Skip to Content.
Sympa Menu

sm-discuss - Re: [SM-Discuss] restricting access to important spells

sm-discuss AT lists.ibiblio.org

Subject: Public SourceMage Discussion List

List archive

Chronological Thread  
  • From: Jeremy Blosser <jblosser-smgl AT firinn.org>
  • To: sm-discuss AT lists.ibiblio.org
  • Subject: Re: [SM-Discuss] restricting access to important spells
  • Date: Thu, 13 Jul 2006 12:13:04 -0500

On Jul 13, Arjan Bouter [abouter AT sourcemage.org] wrote:
> <snip>
> > Let's be clear that it appears we've tracked this change down to someone
> > who was not at all a newbie or someone we don't usually consider
> > responsible with these kind of changes. Lace has *stable* access and has
> > quite legitimate claim to being the/a pam maintainer since he's a security
> > guru, he's no newbie or questionable person. A change requiring only
> > "experienced people" to make changes to things like pam wouldn't have
> > avoided this issue in the least.
> >
> > So let's give him a chance to say what happened and learn from *that*, not
> > a theoretical "an inexperienced person broke something critical" that
> > doesn't appear to have happened.
> >
>
> I didn't say Lace is inexperienced. He wouldn't have stable access
> otherwise.
> I just wanted to get this discussion going and PAM was a good example.
> Lace, if I offended you I hereby publicly apologise.

I doubt anyone is trying to offend someone. Things break. We fix them.

> What I did say was that this kind of problem can be easily avoided by having
> certain spells maintained by people who know more about that particular
> spell.
> In general that is, PAM was just an example.

And my point is that the conclusion doesn't follow from the example. These
changes were made by someone who knows about the spell in question and is
considered plenty qualified to make modifications to it. The things that
would have caught this are testing ahead of committing and code review,
both of which we already practice. If people didn't have/take time to
follow through on those, that's the real issue.

> Also, I started this discussion without the intention of pointing fingers
> at all. I'd like to keep it "in general". Instead of looking for someone
> to blame and discard the original intention.

I'm not trying to blame anyone, this is all meant to be objective. But a
proposed change to solve a problem needs to consider the actual problem
that happened in sufficient detail to determine if the proposal will make
any difference. If a new dev had gone mucking in linux-pam and broken it
you might have a point. The fact that the person that made whatever change
broke it was someone who would still be on any "experienced maintainers
only please" list is pretty core to the discussion.


init has also been mentioned recently I think, and I'll note that the same
applies there: whatever is broken relative to mountroot/WANTS was most
likely part of some changes I made quite some time ago, and which actually
sat in dev for months asking for feedback and never got anything negative
so they went on to test. They've been in test for more months, and now
some people are suddenly having some issues maybe. We don't have a lot of
details yet on even what is broken, but whatever it is that's certainly not
a problem of newbies making changes they don't understand (I don't think I
fall in there anyway ;) or a lack of testing.

Attachment: pgpo1u5B_ZEAt.pgp
Description: PGP signature




Archive powered by MHonArc 2.6.24.

Top of Page