INN 2.3.4 out the door

[Russ Allbery]

Alex Kiernan <alexk at demon.net> writes:

> some changes to tradindexed to check for ESTALE at judicious points when
> you're nfs reader (and the new tradindexed code is so nicely structured,
> that the changes are very non-invasive)

Oh, excellent.  I'm glad to hear that; I tried to structure it well,
although the lock-free reading requires a few layering violations between
the index and data code.

Those changes sound like they'd be reasonable to put in before INN 2.5.

> the binary search code to select the start point for newnews (which I
> was planning on waiting for discussion on changes to the overview API
> before I tried to commit it - right now its implemented as a magic
> global variable which nnrpd and tdx-data are intimate over)

Sounds like a good idea.

> some changes to the nnrpd msg id cache code to make it more effective
> for clients which issue a string of HEADs followed by a string of BODYs

> Do you have any dates in mind for a release?

Since we're doing some additional cleanups, it won't be this month, but
I'm hoping the end of February or the beginning of March.  We may be able
to get it out sooner than that, but generally one can't go wrong expecting
releases to take longer than one hoped.

-- 
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.