Re: ftp-masters
On Thu, Jul 31, 2008 at 00:33, Ben Finney <ben+debian@benfinney.id.au> wrote:
> "Sandro Tosi" <matrixhasu@gmail.com> writes:
>
>> On Wed, Jul 30, 2008 at 15:11, Ben Finney <ben+debian@benfinney.id.au> wrote:
>> > [T]here's no good reason I can see to re-use a release number
>> > after it's uploaded.
>>
>> Because it never reach the archive, so it's better to keep the same
>> version for a REJECTED package
>
> Why is it better? What benefit does it bring to retain the same
> version string, including release number, for two different packages
> that have both been uploaded by the maintainer?
>
>> jumping revision is useless.
>
> It's useful because it clearly distinguishes two different package
> uploads by the maintainer. By definition, they are at least different
> enough that one was REJECTED and the later one hopes not to meet the
> same fate.
>
> That keeps discussions about which one is which easy, and it even
> makes it trivial to distinguish in the package changelog, by
> separating the two releases and describing what was done for the later
> one to make it different.
>
> That's a tangible benefit, not "useless". What is the benefit in not
> doing that?
Because it doesn't add any information for the end users, if not
confusion: it seems that 2 different version reached the archive,
since the users will see 2 different entries in the changelog (for
example using apt-listchanges) when indeed there is only 1 package
uploaded after the one previou in the archive.
So, simply changing the current REJECTED changelog entry
adding/modifing/removing items in it it's clearer.
Ah, did I forgot to remember that "Our priorities are our users and
free software"?
Sandro
--
Sandro Tosi (aka morph, Morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi
Reply to: