Re: Translation statistics: new "diffstat" column
On Fri, Dec 23, 2011 at 07:25:22PM -0400, David Prévot wrote:
> [ Resent: I forgot to CC you this time ]
> Hi Marcin,
> Le 23/12/2011 18:20, Marcin Owsiany a écrit :
> > On Fri, Dec 23, 2011 at 06:13:45PM -0400, David Prévot wrote:
> >> interesting data couldn't be shipped directly at build time. They will
> >> clients, the source data will be picked only once (well, six time a day
> >> to be honest, but only from www-master).
> > Yes, there are obvious advantages to getting the data at build time, the
> > reason I did not go that way was that I was afraid of the potential of
> > overloading the CVS server by the website build process. I don't know
> > how much risk there is in fact, though.
> We could keep a cache of the diff, so it will definitely be an
Right, except for the initial run.
Another thing I thought of now is that the latency multiplied by the
multiple separate CVS server queries might be a show-stopper for the
build... That depends on how many translations are out of date overall,
maybe I'm just worrying too much...
Marcin Owsiany <email@example.com> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216 FE67 DA2D 0ACA FC5E 3F75 D6F6 3A0D 8AA0 60F4 1216