[Gnu-arch-users] Re: Potential flaw in patch-log pruning in proposal

[Stefan Monnier]

> E.g., you'd have files like `patch-10-100.tar.gz' which would contain
> patch-10 through patch-100.  Then the check for patch existance would be

Indeed, and this was discussed here at length already.
I think it's a great idea since it keeps the data around just as before
but solves the "too many inodes" and "too much disk space" problems (I care
more about the inodes since they seriously impact the performance of
every `inventory').


        Stefan


PS: It seemed that bzip2 compresses patch logs significantly better
    than gzip.