(as in chef itself works as documented. but third-party modules (recipes) vary in quality, are often badly documented, and don't quite work as you expect or want)
ocharles: is that the git '/path/to/checkout' do ... end command?
ocharles: this recipe is slowly turning into a bash script :S
Freso
ocharles: It is. (And they should replace upstart with systemd *cough*)
ocharles: I'm curious if we'd have enough users voting/chiming in on stuff for it to be useful though.
But I guess we won't know unless we try. :)
It looks neat either way.
It might also work well for e.g. STYLE changes.
warp
Freso: indeed, IdeaTorrent looks nice.
MiX-MaN joined the channel
djce joined the channel
Freso
Oh, gee.
It's Drupal! \o/
MiX-MaN joined the channel
MiX-MaN joined the channel
nikki
Freso: I'm having trouble imagining how we would fit the various parts of style stuff into ideatorrent. e.g. lots of style things need a wiki page
djce joined the channel
Freso
nikki: I'm not saying the Style documents should go into an IdeaTorrent, but IdeaTorrent might be a better approach than maillist and/or ticket system for discussing changes.
You could still make User:-space wiki pages to showcase proposed changes and link them from IdeaTorrent.
nikki
the thing is that you can make a page, get votes on a change and then change the page that has the votes
then it looks like people have voted for something they didn't :/
Freso
nikki: It's no different than now.
nikki
yes it is. we don't use voting right now
Freso
nikki: +1/-1? :)
nikki
that's not voting though.
Freso
We strive for consensus.
nikki
+1 means the proposal has a seconder
Freso
I don't see why we'd strive any less for a consensus just because we use a different system.
nikki
because getting everyone to agree doesn't work
if it's going to have voting, the natural thing is that you'd aim for >x% in favour, instead of 100% in favour
unless you only use ideatorrent for developing the proposal and then use a separate system for actually proposing the proposal, but that gets complicated and confusing
Freso
I still don't really see the issue, I think. If there are a hundred that say "Yes, the proposal is awesome! Let's do it!" and one that says "Nah, my brother thinks this sounds silly.", we'll go ahead. No different from now. If 12 people say "Great idea! Let's get to it!" and 4 people say "This won't work because of X, Y, Z. Let's rework the proposal or do something else first.", then it will be
sent back to the proposer. Like now.
nikki
Freso: that's not how it works right now though. it's more that there are an unknown number of people who sit silently agreeing while 1 person says "no, I think it won't work"
if you're "lucky", you get one person arguing for and one against and the proposal still goes nowhere
essentially if someone disagrees, you argue with them until they withdraw their veto or you need one of the style leaders to override it
which is not the same as 95% think it's a good idea, so we'll accept it
Freso
If we decide that we strive for consensus, 2% disagreeing with a change with reasonable arguments (ie., not "my brother thinks this sounds silly") can hold a proposal back.
warp
it's called building consensus. it requires level 5 Courage, Diligence, Understanding, and Expression.
(sorry, too much persona 4 for me)
Freso
nikki: If a proposal now gets 95 people giving a +1, three commenting without casting "votes"/"seconding", and two people -2 giving technical arguments as to why a giving proposal won't actually work. What will you do?
nikki
Freso: we don't have 95 people giving a +1
Freso
nikki: I'm not saying we do.
nikki
a proposal needs a +1, most people are lazy and don't respond after that, unless they're really really keen on the proposal
ignoring people trying to do schema changes through style (that's not going to work, veto or not), style stuff largely does not have things like that. it's all about whether people agree that things should be standardised in such and such a way or whether we should add such and such a relationship
MiX-MaN joined the channel
MiX-MaN joined the channel
ocharles
warp: it's sounding like you're trying to use chef for deployment
but it's not really designed for deployment
it's for configuring/provisioning
at least, is my understanding
djce joined the channel
hawke joined the channel
hawke_1 joined the channel
MiX-MaN joined the channel
djce joined the channel
warp
ocharles: yeah, you may be right.
ocharles: so I should replace some of this with fabric?
ruaok joined the channel
ruaok joined the channel
ianmcorvidae
meh, you can't pass the recording to use for a track when doing RE seeding?
warp
ianmcorvidae: you can't?
ianmcorvidae: should be possible.
reosarevok
There are a lot of things that should be possible :)
warp
ianmcorvidae: this whole seeding business is a hack ofcourse. mediums.* is the tracklist page, rec_mediums.* is the recording assocations page.
ianmcorvidae
if it's possible it's not documented
nothing about rec_mediums at all
warp
ianmcorvidae: so it'll be something like rec_mediums.0.associations.0.gid = recording mbid.
ruaok: Also, I'm around for a bit if you want to poke me about #musicbottle :)
ianmcorvidae
it's never reasonable to run flash or java except with explicit approval, and often not even then :P
ianmcorvidae <3 chrome's 'click to run' option
Freso <3 NoScript
warp <3 default linux firefox?
kepstin-work uses flashblock
kepstin-work
er, but that's java
mb-chat-logger joined the channel
Ben\Sput joined the channel
djce joined the channel
Prophet5 joined the channel
Mineo
fun fact: I actually tried to file a bug against openjdk earlier todays but the bug submission form needs longer to load than a 100 disc classical release on musicbrainz :/