[Debichem-devel] What about SVN to Git migration
Andreas Tille
andreas at an3as.eu
Tue Dec 12 04:59:37 UTC 2017
Hi Michael,
On Mon, Dec 11, 2017 at 10:57:39PM +0100, Michael Banck wrote:
> > I would also like to migrate prospective packages which are mentioned
> > in the Debian Med task list. I spotted molekel[1] which I would also
> > like to move to Git ... to ensure that the status will be conserved
> > after the Alioth move.
>
> Which version of it?
When I migrated Debian Med prospective packages I imported the last
upstream version I was able to find.
> Molekel seems rather dead at this point; the
> author moved it to github, but hasn't done any further development since
> then in the last three years.
Thanks for the hint to Github.
> The 4.x version was kinda unique in its abilities, but isn't fully open
> source IIRC.
>
> The 5.x version is (I think) fully open source but was not on-par with
> 4.x in terms of features and stability last I looked.
>
> And there hasn't been any major development in the last 10 years or so.
>
> But that shouldn't stop you from converting it I guess.
Sometimes the motivation to migrate prospective packages from SVN to Git
was just to keep the upstream source available on the net before it
vanished at all. I do not want to spent much time into it - just saving
previous developer work remaining accessible, sometimes conserving
upstream work and keeping our tasks pages complete.
> > I repeat that I have not yet seen any reliable commitment to keep SVN
> > working and I'm really concerned about this move.
>
> OTOH, I haven't heard anything about going forward with gitlab for a
> while, so not sure how urgent this is.
Well, at some point in time SVN will vanish and at some point in time we
need to migrate. I have decided that this point in time for the
packages I'm involved in is now and since these are a lot I will try to
make sure that I do not get in a hurry once the forced point in time
comes closer. I would not rely on the fact that any migration (also the
GitLab one) takes usually longer than expected.
Kind regards
Andreas.
--
http://fam-tille.de
More information about the Debichem-devel
mailing list