On Fri, 29 May 2009, Matt Domsch wrote:
On Thu, May 28, 2009 at 11:46:20PM -0500, Mike McGrath wrote:
> I'm all for pruning, lets have a plan for it though. Anyone see any
> reason not to have these up there?
> Should we come up with some test for what does and does not get removed?
I agree. Here's what I am going by:
a) content that has reached the end of life. This includes:
1) pre-release content (Alpha, Beta, snapshots, ...) that have been
superceeded, and are thus no longer useful for testing.
2) EOL releases that we have moved to archive.fp.o
(I'm open to be swayed on this one...)
b) content which has exceedingly limited seeders and downloaders, and
which has little prospect of increasing those numbers, and which is
> 1 year old. The several-years-old videos fall into this
category, with 0-1 seeder, and no significant increase in downloads
in a while (by visual inspection, ~3000 downloads as far back as I
Content which is still considered "current" (e.g. spins of non-EOL
releases) get to stay.
We haven't traditionally hosted spins elsewhere, such as archive.fp.o
or alt.fp.o, so nuking them removes the only method by which someone
could obtain them. Given we're OK on space right now, there's no good
reason to remove spins even of EOL releases where the non-spins got
FYI +1 to removing stale content from torrent1