[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Back to basic (was: Re: *seconded* Re: Bits (Nybbles?) from the Vancouver release team meeting

On Tue, Mar 15, 2005 at 11:28:29PM +0100, Thijs Kinkhorst wrote:
> On Tue, March 15, 2005 22:50, Stephen Frost said:
> > I'm not sure that we've entirely missed the point as much as we like to
> > think there's a better solution than dropping all but 4 archs.
> Here's where things go wrong in this discussion. I think the original
> proposal was (in retrospect) worded too definitive and too detailed. Many
> people are falling over the fact that the proposal asserts that only 4
> archs remain and most people keep repeating that "we are dropping 8 archs"
> like that's the key of the proposal .

Exactly that. Here's my thoughts of the proposal:

The proposal is a nicely crafted troll, designed to make an end once
and for all to those nice "drop archs to speed releasing!"
threads. With clear-cut rules when an arch drops, there is no
speculation, and everyone knows when that happens.

Now, what will happen is that Etch will release with 11 archs. I bet
one or two might drop, but there's one or two to replace those. Here's
the proof, that in my mind covers every possibility.

1) There are enough resources for the arch

   A) Vancouver meeting was a devious conspiracy

      The porters play against the conspiracy by their own rules. With
      enough resources, they have 100% of packages built with minimal
      amount of buildds, and they have the manpower to spew those
      patches to packages that FTBFS.

   B) Vancouver meeting was not a devious conspiracy

      Non-issue, trivial, etc. The release team sees that the arch can
      release, and that happens.

2) There are not enough resources for the arch

   A trivial case. Not enough resources => shouldn't release. But
   that's not the case for everyone's pet arch, from what I've been
   reading on the list. So much posts about "suddenly dropping
   so-and-so many users and machines and whatnot" cannot mean
   anything else.

Petri Latvala

Attachment: signature.asc
Description: Digital signature

Reply to: