[MERGE] Redo annotate more simply, using just the public interfaces for VersionedFiles.

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Robert Collins wrote:
| On Wed, 2008-07-02 at 00:14 -0400, Martin Pool wrote:
|> Martin Pool has voted tweak.
|> Status is now: Conditionally approved
|> Comment:
|> So this is certainly simpler, which was nice.
|>
|>  From your previous conversation with John it was not clear to me whether
|> this would be slower or not:
|>
|>>> It will be slower primarily because we can't use the cached matching
|>>> blocks. So we have to re-diff every file.
|>> There is self._extract_matching_blocks. Oh, the optimiser is missing
|>> there on knits - I'll do a patch to reinstate that today. But bundle
|>> generation uses that for make_mpdiffs and is still totally generic.
|> It looks like it might be.  Maybe we should measure this on a regular
|> pack repository and if necessary special-case it for non-stacked
|> repositories?
| ..
|
| Sounds reasonable. John - do you have a bench for annotate?

time bzr annotate --show-ids bzrlib/builtins.py > /dev/null is a pretty
good one.

NEWS isn't bad either. Though if you want real work, there are some good
ones in the mysql tree. (sql/sql_table.cc or sql/ha_ndbcluster.cc seem
pretty good.)

|
|> This smells like it should be a separate named method, maybe on the
|> knit?
|
| Sure.
|
|> Aside from that if you think the performance tradeoff is reasonable it's
|> ok with me.
|
| Well, working >> not working I guess :P. But yes, lets check its not a
| total disaster before merging.
|
| -Rob

Yeah, though if you just switched this off/on based on stacking.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkhrCQEACgkQJdeBCYSNAAObrwCgvTD7N3IeIR4jXhQcV3c5Qeo/
pTcAoJ4PusT+5pTJHXM7zm9FBL/Um9LC
=4cn5
-----END PGP SIGNATURE-----