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

[John Meinel]

Matthieu Moy wrote:
> Thomas Lord <address@hidden> writes:
>
>
> > So: how about we modify the process so that we maintain a file
> >
> >         ./src/tla/=merged
> >
> > Each time a (log-pruned) version is merged into tla, it's version name
> > will be appended to that file.   No version name should appear twice
> > in that file.
>
>
> Wouldn't it be better to let (a future version of) tla maintain this
> file, and integrate it in the arch protocol instead of some user
> convention?
>
> (I think you had another proposal on how to avoid log pruning, but I
> don't remember exactly what it was)
>
> If I take the example of the Xtla project, on a freshly checked-out
> project, I get this:
>
> $ du -sh
> 8.7M    .
> $ du -sh \{arch}/xtla/
> 7.4M    {arch}/xtla
>
> So, the patch-log takes 85% of the disk space (7 times more than the
> source files !!). If I take only the list of merged patches in a file,
> it goes down to 120Kb, which is around 10% of the source size.
>
> Another option would be to have a .tar.gz file of the patch-log. On
> Xtla, this makes a 400Kb compressed file, but I suppose this will make
> a big performance problem because to answer the question "is changeset
> X already merged?", tla would have to gunzip the whole file.
>
> --
> Matthieu

I think the idea that the patch name gets kept, but you can optionally 
keep the patch-log itself is a good idea. So the '{arch}/=merged' file 
would contain the equivalent of
tla logs --merges

Or possibly with the summaries as well. (This isn't as important, and 
might cause file size to go up again.)

Then when tla wants to check if a patch has already been merged, it can 
look for the patch-log, and secondly look for the entry in =merged.

It might be nicer if =merged was some sort of indexed file format, so 
that lookups could be faster, but greping even a 120K file doesn't seem 
very expensive.

John
=:->

signature.asc
Description: OpenPGP digital signature