Re: proposal for improvements in indexing

[Milan Zamazal]

>>>>> "DB" == Dirk Bergstrom <address@hidden> writes:

    DB> what's really needed is to move to a better index format, like
    DB> gdbm.  i *think* that would allow us to get rid of the full
    DB> database lock (for most cases).  we would only need PR locks,
    DB> and the gdbm_open() call (which, presumably, does not allow two
    DB> writers on the same dbm file).

I'm not familiar with gdbm, but if it provides the useful features, I've
nothing against using it.  And it should be always better than the
current proprietary binary index format.

I think GNATS should generally move to using more available libraries.
We'll need one for MIME handling, I'm also going to look at glib, since
it could simplify the current internal data structure mess which is a
source of pointer bugs, etc.  Some people might object this would
complicate GNATS installation on proprietary systems, where such
libraries might not be commonly available, but I think this argument
doesn't outweigh the development troubles nowadays.

    DB> of course, that's real work.  but i think it should be
    DB> considered for 4.1 or 4.2...

OK, I'm taking a note about it as a post-4.0 suggestion.

Regards,

Milan Zamazal

-- 
  _/_\_/_  o     _\_/_\_  o     _/_\_/_  o     _\_/_\_  o    BEWARE!
-<_|_|_|_><--  -<_|_|_|_><--  -<_|_|_|_><--  -<_|_|_|_><--   *Bugs* are
   / \ /   o      \ / \   o      / \ /   o      \ / \   o     approaching!