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

Re: woody is getting worse...



On Wed, Oct 17, 2001 at 02:55:01AM -0500, Branden Robinson wrote:
> On Wed, Oct 17, 2001 at 05:29:41PM +1000, Anthony Towns wrote:
> > Well, the alternative is of course that you knew about the bug, and knew
> > that it'd break for everyone as soon as they installed your package,
> > and just couldn't be bothered fixing it.
> "The alternative"?  Your world is either-or?  You never need switch
> statements or "else if"s because your world is all "if-then-else"?

Yes, *the* alternative. Either your testing was inadequate because it
didn't show you the bug in the package you uploaded; or it did show you
the bug and you ignored it.

> What happened was, I tested the new versions of the conffiles on my home
> box as a guinea pig test in response to the initial filing of 113878,
> and lost track of which version of the conffiles I had installed (the
> ones from the package, or the ones I had used for testing).

Which indicates that your testing was inadequate, since you weren't testing
what you uploaded.

But, again, that's not a huge problem. You can't test everything
perfectly, and if making mistakes was a crime, well, we wouldn't have
any developers left anyway.

> I know it's a lot to ask people to read the logs of #113878.

Eh? "Thanks for the report", "Oh, wait a minute, REALSTARTUP is still
broken", "Read -devel". How was someone meant to deduce that you were
testing the wrong conffiles from those logs?

> Fortunately, a lack of concern that isn't warranted in this case, but I
> understand people's need to attribute this incident to malice rather
> than stupidity (or confusion, if one wishes to be charitable, which --
> let's admit -- no one apparently does).

I'd've judged the original bug getting into unstable as bad luck, and
the bug report getting a "normal" severity as carelessness.

I wouldn't expect the former to be particularly avoidable in general
(although I'm sure you're competent and professional enough to ensure
that you won't make the same mistake again), so it doesn't really
interest me. I would expect the latter to be completely avoidable though,
especially in packages as standard as X.

Cheers,
aj

-- 
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.

 "Security here. Yes, maam. Yes. Groucho glasses. Yes, we're on it.
   C'mon, guys. Somebody gave an aardvark a nose-cut: somebody who
    can't deal with deconstructionist humor. Code Blue."
		-- Mike Hoye,
		      see http://azure.humbug.org.au/~aj/armadillos.txt

Attachment: pgpyC38xwYtBN.pgp
Description: PGP signature


Reply to: