[12:32] <davmor2> apw: hey dude https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1374126  confirmed it is still happening on todays image, is there anything else that you could do with log wise?
[12:38] <apw> davmor2, i take it installing brcmwl manually works ?
[12:39] <davmor2> apw: if I hook it up to the internet via cable I can indeed install brcmwl and it works fine then,  It just doesn't seem to be install like it was in trusty from the install 3rd party from ubiquity
[12:40] <davmor2> apw: is it worth seeing if the package is even on the cd?
[12:42] <apw> davmor2, yeah and if software-properties-gtk is on there
[12:45] <davmor2> apw: so without the cd enabling as a repo, I get No additional drivers available on the additional drivers page,  Let me enable the cd now
[12:46] <davmor2> apw: okay now it shows the broadcom driver, so it is on the cd
[12:48] <apw> and what id's does your brcm have
[12:49] <davmor2> apw: Broadcom Corporation: BCM4331 802.11a/b/g/n
[13:01] <apw> davmor2, ok that is pretty standard osunding.  this seems more likely to be a a bug in ubuntu-drivers-common ... but i am not 100% as the last time i looked at this stuff it was called jockey
[13:07] <apw> davmor2, ok what does "ubuntu-drivers list" command say when in the CD ?
[13:08] <apw> booting a livecd ought to be enough i recon
[13:11] <apw> davmor2, and indeed "ubuntu-drivers autoinstall" should install them if available
[13:27]  * apw pokes davmor2
[13:34] <davmor2> apw: sorry Lunch got called
[13:35] <davmor2> apw right give me 2 minutes and I restart the live cd
[13:36] <apw> ack
[13:43] <davmor2> apw: ubuntu-drivers list only displays bcmwl-kernel-source
[13:44] <apw> davmor2, well that is the right one, does ubuntu-drivers autoinstall make wifi work ?
[13:44] <apw> as that is what the installer is meant to run during its life
[13:45] <davmor2> apw: ubuntu-drivers autoinstall throws up No drivers found for automatic installation
[13:45] <apw> !?!
[13:46] <apw> can you paste the output of both of those commands into a comment please
[13:46] <davmor2> apw: sure can
[13:47] <apw> davmor2, and the output of $ cat /sys/class/net/wlan0/device/modalias 
[13:47] <apw> oh no that prolly won't be called that
[13:54] <apw> davmor2, i need the modalias file from the /sys/ directory which is that device
[13:55] <davmor2> apw: right on live cd or the installed system?
[13:59] <apw> davmor2, either
[14:02] <davmor2> apw: is it just class you need or all of sys?  my plan is just to tarball it
[14:02] <apw> davmor2, i just need the modaliases from the wl car
[14:02] <apw> card
[14:02] <apw> just that file, that is what the autodetect stuff uses
[14:03] <davmor2> apw: under net I only see eth0 and lo
[14:03] <apw> are you in the live cd ?
[14:03] <davmor2> yes
[14:04] <apw> then it would be either in /sys/bus/pci or sys/bus/usb
[14:04] <apw> i assume
[14:05] <apw> head /sys/bus/pci/devices/*/uevent
[14:05] <apw> davmor2, something like that should find it
[14:14] <davmor2> apw: I'm going to paste what I have, under driver I see b43, but under devices I only see a bcma which I think is the card reader
[14:15] <apw> smb, do you routinly test with more than one vCPU when you make VMs, i had an install failure friday with 2 vCPUs, which i ignored but have no idea if that was the cause
[14:16] <davmor2> http://paste.ubuntu.com/8507607/
[14:16] <smb> apw, I mostly run with more than 1 
[14:17] <apw> smb, ok cool, ignore that then it is likely just random
[14:17] <smb> apw, ok, but can keep an eye open...
[14:21] <apw> smb, ta ... i have a couple of installs here with 1cpu i guess i'll reboot them with more and see
[14:21] <apw> davmor2, ok there appears to be a device which would match
[14:22] <davmor2> apw: \o/ thank god for that :)
[14:24] <apw> davmor2, waht does "apt-cache show bcmwl-kernel-source | grep Mod" say
[14:29] <davmor2> apw: http://paste.ubuntu.com/8507678/
[14:30] <apw> ok that matches my expectation
[15:48] <apw> tseliot, ok i think you are in the frame for bug #1374126 ... seems you stopped installing bcmwl by defaul, in favour of drivers using linux-firmware-nonfree
[15:48] <apw> tseliot, but ... you have the former on the plastic, and not the latter, and without a network you can't get a network
[15:53] <tseliot> apw: yes, the open driver is usually much better than the blob (which breaks some systems)
[15:56] <apw> tseliot, as ubuntu-drivers wants to install linux-firmware-nonfree, would it be helpful if it at least had a way to tell you it wants to but cant?
[15:56] <apw> ad right now, trusty CDs work, and utopic CDs don't have wifi ... which is rather confusing
[15:58] <tseliot> apw: does that mean that bcmwl is on the cd whereas linux-firmware-nonfree is not?
[15:58] <apw> tseliot, i believe so, davmor2 ^^
[15:58] <apw> tseliot, and i don't think l-f-nonfree can be
[15:59] <tseliot> apw: can't we ship it without installing it and then have u-d-c install it when needed?
[16:00] <apw> i don't think we are allowed to have anything non-free on teh CD are we ?
[16:00] <apw> infinity, ^
[16:00] <tseliot> but how is bcmwl free?
[16:00] <apw> you have a fair point
[16:01] <tseliot> and slangasek ^
[16:02] <apw> davmor2, i'll install l-f-nonfree and at least confirm doing so has some effect
[16:02] <davmor2> tseliot: apw: hmmm so b43-fwcutter and bcmwl-kernel-source both show if I remove the internet connection and do apt update && apt search broadcom
[16:04] <infinity> /pool/restricted/b/bcmwl/bcmwl-kernel-source_6.30.223.248+bdcom-0ubuntu1_amd64.deb
[16:04] <infinity> It's on the CD, yes.
[16:04] <apw> installing l-f-nonfree doesn't seem to fix this box either
[16:04] <infinity> If it depends on l-f-nonfree, why isn't that dependency in the packaging?
[16:04] <tseliot> infinity: in the packaging of what?
[16:05] <infinity> Also, is this actually non-free, or is it only in resitrcted because of sketchy dependencies?
[16:05] <infinity> tseliot: bcmwl-kernel-source...
[16:05] <tseliot> infinity: well, it's not a dependency
[16:05] <apw> infinity, no, this is brcwl being "replaced" by l-f-nonfree and open drivers
[16:06] <infinity> apw: Oh, I guess I've missed some context here.  If it's being replaced, let's tear it out. :P
[16:06] <infinity> apw: That said, free drivers don't get you very far if you still need non-free firmware. :/
[16:07] <tseliot> apw: what does dmesg say with the firmware?
[16:08] <apw> [ 8856.808138] b43-phy2 ERROR: Firmware file "b43/ucode29_mimo.fw" request failed (err=-12)
[16:08] <apw> [ 8856.808144] b43-phy2 ERROR: Firmware file "b43-open/ucode29_mimo.fw" request failed (err=-12)
[16:09] <apw> which i presume is not in l-f-nonfree
[16:09] <infinity> So, those files aren't shipped?
[16:09] <apw> i asusme not, but have not checked, it threw those errors after installing
[16:09] <tseliot> apw: that was already there though: https://launchpadlibrarian.net/185777218/syslog
[16:10] <infinity> Wow, who let linux-firmware-nonfree into the archive? :/
[16:10] <infinity> This is beyond non-free.
[16:10] <infinity> rtg: Double-U Tee Eff, mate.
[16:10] <infinity> apw: Also this is your issue:
[16:10] <infinity> linux-firmware-nonfree (1.15) utopic; urgency=medium
[16:10] <infinity>   * Drop B43.tgz for license concerns. This could cause regressions with existing
[16:10] <infinity>     users. However, there are clear instructions at
[16:10] <infinity>     http://wireless.kernel.org/en/users/Drivers/b43 for how to install this firmware.
[16:11] <infinity>     -LP: #1326776
[16:11] <infinity> Or, I assume it is.
[16:11] <infinity> -- Tim Gardner <tim.gardner@canonical.com>  Wed, 13 Aug 2014 07:50:00 -0600
[16:12] <tseliot> perfect... just perfect... :)
[16:12] <apw> ahh ... sigh ... we're getting to the point where we cannot help them
[16:13] <infinity> To be fair, I'm inclined to drop this whole package "for license reasons".  Mainly, that we have no license.
[16:13] <infinity> This package is one massive copyright violation.
[16:13] <infinity> rtg: Enlighten me if this is somehow not true.
[16:14] <tseliot> well, I suppose I could work around that, and make sure that we install bcmwl only if we get those errors in dmesg (?), so that we don't break systems that work well with the open driver (without the firmware, if this is even possible)
[16:18] <tseliot> OR I can simply restore the original behaviour in ubuntu-drivers-common (less work and a safer solution for 14.10)
[16:30] <tseliot> infinity: so let's keep bcmwl on the cd please
[16:30] <apw> tseliot, i seem to remember jokey popping to tell you do things in the old days
[16:30] <tseliot> not that we have much of a choice
[16:30] <apw> tseliot, could you talk people though installing wl, getting the firmware and swtiching
[16:32] <tseliot> apw: I'm not sure u-d-c supports user interaction as jockey used to
[16:35] <infinity> To be fair, "you need a network to install your network driver" is a Windows installer behaviour I'd prefer not to emulate, if we can avoid it.
[16:35] <infinity> (Though, I realise we may have no choice with hostile upstreams)
[16:48] <tseliot> as I suspected, that kind of behaviour is not part of the current code. IIRC that's by design
[16:53] <tseliot> we'll have to stick with bcmwl :/
[17:02] <davmor2> apw: I assume you no longer need this device right so I can shut it down?
[17:02] <apw> davmor2, indeed thanks, and close that pinhole
[17:02] <davmor2> apw: ta
[19:36] <Kano> hi, can you update kernel wedge to latest debian version