[darcs-users] soc progress 5

[Trent W. Buck]

Petr Rockai <me at mornfall.net> writes:

> I wanted to add [magic to] the index, so we could quickly identify an
> old (or future) version and discard it immediately (triggering a full
> index rebuild). [...] endianity conversion for the on-disk
> format. This can be done together with the magic word.

Suppose Alice and Bob both share a repository (that is, they actually
make changes to the same working tree) accessed over NFS.  Alice uses a
PowerPC system and Bob uses and AMD64 system.  Will they "fight" for the
index file, each time one or the other runs a Darcs command causing the
index to be regenerated?

I readily admit that this is a silly case, because Alice and Bob
shouldn't both be *editing* the same repository.  They should make
branches and then push their work back into the main repository.

> Finally, for the post-2.3 bits. In darcs-hs, I have bitten the bullet and
> flipped all unrecorded-state (basically pristine -> working copy diffing)
> machinery over to Gorsvet's unrecordedState (implemented using Index and
> hashed-storage). This might have introduced some performance regressions,
> sadly.

Are you gonna check, by pushing it through the performance test harness? :-)