#qi-hardware IRC log for Saturday, 2011-05-28

wpwrakoh, so no broken cables ? that's actually good00:00
wpwrak(vga) yeeah ! ;-)00:00
wolfspraulah hold, maybe 1 broken cable, yes00:00
grunthushow small are the solder connections?00:00
wolfspraulit has a small punctured hole in it, very strange00:00
wolfspraulright in the middle of the hinge, no idea how that hold got there00:00
wolfspraulmaybe it was punctured already during assembly, and only failed later00:01
wolfspraulgrunthus: if you are up for it, you can definitely try to disassemble the ben and see whether reseating the cable fixes the issue00:01
wpwrakwolfspraul: so no cables with hairline cracks ? that's the one i'd dislike the most.00:01
wolfspraultuxbrain will take your device back and replace it with a new one even if you 'hacked' it a little :-)00:01
wolfspraulthat's assuming that you be nice to the Ben and really just try to fix the cable, not some wilder hacking, then your Ben is really yours :-)00:02
grunthuswolfspraul: I'll probably try to fix it then. Only not tonight, after a glass of wine!00:02
wolfspraulwpwrak: no I don't remember [hairline cracks]00:02
wolfspraulgrunthus: yes of course, let's do it! That's more fun anyway. If we fail, you get the Ben replaced anyhow.00:03
grunthusI see 3 screws, one in battery comp., two in front base.00:04
grunthusLooks like they will release keyboard surround from base?00:05
wpwrakgrunthus: there's a 4th screw, on the other side of the battery. may be covered by some sticker00:05
grunthusFound the fourth screw.00:08
grunthusMy miniature screwdriver set here is not miniature enough. Will need to get another one tomorrow.00:08
grunthusAre you likely to be around at all?00:09
wpwrakgrunthus: you may find these instructions useful: http://lists.en.qi-hardware.com/pipermail/discussion/2011-April/007766.html00:09
wpwrak#qi-hardware never sleeps ;-)00:09
grunthusOK, got the link and will work on it tomorrow.00:12
wolfspraulwpwrak: in which way did you think Linaro can help us?00:12
wpwrakwolfspraul: they may know sources of financing. they themselves are non-profit yet have a sizeable workforce. maybe they know people who want more than just ARM ?00:14
wpwrakwolfspraul: also, there seems to be ubuntu money in there. you never know where that may want to go. and ubuntu is certainly about long-term strategy, just as we are.00:15
wpwrakwolfspraul: their cto, davis rusling, will be at linuxcon in japan next week. may be a good time to ping him. see what he says. if he sounds interesting/interested, maybe you could meet in person. linuxcon: http://events.linuxfoundation.org/events/linuxcon-japan00:18
wolfspraulhe :-)00:24
wolfspraulok, got it00:24
wolfspraulmy understanding of arm is that it is basically initiated by ARM Ltd, and they are beating the drums to get their customers or other players in the 'ecosystem' to pay money to linaro for support00:25
wolfspraullike a pool00:25
wolfspraulthen Linary spends that money mostly on engineering for the greater good of the platform00:25
wolfspraulI hope they don't fork the kernel, as long as their work is directed towards kernel.org it should be a good thing00:25
wolfspraulI highly doubt they are interested in anything non-arm00:25
wolfsprauleven with arm, the fee to have your device/company supported is typically 3 million USD / year00:26
wpwrakthey seem to focus strongly on mainline00:26
wolfspraulbut nothing bad about extracting some money from big corps who would otherwise be ignorant or wouldn't even know how to feed money to 'upstream'00:26
wolfspraulfrom the people they hire on the engineering side that probably comes more or less automatically00:27
wpwrakthey may not be interested themselves in non-arm, but they may know someone who is00:27
wpwrakheh ;-)00:27
wpwrakso i would just present the qi-hw effort with its highlights (ben, linux support for it, mm1, rtems/partial linux, wpan), and explain our predicament. ask if they have useful contacts, who would be interested in long-term investment along these lines00:31
wpwrakif you consider the meager resources of qi-hw, the results are quite amazing. that may ring some bell.00:32
wolfspraulwith my understanding of linaro, that wouldn't make any sense00:33
wolfspraulthey are essentially an engineering pool setup to funnel resources into 'arm upstream' into kernel.org00:33
wpwraki wouldn't try to get linaro to get involved directly. but see if they can refer us to someone who can be useful00:33
wolfspraulbetween arm ltd. and their large customers, or anyone they can convince to pay x million USD / year into linaro00:33
wolfspraulthen linaro spends it on hiring individual kernel hackers00:33
wolfspraulARM is a highly proprietary chip company, they believe in their IP, for-money licenses to that IP, etc.00:34
wpwrakyes, so they've figured out how to solve part of the problem qi-hw has00:34
wolfsprauloh sure, they have very profitable proprietary ip that can fund engineering of open source drivers/kernel to sell more of that proprietary ip00:35
wolfspraulthat model is easy to understand, and Linaro is doing well, and a good thing for ARM00:35
wpwrakno, i mean linaro, not arm00:35
wolfspraulthat's the same thing. linaro is a pool of money ARM setup, and they get their large customers to pay into that pool00:36
wolfspraulthen the pool spends the money on high-quality upstream-capable kernel hackers00:36
wolfspraulsomething those companies paying into the pool may not be able to do themselves, or they still want to do that in addition to what they do in-house00:36
wolfspraularm has been quite successful in getting people to understand the importance of this and pay into linaro, so that's a good thing00:37
wolfspraulfor arm, of course00:37
wolfspraulwith Milkymist we are all out alone :-)00:38
wpwraki wouldn't try to get into linaro's model. but to see if we can milk them for contacts. they must talk to a lot of people. certainly not all of them are just 100% arm.00:38
wpwrakdon't forget the nanonote ! :)00:38
wolfspraulnobody supporting it on the compiler side, no customers who could pay into a pool, nothing. Just a few people :-)00:39
wpwraktypical for an immature project. mm1 needs probably 2 real product runs before it can become sustainable.00:40
wpwraknanonote needs one more, if it's played well00:40
wolfspraulfirst I was talking about Milkymist the SoC00:40
wpwrakboth are "cheap", as such things go00:40
wolfspraulLinaro is a model to support a proprietary chip better in kernel.org00:41
wpwraksure. you don't want their business.00:42
wpwrakbut they ought to have contacts that are of use for us. contacts they wouldn't mind giving away, if asked nicely.00:43
wpwrakthey seem to understand the principles of doing successful kernel work. you could connect there.00:44
wpwrakthe achievements of qi-hw are, considering the size, quite stellar. in the kernel area, mainly thanks to larsc.00:44
wpwrakthe success story of qi-hw so far is the ben. limited in power, limited in openness, but it works. has top-notch mainline support. no less than two major embedded distributions support it.00:45
wpwrakmm1 is the next thing. risky. exploring into areas no man has ventured before. but consistent with the philosophy.00:46
wpwraknow, the ben needs a successor. badly. we can do better. there has been incremental development. take ben-wpan. everything that happens there is 100% open. no proprietary tools or formats.00:47
wpwrakso, take the experience of ben-wpan, merge with nanonote, and add what we haven't dared to tackle yet, namely case design (and preparing for compliance/certification)00:48
wpwrakthis should a) produce a better product, b) help to keep the community alive and even grow, c) provide a proof of concept that the whole approach works.00:49
wpwraknow, the issue is finding someone who thinks this is worthwhile.00:50
wolfspraulI don't know about the third product yet, I need to really settle m1 first00:51
wpwraktraditional investors probably don't. but there ought to be folks who would, if it was shown to them in the right way. linaro may know such people.00:51
wolfspraulwhich is making good progress00:51
wolfspraulno way, I still didn't get my point about linaro across00:51
wolfspraullinaro is an engineering pool00:51
wolfspraulnon-profit doesn't mean much, they are very much using real money and can only spend as much as they make00:52
wpwrakyes yes. our biz model, as far we can call it than, isn't theirs. perfectly clear.00:52
wolfspraul'non-profit' from ARM's perspective, yes00:52
wolfspraulARM did a good thing with Linaro00:52
wolfspraulremember how so many people tried to explain all sorts of big corps about 'upstream'00:52
wolfspraulwith limited success00:52
wolfspraulLinaro solves that problem outside of those corps00:52
wolfspraulnow they can wire x million USD to somewhere, and the money will be used in a good way00:52
wolfsprauland ARM is pushing it, i.e. inviting their customers to pay into this pool00:53
wolfspraulso it's 'official' and those corps are not worried about wasting their money00:53
wolfspraulMIPS and others should do something similar, even Intel could do something like that00:53
wolfspraulI really like the clarity of the model00:53
wpwrakso, linaro understand how this works. this means that they will also understand why what we do is good. this means that we have basic credibility with them.00:53
wolfspraulLinaro's model00:53
wpwraknow, they talk to people doing hardware. probably lots of them. all sorts of ideas, directions, whatever.00:54
wolfspraulif Linaro is 'relatively' independent from ARM (which I doubt), then Linaro management would probably be open-minded about being approached from Intel about extending the company (Linaro) to cover x86 as well00:54
wpwrakwhat they could do for us is to filter their contacts for those who could have common goals with us. and pass them on.00:54
wolfspraulmaybe the type of engineering linaro offers has less value on x86 than on arm...00:54
wpwrakwe're obviously far out in crazy land to be a direct threat. so why shouldn't they help a distant cousin in arms ?00:55
wpwrak(not ARMs ;-)00:55
wpwraki think the key is to find some common ground. then try to extend it.00:57
wolfspraulI wouldn't know how to talk to anybody from Linaro, well maybe that's a reason to try, in your spirit :-)00:57
wpwrakin this case, the common ground would be the non-trivial understanding of how open source development works. and, in our case, extrapolating it. that makes what we do interesting.00:57
wolfspraulbut I think when you read about Linaro, you only see 'upstream' and then you think there is some overlap00:58
wpwrakyou could just say "we're qi-hw, our objectives are this, we've already done that, but now we need money. we're too stupid to find someone who can help us. do you know someone who could ?"00:59
wolfspraulLinaro wants to find more paying customers. If the Ben had an ARM chip, and we had a few million USD to spare and would generally struggle with upstream, Linaro would be glad to talk to us.00:59
wolfspraulnone of those 3 are true though :-)00:59
wolfspraulwe have no ARM, we have no money that would even cover 1 Linaro business trip, we are not struggling with upstream01:00
wpwrakyes, "upstream" rang a bell. as it should for them. remember, the key is to establish credibility. if the person you contact thinks you're good, you may get a result. if they don't know what gibberish you're talking about, they'll just ignore you.01:00
wpwrakno no. don't invade on their business. make it a favour, lent to a friendly neighborhood enterprise.01:01
wpwrakthe overlap serves to establish credibility. they "grok" upstreaming. so tell them that's what we do for breakfast. on a shoestring. that'll impress them.01:02
wpwrakwhile they're impressed, they (well, the person you're talking to) would be willing to help. not with money, because there's no business in it for linaro, but with contacts. contacts that are a better fit.01:04
wolfspraulI'm wondering about the relationship between ARM/RVDS and Linaro/gcc01:07
wolfspraulif Linaro had the green light to really improve gcc for arm, that would be great01:07
wolfsprauleven better if ARM would donate RVDS to Linaro, as a copy/paste source01:07
wolfspraulARM's proprietary (and expensive) compiler outperforms gcc so much it's not funny01:08
wolfspraul20-30% faster code _and_ 20-30% smaller code01:08
wpwrakmaybe they want to keep it that way ;-)01:08
wolfspraulI'm just reading about Linaro's work, and gcc caught my eye01:08
wolfspraulwell Linaro does work in gcc, that's why I was wondering01:08
wolfsprauland Ubuntu uses the Linaro gcc now01:09
wpwrakbesides, 20-30% squared is just about one year ;-)01:09
wpwrakyeah, there's some line between the two. that's also something i'd consider an interesting path: ubuntu/canonical is about strategic planning. for several years, they've been sinking lots of money without return whatsoever.01:11
wpwrakso qi-hw shouldn't look half crazy to them. that is, if they see a light at the _end_ of the tunnel.01:12
wpwrakagain, linaro could be the conduit to pass such information.01:14
wpwrakbut i don't think it's necessary to push in any specific direction. just gain respect, outline the problem in plain terms, ask if they know someone who can help. i would put that indirection, though. else, they may think you want money from them, like so many other beggars every day.01:23
wpwrak"they" would of course be the person you contact. keeping it narrow avoids the "someone will take care of it so i don't have to" reaction that will happen in 100% of the people who could take care of it.01:24
wpwrakdid i convince you ? :)01:29
qi-bot[commit] Werner Almesberger: prod/doc/atben.fig: cover unused PCB space with a white rectangle http://qi-hw.com/p/ben-wpan/caa8c1401:31
qi-bot[commit] Werner Almesberger: prod/doc/setup.hmac: described PC and Ben software and test script installation http://qi-hw.com/p/ben-wpan/6ac10dd01:31
qi-bot[commit] Werner Almesberger: prod/doc/setup.hmac: filled the intro sections; added dfu-util installation http://qi-hw.com/p/ben-wpan/717f7e701:31
qi-bot[commit] Werner Almesberger: prod/doc/: added scope screen shots to analysis; show crystal as alt GND for clk http://qi-hw.com/p/ben-wpan/c93f6d901:31
qi-bot[commit] Werner Almesberger: prod/doc/: renamed at???-clk*.png to scope-*, to prevent wildcard clash http://qi-hw.com/p/ben-wpan/d884ab501:31
qi-bot[commit] Werner Almesberger: prod/doc/test.hmac: added introductory text for test setups http://qi-hw.com/p/ben-wpan/b58314201:31
larscjust scared him away01:34
wpwrakit's late breakfast time in china. maybe the wolf just went hunting :)01:38
wolfspraulnah, it's 06-01 news day today02:03
wolfspraulhave to start cleaning this up a bit ;-)02:03
wolfspraulLars - you are our Linaro!02:03
wpwrakheh, getting an early start this month :)02:04
wolfspraulnot another embarassment like last month02:04
wolfspraulas usual it's a LOT, I will need to spend some time on it02:04
wolfspraulI think I also want to mention azonenberg's DIY wafer stuff, it's just too exciting02:04
wpwrak"we have something similar to your 100+ man/xxx MUSD company. we call him 'larsc'" ;-)02:04
wolfspraulI am 100% sure if Lars wanted to, he could accept a nicely paid position at Linaro.02:05
wpwrakyeah, cool stuff there, definitely :)02:05
wpwrakpssst !02:05
wolfspraulLinaro seems to be off to a good start, they have money, they seem to be able to convert the money into meaningful progress, all good.02:05
wolfspraulARM is successful in convincing their ecosystem partners to pay into the Linaro pool02:06
wolfspraulI have never seen an upstream support model implemented to successfully by any chip maker.02:06
kristianpaulwell, ubuntu for ARM will call much atention tought..02:06
wolfspraulmost are just singling out a few people to work on that 'upstream thing'02:07
wolfspraulbut inside the chip makers, enthousiasm for that is limited, essentially you just do dirty cleanup work that the best people don't want to do, no recognition from management, etc.02:07
wolfspraulI've seen that many times02:07
wpwrakthe only thing surprising about it is that google aren't well ahead of them. i guess that shows that they've gotten too fat already.02:07
wpwrakoh, it's a system with many local maxima. some choose to get caught in one of them :)02:08
kristianpaulwell, just matter of see what have to offer those xilinx fpga with arm core inside ;-)02:11
kristianpauli remenber larsc pointed they we're working on upstream it?..02:12
kristianpauli mean, i see this like a excuse, to talk then no more :-)02:12
kristianpaulokay now i need some ramdisk.04:12
rjeffrieswpwrak wolfspraul does not show interest into Nanonote follow-on. MM is his focus, period.05:31
kristianpaul"at this time" = once at the time..05:39
kristianpaulso at this time i'm going bed :-)05:39
wolfspraulphew, my vpn machine disappeared for a few hours :-)06:43
kyakdid you think for a second they are coming for you, too? :)06:44
LunaVoraxBallmer said about Windows piracy in China "If you can afford a PC, you could afford the software"09:18
LunaVoraxWell in that case if they can afford a PC, why now afford a Ben ? :)09:18
LunaVoraxOk ok, a bigger Ben with a screen, mouse, USBs ports, Ethernet and a real keyboard ;)09:19
LunaVoraxBut still is :D09:19
vladkorotnevA bigger Ben is a OpenWRT-based PC :P09:19
LunaVoraxI don't see the problem :P09:19
LunaVoraxThere's two computers I want to buy for a long time now09:21
LunaVoraxA Loongson computer09:21
LunaVoraxAnd the next Amiga X100009:21
LunaVoraxBut I'm not rich :'(09:21
kyak"If you can afford a PC, you could afford the software" - a typical thing to say for some rich fuck09:38
kyakwho has no idea about how people survive09:38
wpwrakhe's right, though. linux has nearly zero cost :)09:39
kyakif you can afford bread, you can afford butter. Same thing09:39
kyakwpwrak: i think he meant "you can afford buying"09:39
wpwrakkyak: that linux cd can't be all that expensive either ;-)09:40
kyakwpwrak: sorry, might have missed your irony :)09:40
kyakthis words made me angry09:40
wpwrakwell, what else can he say ? "be copyous and prosper" ? ;-)09:41
kyakhe might say that at his deathbad :)09:43
wpwrakask anyone in traditional textile manufacturing (i.e., the manual one) what they think of those fancy new steam-powered machines and all the hype about that industrial revolution fad :-)09:44
wpwraki'm sure you can find very educated writings from that epoch that clearly explain why such devil's work must fail09:46
wpwraka library search among luddite literature might be fun. you could probably piece together a coherent statement for strong intellectual property rights and against free software only using quotes from 18th century texts09:48
wpwrakwell, or 19th century. back in the 18th, the machines weren't perceived as much of a threat yet09:50
kyakwpwrak: this could mean that there is a chance for humanity on 22nd century :)10:06
viricwpwrak: there is a nice film about that...10:07
viricI should recall it10:07
wpwrakkyak: just imagine the educational and human capital that could be freed if society abolished patents and similar nonsense :)10:08
viricfrom peter sellers...10:09
wpwrakpeter sellers on open source ? that sounds like fun :)10:10
viricmaybe I remember bad... but I remember as if in the film, workers strike against the addition of machines in the factories10:11
whitequarkwpwrak: don't you think that education done by patent trolls won't be even more harmful? :)10:12
kyakwpwrak: yeah, and how much smaller the amount of millionaires would be? :)10:12
virickyak: no matter what you do, the amount of millionaries decreases, those resting getting far more rich10:14
kyakso true, they are becoming billionaries :)10:15
viricthat film from Sellers got forbidden in the Spanish dictatorship, iirc, because it showed too much strength by the unions10:17
wpwrakhmm, i'm kinda anti-union myself. too much potential for abuse. i guess i'm an anarchist at the bottom of my heart. so no to kleptocrats of whatever couleur :)10:29
qi-bot[commit] kyak: luit: Locale and ISO 2022 support for Unicode terminals http://qi-hw.com/p/openwrt-packages/7f2174a10:38
qi-bot[commit] kyak: luit: Locale and ISO 2022 support for Unicode terminals http://qi-hw.com/p/openwrt-packages/f370d5c10:43
rohwpwrak: yes, qcad13:06
vladkorotnevhello everyone13:12
vladkorotnevif /tmp is a ramdisk, does it use swap partitions as well?13:12
viricvladkorotnev: I always forget about it13:14
vladkorotnevviric: it seems that a prog I am trying to compile uses too much space in /tmp, so it crashes with a 'No space left on device' error.13:15
viric'mount' will tell you the size of that filesystem. Or 'df' too13:15
viricyou can use remount to change it13:15
viricI don't know if it can use swap or not13:15
vladkorotnevand also, is there any way to make the program think that /data/folder is / ?13:15
viricvladkorotnev: chroot13:16
vladkorotnevit is shit and doesn't use makefiles13:16
vladkorotnevviric: won't chroot affect the possibility to run gcc and such?13:16
viricwell, you will need gcc in that /13:17
vladkorotnevviric: will symlinks work as well?13:18
vladkorotnevviric: how do I have to compile it if I don't have that much space on my / ?13:20
kristianpaulcross compile ? :_)13:20
viricWho wants to put things on /? The compiler?13:20
viricif you have the source, you can make it do whatever you want13:20
vladkorotnevviric: the part of the shell script that is installing the binaries13:20
viricvladkorotnev: change that13:21
vladkorotnevkristianpaul: to cross compile I need to build the image, which takes 10hrs13:21
kristianpaulyou can build the minimal image, not more than 3 hrs i think13:21
vladkorotnevkristianpaul: but it doesn't use makefile D:13:22
viricyou will cross-build instead of changing the install script?13:22
vladkorotnevviric: I think I'll have to13:23
vladkorotnevkristianpaul: also I'm on a mac so I don't have the stuff to cross-compile13:24
viricvladkorotnev: well, I can't fight against religion :)13:24
Action: kristianpaul :D13:25
kristianpaulyou can try a virtual machine, i bet there are several methods13:25
viricI'd rent a remote linux system13:25
vladkorotnevkristianpaul: I would, but 3 gigs free on my HD make me cry :(13:25
vladkorotnevbtw, is there nurses library for the NN?13:26
kristianpaulwell.. you need to do  sacrifices from time to time13:26
vladkorotnevwell, it compiles...13:40
qi-botThe build was successfull, see images here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.trunk-05272011-1129/14:27
wpwrakqi-bot should announce such things with "*** FRESH MEAT !!! ***" ;-)14:29
wpwrakviric: (rent) i'd get a netbook and end the suffering right there and then ;-)14:31
wpwrakto paraphrase balmer, if you can buy a mac, you can afford a netbook :-)14:32
vladkorotnevIs it safe to rm -rf /tmp/* ?14:32
wpwrakusually not14:32
wpwraksome programs may keep files there they intend to open again. so you should only do this when the system is completely quiescent14:33
vladkorotnevI cleared some files from a failed build14:34
wpwrakalso, X has files named /tmp/.X*14:34
wpwrak(and gets very unhappy if these should vanish)14:34
vladkorotnevNow I need SDL/SDL.h and so on :P14:34
viricwpwrak: I was joking. :)14:34
vladkorotnevany SDL headers for the NN?14:35
wpwrakvladkorotnev: at least the Jlime SDK lets you install -dev packages into the SDK, with opkg-target. this may also work with openwrt14:37
vladkorotnevwpwrak: I compile directly on the NN :P14:38
wpwrakoh :) well, then just install the -dev package ;-)14:38
vladkorotnevwpwrak: got it, it wasn't in the 24-05-2011 pkg list :P14:39
qi-bot[commit] Werner Almesberger: prod/doc/setup.hmac: explicitly mention SDL_gfx as a prerequisite http://qi-hw.com/p/ben-wpan/2bb2e7b16:23
qi-bot[commit] Werner Almesberger: prod/: moved more configurable items to variables http://qi-hw.com/p/ben-wpan/0372d7d16:23
qi-bot[commit] Werner Almesberger: prod/doc/setup.hmac: completed test script customization section http://qi-hw.com/p/ben-wpan/c78423b16:23
qi-bot[commit] Werner Almesberger: tools/atrf-path: added interactive dumping with D http://qi-hw.com/p/ben-wpan/b82472b16:23
qi-bot[commit] Werner Almesberger: atrf-path: added visual history of dumps http://qi-hw.com/p/ben-wpan/0eec74716:23
qi-bot[commit] Werner Almesberger: prod/Makefile: target "spectrum" to generate reference spectrum; target spotless http://qi-hw.com/p/ben-wpan/680086b16:23
qi-bot[commit] Werner Almesberger: prod/doc/: completed description of signal strength measurement http://qi-hw.com/p/ben-wpan/33f84b416:23
viricdear electronic hackers17:38
viricI've a solar charger here...17:38
viricfrom Solio.17:38
viricfor mobile phones and so.17:38
viricIt has four output pins.17:39
viricVcc and GND for the mobile phone to charge, and between the other two, a resistor.17:39
viricdifferent for every kind of mobile phone connector17:39
viricWhat do you suspect from that?17:39
viricIt did not look as the resistor determining a tension supply18:02
viricmaybe it's for a current supply18:02
wpwrakor maybe it adjusts the output voltage ?18:05
viricwith a votlmeter, it does not look like so18:07
wpwrakno ... the resistor could be for setting a regulator in the supply18:08
viricthat may not work at zero current?18:12
viricI see a PIC on the board...18:13
viricand I'm looking at other components18:13
viricwhat a bad sight, this mine!18:27
viricVA7205CH (a 8 pin chip)18:27
viricE2FE (a 3 pin chip, one of the three to a heatsink part of the circuit)18:28
viricGS7H (a 3 pin chip)18:29
viricGSY7H I mean18:29
viricor 65Y7H  (with a _ over the 7)18:30
viricva7205 is for charging a battery the device has.18:43
whitequarkviric: I have an idea of what it may be19:13
whitequarkminiUSB has an ID pin which should be connected to GND directly or via a resistor19:14
whitequarkand some phones, e.g. openmoko, measure the resistance and determine if it is the charger (in which case it will draw all the current) or just PC USB (then it'll get 100/500mA)19:14
whitequarkmaybe here PIC measures the resistance too, detects charger type and sets some settings accordingly19:15
lunavorax_frizzlwhitequark, just like the iPhone :)19:29
lunavorax_frizzlWait, the iPhone is closed stuff.19:29
lunavorax_frizzlHey, I'm wondering something.19:29
lunavorax_frizzlWhy does the some laptops HAVE to be filled ONLY with chips that has NO open drivers.19:30
lunavorax_frizzlLike my MacBook Pro19:30
lunavorax_frizzlWhy does the chip is a Broadcomm one19:30
lunavorax_frizzlAnd not an Intel one19:30
lunavorax_frizzlIs it so had to fill in the case or what ?19:30
whitequarka funny thing happened when I replaced Broadcom WiFi with Atheros chip19:31
whitequarkin my notebook19:31
lunavorax_frizzlAhteros is... ?19:31
whitequarkbefore the replace, it had worked ~3 hours from battery (even with wifi turned off with killswitch)19:31
lunavorax_frizzlclosed or open ?19:31
whitequarkand after, it is working more than five19:31
whitequarkatheros is a nice vendor of RF parts who opens drivers and even firmware19:32
whitequarkmoreover, the new card finds much more networks than the old one19:32
whitequark(it'd be very interesting if anyone familiar with RF stuff here would explain the behavior to me)19:33
lunavorax_frizzlwhitequark, I dreamed very often of starting my own computer company made exclusively of open hardware.19:34
whitequarkwe all do19:35
lunavorax_frizzlSo far the loongson company is doing so, but their stuffs are "cheap".19:35
lunavorax_frizzlAnd, AND to CARE about the battery life.19:35
lunavorax_frizzlIt looks like only Apple do care about the battery life of their laptops (sad fact)19:35
lunavorax_frizzlThe other consturtors are saying "Hey it's a 3/6/9 cell battery"19:36
whitequarkapple stuff definitely has great hardware19:36
lunavorax_frizzlWithout specifying19:36
lunavorax_frizzlOr worse "Hey it last for 10hours" and bang you discover that you have to buy the oversized battery to achieve so.19:36
lunavorax_frizzlwhitequark, I agree 500%19:36
whitequarkwhen they'll do their netbook, I'll buy it, whatever the cost will be19:37
whitequarkeven through it is closed19:37
lunavorax_frizzlwhitequark, that's my dream, Apple quality, but open.19:37
whitequarkthat won't happen in near future. you should have all the experience Apple has19:37
lunavorax_frizzlwhitequark, they already do netbooks, the MacBook Air 11' is one.19:37
lunavorax_frizzlwhitequark, I know, but manufacturer should also stop to make shitty stuff.19:38
lunavorax_frizzlwhitequark, like Compaq seriously stop to f**k around with their computers19:38
lunavorax_frizzlseriously need to stop*19:38
lunavorax_frizzlIronically, HP is doing good things imho19:39
whitequarklunavorax_frizzl: the 11" air is somewhat bigger than my current 10" netbook, and that's critical for me19:39
lunavorax_frizzl(ironically cause HP owns Compaq)19:39
whitequarklet me find my caliper19:39
whitequark(hp) i've had very, very bad experience with their tablets, especially tx100019:39
lunavorax_frizzlOh ok whitequark but I don't know if Apple is going to go smaller.19:39
lunavorax_frizzlwhitequark, I don't do tablets19:40
lunavorax_frizzlwhitequark, I consider them other things19:40
lunavorax_frizzlAsus is making good netbooks so far. But seriously, there's too many products. Why cant' they focus on 3max or 5 ?19:41
whitequarklunavorax_frizzl: I won't agree on that, eee series is cheap crap19:42
lunavorax_frizzlAnd we're back with the Apple comparaison with very few products (you can customise).19:42
lunavorax_frizzlYes well I suppose the using experience is relative.19:42
lunavorax_frizzlwhitequark, when i compare my Dell Latitude 2110 with the Asus 1015PEM of my girlfriend, the Asus is definitely smaller, lighter and maybe even faster.19:44
lunavorax_frizzlwhitequark, and it doesn't feature the annoying fan technology Dell is using19:45
lunavorax_frizzlwhitequark, the mad fan that goes crazy when  it have a little too much calculus.19:45
whitequarklunavorax_frizzl: the fan on my samsung is not heard at all when it's idle and is very, very quiet on 100% cpu load19:46
whitequarkand the bottom is moderately hot (not like some other notebooks when you touch it and get some skin damage)19:46
whitequarkI'd say that samsung N250 (IIRC) is like 80% good. anyway, that is impressive for its cost19:47
lunavorax_frizzlwhitequark, isn't Samsung not very Linux freindly ? (or open in general ?)19:47
lunavorax_frizzlwhitequark, oh yeah that's the model I wanted to buy fo a time. I should have bought it.19:48
lunavorax_frizzlwhitequark, but I seriously can't handle 1024 x 600 it's definitely the worst resolution ever.19:48
whitequarklunavorax_frizzl: well, I'm not a fanatic. their hardware works great, and does not die after half a year (my previous netbook, by Toshiba, did that. and it was very crappy anyway)19:48
whitequarklunavorax_frizzl: they have a new model with 1366x768, but there's DDR319:49
whitequarkDDR3 cuts the battery time by around 2x19:49
viricwhitequark: I thought that the pic is measuring the resistence19:49
lunavorax_frizzlOh ok19:49
lunavorax_frizzlwhitequark, yeah Toshiba sucks19:49
viricBut I'm now thinking of the resistor being part of an analog circuit involved in the charging19:49
viricI've to check the exact pinout19:50
lunavorax_frizzlwhitequark, anyway, the Intel Atom chips are crap. I would give a try to the ULV ones if I could gets my hands on.19:50
viricdoes anyone know anything about supercapacitors?19:51
viricit's something I've been waiting for a long time, and I don't know if I can start making use of them19:51
lunavorax_frizzlwhitequark, so far I've been satisfied with the use of Atoms only with a netbook recycled as a web server.19:51
whitequarklunavorax_frizzl: but atom netbooks may work more than 4 hours19:52
whitequarkand I never seen regular (Core, etc.) notebooks work more than 2 hours19:52
whitequarkthat's another critical point for me19:52
lunavorax_frizzlwhitequark, I think I still would prefer a processor that works well rather than a processor that last long19:53
whitequarkand I prefer the former -- that's why I bought a netbook19:53
whitequarkit's quite frustrating (I'm going mad trying to compile gcc on it), but the autonomous time is great, and that's good19:54
whitequarkand the intel graphics card, which they should not dare to call a graphics card at all...19:55
whitequark(they should rather call it "3D decelerator")19:56
lunavorax_frizzlwhitequark, indeed BUT it's good to have cards that have real open drivers19:56
whitequarklunavorax_frizzl: yes. and these open drivers sometimes mess up all the console (that happens frequently), and sometimes crash X19:57
lunavorax_frizzlwhitequark, bah ;)19:57
lunavorax_frizzlwhitequark, and yeah I know about X crashing19:58
whitequarkat some moments, I lose belief in open-source...19:58
whitequark(now someone should ban me on this channel, right? :)20:00
lunavorax_frizzlwhitequark, I know that, but it's better to have a crashing X server rather than using Windows or Apple20:00
Action: lunavorax_frizzl remember the time when he was an Apple fanboy... oh god20:01
lunavorax_frizzlOh well, at least they have an open kernel :P20:01
whitequarkyeah. and no one can ever compile that into something executable20:03
Action: whitequark just attached a CPU fan using a duct tape and feels something wrong20:04
DocScrutinizerwhitequark: don't mess with this, I had to ban somebody today according to his own explicit decision20:06
DocScrutinizer(not this cahn though)20:07
DocScrutinizera rather sad case...20:08
whitequarkDocScrutinizer: oh. why has that happened?20:09
DocScrutinizerhe said he feels like he can't stop trolling and I shall ban him to make it easier to stay away from that channel20:10
DocScrutinizerI think he got a lethal brain injury of some kind, stroke or cancer or dunno what20:11
DocScrutinizerbeen a member of #maemo since years20:12
lunavorax_frizzlWhat ?20:46
lunavorax_frizzlTrolling is ok if you have constructive arguments and submit ideas to criticism.20:46
lunavorax_frizzlWait, that's not the definition of trolling, right ?20:46
lunavorax_frizzlAnyway, for my part I believe that open-source is never gonna make it in the user world20:47
lunavorax_frizzlAnd by open-source I mean 100% open-source20:47
lunavorax_frizzlNot like "oops, I did it again, you must install the proprietary driver"20:48
lunavorax_frizzl'cause some way, idk, things have to be "cool" for the lambda people20:48
lunavorax_frizzlHave an iBigbrother is so hype, see ?20:48
lunavorax_frizzlBut having a phone that repsect your freedoms.20:49
lunavorax_frizzlNah, that a completely different idea.20:49
lunavorax_frizzlAnd understanding that it's not a phone anymore but just a regular computer with a battery.20:49
lunavorax_frizzlAch, another story again.20:49
lunavorax_frizzlWith facebook + twitter + "smart"phones20:50
lunavorax_frizzlIt feels like 1984 made it.20:50
lunavorax_frizzlIn some way, the soft way, 'cause it's fun and no one cares.20:50
whitequarkcan anyone here explain me how SDRAM works? looks like I confused a lot of people with stupid ideas :/20:52
whitequarkthe problem is:20:52
whitequarkin my device, there is just one CS. Then, I have 15 address lines A0-A14 on CPU, of which A0-A12 go to A0-A12 lines of SDRAM chip, and A13-A14 are wired to BA0-BA120:54
whitequarkthe chip is a 32MByte 16-bit SDRAM20:54
whitequarkis there a way to add another 32 MBytes of memory by either replacing it with a 64 MByte one (are 14 address lines enough?), or adding a second 32M chip on top (I thought that absence of second CS makes this impossible)20:55
grunthusHi! I've just started opening my Ben22:40
grunthusHinge area near speaker resisting.22:43
grunthusGot it. Just needed some patience and gentle coaxing.22:44
grunthusThe black connector for the LCD was 1mm loose on left side.22:51
grunthusSlid back in to place easily. 22:52
grunthusPower button cycles fine (watch lsusb), but still no LCD :(22:53
grunthusTouching LCD connector with pencil eraser - LCD activates!!22:55
wpwrakgrunthus: black connector ... hmm, what's that ?22:56
grunthusThere is a black plastic strip between the connector on the PCB and the end of the contacts in the LCD cable22:57
wpwrakgrunthus: do you mean the brown bar on the beige connector ?22:57
grunthusyes, looks black to me, but yep.22:57
wpwrakokay, black-ish enough :)22:58
grunthusLCD is on now, but if I place the keyboard surround back, it either cuts out or becomes horribly interlaced in appearance.22:58
wpwrakso it was loose. that's good bad news. did you reseat the cable as well ? or just lock the connector ?22:59
grunthusNow has progressed to white display with vertical gray blocks22:59
wpwrakyeah, that's still bad contacts22:59
grunthusI didn't remove and reseat, just eased the bar back down23:00
grunthusI'll power off, remove and reseat.23:00
wpwrakthe procedure would be as follows: power down completely. gently pull the bar away from the connector. (moves by about 1 mm; make sure the movement is parallel)23:00
wpwrakto reseat, gently push the cable with a fingernail. don't push down hard on it or force it. it should slide in effortlessly.23:01
wpwrakif it's pulled out, it can get caught on the bottom edge of the connector and then won't move in. may need a little prodding.23:01
grunthusThe brown bar is now pulled back by 1mm23:02
wpwrakonce it's in completely, push the black bar back such that it's flush with the beige connector23:02
grunthusA question: having pushed black bar out 1mm, does that constitute removal of cable?23:03
wpwraknot, it just releases the cable. the cable doesn't move in the XY plane of you change the lock.23:04
wpwrakit may move a tiny bit in the Z direction, as the force pushing on it is removed/added23:04
wpwrakwell, pushing on the contacts23:04
wpwrakwhat happens is that the black bar changes the pressure on the contacts vertically pushing on the cable. without pressure, the cable can slide in and out freely.23:05
grunthusOk, I am getting that sort of behaviour. Do I need to remove the cable? Presumably not?23:05
wpwrakwith pressure, it should stay in place.23:05
wpwrakyou shouldn't need to remove it. just make sure it goes in completely. it's probably only half-inserted or maybe even angled.23:06
wpwrakif you remove it, it's a little harder to insert than it is to push in an already partially inserted cable.23:06
wpwrakhmm, i wonder if we have a picture somewhere of what would be "right"23:06
grunthusIt is already parallel. I photographed it.23:07
wpwrakhere's a good indicator: the white end of the "branch" that goes outside the connector should hit the capacitor. (or be extremely close)23:09
grunthusAh, bit blurry.23:11
wpwrakyeah. in any case, the white tab should hit the capacitor. then you're good and can lock the connector.23:12
grunthusLocked, powering up.23:14
wpwrakbetter ?23:17
grunthusI'm sorry to say not. Power is on (lsusb) but no LCD23:18
wpwrakhmm. maybe try this: power down, then unlock the connector again. then gently lift the tab hitting the capacitor. then see if you can push the cable into the connector some more.23:21
wpwrak(again, almost no force should be needed)23:21
wpwrak(lift = lift so that it can go above the capacitor, if the cable wants to go deeper)23:21
wpwrakif that still doesn't help, then the next step would be to power down, unlock, pull out the cable completely, if you have "compressed air" (isn't not really air, but that's what these cans are called) give the inside of the connector a blow, then put in the cable again, lock, and see if the stubborn thing finally cooperates.23:23
grunthusi'll try removal23:26
grunthusOK, eased connector out.23:29
wpwrakyou got some "compressed air" around ? (the canned stuff used for blowing dust out of computers and such)23:37
grunthussadly, not here23:38
wpwrakwell, then let's hope for the plain magic or removal and insertion to be enough23:39
kristianpaulit should23:41
grunthusSadly again, no joy!23:42
kristianpaulhum :(23:42
grunthusGiven the extreme sensitivity the cable had to any pressure, I think it's a break in cable23:43
grunthusNow, the LCD doesn't power on with a touch to the cable.23:44
wpwrakyeah, sounds a like a return case :-(23:44
kristianpaul i have a ben board lus lcm at home and the most suffering part is that cable, why this things happens..23:44
grunthusI think I'll ask for return now. Thank you everyone.23:45
kristianpauli hope all go okay for you and back for the ubb-vga hacking :-)23:45
kristianpaulwich remenber me i should had solder something..23:46
wpwrakwolfgang mentioned that the connector can sometimes also be badly soldered, while the cable itself rarely breaks. (which is different from the experience we had at openmoko, where a somewhat similar cable had no end of troubles with hairline fractures)23:46
wpwrakkristianpaul: (solder) yes ! :-)23:46
grunthuswpwrak: It felt like a hairline break in cable. It was sooo sensitive to touch.23:47
wpwrakhairline cracks are the nightmare of all QA. there's just no way to spot them :-(23:49
kristianpaul[not related to qi] but that fractured cable is so often for freerunner? i have a friend asking me to check his neo* because the screen just went off..23:49
kristianpauleven with X ray? :-D23:50
grunthusAll snapped back together nicely.23:50
wpwrakgrunthus: you said the connector also wasn't properly locked when you opened the device ?23:50
grunthusThat is correct, the left side of black lock bar was approximately 0.5-1.0 mm out from white connector.23:51
wpwrakkristianpaul: the cable we had tons of trouble with was the debug board cable. i don't know of issues with the lcd cable. i never experienced any myself.23:51
wpwrakkristianpaul: (x-ray) hmm ... :)23:51
grunthusThe cable itself was parallel to the connector, and the white tab butted against the cap.23:52
kristianpaulfrom that pic btw the "black thing" that hold the cable looks loose23:53
wpwrakgrunthus: interesting correlation. i wonder if it points to a common cause. e.g., operator accidentally pulling the cable or such, making the connector partially unlock and cracking a trace in the process.23:53
kristianpaulare you sure you secured it?23:53
grunthuskristianpaul: definitely, 3 times (moved flush, then released and relocked, then released, removed cable, reseated and relocked, all parallel.23:55
kristianpaulany case, be carefull i broke that thing when hacking my ben.. so the damn tape loose once at month..23:55
grunthusMine is new <1 week, problems from start.23:56
kristianpaulwell, time to call tuxbrain red phone ;)23:56
kristianpaulyeah factory issue, part of that 1~..23:56
grunthusYep, support has been good, here and via email tuxbrain23:56
grunthusThank you wpwrak in particular for your time on this.23:57
wpwrakgrunthus: no problem. sorry that it didn't work in the end :(23:58
grunthusI enjoyed taking it apart, interesting. I'm working with Arduino stuff too, so interesting to see about LCD connectors23:59
--- Sun May 29 201100:00

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