[05:55] Does d-i allow you to install to a degraded md RAID array? [05:56] e.g. can I install onto a one-disk RAID1 or a two-disk RAID5, while I wait for the last disk to arrive [10:44] michaelforrest1: http://people.canonical.com/~evand/tmp/timezone_menu_button_icon.png [10:50] and just like that, the menu is getting removed entirely [10:51] michaelforrest1: for posterity: http://people.canonical.com/~evand/tmp/timezone_menu_open.png [10:56] ubiquity: evand * r4153 maverick-redesign/ (4 files in 3 dirs): [10:56] ubiquity: * Add a new menu button widget. [10:56] ubiquity: * Wire up the timezone model to the menu button's menu. [11:02] ubiquity: evand * r4154 maverick-redesign/ (4 files in 3 dirs): [11:02] ubiquity: * Get rid of timezone menu button. We had three ways of doing exactly the [11:02] ubiquity: same thing (click on the map, type in an entry with autocomplete, or click [11:02] ubiquity: on the menu button). We were exposing the raw data in a way that was [11:02] ubiquity: unnecessary. [14:52] ev: bug 524803 - am I right in thinking that it would simplify usb-creator if all our CDs were hybrid? [14:52] Launchpad bug 524803 in syslinux (Ubuntu) (and 1 other project) "isolinux hybrid mode should be used - all other major distributions do so since last year (affects: 4) (heat: 18)" [Undecided,Fix released] https://launchpad.net/bugs/524803 [14:59] the cost is an extra 100K on CD images, and some considerable thought about how to handle jigdo integration === ara_ is now known as ara [15:03] slightly confused, I thought we were dropping syslinux for grub2 [15:04] (which would also simplify usb-creator) [15:04] cjwatson: ^ [15:05] we are ultimately, but it's the same principle [15:05] that being to have the CD be bootable directly as a USB stick rather than requiring conversion [15:05] of course usb-creator would still need to do stuff in order to enable persistent storage [15:05] indeed [15:06] I suppose it comes down to how long its going to take us to move to grub2 for the CD bootloader. If it's coming this cycle, I don't see why we should waste the cycles on syslinux work. [15:06] it's at risk [15:08] ah [15:08] well then [15:09] that said, experimentation suggests that right now we get to choose between fast jigdo generation and hybrid images; or possibly some hideous hack to tweak the template file post-generation [15:10] I'd ideally like to avoid doing any usb-creator work prior to FF as I'm racing to finish the installer redesign. I'm planning on dropping my other usb-creator work items. So ideally if we're going to do this, I'd like to make the usb-creator changes as a bug fix in late August. [15:10] at least until somebody gets http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=567636 done [15:10] Debian bug 567636 in src:libisofs "port jigdo template engine to libburnia" [Wishlist,Open] [15:10] so, if we can see a way round the other problems, we can do isohybrid without requiring any work in usb-creator [15:10] it would just *permit* usb-creator to be simplified [15:11] ahh, right [15:12] it's still a valid USB image [15:12] er, still a valid CD image [15:12] indeed [15:13] ev: so you noted I got it installed? [15:14] seems the nodmraid flag didnt do anything. when I did a simple apt-get remove dmraid (which did give an error, but still removed it) and opend the installer, I had disks detected. [15:17] jussi: odd [15:18] ev: I agree odd. the above solution was mentioned in a related bug I found, so I gave it a go. [19:06] Question, when Ubuntu-cdimage is building iso for lucid where does it pickup which version of isolinux and menu to use? [21:48] What is the purpose of the apport. It seems if at least one of my seeds does not include apport, installation fails. [22:06] cjwatson, in scroll back i saw a mention that grub for cds is at risk, is that to lead me to think uefi is also at risk too, or are you treating them as mutually exclusive and still looking at possibly being able to have efi grub and syslinux both on the cds then? [22:07] and more importantly, is there anything I can offer to help alleviate risk? [23:08] superm1: the bit that's at risk is replacing gfxboot; I plan to get uefi-support done either way [23:08] cjwatson, ah i see, okay :) [23:09] it might just involve two boot loaders on the CD and not having gfxboot in the uefi case [23:32] ubiquity: superm1 * r4134 ubiquity/ (5 files in 2 dirs): Allow setting a custom title via ubiquity/custom_title_text.