Re: [Gnu-arch-users] [PATCH] arch speedups on big trees

[Tom Lord]

    > From: Chris Mason <address@hidden>

Boring "point-by-point" stuff omitted but I think we can agree about:

1) A locking mechanism for revision library entries.
   The purpose of this is to permit a tla process to take
   away a revlib tree for purposes of turning it into a 
   different revision.    We can talk about the locking protocol
   separately, if you like.

2) "sliding" libraries -- that try to keep the number of revisions in
   a version constant by recycling old revisions (honoring the locking
   protocol)

3) "in tree" libraries.   So that if a special string occurs in my
   library path, that implies a created-on-demand greedy, sparse,
   sliding library in my current project tree (if there is one).

4) A default library path which is nothing more than that special
   string.

5) Remove all code related to pristine trees.   The above features 
   turn that into dead code, more or less.

6) If users balk at the edge cases resulting from having a non-default
   library path that doesn't include the "special string" for in-tree
   libraries, then perhaps add that implicitly to the end of the
   add-path.


I think that will give you 90% of what you want and it also truly
kills pristines (though introduces in-tree libraries).

It's also a nifty plan because it's (finally!) an occaision where we
can probably wind up _removing_ more code from tla than is being added
:-)

add-pristine can, indeed, take a --link option.  Seems pretty obscure
to me but there's no harm in it.  Perhaps that adds up to 95% of what
you want?

And, please give up on "maintaining an inventory", reverse-mapping or
no, in project trees.  It just plain won't work.

Does that make sense to you?   Would you like to work on any of that?


-t