/srv/irclogs.ubuntu.com/2008/03/25/#ubuntu-installer.txt

joshkcjwatson, evand: ping?00:02
TheMusojoshk: cjwatson will not be around this week, and I'd say its the end of evand's day, so he may not respond for another 12 hours at least.00:16
joshkthat's okay - i've emailed the list instead00:26
joshkbut thanks for the info00:26
evandcjwatson: On bugs like https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/205762 , is a null modem cable and serial port our only option for further diagnosing the problem?04:24
ubotuLaunchpad bug 205762 in ubiquity "hardy install ubiquity freezes during partitioning stage" [Undecided,New]04:24
superm1netconsole?04:26
superm1depending on when exactly the freeze comes down, netconsole is at least worth a shot04:27
evandooo04:33
superm1never used it?04:37
evandnope, this is the first I've heard of it04:38
superm1I was using it to grab a kernel panic, but unfortunately realized that the panic was in the network driver, so moot point04:39
evandheh04:39
superm1i'm hoping their is a BoF at UDS regarding debugging improvements.  I'd like to see the usability of turning on things like kdump, netconsole, and serial console to be much easier for people04:40
superm1something like going into /etc/default and setting something to true/yes and rebooting04:40
superm1s/their/there/04:40
superm1would probably go the extra mile for people like bdmurray and yourself when triaging through bugs without having to describe complex steps to folks that won't necessarily understand them04:42
evandindeed.  I think heno would be the person to talk to about getting that scheduled.  aiui, the team leads make the schedule for each track, though you can probably make a spec, propose it for UDS, and then ping one of the people who have access to the scheduler (cjwatson and keybuk, I believe).04:45
superm1so just writing the spec and proposing it isn't enough then normally?04:46
evandIt might be enough, but the fact that lots of specs get created and proposed for UDS always worries me that anything I propose will get buried.04:47
superm1well i have a few more that I had in mind too, but my other worry is that for each one that I propose, i'm not sure that a whole "timeslot" needs to be allocated04:47
superm1so it's hard to judge whether it's worthy of even propsing04:48
evandhrm, not sure what the best approach there is, there might be sufficient break away room space.  Though that's probably best answered by the jono, jorge, and the rest of the community team as I believe they're in charge of the event itself.04:51
superm1well at least in the interim i'll just have to remember to keep adding these to LP so I don't forget.04:53
joshkevand: still around?07:04
PecisDarbsHi people, question about translation of Live CD installer - I have translated Live CD menus and help (the beginging in console) - I have problems with fonts, where should I report this?09:58
james_wHi all. Does this change look ok?13:56
james_whttp://paste.debian.net/5180113:56
james_wfor https://bugs.launchpad.net/ubuntu/+source/grub-installer/+bug/4201913:56
ubotuLaunchpad bug 42019 in grub-installer "installer: grub password not verified" [Medium,Confirmed]13:56
james_wIs there an easy way to test this?14:00
evandjames_w: looks good, and you can test it by building a new d-i with your modified udeb in localudebs.14:10
james_wevand: have you a pointer to how to do that?14:13
james_wapt-get source debian-installer; fakeroot debian/rules binary?14:13
james_whow do I make the udeb as well? just build grub-installer?14:14
evandSure.  First, build your modified grub-installer with debuild, then apt-get source debian-installer and copy the udeb file to localudebs.  Then in debian-installer/build type fakeroot make netboot to build a new iso.14:15
james_wevand: great, thanks.14:15
evandyou're welcome14:16
james_wI've only made the grub changes on x86, I presume I need to do it for other arches as well, is that correct?14:17
james_wnetboot isn't a valid target.14:18
evandnope, the grub-installer source package handles all the architectures that it supports.14:18
evandsorry, build_netboot14:18
evandmake without arguments will give you the full list14:18
james_wcool, thanks.14:19
james_wI checked out svn://svn.debian.org/d-i/trunk/packages/arch/i386/grub-installer/ though, so I assumed that it was x86 only14:20
evandah, fwiw, we have a bzr import of grub-installer trunk here: http://bazaar.launchpad.net/~ubuntu-core-dev/grub-installer/ubuntu14:22
james_wah cool, thanks.14:23
james_wis it likely that you will merge from Debian again before release? i.e. is just submitting their ok, or would you like me to push to both?14:24
evandany changes to grub-installer will end up there.  You need to be in core-dev to commit, but let me know if your patch passes testing and I'll sponsor it for you.14:24
james_wevand: cool, thanks.14:25
evandI don't think we'll pull from Debian again this release cycle.14:25
evandjoshk: I'm back now.14:26
evandholding off on releasing a new ubiquity until I unbreak interaction between hal-lock and the noninteractive frontend.  Working on that now.16:34
james_wIt didn't change anything in the installer18:39
james_w(took me a while as I forgot to go to expert mode first time).18:39
evandas in the fix worked without affecting anything else or your CD didn't include the new grub-installer?18:41
evandif it's the latter, did you bump the debian version before calling debuild?  Somewhere in the dest folder there should be a manifest file.  That will have the version of each package included on the CD.18:47
james_wevand: ah, I didn't change the version number of grub-installer.18:51
evandah, sorry for not mentioning that.18:51
james_wno problem, I did wonder as I was doing it, I should have asked.18:52
mario_limonciellevand, in the DVD generated on 3/22, it's looking like ubiquity noninteractive isn't spawning on its own.  Is this a known issue?20:58
mario_limoncieller on both DVDs generated 3/22 that is.20:58
evandvery odd.  I'll look into it on the dvds I'm going to generate tonight (once the new ubiquity leaves the buildds and hits the archive)20:59
mario_limonciellwell it looks like that behavior is changing anyhow with these changes in capser 1.1.27 and ubiquity 1.8.121:00
mario_limonciellso perhaps it will just absolve it self21:00
evandthat's what I'm hoping for :)21:00
joshkevand: did you see my post on ubuntu-installer?21:01
joshki was wondering about the behavior of is_supported_fs and why it doesn't support iso966021:02
evandjoshk: did you catch Agostino's follow up?21:05
HopsNBarleyhi gang - can anybody help me get rid of the prompt for netcfg/get_hostname in a 7.10 install?  i've even got partitioning working with preseed, but can't get rid of this!21:07
mario_limonciellevand, fyi it just published a few minutes ago.  https://edge.launchpad.net/ubuntu/+source/ubiquity/1.8.121:11
evandHopsNBarley: Have you read through section B.4.2 of the install guide?  Are you sure it's asking netcfg/get_hostname despite you preseeding it?21:13
evandmario_limonciell: thanks, still have to wait for the built packages to hit the archive and then some padding for the livefs buildds to catch up, but I'll keep an eye on it.21:15
HopsNBarleyevand, this is a network install - i'm supplying the hostname (now) two ways: dhcp and preseed.  it is picking the dhcp name.21:21
evandright, dhcp takes precedence.21:23
HopsNBarleyin either case, i get the dialog, and have to "select" it to continue.  this is the only part of my auto install that i have to hit "return" on, and i'd like to get rid of it  (-;21:23
HopsNBarleyi tried adding the flag "seen true", but that doesn't do it.21:24
evandhrm, I seem to recall a longstanding bug about this.21:24
HopsNBarleyshould i supply a null setting to the kernel, as i've seen in some examples?21:24
HopsNBarleyevand, hey, i appreciate the help.21:25
xivulonevand I am back21:26
xivulonwill update wubi-installer.org soon21:26
xivulonfirst beta will be rev 457 with new translations21:26
HopsNBarleyevand, lemme try adding this preseed/run file.21:28
evandok21:29
evandHopsNBarley: shot in the dark, but try preseeding the domain as well, or preseeding it to a null value.21:30
evandxivulon: great!21:31
* xivulon is writing the faq while waiting for rosetta downloads...21:34
HopsNBarleyevand, i'm already preseeding the domain - and it doesn't ask me for it!21:37
evandheh21:40
joshkevand: yes, i saw it.. but it didn't answer my questions21:49
joshkoh oops i think i forgot to group-reply my second email21:49
* joshk bounces21:50
joshkxivulon: so can you explain about is_supported_fs?21:51
joshkwhy doesn't it support iso9660?21:51
joshkdoes is_supported_fs mean is_supported_writable_fs?21:51
joshki don't see why it's necessary because it is just used to copy off the custom-installation dir to the ramdisk21:52
joshkanyway, this is important because i would like the vmware feature that i'm working on to support 8.0421:53
mario_limonciellevand, i see that the new one was generated, but for some reason or another gawk is no longer in the pool directory on the list of packages on the DVD?21:54
mario_limonciellsometime between http://people.ubuntu.com/~ubuntu-archive/cd-build-logs/ubuntu/hardy/dvd-20080318.log and http://people.ubuntu.com/~ubuntu-archive/cd-build-logs/ubuntu/hardy/dvd-20080320.log it dissappeared21:59
xivulonevand, pushed rev 457 can you give it a quick go before I upload?21:59
james_wgrr, I failed again.22:02
mario_limonciellactually i'll bug slangasek about that, it's likely a seed change22:02
xivulonjoshk, simply put when I wrote lupin I wasn't thinking about searching a CD to find an ISO/preseed...22:06
xivulonbut there is no intrinsic reason why iso9660 cannot be added22:06
xivulonsimilar argument goes for floppies...22:06
xivulonIf memory does not fail me, there might have been some complaints about users getting stacked on floppy access22:07
joshkxivulon: would it be risky to add iso9660 to is_supported_fs list for hardy?22:08
joshkif so, i don't even care about floppy support22:08
joshki was only using that to get around is_supported_fs, because the floppy was vfat22:08
xivulonjoshk I wouldn't think there would be any problem in adding iso966022:09
xivulonbut it's feature freeze so the ff exception rules apply22:09
xivulonas far as I am concerned I have no problem in changing that22:10
joshkI can help you with the appropriate testing if you need me to22:10
xivulonevand is the above ok?22:10
HopsNBarleyevand, turns out a "priority=critical" boot flag fixes the problem.22:10
joshkand.. i can come up with a compelling reason. Canonical is allied with VMware and we want to support it in our product :)22:10
james_wcould someone tell me what I am doing wrong here please?22:14
xivulonjoshk can you add a bug to launchpad if you haven't done so already?22:14
joshkyeah, i'll do so today22:14
joshkFirst i have to go get a haircut22:14
xivulonthanks22:14
joshkwhat should i file it against?22:14
joshklupin?22:14
xivulonyes22:14
joshkokay22:14
james_wI am trying to create a modified image that contains a changed grub-installer. I made my changes to an svn checkout of grub-installer, ran dch -i and added a changelog, then ran debuild.22:15
james_wI then took the udeb from the parent dir and dropped it in build/localudebs in the source package of debian-installer.22:15
james_wI then ran fakeroot make build_netboot22:15
james_wand then booted vbox from the mini.iso in dest/netboot/22:16
james_wand it appears that my modified package is not used.22:16
james_w dest/MANIFEST.udebs doesn't list grub-installer at all.22:18
evandjames_w: ah, my mistake.  You probably need to add it to pkg-lists/netboot/common.  I'm confirming this with a test install now.  Sorry about that, I had assumed it was already in there.23:06
evandxivulon: yes, that's ok23:06
james_wevand: it's no big deal23:07

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!