[13:20] <xnox> ogra_: as per "Documentation, firmware and drivers" section in http://www.ubuntu.com/project/about-ubuntu/licensing I see no problem with keeping panda images as "Ubuntu Desktop"
[13:21] <xnox> it has been re-inforced by TB in the past that on case-by-case we do include firmware and drivers provided that they "do not restrict our ability to make Ubuntu available free of charge, and that you can continue to redistribute Ubuntu."
[13:22] <ogra_> xnox, right, and the TB decision that pitti referred to explicitly stated the deriver exception
[13:22] <ogra_> *driver
[13:22] <xnox> yes.
[13:23] <ogra_> but we still should make users aware of the fact that there are binary drivers included
[13:23] <xnox> https://bugs.launchpad.net/ayatana-design/+bug/723831/comments/19
[13:23] <ubot2> Launchpad bug 723831 in ubiquity "Installer – The option to 'install third-party software' when installing Ubuntu should be selected by default (aka "make Youtube work")" [Wishlist,Won't fix]
[13:23] <xnox> he even wrote the comment himself =)
[13:23] <ogra_> (not only to warn them but also because they will love us for that ;) )
[13:23] <xnox> I can totally see "love" comming our way
[13:23] <ogra_> :)
[13:25] <ogra_> stgraber, so playingg with the zatab i can see it initializes the LCD backlight when u-boot comes up (still fiddling with the kernel) so something is definitely working right here
[13:25] <ogra_> i wonder if the kernel just defaults to HDMI ;)
[13:28] <stgraber> ogra_: AFAIK getting a working kernel framebuffer before X is quite tricky, maybe rbelem managed it now. My initial try was to mess with my setup until I'd get something to appear in the logs on the microsd, though without much success.
[13:29] <ogra_> well, zareason shows pics of ubuntu running on the tab on their facebook page
[13:29] <ogra_> but thats attached via hdmi
[13:29] <ogra_> they should just make the images available they have ...
[13:29] <ogra_> even if they are crappy
[13:30] <ogra_> oh, btw, there is no /proc/config.gz in the default android ... seems they disabled that :/
[13:38] <ogra_> stgraber, hmm, seems *everything* related to graphics output is a module in the defconfig
[13:56] <ogra_> stgraber, where did you get the script and bootscr from again ?
[13:59] <stgraber> ogra_: script.fex is a clear text version of script.bin from /dev/nand1. On our model it's: http://paste.ubuntu.com/1214832/
[13:59] <ogra_> thx
[13:59] <stgraber> ogra_: the boot.cmd is a fairly standard: http://paste.ubuntu.com/1214835/
[14:02] <ogra_> ok, this kernel is definitely doing something
[14:25] <ogra_> heh, framebuffer console is disabled by default
[14:36] <ogra_> grmbl, so i'm pretty sure the kernel comes up fine
[14:37] <ogra_> the backlight level changes definitely after a few seconds when booting
[14:37] <ogra_> (which doesnt happen if i remove uImage)
[14:59] <zumbi> hi!
[15:00] <zumbi> do you know what are default user/passwd for precise armhf images?
[15:00] <ogra_> ??
[15:00] <ogra_> the same as on official debian images (read: there are none)
[15:01] <ogra_> we expect the initial user to be created by the installer :)
[15:01] <ogra_> zumbi, what images are you trying
[15:01] <zumbi> http://cdimage.ubuntu.com/releases/12.04/release/
[15:01] <ogra_> what arch ?
[15:02] <zumbi> ogra_: on VGA monitor it brings you to desktop and asks questions, but I got a login on serial tty, after enabling it
[15:02] <zumbi> ogra_: armhf
[15:02] <zumbi> ogra_: for mx5
[15:03] <ogra_> well, normally it should run oem-config (subset of ubiquity) and create a user
[15:03] <zumbi> ogra_: right, it does that, I just dont have USB keyboard around to enter the details
[15:03] <ogra_> after all that is done that user should exist on the system and you should be able to log in with these credentials
[15:04] <ogra_> beyond that, there is no user at that point, the rootfs is competely virgin
[15:04] <zumbi> not even for root?
[15:04] <ogra_> ubuntu doesnt make use of root ;)
[15:04] <ogra_> sudo by default
[15:04] <ogra_> so the root account is always locked
[15:04] <zumbi> right.. that's a bit crazy
[15:05] <ogra_> i would suggest using a server image ...
[15:05] <ogra_> they default to serial for everything
[15:05] <zumbi> there are no server images for mx5
[15:05] <ogra_> (and tasksel offers ubuntu-desktop too if you want it)
[15:05] <zumbi> ogra_: well, thanks, I think I'll hack a user/pass with qemu
[15:05] <ogra_> hmm, i thought we had them
[15:06] <ogra_> zumbi, non need for qemu ...
[15:06] <ogra_> just sedit /etc/shadow on the SD
[15:06]  * GrueMaster had suggested netboot images as a bare minimum for all hardware, but ....
[15:06] <ogra_> but be aware that you miss configuration indeed
[15:08] <zumbi> ogra_: what do i need to change in /etc/shadow?
[15:08] <ogra_> the exclamation mark for the root pw
[15:08] <ogra_> just remove it
[15:08] <zumbi> root:*:15453:0:99999:7:::
[15:09] <zumbi> there is no !
[15:11] <ogra_> then the *
[15:12] <ogra_> should end up like a debian install where you set no rootpw
[16:59] <xnox> booting panda desktop image....
[17:02] <RoyK> should work...
[17:43] <xnox> ogra_: can you boot desktop panda sd card for installation.... but do not attach any other hard-drives
[17:43] <xnox> what is your expectation from ubiquity?
[17:44] <xnox> currently it goes straight from wifi setup -> manual partitioning screen
[17:45] <xnox> and finish button can be clicked dispite not doing anything useful (well crash)
[19:16] <ogra_> xnox, hmm, it should pop up the notice about installing to source media and let you do it
[19:27] <xnox> ogra_: well installing back on to the sd card should be supported.
[19:27] <xnox> ogra_: and stgraber said on #-release that why did it not get stuck at the prepare page - not enough disk space
[19:27] <xnox> ogra_: and if it didn't get stuck there, why didn't d-i offer auto-partitioning
[19:28] <xnox> ogra_: see bug 1053030
[19:28] <ubot2> Launchpad bug 1053030 in ubiquity "highly confusing UI on desktop when installation media is big enough and no external storage is attached" [High,Confirmed] https://launchpad.net/bugs/1053030
[19:31] <ogra_> xnox, i think simply because mmc support in d-i is really not across the board
[19:31] <ogra_> when we started with arm it didnt have any at all
[19:32] <ogra_> and we only implemented bits on the go where we needed them ... the partitioner never was among these
[19:32] <ogra_> since we never used Sd as target before
[19:32] <xnox> yeah, only in the pre-installed case.
[19:33] <xnox> but no ubiquity in pre-installed case.
[19:33] <infinity> sd wasn't a "target" in preinstalled, either.  It was just, well, pre-installed. :P
[19:33] <xnox> =))))) lolz
[19:33] <infinity> (And it totally used ubiquity)
[19:33] <ogra_> yeah
[19:33] <xnox> infinity: ok, no partman =)))))
[19:33] <infinity> Right.
[19:33] <ogra_> ac100 images still work that way today
[19:33] <infinity> Because no installing.
[19:33] <ogra_> right
[19:34] <xnox> right. So who wants to implement partman-auto support for mmc
[19:34] <xnox> in particular biggest_free should be doable.
[19:34] <infinity> I don't honestly see why it should be different in any way from the /dev/sdX support.
[19:34] <ogra_> right
[19:34] <infinity> Except that the device nodes are named differently.
[19:35] <infinity> Otherwise, the partition tables are the same, the high level interfaces to talk to it all is the same.
[19:35] <ogra_> and its likely just some kind of case statement )or equivalent) that needs to be enhanced
[19:35] <xnox> for some reason it is. I will poke it a bit.
[19:35] <xnox> For some reason it none of the autopartitioning recipes qualify and it drops into manual partitioning, e.g. partman/choose_partition
[19:36] <xnox> which breaks ubiquity's state-machine =)
[19:36] <infinity> (Maybe a bit of sauce to make /dev/mmc* device filesystems be mounted with flash-friendly options would be nice, but that's actually a more widespread problem that also affect installing to USB sticks on /dev/sd* as well)
[19:36] <ogra_> there are several bugs about that
[19:36] <ogra_> TRIM support etc