawer rate is high, no doubt01:19
wolfspraulI think you slowly understand our priorities though, even though I am painfully slow in making it clear, in light of your very reasonable feedback/questions.01:20
rjeffriesI do not share the religious fervor that I admire in you. it taes unreasonable men to get things done01:21
wolfspraulI see software on the Ben improving, and a whole group of open efforts around it, like ben-wpan, a similar hoperf project, openwrt, jlime, werner's ben case work, roh's m1 case work and maybe ben-wpan case work, and so on01:21
rjeffriesBut I am not unfamiliar with producy managment and the practicalities of maing a buc with systems01:22
wolfspraulno religious fervor, just follow what feels right01:22
wolfspraulhow do you like Milkymist One?01:22
rjeffriesbuc (as in $UASD) aka profit ;)01:22
wolfsprauloh it will be hugely profitable01:22
wolfspraulSebastien is waiting in awe for your first comments on Milkymist One or Milkymist the SoC :-)01:23
rjeffriesMilkymist is technically very interesting. the designer ,ay welll be at genius level01:23
rjeffrieshis challenge will be to find the fisrt 50 people who will invest the amount of sale when volumes are low01:24
wolfspraulI've sold 25 until today01:25
rjeffriesI have no idea how big the VJ niche is, or whether MM will exceed what can be done with a fast lappie01:25
wolfspraulI have 5 left from the rc2 run.01:25
rjeffrieswolf... but one assume you sold those to his distributer01:25
wolfspraulnot even rushing to sell them, as my main priority is to get to people who may really contribute back, avoid the drawers...01:25
wolfspraulnot really, most of them are sold through01:25
rjeffriesgood stuff01:26
wolfspraul10 for fosdem, they are not sold yet01:26
wolfspraulnot sold through yet01:26
rjeffriesbriefly, a new topic?01:26
wolfspraulare you writing a new topic? or you are asking whether m1 is a new topic?01:29
wolfspraulbtw, one massive priority I personally want to work more on this year is to get free data onto the Ben, and m1 once we have established some way to share visual patches and other artworks01:34
wolfspraulwe need to connect the out of the box experience to high quality free data resources on the web01:34
wolfspraullike wikimedia foundation content (wikipedia, wikibooks, wikispecies, wikimedia commons)01:35
wolfspraulproject gutenberg, maybe some spoken stuff - I love audio books for example01:35
wolfspraulmaybe some animation series, if any free ones exist.01:35
wolfspraulso the Ben out of the box should allow you to just download the Beethoven symphonies from Wikimedia Commons, and play them.01:36
wolfspraulmaybe we can do something similar with jamendo for newer music (wikimedia commons is only collecting classical music right now)01:37
wolfspraulmy point is that all of this may already be possible theoretically now, but I think we should make it very easy, out of the bo01:37
wolfspraulthe Ben should be a window to all sorts of free data in the world, 'preinstalled' (= the correct URLs and viewers/download apps preinstalled)01:38
rjeffriesthat sounds cool01:39
wolfspraulwith m1 we first need to establish some forum to share patches, background pictures, etc. I very much look forward to working on that.01:39
rjeffriesone last question before I am off to bed01:39
wolfspraulthat's the question?01:40
rjeffriesis Carlos actively continuing to develop what used to be SAKC (I can;t eep all the new meaningless model names straight;)01:41
wolfspraulyou are always asking the good questions01:41
wolfspraulI respect Carlos work and perspective, but right now we can say he 'suspended' his activities in Qi :-) I think that's fair to say...01:42
wolfspraulhe was very pissed about the discussions about Jane's wonderful crochet work01:42
wolfspraulthen he proudly unsubscribed01:42
wolfsprauland didn't even read the FABULOUS replies he got to his rant mail01:42
wolfspraulwhich he is also proud of (to not have read them)01:43
wolfspraulso we leave it like that for now01:43
wolfspraulI think he will be back eventually.01:43
rjeffrieshis card is interesting. very different target audience than Milymist01:44
wolfspraulthere are tons of dead open hardware projects in the world01:44
wolfspraulsoon can be dead_open_hardware_project++;01:45
wolfspraulyou need to ask Carlos for proof of live, not me.01:45
rjeffriesthe camera project also has merit, especailly since it MUST include Werner;s radio crcuit so w ecan control remotely01:45
wolfspraulhe didn't like David's part (marketing) either, so David/Tuxbrain dropped out of SIE/SAKC.01:45
wolfspraulthe camera project is beautiful, I agree.01:46
wolfspraulbut I need to see how I can make it a reality.01:46
wolfspraulfocus is on m1 now.01:46
wolfspraulrjeffries: one last word today. you give us very valuable feedback.01:46
rjeffrieswell, good night and best of luck01:46
wolfspraulbecause you don't come down from your 10,000 feet view, you don't see many of the details we are working on.01:46
wolfspraulso you need to accept that as a limitation of your arguments01:47
wolfspraulI'm not trying to sell you a Ben anymore.01:47
wolfsprauljust get an iPad for now and wait :-)01:47
rjeffriesmy background has also included the down and dirty practical part01:47
wolfspraulthere are many many worthwhile details we are working on, and those details are essential/mandatory/required for the next steps.01:47
rjeffriesactually I do not buy any Apple products01:47
wolfspraulyes sure, but if you would have gotten a Ben or M1, you would see priorities different.01:48
wolfspraulI love Apple.01:48
rjeffriesbut when the android 7 inch tablets get real, maybe01:48
rjeffriesI admire apple, yes01:48
rjeffriesJobs is a genius01:48
wolfspraulthere is no company in the world that cares less about the amount of destruction they cause than Google.01:48
wolfspraulI'm so happy the Ben has the Android immunization, in the form of 32 MB memory.01:49
rjeffriesGoogle is very powerful. I like what they are doing. it keeps apple from becoming a monopoly01:49
wolfspraulit wasn't even on purpose, but man it worked!01:49
rjeffriesthere ya go.01:49
wolfspraulnobody, even the most insane people, suggest we should port Android to the Ben01:49
rjeffrieswell, must get toe bed. nice chatting!01:49
wolfspraulor have you read it?01:50
rjeffriesN8 is the phone I would buy were I to buy a smart phone (becaus eI love the camera01:50
rjeffriesbut I do not use a smart phone as I consider data plans to be too expensive01:51
kyakxiangfu: thanks for helping with git, it worked fine!02:21
xiangfukyak: Hi, I find another way to enable qipackages libiconv, four very simple commands:02:47
xiangfu./scripts/feeds uninstall libintl02:47
xiangfu./scripts/feeds uninstall libiconv02:47
xiangfu./scripts/feeds install -p qipackages libiconv02:47
xiangfu./scripts/feeds install -p qipackages libintl02:47
xiangfuafter 'make distclean' then make the whole openwrt, sometimes it's still using the openwrt stubs libiconv,02:48
xiangfuso I add those four command to build script file02:48
kyakxiangfu: good find! however, i think qipackages should still be on top of feeds.conf02:48
kyakthere is glib2, that's also overriden -\02:48
xiangfukyak: yes. it's still on the top.02:49
adamw_xiangfu, when I 'git push', I got 'fatal: The remote end hung up unexpectedly'. What possible reason will cause this?02:50
xiangfuadamw_: maybe you don't have write access, what is the output of 'git config -l | grep origin.url"02:51
xiangfuadamw_: the 'git@*** : ***.git' have write access, 'git:// *** / ***.git' only have read access.02:52
xiangfukyak: then I will add another command to my build scripts file :)02:52
adamw_does this mean that I need to git clone with 'git@***: ***.git'?02:54
xiangfuadamw_: or just change that  by:02:55
xiangfu1.git remote rm origin02:55
xiangfu2. git remote add origin git@projects.qi-hardware.com:m1.git02:56
qi-bot[commit] Adam Wang: Matching part number& datasheet of P1 & P2 for jtag/serial cable pod run1: http://qi-hw.com/p/m1/60f629802:59
adamw_xiangfu, tks.03:01
xiangfuadamw_  :)03:03
wolfspraulkyak: sorry to bug you. do you know whether this channel (#qi-hardware) indicates its charset preference somewhere?04:16
wolfspraulI am trying to cleanup the irclogs, and somewhere the utf-8 gets garbled04:16
wolfspraulit seems eggdrop writes the log files in 8859-1, but I don't see how I can tell eggdrop to write them in utf-8 instead.04:16
wolfspraulmaybe eggdrop decides it should log in 8859-1 because it thinks/sees that the charset on #qi-hardware is 8859, not utf-8?04:16
wolfspraulI'm wondering whether there is a flag somewhere to say which charset a channel has, and then how we can indicate that #qi-hardware should be utf-8...04:17
wolfspraulI keep googling :-)04:17
kristianpaulmaybe some animation series <- http://mimiandeunice.com/ http://www.archive.org/details/NinasAdventures - Like Comics?? :-)07:02
kristianpaulben delighfull <- ah?. is a working ogg player, wikireader* image viewver (yeah missing content on this part), calculator, note keeper. :-)07:04
kyakwolfspraul: hey. this channel does support utf-8; the problem is, eggdrop doesn't07:06
kyakto enable utf-8 in eggdrop, you need to follow these instructions http://eggwiki.org/Utf-807:06
kyakactually, it's not the channel that suports utf-8, but the network.. and freenode is utf-8 aware07:07
kristianpaullets see ññññññ07:07
kyakmostly all the symbols here are 8859-1, except for these - 3K3K3K07:08
kyakor this V07:08
kyakwolfspraul: after patching eggdrop, you need to start it in utf-8 locale, too07:10
kyakthough it's mentioned to use this patch with care, it solves all utf-8 problems for me :)07:11
wolfspraulkyak: ok thanks for pointing me in the right direction!07:18
wolfspraulI will probably not build eggdrop from source now, hopefully they will add this in a future release...07:19
wolfspraulfrom the 2 steps in the wiki page, #2 looks quite crude. If you look at the source that one shouldn't be necessary, right before is the code to check LC_ALL or a few others.07:21
wolfspraulso it must be step #1 then that makes the difference07:21
Jay7who said '3K3K3K' here? :)07:24
Jay7I'm sure he is russian ;)07:24
qi-bot[commit] Xiangfu Liu: nlove, offrss: fix typo and depends http://qi-hw.com/p/openwrt-packages/1fca17c07:40
qi-bot[commit] Xiangfu Liu: fix typo http://qi-hw.com/p/openwrt-packages/35f03d907:51
kyakJay7: indeed he is08:30
wpwrakgrmbl. cablemodem disconnected and took a LOT of talking to to come back again08:42
wpwraklet's see what i missed ...08:42
wpwrakwolfspraul: (heavy mail) i guess i hid the 8:10 card issue well ;-)08:45
wpwrakwolfspraul: (ya) i think a new device that's not too radically different from the old one would help to motivate and grow the community08:45
wpwrakwolfspraul: show that there's not only revolution but also a maturing process. build confidence that existing problems will be fixed, now just superseded by cool new ones08:46
wpwrakwolfspraul: i don't think the MM1 core is really ready for this yet. if i interpret what sebastien writes right, linux support is still in very bad shape, with issues not only in the kernel but also in the toolchain and probably lots of problems also in creating a distribution08:48
wpwrakwolfspraul: so i think nanonote and mm1 should run in parallel for a bit. that would also create an opportunity to resolve more of the openness issues that still plague the ben, in particular the case08:49
wpwrakwolfspraul: for the cpu, i don't know. even if the 4760 isn't available or needs an excessive amount of new development work, some of the older ingenic cpus don't look too horrible either08:50
wpwrakwolfspraul: for kernel work and such, i would actually hope that a migration to a different ingenic cpu should be relatively easy08:51
wpwrakwolfspraul: agreed on not inviting crazy amounts of work on what's ultimately a proprietary design, though08:52
wolfspraulyup. all agreed.08:53
wpwrakof course, there are certain obstacles to get a ya project rolling, like manpower and - i would assume - financial resources.08:54
qi-bot[commit] Xiangfu Liu: offrss, fix description and add -liconv http://qi-hw.com/p/openwrt-packages/404b3fd09:41
qwebirc25598I have a problem unbricking nanonote09:43
qwebirc25598and I do not want to solder the two pins09:43
wpwrakroh: btw, the usb atmel might actually be a choice for a future version of atusb. the transceiver can output its clock and the atmel would then have a very accurate external clock.12:47
rohah. nice13:00
rohwpwrak: i wonder how this works  in detail (henn/egg)16:50
rohmeaning.. what enables and configs the transciever to output the clock to enable the avr to work?16:51
wpwrakany kind of reset does :)16:54
rohi see. they planned it right16:55
wpwrakyeah, they put their brains to good use :) default clock frequency is only 1 MHz, but that's plenty to get things running16:56
qi-bot[commit] Werner Almesberger: atrf-txrc.c (usage, main): changed -P to -E to free -P for other use http://qi-hw.com/p/ben-wpan/fc226b219:09
qi-bot[commit] Werner Almesberger: atrf-txrx: new option -P to select ping mode http://qi-hw.com/p/ben-wpan/f16224f19:09
qi-bot[commit] Werner Almesberger: atrf-txrx: new option -r to select data rate http://qi-hw.com/p/ben-wpan/705864d19:09
rjeffriesgood day wolf19:42
zrafawpwrak: great atben/atusb stuff man!20:15
zrafawpwrak: it is really great for the nn future.. wow, really nice project (the whole qi projects family)20:17
wolfspraulzrafa: you on drugs?20:18
wolfspraulrjeffries: hi Ron20:18
zrafawpwrak: would be it too complicated to offer a bridge some day?.. For example, another tiny similar device (no atben, no atusb) which you just put somewhere to reach more distances?20:18
zrafawolfspraul: why?20:19
zrafawolfspraul: no me.. werner and you maybe20:19
kristianpaullol :-)20:21
zrafawolfspraul: bridge.. dont know the exact word .. repeater?20:22
zrafawpwrak:        ^20:22
zrafawolfspraul: sorry, it was for werner20:22
zrafawolfspraul: I gave some friend, as a gift, the nn Miriam and I bought in Spain.. That guy asked me to buy some wifi stuff.. and I just told him to wait for a few months.. something will appear on the qi world for that :)20:25
kristianpaulwpwrak: access point?20:27
kristianpaulyou mean atusd i guess?..20:27
wpwrakzrafa: hey, thanks ! i must say that i'm rather pleased with it myself :)20:29
wpwrakzrafa: (repeater) yeah, you could even make an atusd do that. of course, this also reduces your total bandwidth, because each packet then needs two transmissions20:30
wpwrakkristianpaul: atusd is now called atben :)20:31
kristianpaulatusb still same?20:32
wpwrakyup. didn't you read my mail announcing the changes ? :)20:32
wpwrak"the 8:10 card, [...]"20:32
wpwrakthe mail wolfgang calls too heavy to digest ;-)20:33
wpwrak(and i still have the one on the cam stuff sitting around ... some 120 lines by now :)20:33
kristianpauli still reading it too btw :-)20:33
wpwrakc'mon, it wasn't *that* bad :)20:34
kristianpaulno no, was my fault, i was some busy this week, i have others mails to read too20:35
kristianpaul did a quick read, now full mode  :-)20:36
kristianpaulahh was short mail actually ;-)20:36
kristianpaulSo i can address two atben in the same ben nanonote? (that goes for the FAQ)20:37
wpwraktwo atben ? with one slot ?20:38
kristianpaulwell some sort of mix wires and sharing buses may be?..20:39
kristianpaulI just was thinking on the "reapeater"20:39
wpwrakah .. well, you could plug two atusb into a pc20:40
kristianpaulfor example i'm aware of for RFMB12 you need at least two chips for fullduplex comunication, i'm not aware of this situatation with atben20:40
kristianpaulPC i see20:41
zrafawpwrak: I mean some independent device.. for example something powered externally, no from ben or pc, which you plug on the wall for example20:41
wpwrakwpan has fairly narrow channels, so you can have several devices usefully communicate in parallel.20:41
kristianpaul                                                                         yeah^^^20:41
wpwrakzrafa: take an atusb, teach the firmware how to be a repeater, then plug it into a usb charger. done :)20:42
zrafawpwrak: ah.. right :)20:42
kristianpaulhmm thats interesting setup20:42
kristianpaulthat respond my question too20:43
wpwrakzrafa: usb chargers should be a few cents - if you buy the cheap junk openmoko had ;-) (i don't think any of these actually made it to customers)20:43
wpwrakkristianpaul: wpan communication is half-duplex. multiple chips for multiple cannels would be an unusual configuration.20:43
kristianpaulwpwrak: whats the power consumption for atusb? i wonder if can feed it with solar light panels too20:44
kristianpaulwpwrak: ethernet is full duplex, why not wifi as well? :-)20:44
kristianpaulany way just saying...20:44
wpwraki haven't measured yet. i would expect < 50 mA peak. the current firmware is crazily inefficient, so average would be ~20-30 mA. but i can fix that with time.20:45
wpwrakfull duplex = you need two channels. usually not worth the trouble.20:45
wpwrakmost communication is mainly one-way anyway. well, bulk data in one way, a few acks in the other way.20:46
rohwpwrak: nice graphic ;)21:20
rohdoes that mean if you put the atusb into your bedroom it could work in the entire flat?21:20
wpwraki could probably reach the bathroom, yes. well, in the bedroom, i only have a laptop, so it may perform a little worse than the pc (size of ground - the antenna apparently likes to be close to a big chunk of metal)21:22
DocScrutinizer:nod: I seem to remember ground plane antenna design to be the most easy and efficient, from my time with citizen band23:15
DocScrutinizerbasically there's no such thing like an antenna uni-pole. Antenna always is a dipole, as some learned the hard way with GTA02 buzz issue23:16
DocScrutinizer-x- and lx-  are the two basic antenna designs, (plus of course closed dipole which is a special form of first geometry). groundplane being the second of those 2 geometries23:21
DocScrutinizerin real life it's usually -------x-  and lllx-23:21
DocScrutinizerso to make things simpler to calculate, you usually pretend the "gnd" half of dipole to be infinite size23:22
rohheh. i wonder if i could make antennas by lasering a '8' into acryllic plastic on one side and sheet the other one with some metal mesh and then 'lay a wire' into the lasered groove23:28
rohwould make reproducability (precision) high as well as cheap protoyping23:28
rohi just dont have any clue about the math needed to make it useable23:29
rohi wonder if one of these metallic foils 'rettungsdecke' does make a good enough conductor to be used as an antenna23:30
rohcould laminate that and make flexible antennas23:31
--- Fri Jan 21 201100:00

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