<qi-commits> Xiangfu Liu: remove u-boot under target/linux/xburst/image, use package/uboot-xburst instead http://qi-hw.com/p/openwrt-xburst/c0f7879
<qwebirc76045> kristianpaul: ping
<tuxbrain_too> kristianpaul: ping
<tuxbrain> kristianpaul: ping
<kristianpaul> tuxbrain: pong
<tuxbrain> hey man are you seaching for me?
<kristianpaul> yeah i was
<kristianpaul> about where you get 3.3v from nano board
<tuxbrain> yes?
<kristianpaul> is not clear for my in the picyres
<kristianpaul> but i wired the 3.3v from a label with that name
<tuxbrain> and it works?
<tuxbrain> in this one http://en.qi-hardware.com/w/images/0/03/Hwmod_serial_2.JPG , the black wire is attached to the first "big" capacitor counting left to right
<kristianpaul> i ditn test yet, hope i'll post as soon i got it working
<kristianpaul> but i just can said i avoided make holes in one side of the nano
<tuxbrain> mmm intrigued to see your solution :)
<kristianpaul> hehe
<steve|m> kristianpaul: this is related to osmocom-bb, wolfspraul bought the phones, since some of us want to get familiar wth the MediaTek chipsets ;)
<tuxbrain> yummy, isn't it? also Harald Welte has an eye on it , but don't hold your breath too much this is a long time project if even started
<tuxbrain> steve|m: I hope you can fit osmocom-bb on it :)
<kristianpaul> steve|m: good to know :)
<steve|m> tuxbrain: I'm currently adding support for the MTK romloader in our 'osmocon' tool which handles the serial connection to the phone, so a simple hello-world program on the mtk basebands should run soon
<kristianpaul> steve|m: yeah i was linked by laforge blog
<steve|m> kristianpaul: yes, harald wants a phone with jtag, I'm currently using this one, which is of better quality: http://bb.osmocom.org/trac/wiki/MobistelEL570
<steve|m> (which seems to have test-pads for jtag, too)
<tuxbrain> steve|m: I have not understand very well due my hardware knowledge is near to 0 , but it sound really good news :)
<qi-commits> Bas Wijnen: usb-booting alarm clock, with split gui http://qi-hw.com/p/iris/8c7cac3
<qi-commits> Bas Wijnen: removed old device handling http://qi-hw.com/p/iris/a4b5d94
<qi-commits> Bas Wijnen: before device-removal http://qi-hw.com/p/iris/62ac5f1
<kristianpaul> tuxbrain: http://identi.ca/notice/31639411
<gbraad> Nice to see osmocom picking up; had seen those CECT3100 very often in Beijing :P.
<kristianpaul> seems is cheap (13usd?)
<gbraad> you mean the phone? yes... about 115 Chinese yuan
<kristianpaul> yes
<gbraad> would be around 15 or 16 usd
<steve|m> but hell.. there are tons of better quality mtk based phones, which don't cost much more.. its just that not all of them have jtag
<gbraad> probably... but is JTAG exposed on those?
<steve|m> gbraad: not on all phones, but this is only needed for hardcore-debugging... there is a rom-bootloader, just like in the TI Calypso we currently support, so you can upload code via the uart to the ram, and run it
<gbraad> I understand. Will keep my eye on it. Soon I will be in China again... I have a lot of friends in the H/W community of Beijing.
<gbraad> but I will be off now. see you later...
<tuxbrain> kristianpaul: hey put that one in the wiki page! :) ,
<kristianpaul> tuxbrain: yes i will
<tuxbrain> steve|m: you see possibilities on run a linux in those phones(long term of course)? how much ei570 costs?
<steve|m> tuxbrain: this should somehow be possible, the EL570 costs about 80 euros
<steve|m> tuxbrain: this is basically what the guys of http://peeklinux.com are trying on the TI Locosto chipset.. which is ARM7, too, like the calypso and the mtk chips we're talking about
<steve|m> tuxbrain: but mabe it's better to use a phone with separated baseband/application processor then, there are some mtk designs that work this way.. and on the calypso-side the OpenMoko devices of course
<steve|m> well anyway, good night
<tuxbrain> steve|m: good knight and thanks for share this great info
<tuxbrain> knight->night
<steve|m> tuxbrain: one more thing, the approach of the guys is very interesting: we have deep documentation for the TI calypso since the beginning of the project, but for the Locosto there was almost nothing, so they have added support to qemu to emulate the locosto/peek and all its peripherals, and they're making good progress.. serial output and keypad is already working and so on
<steve|m> tuxbrain: they basically run the original proprietary firmware in qemu, and observe what it does :)
<tuxbrain> I just admire guys than can open the plastics of a device and can guess things just with that
<steve|m> tuxbrain: well, a scope/logic analyzer sure is of a great help, this is how I wrote e.g. the motorola C155 display driver for osmocom.. since you don't know the controller of the display, sniffing the commands, and then googling after them to find the datasheet is almost the only way (besides the approach of the peeklinux guys, or completely disassembling the firmware with IDA or something)
<tuxbrain> by the specs peeks seems basically a rigid nanonote with simcard :P
<steve|m> hehe
<steve|m> well, but gn8, this time for real ;)
<tuxbrain> steve|m: ok , idem, going to sleep cu
<gbraad> same here, for real too
<gbraad> 5 hours till alarm
<tuxbrain> 1 more than me :P zzzzzzzz
<methril_> someone up?
<wolfspraul> methril_: me
<methril_> i was guessing if i could help with the kernel side
<methril_> but most of the devs are sleeping :)
<wolfspraul> wah of course you can help
<wolfspraul> last I heard from Lars and others was DMA for MMC transfers, for example
<wolfspraul> or pushing our patches upstream
<methril_> this could be hard
<methril_> (pushing patches upstream ;)
<wolfspraul> I think Lars has done a lot of work already to cleanup and prepare for that.
<wolfspraul> not sure what you are most interested in, there is work everywhere
<methril_> i see a pwm driver
<wolfspraul> outside the Linux kernel, there are still many things in usbboot/xbboot and u-boot that could be fixed
<methril_> i did some pwm drivers before (in other archs)
<wolfspraul> ok
<methril_> i'm familiar with u-boot too
<methril_> i'm going to take a look
<wolfspraul> in u-boot, one long outstanding bug is that when booting from sd card, the kernel is read from a vfat partition, the rootfs from ext2
<wolfspraul> but for some reason the kernel cannot be read from the ext2
<methril_> it's an upstream bug?
<wolfspraul> don't know
<methril_> ok
<wolfspraul> just off my memory
<methril_> i'll take a look
<methril_> xiang or Lars connects to IRC?
<wolfspraul> also, I think u-boot currently does not support booting from multiple rootfs'es on the SD card
<wolfspraul> yes definitely
<wolfspraul> both
<methril_> i'll try later
<methril_> we are not in the same timezone
<wolfspraul> irclogs are here: http://en.qi-hardware.com/irclogs/
<methril_> and get harder to chat with them :)
<wolfspraul> so you can just write some questions to xiangfu or lars here, maybe they reply later at least in the irclogs you can see what they said :-)
<wolfspraul> methril_: another u-boot thing would be the ability to boot into Iris from SD card
<wolfspraul> that would allow people to easily try Iris
<methril_> uhm!! i like this one
<wolfspraul> I think Bas may have started with that but not sure.
<methril_> i see the idea of Iris
<wolfspraul> well Bas would be thrilled if you could help
<methril_> but i didn't find the time to help with it jet
<wolfspraul> well good opportunity, did you see Bas' mail on the list?
<wolfspraul> he just published an update
<methril_> yes
<wolfspraul> but I think it's hard for people to try out, at least I don't know a place that clearly says how to install it
<wolfspraul> if we even have binaries anywhere (I doubt it)
<methril_> i'm going to reply it
<wolfspraul> I think this only runs on 1 Nano in the world right now :-)
<wolfspraul> the best would be to allow u-boot to boot Iris from microSD
<methril_> then i could start writing a getting starting with Iris
<wolfspraul> that would give people an easy way to try Iris without fiddling with what they have in NAND
<wolfspraul> yes that would be awesome!
<methril_> i'm going to try it a little bit before to write , and i tryto help Bas
<wolfspraul> great, he will be very happy
<methril_> i'm going to prepare the toolchains before going to sleep