Gunnar Wolf wrote:
> It's been a while since I last made a complete 'svn up' from our base
> repository. We have now 197 packages in main [1], and looking quickly,
> we seem to have 220 packages in our repository [2]. Operations become
> long.
_What_ operations?
> How would you feel towards moving to a structure similar to the pools
> in the Debian archive? We could add one more level to the archive,
> separating according to the module's first letter:
> [...]
> It is a trivial move, but AFAICT it will make faster our 'svn up' or
> 'svn cleanup' commands. And will make life easier as our repository
> continues to grow.
I think most of the operations that are slow today (like, checking out or
updating the whole tree) will remain slow regardless of any reorganiza-
tion of the repository itself.
The real question should be IMO how the operations that actually need to be
performed can be optimized. We shouldn't have to check out the entire
tree or do similar things -- there ought to be more intelligent solutions
for working on our packages.
Joey Hess wrote:
> I wasn't aware that it was possible to check out all pkg-perl packages
> in one checkout. How do you do it?
How to do it should be obvious (svn co svn+ssh://user@svn.debian.org/svn/
pkg-perl/packages), but why would one want to do that in the first place?
If you just wanted to check out all the trunks, you could always do the
following:
$ export url=svn+ssh://user@svn.debian.org/svn/pkg-perl
$ for pkg in $(svn ls $url/packages); do \
svn co $url/packages/$pkg/trunk $pkg; done
Attachment:
pgpzKKlizXiJb.pgp
Description: PGP signature