#qi-hardware IRC log for Sunday, 2010-09-26

ebausxiangfu: Just follow all the instructions to install debian on a SD CARD but no luck because I can not even turn on my nanonote00:35
ebausIs there any way to reset the device ?00:36
kyakxiangfu: hi01:34
kyaki already mentioned once, there's an empty dir bin/xburst/uboot-xburst-qi_lb60/ after openwrt build01:35
kyakis there some use for it?01:36
xiangfukyak: Hi. I just back to my compile. I will try to look into today.  it's useless.03:04
kyakxiangfu: thnkas!03:11
xiangfukyak: sorry for my delay work. I just found which line create this folder. now testing :)03:12
kyakxiangfu: no problem!03:13
kyakMplayer runs flawlessly with -nosound03:21
kyakseems that built-in mp3lib is causing problems03:21
xiangfukyak: which system ? debian or openwrt ?03:22
kyakxiangfu: do you know if we can port lame or some other mp3 library?03:22
kyaki mean, wouldn't it cause patent problems?03:22
kyakxiangfu: openwrt03:22
kyaksee openwrt-packages, i commited it yesterday03:22
kyaki do it like this:03:24
kyak1. prepare your avi files for Ben:03:24
kyakfor i in $(echo "*.avi"); do /usr/bin/mencoder -o "$i.mpg" -vf scale=320:-2 -of mpeg -ovc lavc -lavcopts vcodec=mpeg1video:vpass=1:vbitrate=200 -ofps 24000/1001 -oac lavc -lavcopts acodec=mp2:abitrate=64 "$i" && /usr/bin/mencoder -o "$i.mpg" -vf scale=320:-2 -of mpeg -ovc lavc -lavcopts vcodec=mpeg1video:vpass=2:vbitrate=200 -ofps 24000/1001 -oac lavc -lavcopts acodec=mp2:abitrate=64 "$i"; done03:24
kyak(thanks tuxbrain)03:24
kyak2. play your files with Ben:03:24
xiangfukyak: we can add it to openwrt-package.git. but we can not ship the image that include mip3 library with NanoNote03:24
kyakmplayer -vo sdl -vfm ffmpeg -nosound file.mpg (it only consumes 50 % CPU)03:24
kyakxiangfu: i see, but we can still offer a convenient way for users to do opkg install lame?03:25
kyakor they would have to build from source themselves?03:25
kyakanyway, need to port lame or mpg123.. or encode sound to something else03:26
xiangfukyak: I think opkg install is ok.03:27
xiangfukyak: agree.03:27
kyakxiangfu: btw, you should try fbterm+w3m -> should be able to display Chinese :)03:30
xiangfukyak: oh.. I only tried fbterm. yes. I will try it later . thanks :)03:31
xiangfukyak: it should be jfbterm.03:40
kyakboth work well03:41
qi-bot[commit] Xiangfu Liu: remove empty folder, we copy uboot to bin/xburst http://qi-hw.com/p/openwrt-xburst/a08dad403:53
qi-bot[commit] Xiangfu Liu: remove empty folder, we copy uboot to bin/xburst http://qi-hw.com/p/openwrt-xburst/7f4d50903:55
xiangfuupdate the commit message. :)03:56
xiangfukyak: fixed ^ :)04:01
kyakxiangfu: thanks :)04:02
kyakhm, i'm concerned.. playing ogg file with mplayer and libvorbis is still very slow04:02
kyakin comparison to gmu, which is using the same libvorbis04:02
kyaki thought there would be problems with video decoding, but it turned out that audio decoding is a bottleneck04:03
qi-bot[commit] Xiangfu Liu: mtd-utils goto upstream, don't need ubi-utils anymore http://qi-hw.com/p/openwrt-packages/398002e04:11
kyakxiangfu: in fact, lame is already ported in openwrt04:26
kyaknow to enable it in mplayer04:27
xiangfukyak: w3m compile fine after "config.full_system" build. but there is some error in a "config.minimal" build system.04:31
kyakit might lack for some dependency04:34
xiangfukyak: I will test the new packages in opewnrt-xburst.git . then try to add them to config.full_system tomorrow.04:34
kyakwhich is pulled by some other package from config.full_system04:34
kyakxiangfu: ok!04:35
kyakxiangfu: what was the error in config.minimal build?04:36
xiangfuyes. the dependency is always hard to find out.04:36
kyakneed to go now, see you later04:38
xiangfukyak: see you.04:39
xiangfukyak: the error mesage: http://pastebin.com/ZhBfURbA. you can take a look when you back.04:39
wejpkyak, Gmu is not using libvorbis for ogg vorbis decoding05:18
wejpif mplayer uses libvorbis it has to be very slow on the nanonote05:18
kyakwejp: what does gmu use then?07:08
wejplibvorbis is useless on devices without a floating point unit07:09
kyakwejp: i see.. mplayer can also use tremor (by default)07:09
wejpyes it can07:09
kyakand it is terribly slow07:09
wejpyou have to enable that when compiling mplayer07:09
kyakwejp: why the depenedecy for libvorbisidec in gmu?07:10
kyakwejp: i did07:10
wejplibvorbisidec is tremor07:10
kyakmplayer has some built-in tremor07:10
kyakit is not the same as libvorbisidec then?07:11
wejpif it is built-in it is not the same as libvorbisidec, but is most likely derived from it, when they say it is the tremor decoder07:11
kyakok, will have to have a look07:12
wejpalso you should explicitly disable the normal libvorbis in mplayer07:13
kyakit doesn't have "normal" libvorbis07:13
wejpyes it does07:14
kyakit has the tremor07:14
kyakhm explain please07:14
wejpit can also use floating point vorbios decoding07:14
wejpmplayer usually does NOT use tremor for vorbis decoding07:14
wejpyou can enable that, but by default it doesn't07:14
kyak --disable-tremor-internal disable internal Tremor [enabled]07:15
kyakenabled by default07:15
wejpbut it is still not used by default07:15
wejpthat's why i said you should disable libvorbis support07:15
wejpit is useless anyway, so there is no reason to have it in mplayer on the nanonote07:15
kyakok this is clear07:16
wejpsame is true for other decoders like mp3 (if you are going to build that)07:16
kyakhow to link against libvorbisidec?07:16
kyakhow to make mplayer use it?07:16
wejpthe internal tremor should be very similar to the external libvorbisidec. i would be really surprised if it wasn't07:17
kyakwejp: and another question, is there an mp3 decoder that doesn't need a fpu and therefore will work on Ben?07:17
wejpyes, of course there is. mpg123 can be used in integer mode. also libmad does not use floating point at all (but is still slower than mpg123 and has a horrible api)07:18
kyakwejp: i'll try again with that internal tremor.. when it was enabled, mplayer played ogg files with "ffvorbis"07:18
kyakwejp: do you know if lame can be used in integer mode?07:19
wejpffvorbis doesn't sound right, but i'm note sure07:19
wejpno idea for lame, probably not07:19
kyakok, thanks for th ehints07:20
wejpyou're welcome07:20
kyakmpg123 would be useful to port anyway, also for use with gmu07:23
wejpyou can easily use it with gmu, but keep in mind that things with mp3 support will never be supported by the default nanonote firmware due to patent issues07:24
wejpwell at least as long as those patents are still valid07:25
kyaki understand07:25
wejpthe last of the mp3 patents are going to expire in the end of 2012 in most countries, so maybe then it can be included ;)07:25
kyaki;m not aiming for default firmware, but to make it available for those who wish it07:26
kyakthey say the world is going to end in 2012, so who cares about m3 :)07:27
bartbesI do07:27
bartbesif I go to hell I might as well listen to mp3s there07:27
kyakyou will be made listen to Spears mp3s :)07:28
kyak32 kb/s, mono07:28
kyakyes, this is hell07:28
bartbesyou are wrong07:28
bartbeshell is where you listen to full-quality uncompressed recordings of the village people07:29
kyaktheir screams?07:29
wpwrakbartbes: i thought in hell, they had karaoke ?08:05
wpwrakwolfspraul: aha ! qi-bot doesn't look for cae-tools commits yet08:11
rafawolfspraul: do you have pictures of the idbg installatino day? I need to prepare the documentation with good pictures, for example, when werner used tht dremel, etc08:13
wolfspraulrafa: yes I uploaded them recently, one sec08:18
wolfspraulrafa: http://en.qi-hardware.com/w/index.php?title=Special:NewFiles&until=2010092515355208:19
wolfspraulalso the page before ("older")08:19
wolfspraulI need to embed them nicely in a page...08:19
rafawolfspraul: thanks a lot08:20
rafawolfspraul: I will write some documentation about all the installation processes08:20
wolfspraulwpwrak: cae-tools added08:26
qi-bot[commit] Werner Almesberger: Adjust button size when switching between views. http://qi-hw.com/p/cae-tools/2997c6b08:40
wpwrakwolfspraul: thanks !08:40
tugasofthello all08:40
tugasofti've been using the nanonote for a few weeks, with debian from pyneo08:41
tugasoftand after a few reboot while a was trying to upgrade debian, it never powered up again08:41
tugasoftdid anyone encouter the same problem?08:41
rafalarsc: did you do the port of the wifi SDW-823 microSD driver to 2.6.32?08:41
wolfspraulwpwrak: what's the model or your Roland 3D scanner again?08:42
rafatugasoft: upgrade... software upgrade? using apt?08:42
wpwrakwolfspraul: it's an MDX-1508:42
rafatugasoft: have you tried to remove usb cable and battery for a while (in order to be sure that you will not be into usbboot for example) and then to try again?08:43
wolfspraulwpwrak: thanks!08:44
tugasoftrafa: yes, tryed this.08:46
tugasofti was checking dmesg and lsusb at each try08:47
rafawpwrak: ah.. you have a modest MDX-15.. in china I buy cigarrettes and they will give me a mdx-15 as change if they do not have coins :)08:47
tugasoftthe device is not detected at all.08:47
rafatugasoft: have you tried another usb cable?08:48
rafamaybe the battery needs power and you cable is not okey08:48
rafayour cable08:48
wpwrakrafa: ;-))))08:48
tugasoftred led turns on when I  plug the usb08:49
wpwrakrafa: the mdx-15 is indeed the lowest-end model they have. there's a slightly bigger one with the same features, the mdx-20. luckily, the ben has just the right size for the mdx-15 ;-)08:49
tugasoftbut battery seems to be empty, indeed, it doesn't even warms up08:50
tugasoftdo you know if it's possible to get a new motherboard for the ben?08:50
tugasofti'm working in an arduino project, and now it is blocked :(08:50
wpwraktugasoft: you should be able to run it just from usb, without battery08:50
rafatugasoft: I would try with another usb cable as well.. and yes, as wpwrak says.. you can try without battery08:51
tugasoftif i plug usb only, the red led flashes very shortly and then nothing, i've tried several times08:51
tugasofthaven't tried other cable08:51
rafatugasoft: so you upgraded debian by software right? with apt or something like that08:51
tugasofti will try now, thanks for the suggestion08:51
tugasoftno, i just wanted to flash with new debian08:52
rafatugasoft: anyway.. you should see something in dmesg08:52
tugasoftno apt-get08:52
tugasoftwanted a clean install08:52
rafatugasoft: a.. so maybe boot loader or kernel is broken.. and you need to reflash again08:52
wpwraktugasoft: have you tried this ? remove usb and battery. wait 30 seconds. press and hold down the "U" button. plug usb. see if something shows up08:52
rafatugasoft: yes, wpwrak suggestion is good.08:53
rafawpwrak: if you did not flash properly bootloader.. U+button will work?08:53
rafaor he will need to use the nice usbboot pins under battery?08:54
wpwrakrafa: if the boot loader is broken, it won't work. but it's worth a try, considering how many difficulties people had with the pins08:54
rafayes :(08:55
rafawpwrak: some day I will try for one hour or less.. but I was thinking that nn was broken.. so you can imagine how many tries I did until it worked :D08:56
rafaand I was..*08:56
wpwrakrafa: yeah, it's a highly unreliable process that gets used in an unclear and disturbing situation. not a good combination.08:57
tugasoftwpwrak: will try the U button08:57
tugasoftand another usb cable too08:58
rafaand check dmesg all the time or lsusb or whatever you use to realize that something is pluged08:58
tugasofti really loved the nanonote, and i'm requesting for a few more. but if this first one is broken i won't be allowed to buy more :(08:58
rafatugasoft: I guess that it is not broken.. maybe something was wrong when flashing.. and if it was the bootloader which is not in a good shape you will need to use the pins under battery to get usbboot and then to flash again09:01
rafatugasoft: I guess that all of us feel you feel now at least once :)09:01
rafaus felt*09:02
wpwraktugasoft: also the USB port on the host can be a potential source of problems. the connectors sometimes wear out. (mini-USB surprisingly seems to be tougher. maybe because it's more fragile, so people automatically pay more attention)09:02
wpwrakrafa: after a while, you get very nonchalant about electronics that pretend to be broken ;-)09:03
tugasofti've just tried to boot using the U key, then by shorting the pins under the battery, none worked.09:05
tugasofti will keep trying, but I would like to know if there is a way to know if this is broken or not09:06
tugasofttried the host usb with  a webcam and a usb hard drive, both work well09:06
wolfspraultugasoft: first order of business - relax :-)09:07
rafatugasoft: I have tried shorting pins around 30 times until it worked for me :(.. but if you know well that maybe a few tries are okey09:07
tugasoftok, I will keep trying09:07
wolfspraultugasoft: let's do this: remove battery, unplug USB cable09:07
rafatugasoft: you need to try and to know how all the stuff work09:07
wolfspraultugasoft: let me try to help you09:07
rafatugasoft: so you know which could be the problems09:07
wolfspraultugasoft: remove battery, unplug USB cable09:07
wolfspraulfirst step09:07
wpwrakrafa: did you try with that carbonized rubber ? (i would just solder the darn thing after a few misses :)09:07
rafatugasoft: there are a lot that you will learn with it ;)09:07
tugasoftwolfspraul: i'm reading you, let's try again09:07
tugasoftbattery removed, ubs cable removed09:08
wolfspraulclose the nanonote, turn it upside down (so that you are looking at the open battery compartment)09:08
wolfspraulyou see the "usb boot pins" opening in the label there...09:08
tugasoftyes, see them09:08
rafawpwrak: yes, I was trying with the carbonized rubbber.. but because it was the first time using that I was not being sure how many seconds to use it before button.. if I needed to release that before button.. a lot of questions09:08
wolfsprauldo you have the small carbonized rubber button that came with your NanoNote?09:08
wpwrakrafa: ah yes, that adds to the confusion09:09
tugasoftgot the carbonized rubber09:09
wolfspraulyour nanonote should now be lying flat (it's closed, battery compartment up), so you just put the carbonized rubber button right over the pins, and press down solidly to short the pins09:09
wolfspraulyou can press really firm, obviously you don't want to crush the device but other than that you can be firm.09:10
wolfspraulreally firm :-)09:10
wolfspraulthen, still holding that button down, you plug in the USB cable from the side (no need to open the nanonote, just leave it lying there with the usb boot pins shorted by pressing down)09:10
wolfspraulsorry we are walking backwards a bit :-) on your host, you could have 'watch lsusb' running...09:11
wolfspraulso the correct order is this:09:11
tugasoftgot the idea09:11
wolfspraul1. run 'watch lsusb' on your notebook/desktop09:11
wolfspraul2. plug the USB cable into the notebook/desktop already, but not into the nanonote09:11
wolfspraul3. unplug the USB cable on the nanonote side, remove the battery09:11
wolfspraul4. put device face down, press the carbonized rubber button really firm onto the usb boot pins09:12
wolfspraul5. plug the usb cable into the nanonote from the side, leaving the device closed09:12
tugasofti'm used to small electronic devices, I know how be frim yet gentle with them :)09:12
wolfspraul6. in the watch lsusb display, see whether 601a:4740 shows up09:12
tugasoftfirm yet gentle, sorry09:12
wpwraktugasoft: what's important here is that the contact must be made _while_ the system resets (i.e., while it starts receiving power up from usb)09:12
wolfspraulif 601a:4740 does not show up, you can try to clean the usb boot pins with a bit of alcohol (just a wet cloth, no need for a whole drink for your nanonote...)09:13
wpwraktugasoft: i.e., the rubber button does not send a reset impulse or such but affects how the reset is executed09:13
tugasoftwpwrak: by contact you mean short the usb boot pins?09:13
wpwrakwolfspraul: pina colada ? :)09:13
wolfspraultugasoft: don't worry in my procedure everything necessary is already included09:14
tugasoftscotch on the rocks for me, only09:14
wpwraktugasoft: yes. (or it would be the same for the U button i suggested earlier. wolfgang's procedure is a bit more low-level, so it works in even more desperate cases)09:14
wolfspraulI owe you one if you can unbrick your NanoNote!09:14
tugasoftok, i will do the test, be back in a minute09:14
wolfspraulmy procedure works in all cases unless there really is a hardware defect, or the usb boot pins are unshortable09:15
tugasoftgot a question, my linux box has only usb 1 ports, is this an issue?09:15
wolfspraulin tugasoft's case, I am 99.9% sure it's just a simple case of u-boot overwritten09:15
wolfspraulhis problems started after a (failed) Debian upgrade09:15
wolfspraultugasoft: shouldn't be. just use that one port.09:15
wpwrakwolfspraul: i think the procedure should include also releasing the button. i suspect that a lot of people automatically think of it as a "reset button" and then diverge09:16
wolfspraulfair enough. it needs to be clear that the usb boot pins need to be held short while the USB cable is plugged in.09:16
wpwrakwolfspraul: (probably just software) yup. almost certainly nothing more09:16
wolfspraultugasoft: for safety, press down firmly even 1-2 seconds after plugging the usb cable in from the side09:17
tugasoft\o/ yahoo!!!! it came back to life09:17
wolfspraulfor the pressure, I never measured but I think even 1-2kg or so should be fine09:17
Action: wpwrak really needs to make the next, simpler, version of idbg :)09:17
tugasoftguys I owe a beer or an orange juice for each one of you09:17
wolfspraulI don't understand why so many people end up with bricked Nanos after Debian installs/upgrades.09:17
wolfspraulthose procedures don't seem to be very reliable...09:17
tugasoftnow I will (for sure) have flashing issues, but I think i will be able to handle them09:18
wolfspraultugasoft: your adrenalin can go down now first09:18
tugasoftlauched a successful09:18
tugasoftusbboot -c "boot"09:18
wolfspraulNanos may break, may die. but it's rare. and even then we replace them, of course! no worries.09:18
wolfspraulfrom the 1000 we sold so far, we had about 5 or so genuine returns.09:19
wolfspraulover time there will be more, but trust me if there is one thing I am motivated about it's customer support!09:19
wolfspraulI love it!09:19
wolfspraulthose 5 all got a brand new NanoNote, of course.09:19
wpwrakwolfspraul: did you ever find out what happened with the one where the cable seemed to be broken at the lcm side ?09:19
wolfspraulwpwrak: those 5 we had were almost all 'lcm stays dark' cases09:20
wpwrakwolfspraul: (customer support) worried about your karma after om ? :)09:20
wolfspraulso the NanoNote boots up, you can ping/ssh into it, but screen stays dark09:20
wolfspraultugasoft: if you tell me that that's what you see I would be more worried :-)09:20
wolfspraulI always loved customer support.09:20
wolfspraulbrings me down to earth, it's just great09:20
wolfspraulwpwrak: so in those cases, I think mostly it has to do with the FPC or connector on the mainboard09:21
wpwrakwolfspraul: i mean the one where the display worked intermittently when he pressed down on the lcm case. you did some support session about that one here on irc.09:21
wolfspraulone time there was a little hole in the FPC09:21
wpwrakoh, cute :)09:21
wolfspraulwpwrak: yes sure, we just replaced it with a new one.09:21
wolfspraulwe have not yet done detailed analysis on this, it's currently with David in Barcelona, 'out of service'09:22
tugasoftgot an error that says: can't read bulk data from Ingenic device:6409:22
wpwrakwolfspraul: ah, okay. should be fun to find out what went wrong there.09:22
wolfspraulthe fpc through the hinge is definitely the #1 reason for hardware problems so far, but even then it's rare, like I said about 5 for the first 1000 we sold09:22
wolfspraultugasoft: which commands did you run in usbboot?09:23
wolfspraulI only use usbboot to flash u-boot, the rest I do after booting Linux from microSD09:24
wpwrakwolfspraul: i don't think 99.5% is so bad a yield. of those 5, how many were actually broken and how many just needed a re-seating of the cable ? (well, the one with the hole sound like really dead.)09:24
tugasoftif i may allow myself to share my history with you, I also work in customer support, I'm  a support engineer :) for  a French company09:24
wolfspraula company that listens to their support engineers will be a successful company09:25
wolfspraulI hope they listen to you then!09:25
wolfspraullet's try to get your Nano back to life first...09:25
tugasoftthe have to, we do a lot of noise :)09:25
wolfspraulwhen did you get that bulk data error?09:26
wolfspraulyou ran reflash_ben.sh09:26
tugasoftusbboot -c "boot"09:26
tugasoftusbboot -c "nprog 0 openwrt-xburst-qi_lb60-u-boot.bin 0 0 -n"09:26
tugasoftthe second command gives me the error09:26
wolfspraulif you don't mind erasing the entire NAND, try "nerase 0 4096 0 0" after the "boot" command09:27
wolfspraulfirst turn the NanoNote off (unplug usb cable), then the usb-boot pin drill again...09:28
tugasoftthere's the output09:28
tugasoftif anyone is interested09:28
wolfspraulif that bulk data error is still there, maybe you need to try on another Linux machine09:28
wolfspraulunfortunately yours only has 1 USB port09:28
wpwraktugasoft: did you do the nerase ?09:29
tugasoftwpwrak: not yet09:29
wolfspraulif you have a USB hub, you can also try to plug the NanoNote in _behind_ the USB hub. just to compare.09:29
rafatugasoft: the nerase is important I guess09:29
tugasofti will try it09:29
wolfspraulrafa: but the 'bulk data' error is a usb problem09:29
wolfspraulshouldn't matter nerase or not09:29
wpwraktugasoft: nerase should come before the nprog.09:29
wolfspraulif you have a usb hub, try to put your NanoNote behind the usb hub09:29
rafawolfspraul: IIRC I got that messag when flashing the bootloader09:30
wolfspraulif you have another Linux machine, try to run usbboot on that machine09:30
wpwrakwolfspraul: so an nprog failure is reported properly, not just by "hanging" up ?09:30
wolfspraulsure, tugasoft : try inserting the 'nerase' after boot first...09:30
rafawolfspraul: and then I was thinking that it was not able to check because few data is read.. or something like that09:30
rafawolfspraul: then I continue with the flashing process... kernel, rootfs09:30
wolfspraulwpwrak: I have this 'usbboot' thing, I looked at the sources too much.09:30
rafaand then all the read/check are okey09:30
rafawolfspraul: and finally my nn boots okey09:31
rafawolfspraul: I will try to flash to report that message09:31
wolfspraulbut the alternatives are not ready yet, let's see that we get tugasoft off the ground09:31
rafawhen flashing the first file09:31
tugasofthere's the output fron nerase09:31
tugasoftit ends like: Force erase, no bad block infomation!09:32
wpwraki think this looks good09:32
tugasoftok,  i will try the next step09:32
wpwrakthe output is indeed quite confusing ;-))09:32
tugasoftgot the same error :(09:33
wpwrakargh :-(09:33
tugasofti will try will a newer laptop09:33
tugasoftthe first flash was made with a fresh debian in a newer laptopt and went well09:33
tugasoftso i will have to wait until tomorrow09:33
wpwrakif it's really a usb problem, changing cable or just retrying a few times can help09:34
tugasofthowever, you gring back my nano to life :)09:34
wpwrakchanging cable also means you unplug and replug, which occasionally also solves problems09:34
rafatugasoft: that is why I said that all of us felt that at least once :) (that our machine was broken).. I do not why you geeks do not have patience :D09:35
tugasoftwill try with the original cable, but it think now the problems comes from the usb host, it's a very old laptop that I use as a webserver and has and old debian install, old kernel...etc...09:35
tugasoftrafa: you are 100% right09:36
tugasoftgeek means no patience09:36
tugasoftallways full throttle09:36
wpwraktugasoft: (usb host) okay. sometimes, very old stuff can act a little funny indeed.09:37
rafatugasoft: ;)09:37
rafatugasoft: but then .. some guys like wolfspraul or wpwrak help to remove doubts :)09:37
wolfspraultugasoft: if you have a minute, try this09:37
wolfsprauldownload http://downloads.qi-hardware.com/people/xiangfu/tmp/openwrt-xburst-qi_lb60-zImage-initramfs.bin09:38
wolfspraulthen get your nanonote into a fresh usb-boot state (0x601a:4740 showing up)09:38
wolfspraulthen run "xbboot -u 0x80600000 openwrt-xburst-qi_lb60-zImage-initramfs.bin"09:38
wolfsprauland let us know what happens...09:38
wolfspraul(maybe nothing, but since you are at it maybe you can try real quick)09:39
rafawolfspraul: did you get some way to boot a kernel with a initramfs?? I tried a time ago.. without success.. I wanted to have an initramfs with a boot menu and kexec.. to choose the system to boot and also to upgrade the rootfs of NAND or sd..09:39
wpwrakwolfspraul: btw, which is the nicer one again, usbboot or xboot ?09:40
wolfspraulyes sure it works sometimes (just pasted a link). but it's not robust and well documented enough yet.09:40
wolfspraulxbboot is a subset, it only supports loading a binary into ram and executing it09:40
wpwraksame code base ?09:40
wolfspraulin the long run that's much nicer, we could have a nice rescue initramfs and then do the rest from there09:40
wolfspraulsome codes are reused, but mostly from scratch (xbboot is very short)09:41
wolfspraulit's just a few hundred lines of code09:41
wpwrakah, nice09:41
rafawolfspraul: yes.. that was the idea I wanted to use.. an initramfs as rescue, upgrade, system chooser09:42
tugasoftgot no xbboot09:43
tugasoftdo you have a link for a .deb file?09:43
tugasofti can search by myself too09:44
wolfspraultugasoft: how did you install usbboot?09:44
wolfspraulmaybe you already have the xbboot binary09:44
wolfspraulif not, never mind, it's just a fun test since you are here and responsive and friendly :-)09:45
tugasoftwolfspraul: dpgk -i usbboot.......09:45
wolfspraulare you sure you dont' have xbboot?09:45
wolfspraulI think it's in the same package09:46
wolfsprauljust type 'xbboot' - what happens?09:46
tugasoftnot found09:46
tugasoftbut I'm curious, I want to give it a try09:46
tugasofti read about this a few days ago09:46
wpwraktugasoft: did you type  xboot   or  xbboot  ? (here on irc, you did it right, but that may be an easy typo)09:48
tugasoftok, just figured out, I got and old usbboot .deb09:48
tugasoftwill try tu upgrade to 20100409:48
wolfspraul201009 seems the lates09:49
kyakwolfspraul: does it mean we are close to having that boot menu?09:49
wolfspraulwhich boot menu?09:50
wpwrakif the usbboot is ancient, could that also cause this sort of usb error ? we know that the 201004 has a command-line parsing problem that also caught a few people.09:51
kyakthe one rafa mentioned, with boot choises and rescue09:51
wolfspraulI think it will take more time, so 'not close'.09:51
kyakbut according to what you described, it already works - only needs some user interface09:52
wpwrakthat should be a worthwhile task for the forth aficionados. that's the sort of things forth is actually good for :)09:53
wolfspraulI think it's quite buggy still.09:53
wolfspraulwpwrak: I have a question about some idbg pictures09:53
wolfspraulwhat is the difference?09:54
wolfspraulit seems this one pad at the top right is mostly scratched/washed off in top309:54
wolfspraulis that on purpose?09:54
wolfspraulis that what 'top3' should document?09:54
qi-bot[commit] kyak: MPlayer: added mp3 support via libmad; see README http://qi-hw.com/p/openwrt-packages/a4873ed09:54
qi-bot[commit] kyak: Merge branch 'master' of projects.qi-hardware.com:openwrt-packages http://qi-hw.com/p/openwrt-packages/bf179f809:54
wolfspraulor is it just a failed attempt and we can delete that picture?09:54
wpwrakno. that was an etching problem09:54
wpwraknot sure what it should document ;-) maybe you just snapped picture of all the boards within reach :)09:55
wpwrakthe boards both look usable. they're of course v1, not v2. so the contacts are different09:55
wolfspraulkyak: if you are adding mp3 support please make sure it's properly disabled unless CONFIG_BUILD_PATENTED is on09:56
wolfspraulI think I got those pictures from you, I just didn't know whether this 'removed pad' was intentional or not09:56
wpwrakyou see the different arrangement here: http://www.almesberger.net/misc/ben/idbg-v2-install.jpg09:56
wolfspraulif it's not, I will delete all picture that only differ in this 'removed pad' from another picture we have09:56
wpwrakah, so i was just snapping pictures of all the boards ;-)09:57
tugasoftxbboot gives this: Info - found XBurst boot device.09:57
tugasoftset_addr 80002000h09:57
tugasoftbulk_write successfully wrote 2484 bytes.09:57
tugasoftstart1 80002000h09:57
tugasoftVR_GET_CPU_INFO JZ4740V109:57
tugasoftset_addr 80600000h09:57
tugasoftError writing 3825664 bytes (result -110).09:57
tugasoftit repeats endlessly09:57
wolfspraulhmm, OK09:57
wolfspraulit uses the same stage1/stage2 as usbboot09:57
wolfspraulnext idea: try another Linux notebook/desktop09:57
wolfspraulthanks for trying!09:57
kyakwolfspraul: good point..09:58
tugasoftthanks for suggesting it09:58
wpwrakit could be that i made a picture of each board as it went through production, to help with analyzing problems later, if any would creep up. (such as a bad trace at a place where i can't see it)09:58
tugasoftmy other laptopt is in the office, so i will suspend the tests until tomorrow09:58
tugasoftthank you very much for you support09:58
wolfspraulkyak: ah definitely. Please help to keep mp3 out unless the BUILD_PATENTED flag is on.09:58
tugasoftsee you09:58
wolfspraulit was introduced just for that purpose into OpenWrt09:58
wolfspraultugasoft: thanks for you patience! cu tomorrow...09:59
qi-bot[commit] kyak: MPlayer is too good not to depend on @BUILD_PATENTED http://qi-hw.com/p/openwrt-packages/cebe75410:00
kyakwolfspraul: let it be the whole MPlayer so far10:00
kyakmaybe later it can be built with reduced options, patent-free10:00
wolfspraulkyak: isn't mplayer already included in another openwrt package feed?10:10
kyakyes, no :)10:13
wolfspraulwpwrak: ok, so we delete the pics with etching problem? at least I will add a comment that documents this as an unintentional etching problem10:14
wolfspraulare those boards still functional? or broken?10:14
wpwrakwolfspraul: i think you can just delete the slighly defective ones, yes. they don't show anything interesting. i think the board was still usable.10:18
wolfspraulok will delete10:19
kyakwolfspraul: oh, i just noticed:10:27
kyakhooks/post-update: line 24: echo: write error: Broken pipe10:27
kyakafter git push to openwrt-packages10:28
kyakbut push went fine, as well as announce here in IRC10:28
wolfspraulfirst time I hear that10:32
wolfspraulthanks for reporting10:32
wolfspraulkyak: can you keep an eye on it? let's see whether it happens again, I do nothing for now :-)10:33
kyakwolfspraul: sure10:38
kristianpaulIs it mandatory erase all NAND when reflashing a Ben with a  new openwrt/xbusrt vesion?10:47
wolfspraulno, actually we should never do that. in a perfect world the oob and bad block data would always be preserved, starting from production of the NAND chip.10:49
wolfspraulbut we are far from a perfect world :-)10:50
kristianpaulok so i add that to the wiki in order to advice?...10:51
kristianpaulbut then how to get new bad block data info?10:53
wolfspraulover time as bits fail10:54
kristianpaulsomething is woring with xbusrt backfire11:10
kristianpaulor not11:10
kristianpauli knew it11:17
kristianpauli donmt why in the last backfire the /etc/profile is not same as the one stored at qi_lb60/files/etc11:17
kristianpaulis that okay this way?11:25
Action: kristianpaul bettwe writes a mail11:26
kyakwhat do you mean?11:26
kristianpauli built the last branch of backfire for the xbusrt11:27
kristianpauli noticed that the mgnu2x and other SDL apps not started11:28
kyakyou have to copy manually those files to your Ben11:28
kyakthese are not included into the image11:28
kristianpauli dont have too11:28
kristianpaulif is a the stable branch, isnt?11:28
kristianpaulor where is the stable branch? (in source code not images and flashing scripts)11:29
kyaki would call it development11:29
kristianpaulis no sense devel an app when the source code differ with an image thats seems to be stable11:29
kristianpaulat least for me11:29
kristianpaulsure i can copy11:30
kristianpaulbut i will not11:30
kyaksorry, i don't seem to understand you at all11:30
kristianpaulat least you know how to tell openwrt include those files11:30
kristianpaulafaik nobody does :(11:30
Action: kristianpaul back to write the mail11:30
kyakbtw, i have a small script11:32
kristianpaulanywaay thanks for helping kyak  .)11:32
kyakwhich i upload to Ben and run every time after flashing11:32
kyakit would fetch all latest files from git and also perform some other settings via uci11:33
kyakthis way i don't have to setup Ben manually every time after i reflash it11:33
kristianpaulbut why aditionals scripts, is that i cant be done in the openwrt .config?11:33
kristianpaulor your script is related with the openwrt .config ?11:34
kyakbut i have mu very own custome settings11:34
kristianpauli see11:34
kristianpaulis this script public? i'll like see it11:34
kyakhm, let me check.. i'm not at home right now11:35
kristianpauloh np when you able too11:35
kyakno, can't get it right now.. maybe in several hours11:35
kyakit's pretty dumb11:36
kyakit sets my tclshrc, nighskyrc, so on11:36
kyaksets some useful links, enable mounting of card and ubifs (data)11:37
kristianpaulplease share it in the wiki if you want11:38
kyakFH44 2I11:38
kyakwill do11:38
wpwrakkristianpaul: http://downloads.qi-hardware.com/people/werner/tmp/main.png11:47
kristianpaulwpwrak: nice but, what are the color balls for?11:49
kristianpaulwpwrak: where is that STL i want try print it11:50
kristianpaullooks really nice11:50
wpwrakthey mark the coordinate axes. green is x at 10, 20, 30 mm. blue is y. yellow is z.11:50
wpwrak(stl) ah, there's one little complication :) so far, this is pov-ray only11:51
kristianpaulokay no hurries11:51
wpwraki don't know the other formats. in theory, it shouldn't be too hard to generate them, having the pov-ray generator.11:52
wpwrakof course, i'm hoping that someone already familiar with STL or such would step in :)11:52
kristianpaulah you took the pij file to pov ray?11:56
kristianpaulwow this nanomap is really fast, is because qt?11:56
wpwrakkristianpaul: i took the .csv files, then aligned the faces with solidify (in the cae-tools project), and finally generated .pov with solidify11:58
bartbesokay, since #jlime seems to be quiet12:01
bartbescan jlime mount fat32?12:01
kristianpaulnot sure if there is support in kernel for that12:01
kristianpaulbut, are you running jlime from NAND?12:02
bartbeswell, I am not running it at all atm12:02
bartbesbut I will be running it from sd12:02
kristianpauland you have othe fat in the sd12:02
kristianpauli see12:02
bartbesas 3rd partition12:02
kristianpauli think is a kernel thing12:02
bartbesgoing to use it as home, and share it between owrt and jlime12:03
kristianpaulnot sure is rafa enaable it12:03
bartbeswell, there's an easy way to find out ;)12:03
bartbesalso, do I need the uimage?12:05
bartbesit isn't mentioned in the guide12:05
kristianpaulis all package in the tarball12:08
kristianpaulbut make sure you have last uboot12:08
bartbesthe wiki lied12:10
bartbes126 mb isn't enough12:10
bartbeswait a sec.. it was 129..12:10
bartbesah well12:10
kristianpaulwpwrak: http://en.qi-hardware.com/wiki/File:Roland_scanning_nanonote.jpg is familiar with you?12:10
kristianpaulwow long process12:11
wpwrakkristianpaul: very :-)12:11
kristianpaulguess cheaper?12:11
wpwrak(very) familiar and long, actually :)12:12
wpwraki think it's about usd 3000 in the us these days.12:12
kristianpaulcheck this (http://wiki.makerbot.com/cyclops) is not so accurate but you think could be?12:12
kristianpauland i can buy it ;)12:12
kristianpaulif i want and save some money sure12:13
wpwraki think the resolution is quite low. but yes, this sort of optical system has a number of benefits as well. e.g., you have zero contact force, you can easily change directions, and it's fast12:15
wpwrakthe sort of contact scan i'm using is a somewhat obsolete technology. alas, very accurate optical scanners aren't cheap either.12:16
wpwrakwhen i bought the machine, i actually didn't care much about the scanner. i only undusted the scan head when wolfgang mentioned that he didn't have any cad data :)12:17
kristianpaulhey wolfgang took great pics from shenzhen  http://en.qi-hardware.com/wiki/Special:Contributions/Wolfgang_Spraul12:17
qi-bot[commit] Werner Almesberger: Added auxiliary projections to help mis-levelling with more confidence. http://qi-hw.com/p/cae-tools/a3f17c512:18
qi-bot[commit] Werner Almesberger: solidify/batcvr.sfy: found a better inclination for the top face http://qi-hw.com/p/cae-tools/f635ac912:18
qi-bot[commit] Werner Almesberger: Rendering improvements. http://downloads.qi-hardware.com/people/werner/tmp/main.png http://qi-hw.com/p/cae-tools/32a4d4a12:18
kyakkristianpaul: http://downloads.qi-hardware.com/people/kyak/tmp/ben.sh13:19
kristianpaulkyak: thanks13:21
kristianpaulbtw xianfus respond my mail it just was need a simbolic link13:25
kristianpaul files -> data/qi_lb60/files/13:26
kristianpaulthats it !13:26
kyakwith libmad, mplayer is consuming only 75% CPU.. runs very nice and smooth13:31
kyakstil requires some polishing, like enabling osd13:32
kyakbut in genral, it works!13:32
SiENcEwhats the best dist for nanonote?13:32
kristianpaulit depens on you?13:33
kristianpaultell us yours needs13:33
SiENcEi mean, where do i have good apps like mplayer aso.,  and whats the best dist to compile and develop nativ on nanonote?13:33
kristianpauljlime !13:33
SiENcEbut its not compatible with dingoo?13:34
kristianpauldont thinkso13:34
kristianpaulbecause libc mainly if i remenber well13:34
rafakristianpaul: we checked and dingoo binaries run.. dont ask us why :D13:35
lekernelkyak, for playing mp3s?13:35
kristianpaulrafa: ah?? :O13:35
kyaklekernel: no, playing video with mp3 sound13:35
rafakristianpaul: no idea why13:35
rafakristianpaul: we tested the playstation emulator and that kind of stuff that dingoo has13:36
kristianpaulSiENcE: great for you :)13:36
SiENcEthats great13:36
kristianpaulkyak: or playing ogv?? ;)13:36
rafaSiENcE: if you want to build in nn, kristoffer, already explained what to install at : http://www.jlime.com/wiki/documentation/user/nanonote/muffinman .. "Getting Development Environment Going"13:37
SiENcErafa, thanks13:37
kyakkristianpaul: if i ever see ogv file, i would try13:37
SiENcEi really want to adjust my ports for dingoo to nanonote13:37
rafaSiENcE: that would be great.. zear wants to do that as well I guess13:38
kristianpaulkyak: http://ia331209.us.archive.org/0/items/Patent_Absurdity/Patent_Absurdity_VLQ_176kbit.ogv13:38
kristianpaulkyak: lets do ti !13:38
SiENcErafa, sounds good. what about an future nanonote with usb-host to use wifi?13:39
kristianpaultalking about usb-host is it posible implement it with linux, larsc ?13:40
rafaSiENcE: no, sorry13:40
SiENcEthats all i want on nano, dingoo etc.13:41
rafaSiENcE: there is some weird wifi SD which works13:45
kyakkristianpaul: only plays sound.. i would need to enable libtheora13:46
rafaSiENcE: http://en.qi-hardware.com/wiki/Ben_NanoNote_Wi-Fi13:46
kyakluckily, it's in openwrt13:46
SiENcEah na thats no solution13:47
SiENcEtoo expensive, too buggy13:47
rafaSiENcE: werner is working in some kind of wifi.. making the whole stuff (hw/sw)13:48
kristianpaulkyak: but the image is okay?13:48
SiENcErafa, ok13:49
kristianpaulahh just sound13:49
SiENcEthanks for the infos13:49
SiENcErafa, thx. Jlime works good :)14:15
SiENcE:(  usb ethernet has rndis14:17
SiENcEwithout rndis it would work also on windows14:18
wpwrakSiENcE: well, it's not quite wifi. it's ieee 802.15.4, the stuff underneath things like zigbee. in terms of real-life functionality one can expect, i'd say something in the range of a wireless keyboard or a wireless remote control. of course, you'll get to run tcp/ip over the wireless link, so you can do more interesting applications than just sending a few button-presses.14:20
kristianpaulwpwrak: have you in mind a usb dongle for the other side?14:21
kristianpaulor something like an acess point maybe?14:21
wpwrakkristianpaul: you mean something like this ? http://www.almesberger.net/misc/ben/wpan/board-100813.jpg14:22
wpwrakkristianpaul: well, change the connector and rotate things. then you have a dongle.14:22
kristianpaulahh that usb is for the compter?? i tought was a debuging boarad..14:22
kristianpaulhehe yes wpwrak14:23
SiENcEwpwrak, but where to attach? does nano has usbhost?14:23
wpwrakSiENcE: for the ben, you need this fellow: http://downloads.qi-hardware.com/people/werner/wpan/atusd/atusd-20100908-inserted.jpg14:23
SiENcEi see14:24
wpwrakdifferent view: http://downloads.qi-hardware.com/people/werner/wpan/atusd/tethered.jpg14:24
SiENcEdoes jlime has mc ?14:25
kristianpauluSD ?14:27
SiENcEmidnight commander14:27
kristianpaulmay be not by default but can be installed14:27
SiENcEhow do i install stuff on jlime?14:27
kristianpaulfirst it al14:28
kristianpaulopkg update14:28
SiENcEsorry for my questions. its the first time using jlime14:28
kristianpaulthen opkg install foo14:28
SiENcEopkg :-) ok14:28
kristianpaulopkg search foo | grep boo14:28
SiENcEi know14:28
SiENcEi searched the name of the package manager14:28
kristianpaulSiENcE: you should  join #jlime too14:46
kristianpaulas others here14:46
qi-bot[commit] kyak: MPlayer: enable OSD support http://qi-hw.com/p/openwrt-packages/2d9742e15:05
SiENcEmh dingux apps dont work on jlime15:22
Ayladoes jlime uses uclibc?15:23
qi-bot[commit] kyak: MPlayer: enable volume control via Ben volume keys http://qi-hw.com/p/openwrt-packages/7c7d36515:23
rafaSiENcE: I ran psx4all in jlime from dingux16:05
SiENcEmh i have to test this16:12
SiENcErafa, where is the correct manual?16:25
SiENcEi cant find it16:26
rafaSiENcE: http://www.jlime.com/wiki/documentation/user/nanonote/muffinman16:34
SiENcErafa, ok i know this. i though there is a bigger one16:42
qi-bot[commit] Werner Almesberger: Experimental alignments for solidify. http://qi-hw.com/p/ben-scans/4226d0b21:09
qi-bot[commit] Werner Almesberger: Correct direction of x inclination of z0. http://qi-hw.com/p/cae-tools/517041021:09
qi-bot[commit] Werner Almesberger: Make new/run/pov/disp are now configurable when invoking make. http://qi-hw.com/p/cae-tools/ffb0d9921:09
qi-bot[commit] Werner Almesberger: Moved batcvr.sfy over to ben-scans. Removing it here. http://qi-hw.com/p/cae-tools/1c183e021:09
aisaI'm trying to flash my nanonote for the first time, and I need someone to brainstorm with.21:10
aisaIs anyone available to listen to my setup and offer advice?21:11
xiangfuaisa: yes.21:30
aisaxiangfu: I believe I might have solved my problem, but here is what I'm dealing with.21:31
aisaI'm trying to flash my nanonote for the first time.  This page:21:31
aisawarns not to use a linux virtual machine.21:31
aisaAlas, my only computer is a mac, and I have VMWare for my linux box.21:32
aisaSo I have purchased a microSD card, for which I want to use to update my image.21:32
aisahowever, I cannot get this microSD card to mount through my virtual machine.21:32
aisaso I cannot run mke2fs....21:32
aisawhat I have decided to do, is mount the micro sd card on my nanonote,21:32
aisaand copy the latest image to the device by transfering it to my nanonote.21:33
aisathen I will update my nanonote from the microsd an normal.21:33
aisaIf this fails, I won't be able to try again...21:33
aisabecause I won't have any way to read the ext2fs micro SD card except on my nanonote.21:33
aisabut in this case, I will find an old computer which I can install debian on and go from there.21:33
aisaI am curious why I can't update the ben NanoNote from a virtual machine.21:34
aisaThe wiki tells me not to, but doesn't explain why.21:34
mththere is an ext2 fuse package for macports, but I don't know if it also includes mke2fs21:37
aisaGood, then I can try this in preference to hunting down a computer to install linux on, if my flash fails.21:38
mthalternatively, you could create a disk image of the same size as the ext2 partition and format that in the VM21:38
mthand after that, shut down the VM and convert it to a raw image21:39
aisanow that is a fantastic idea, then I could dd it to the micro sd.21:39
mthqemu has a tool to convert between vmware image formats and raw images21:39
xiangfuaisa: why we can not use VM, because the VM usb port give strange error. some VM work , some not. if you relfash in VM maybe all command success. but the NanoNote will not boot.21:40
wpwrakaisa: the VM warning is probably because some VMs can get USB wrong. but if you have a fallback solution, i think you can just go ahead and give it a try. worst case is that you'll join the ranks of those fumbling with the carbonized rubber button :)21:40
mtha third option would be to find a linux live CD that can boot on a Mac21:40
xiangfuanother option is install linux in Mac :)21:41
xiangfuaisa: I remember someone success build the "xburst-tools" in mac.21:41
aisamth: a quick search reveals that there might be an ubuntu live cd that I can boot on my powerbook, this is fantastic.21:42
aisaxiangfu: This is my employer's computer, and I would have a terrible time testing our software if I couldn't run Mac OS.  Without this problem, I would have already installed linux :-)21:42
aisaxiangfu: I had not thought to try this, but it makes sense.21:43
aisaI may eventually do that to make it easy on myself.21:43
xiangfuI found a carbonized rubber button picture : http://image.itmedia.co.jp/l/im/pcuser/articles/1004/01/l_og_nano_034.jpg21:51
qi-bot[commit] Xiangfu Liu: update the distribution to [unstable] http://qi-hw.com/p/xburst-tools/f1bc77a23:13
--- Mon Sep 27 201000:00

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