[17:33] can armhf do gpt? [17:33] why should amd64 & i386 still semi-default to msdos? [17:33] why don't we switch everything to gpt? [17:33] * xnox assumes that e.g. uboot supports gpt [17:51] We don't switch x86 to GPT wholesale because most old BIOS implementations won't boot it. [17:52] Not all uboot implementations will either. [17:53] recent ones do though ... [18:08] ogra_, i think arm64 defaults to gpt, but not armhf. obviously one can override it. [18:08] infinity, why is lucid still on archive.ubuntu.com mirrors? [18:10] xnox, "defaults" is such a broad word :) [18:10] * ogra_ just booted an arm64 dragonboard via uboot with GPT ... [18:10] i booted it with MBR before though ... its a matter of hacking ;) [18:12] ogra_, partman-partitioning default_disk_label() function. This is #ubuntu-installer =) we get to pick what we lay down by default =) [18:12] lol, right, i missed what channel i'm in ... for me itz was partition.sh and flash.sh :P [18:13] ogra_, hippie =)))))) [18:14] lol [18:19] * xnox ponders why popularity-contest is configured [18:20] because otherwise all systems with it installed end up with the same "unique" id [18:21] at least for ones installed from a livefs base [18:21] one of several packages that require that kind of handling [18:23] lol [18:23] well, we had all machines using same dbus-id and systemd machine-id for quite some time =) nobody complained [18:24] err, yes they did :) [18:24] Quite. [18:24] though the result of that is fairly subtle [18:25] I vaguely recall the dbus thing being fixed a day or two before release in London. [18:25] May have been the same release sprint where some crazy person was debugging grub by attaching gdb to qemu. [18:25] That was a sad time in my life [18:26] well i recall complaining about machine-id but it was just a grumpy rpm loving intel employee pointing fingers rather than it being a real issue i experienced from. [18:27] Havoc Pennington complained about it, in fact :P [18:27] https://bugs.launchpad.net/ubuntu/+source/livecd-rootfs/+bug/322830 [18:29] and this is the sort of thing that probably keeps re-occuring in e.g. touch, snappy, cloud-image variants, lxd, etc. [18:29] a class of $must-be-unique bugs [22:17] anyone know what/who creates the officially released ISOs, and specifically the mini.isos?