This is the mail archive of the
cygwin-apps
mailing list for the Cygwin project.
Re: Deleting old versions of packages
- From: Christopher Faylor <cgf-use-the-mailinglist-please at cygwin dot com>
- To: cygwin-apps at cygwin dot com
- Date: Mon, 28 Oct 2013 20:03:36 -0400
- Subject: Re: Deleting old versions of packages
- Authentication-results: sourceware.org; auth=none
- References: <5266FCC6 dot 7030405 at tiscali dot co dot uk> <20131023021452 dot GC8111 at ednor dot casa dot cgf dot cx> <20131028193452 dot GA5968 at ednor dot casa dot cgf dot cx> <87ppqpqh7m dot fsf at Rainer dot invalid>
- Reply-to: cygwin-apps at cygwin dot com
On Mon, Oct 28, 2013 at 09:21:49PM +0100, Achim Gratz wrote:
>Christopher Faylor writes:
>> I don't like the clunkiness of creating anti-files in the sftp directory
>> to force deletions but I can't think of any other way to do this short
>> of actually making it look like the entire release directory is in the
>> user's sftp area.
>
>What we've got is what I'd call a transactional procedure for
>maintaining our packages. The positive part is implemented: if the
>transaction gets committed (!ready) then any files get transferred to
>the release directory.
>
>Removing files is difficult for two reasons: a) we don't easily know the
>current state of the release directory and b) setup.hint may allow
>multiple states depending on what is already in the release directory.
>
>Keeping the deletion part a manual operation is possible, but probably
>goes against your intentions and a !delete file is iffy. So my proposal
>would be that the current state of the release directory is mirrored
>with (zero-sized?) files in the sftp area. A maintainer could then just
>delete any of those placeholders and the corresponding file would be
>deleted from the release directory. Crucially, the upload and the
>deletion would not necessarily need to happen with the same transaction.
So your proposal is basically to do "I can't think of any way to do this
short of..."
That's not a very inspired proposal given the fact that I already
mentioned the idea and implied that I wasn't keen on doing it that way.
cgf