sm-discuss AT lists.ibiblio.org
Subject: Public SourceMage Discussion List
List archive
Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review)
- From: Seth Alan Woolley <seth AT positivism.org>
- To: sm-discuss AT lists.ibiblio.org
- Subject: Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review)
- Date: Thu, 17 Nov 2005 11:47:46 -0800
I'll try to clarify :)
On Thu, Nov 17, 2005 at 01:13:56PM -0600, Jeremy Blosser wrote:
> Moving this to a ML...
>
> On Nov 17, Seth Alan Woolley [seth AT tautology.org] wrote:
> > Here's my plan:
> >
> > We're using local keys now to publish grimoires now as a matter of
> > unimplemented policy.
>
> We're using locally-stored keys on the other grimoires only as a matter of
> (incidental) fact, though there has been discussion about making it policy
> for other reasons. But right now we're using the same keys we've been
> using and indeed have to keep using until any other (individual) keys can
> make it to the users in a reasonable time.
Accurate.
>
> > So I'll start implementing with stable-rc.
> >
> > I'll make a how-to for people who need to create personal keys for the
> > grimoire based off the process I follow for creating my personal
> > stable-rc key. When we've all had a chance to look at the process, we
> > can immediately republish stable-rc and work on intregrating new keys
> > into the stable and test/games/z-rejected keysets.
>
> I don't think we want to see stable-rc tarballs held up while we come to
> consensus on new policy. We didn't elect to stop doing what we were doing
> before, the server broke. It's fine to make incidental improvements while
> we're rebuilding anyway but we shouldn't make policy change discussions a
> dependency for getting stuff going again.
I wasn't going to make a policy change discussion. I was just going to
"do it".
> I don't know what kind of
> timeline you were thinking there, but IMO anything that takes longer than
> this weekend to get a stable-rc tarball generated again is too long.
>
> We started to make good progress for a while there and had some monthly
> goals going toward 1.0, I hate to see us ending up with months between
> stable ISO and grimoire releases again.
I think there was some confusion of whether or not we wanted Arwed or
myself to handle the stable-rc key generation. He mentioned he was
going to do it, now I'm told I'm going to do it. Since it came from
Eric, I figured that's pretty much the go-ahead-and-get-it-done command.
>
> > By this weekend or next I should be able to tell if I can run the test
> > publishing on my own server as emrys wanted that to be done off-site and
> > pushed to the main site.
>
> I'm not trying to prescribe anything there, just giving my .02. The
> process and policy for how test tarballs get pushed is way more up to Arwed
> IMO. Just keep in mind that we are hosting these from ibiblio and don't
> even have a p4 sandbox there (that I know of).
That's why I was going to be pushing them up instead. This policy
change wasn't really a publicly visible policy change, so I thought it
was something that the people who would manage the keys should be able
to do without having to have it go through the lists, and we have talked
about how it would be better anyways, in public from what I remember
with no negative comments. I considered it non-controversial and
affecting primarily the parties who were going to be doing the
cronned-automatic and non-automatic uploads.
>
> > Ruskie could also push games (instead of me) to the server, in fact,
> > since he does most editing it might be best if he does that.
>
> If we decide we don't care if we use a server-side key to sign
> test/games/z-rejected, I can set up a p4 sandbox on the dedicated box and
> push them to ibiblio from there. If we decide we want to keep the keys on
> developer machines only regardless of grimoire, that doesn't work. I'll
> suggest that's Arwed's decision, but I don't know what Eric's and other
> opinions are.
Well, it's a "release engineering" decision. QA had been doing release
engineering (for grimoire, but not sorcery) in the past, but for the
rebuild, it's gone more toward Arwed, the grimoire lead. I just think
once somebody has said they are going to do it, they need to just do it
(it doesn't take that long). I don't think I have to do this, I don't
think arwed has to do it, I just think one of us has to be ultimately
responisble and we both must be able to (as backup for each other).
As far as the "policy change", I don't really consider it needing public
discussion unless somebody doesn't like it. If we're going to nit-pick
how the rebuilt process happens on every change, it would never have
gotten back up and running.
Regarding the timeline of "this weekend or next", I'm still undergoing a
number of job interviews and I didn't want to commit to the use of my
server as a long-term upload station until I knew what my arrangements
would be for the future (if I had to relocate). It's really two
different issues (maybe I wasn't clear), one was automating test
updates, the other was pushing non-automated updates. The timeline for
stable-rc being generated and on the server will be this weekend at the
absolute latest, as you say. The timeline for test updates, as has been
discussed on the list, I was saying when I could commit to it, but I'm
not opposed to arwed pushing test grimoire updates himself or any other
developer if you don't want me doing it. I was just saying if nobody
else wants to do it what my timeline was for that.
--
Seth Alan Woolley [seth at positivism.org], SPAM/UCE is unauthorized
Quality Assurance Team Leader & Security Team: Source Mage GNU/linux
Linux so advanced, it may as well be magic http://www.sourcemage.org
Key id FDCEE733 = 5302 B414 64C4 6112 3454 E082 99F0 69DC FDCE E733
Attachment:
pgpuI9ipOS9WC.pgp
Description: PGP signature
-
Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review),
Jeremy Blosser, 11/17/2005
-
Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review),
Seth Alan Woolley, 11/17/2005
-
Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review),
Jeremy Blosser (emrys), 11/17/2005
- Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review), Arwed von Merkatz, 11/18/2005
-
Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review),
Jeremy Blosser (emrys), 11/17/2005
-
Re: [SM-Discuss] publishing grimoire tarballs (was: PERFORCE change 69559 by Eric Sandall for review),
Seth Alan Woolley, 11/17/2005
Archive powered by MHonArc 2.6.24.