[00:51] <bigjools> roaksoax: how close to the quantal SRU are we?
[00:54] <roaksoax> bigjools: i uploaded a new verrsion today for quantaul, haven't uploaded a precise one yet
[00:54] <roaksoax> bigjools: but there's 1 more packaging change i need to make
[00:54] <roaksoax> and we should be good
[00:54] <bigjools> roaksoax: \o/
[00:54] <bigjools> we're testing the precise package in earnest
[00:55] <bigjools> a few bugs, but getting there
[00:55] <bigjools> I am upload a new one ATM
[00:55] <bigjools> uploading*
[00:55] <roaksoax> bigjools: right, so the most tedious thing is filing SRU bugs
[00:55] <bigjools> roaksoax: yeah :(
[01:55] <smoser> bigjools, awake ?
[01:55] <bigjools> smoser: of course
[01:55] <smoser> https://docs.djangoproject.com/en/dev/ref/templates/builtins/?from=olddocs
[01:55] <smoser> i'm trying to do:
[01:55] <smoser>  {{node.architecture|default:"nothing"}}
[01:55] <smoser> in a preseed
[01:55] <smoser> and it says
[01:56] <bigjools> we don't use django templates in preseeds
[01:56] <bigjools> http://pythonpaste.org/tempita/
[01:56] <smoser> gah
[01:56] <bigjools> :)
[01:56] <smoser> that would be why the trace said 'tempita' :)
[01:56] <bigjools> haha
[02:01] <smoser> bigjools, so... 2 things iw ant to do.
[02:01] <smoser> it seems that node.distro_release might be None when evaluated (which generally seems wrong to me)
[02:02] <smoser> and second i want the architecture portion of 'node.architecture'
[02:02] <smoser> (which is 'amd64/generic')
[02:02] <smoser> suggestions on how to do that?
[02:02] <bigjools> I don't know why release would be None
[02:03] <bigjools> for arch, you can include a python block
[02:03] <bigjools> so {{py: .... }}
[02:04] <bigjools> you can split on the /
[02:04] <smoser> not None, but renders to empty string
[02:05] <bigjools> hmm
[02:05] <bigjools> can you see if it's set in the DB?
[02:07] <smoser> bigjools, well, from http://localhost:8080/MAAS/nodes/node-628a5c14-24db-11e2-84ac-d4ae527ac125/preseedview/
[02:08] <smoser> i'm using {{node.distro_series}}
[02:08] <smoser> how easiest to check the db ?
[02:08] <bigjools> sudo maas shell
[02:08] <bigjools> from maasserver.models import Node
[02:08] <smoser> thats for an un-allocated node
[02:09] <smoser> one that is "ready"
[02:09] <smoser> for one that is deployed, it does have one
[02:09] <bigjools> oh ok, then that makes sense
[02:09] <bigjools> if it's not deployed it doesn't have any OS
[02:10] <smoser> it should have the default.
[02:10] <bigjools> why?
[02:10] <bigjools> what are you trying to do?
[02:15] <smoser> i'm trying to use it.
[02:15] <smoser> and the rendered view of the preseed is broken
[02:15] <smoser> because it doesn't have a value.
[02:15] <smoser> but there is a default value
[02:15] <smoser> so it should have that
[02:17] <bigjools> I don't think it's broken based on the intended usage
[02:17] <bigjools> what do you need the preseed for, before it's deployed?
[02:18] <smoser> if i dont need it, then why does it show it to me?
[02:18] <bigjools> that's not what I asked :)
[02:18] <smoser> if it weren't nicely displayed in the UI, i'd have no argument.
[02:18] <bigjools> the info presented is correct
[02:19] <bigjools> you could argue that the distro series is not useful though - but it would also be incorrect if it showed the defaut and you then deployed a different series
[02:19] <smoser> well that would clearly be wrong.
[02:20] <smoser> but if it displayed the default, and then deployed the default (because i didn't specifically request something else) that would be right.
[02:20] <smoser> but if i change it, then, well, i changed it, didnt i
[02:21] <smoser> not a terribly large deal, but if this is something that is surfaced to users (which it generally ise) then it is confusing at best to have "" there. when there is (i think) a system wide default.
[02:21] <bigjools> I think your expectations are different to mine here.  That's not to say you're wrong, just different.  It's an interesting problem.
[15:45] <rvba> roaksoax: thanks for the backport of raphael js.  I've tested the package in -proposed.  It fixes the apparmor problem all right.
[15:46] <roaksoax> rvba: awesome then! I can start taking care of filing SRU bugs then
[15:47] <rvba> roaksoax: the package of python-django that Julian has put in the experimental ppa is also fine.
[15:48] <roaksoax> rvba: alright, I need to review that and make sure we can SRU it
[15:48] <roaksoax> rvba: i believe he put a patch which is upstream and not fixed in raring?
[15:48] <rvba> roaksoax: I hope you'll succeed because we need it ;).  Well, technically, there is a last resort option but the SRU would be much better.
[15:49] <rvba> roaksoax: he has backported a new utility method 'prefetch_related' that we need to optimise SQL queries.
[15:50] <rvba> roaksoax: he has backported the fix for 16937.
[15:50] <rvba> roaksoax: https://code.djangoproject.com/ticket/16937
[15:55] <roaksoax> rvba: right, nt is that right but is that part of django in precise?
[15:55] <roaksoax> err
[15:55] <roaksoax> i mean in raring?
[15:55] <rvba> It's part of Django 1.4 which is in Quantal
[15:56] <roaksoax> rvba: oh ok, so it is easy to SRU then
[15:56] <rvba> Cool
[18:53] <hazmat> greetings maas devs