pre-2.4 (was Re: buffindexed jumbo patch)

Russ Allbery rra at stanford.edu
Tue Dec 24 06:51:41 UTC 2002


Jeffrey M Vinocur <jeff at litech.org> writes:

> No strong opinion; is there any advantage to putting it off?  (For
> example, if something scheduled for 2.5 will require a complete rebuild,
> then it would make sense to delay it.)

So far, I don't know of any changes that are planned that would require
people do rebuilds for 2.5.  I've thought from time to time about redoing
the format in which tradindexed data is stored on disk so that it was
endian-neutral and portable between operating systems, but that would be a
decent amount of work and would have performance implications for the
code, so I'm not sure it's actually worth it.

If that were done, that would require a rebuild, although there would
probably be a conversion tool.

The things I mostly have in mind for 2.5 are replacing more of the
configuration parsers with the new one and possibly redoing some of the
internal APIs.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>

    Please send questions to the list rather than mailing me directly.
     <http://www.eyrie.org/~eagle/faqs/questions.html> explains why.


More information about the inn-workers mailing list