[26263] in Source-Commits
Re: /svn/athena r25476 - trunk/debathena/scripts
daemon@ATHENA.MIT.EDU (Jonathan Reed)
Mon May 7 15:39:25 2012
Mime-Version: 1.0 (Apple Message framework v1084)
Content-Type: text/plain; charset=us-ascii
From: Jonathan Reed <jdreed@MIT.EDU>
In-Reply-To: <alpine.DEB.2.00.1205071225160.19587@dr-wily.mit.edu>
Date: Mon, 7 May 2012 15:39:22 -0400
Cc: source-commits@MIT.EDU
Message-Id: <1BC924D9-7D75-496F-BF7C-04D4429E1201@mit.edu>
To: Geoffrey Thomas <geofft@MIT.EDU>
Content-Transfer-Encoding: 8bit
> Speaking of which, to be clear, the format here is space-delimited release names? (We should be more clever about releases whose names are substrings of other releases, at the very least so we can do things like precise-backports if we ever want to support something like that, but that's not a regression from the current code.)
Correct. I did not change the format from the nobuild files. I agree we should support that in the future, but this should not block on that.
>> + if echo $i | grep -q '\.dsc$'; then
>> + dscfile=$i
>> + fi
>> +done
>
> This seems like a kind of strange algorithm. Why are we not shifting out the other arguments?
Because then I get to shift out -A and re-add it? This algorithm picks the last .dsc file specified on the command line. It's not clear to me that shifting out the arguments, and then re-adding them is noticeably better.
-Jon
