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

Re: Let's stop using CVS for debian.org website



Hi,

On Thu, Nov 24, 2016 at 11:16:30AM +0100, Rhonda D'Vine wrote:
> * Mathieu Parent <math.parent@gmail.com> [2016-11-23 20:13:01 CET]:
> > 2016-11-23 19:48 GMT+01:00 Andreas Tille <andreas@an3as.eu>:
> > > Hi,
> > 
> > Hello,
> > 
> > > On Mon, Nov 21, 2016 at 03:52:47PM +0000, Neil McGovern wrote:
> > [...]
> > > I can ensure that ~260MiB is quite much for my Vietnamese friend Minh
> > > who just started with the translations.  She is doing this on an old
> > > netbook I have given her and her connection is weak as well.  It would
> > > be really good if partial checkouts would be possible.
> > 
> > git clone has "--depth" flag. Setting it to 1 will reduce the
> > bandwidth and used  space.
> 
>  Not by enough that it would make a proper difference.  And AIUI commits
> from such checkouts can't be done (maybe that changed since I
> investigated shallow clones last time).
> 
>  I would rather see a possibility to request and submit single files
> through other means than needing a git checkout at all, through some
> github like edit/patch possibility (which was one of the reasons brought
> forward for the switch anyway).  Anything else I don't see to really
> work out.
> 
>  And yes, I totally agree with that the whole repository burden is an
> obstacle for translators that we need to think about.  If we can offer
> something like a webbased edit possibility (definitely
> authenticated/authorized unlike what DDTSS offers which is a real PITA,
> we need that information for the --author anyway) that could be very
> useful.

What about splitting repo into multiple submodules for each language.

I think that should make each git checkout smaller.

Osamu


Reply to: