=== alexpilotti_ is now known as alexpilotti === harlowja_away is now known as harlowja === harlowja is now known as harlowja_away === harlowja_away is now known as harlowja [18:06] smoser u see all my changes?? lol [18:06] i think https://code.launchpad.net/~harlowja/cloud-init/changeable-templates was what u were thinking of [18:08] harlowja, i did see that go by. i haven't really lookd at it. [18:08] np [18:08] i'd like to have that for 14.04, as a way to tell pepole "use this for backwards compat to 16.04" [18:08] but i need a FFE to get it. [18:09] ah [18:12] bah [18:12] I need a brain reboot this morning. [18:12] reboot! [18:13] smoser let me know if i can help, i can say it needs a FFE [18:13] lol [18:14] seanwbruno hey, the whole whatsapp stuff ran on freebsd, there u go :-P [18:14] indeed it did [18:15] a lot of ex-Y! people over there. [18:15] swimming in $$ now, lol [18:16] i loved/love whatsapp. [18:16] but fear where facebook will take it. [18:17] i surely don't wantto be inundated with pics of harlowja everytime he posts on facebook. [18:17] lol [18:17] smoser u know u want it [18:17] lol [18:18] * harlowja should now be quiet or else i get in trouble [18:31] harlowja, one thing in that changeable template... [18:31] sure [18:31] it'd be nice if it worked without cheetah [18:31] hmmm [18:32] as in, it wouldn't work in backwards compat mode [18:32] but it woudn't stck trace [18:32] like it would just use mako? [18:33] well, i guess what i'd really like would be: [18:33] or are u thinking some basic simple impl? [18:33] a.) update all existing templates to "future" [18:33] b.) try loading cheetah, silently swallow error [18:34] c.) if need cheetah becaus eyou find old format, then either use what you imported, or cry loudly at that point. [18:34] gotcha [18:34] sure, smoser makes sense [18:34] 'c' could be improved in the future to say "if i can render this in hack-mode, then i will still be ok" [19:32] smoser: https://code.launchpad.net/~harmw/cirros/cirros-sysinfo [19:32] thoughts? :) [19:35] harmw, i'd rather read /proc/partitions. [19:37] ok [19:38] its all about warpspeed, mr Worf [19:38] :) [19:41] http://paste.ubuntu.com/7029317/ [19:42] nice [19:42] harmw, do we have lsblk in there ? [19:42] yep [19:42] pff lol [19:42] thats easy output as well [19:42] yeah. [19:42] $ lsblk [19:42] NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT [19:42] vda 253:0 0 1G 0 disk [19:42] `-vda1 253:1 0 1011.9M 0 part / [19:43] oh. the other hting to be awware of is running in lxc [19:43] i dont want a bunch of silly error messages that are not relevant. [19:43] lsblk has nice formatting options too [19:44] I see now [19:44] still favor /proc/partitions? [19:44] well, the above is really nice. [19:44] and gets you LABEL easily too [19:44] http://paste.ubuntu.com/7029329/ [19:45] see how i use it in _lsblock there [19:45] i'm not saying we want all of that, but for sure, or even if we want it (since that ascii art is so nice to humans) [19:45] one thing i'd think about (and yes, i overthink just about everything) [19:46] is that i want to enable machines to read this stuff. [19:46] in additionto humans. [19:46] and if i had to choose between humans and machines, i'd probably pick machines. [19:47] uhm, how does that relate to cirros-status? [19:47] (I'm not sure I'm getting it) [19:47] well, what i'm afraid of is that if we put something in like that. [19:47] in any format [19:47] then people will make scripts/programs that expect it in that format [19:48] and thn i'm tied to an api of sorts. [19:48] ah [19:53] harmw, trying to do too many things. [19:53] cirros-status runs to stdout by default, right? [19:53] yea [19:53] k. [19:54] it's ran at the end of booting cirros, pushing all the info down to the console.log [19:54] right. [19:54] so i'm not opposed to having status output mostly human readable at this point. and takign some patch like you have there. [19:55] i thin i'd like lsblk if it put label output. that'd be great. [19:55] like this? [19:55] $ lsblk -d -o NAME,SIZE [19:55] NAME SIZE [19:55] vda 1G [20:12] harlowja: i've submitted the fbsd static network stuff btw [20:15] harmw yup yup, seanwbruno can u check it out? [20:15] if u don't mind boss [20:16] https://code.launchpad.net/~harmw/cloud-init/freebsd-static-networking/+merge/208973 [20:28] harmw, i'd like to see LABEL in there too. [20:35] smoser: pushed [20:40] eh? [20:40] oh [20:40] one sec [20:42] seanwbruno if u don't mind, harmw has been doing some stuff to write out the freebsd network config [20:42] so since u likely know that format, u might be a good reviewer ;) [20:43] seanwbruno v [20:43] *oops [20:43] seanwbruno https://code.launchpad.net/~harmw/cloud-init/freebsd-static-networking/+merge/208973 [20:43] stupid paste key, lol [20:57] harmw, how does it look if you dro pthat '-d' [20:57] ? [21:03] $ lsblk -t -o NAME,SIZE,LABEL [21:03] NAME SIZE LABEL [21:03] vda 1G [21:03] `-vda1 1011.9M [21:07] harmw, so i like this: [21:07] sudo LANG=C lsblk --ascii --list --bytes -o NAME,MAJ:MIN,SIZE,LABEL,MOUNTPOINT [21:07] you can drop the LANG=C i guess and sudo. [21:08] i like the output other than i wish it would print a '.' or somethign for "no label" [21:08] as parsing that is crap [21:31] ah, with sudo I get a label returned [21:31] wicked [21:40] harmw, well, i added LABLE to output there. [21:40] but reading label may require root. if it wasn't using blkid and that data already cached. [21:44] indeed [21:44] I'm changing it to the above [21:44] without LANG=C though [21:45] and yes, something like no-label would've be nice [21:45] perhaps something similar for mountpoints [21:45] for the sake of parsing === harlowja is now known as harlowja_away === harlowja_away is now known as harlowja