For 1.5.9: attachment counting for index display
http://home.uchicago.edu/~dgc/sw/mutt/patch-1.5.8.dgc.attach.5
Thomas, I know you've previously stated that you won't include this
patch, citing (a) concerns about the performance hit to the "extra"
MIME tree parse, and (b) memory concumption with the parsed data
accumulation. Since a MIME tree parse is crucial to the feature, I
can't really argue with that. However, it performs fine to me, and the
tree parse doesn't occur if you're not using the feature. (That is,
if you have no %V elements in $index_format.) So if there's any harm
done, it's only to people who expressly sign up for it. Moreover, the
cached tree-parsing data is tossed when this patch no longer needs it,
so memory usage remains tame. This impedes performance of the ~V search
function and of page display when you use %V, but it saves the memory.
I've been asked to try again, though -- see another thread -- so please
review this for consideration.
This patch has been stable (at version 5) since March 14, 2002, for
mutt-1.3.28. All updates since then have just been refreshes for new
mutt releases, with no changes to the implementation itself. I have no
firm plans to make significant changes to it, unless doing so puts it
closer to CVS.
I've considered modifying mutt's envelope structure to retain the number
of attachments, so that I can still be conservative about memory (see
above) while not damaging performance. Then a MIME parse would only be
needed (a) for new messages, or (b) when the parameters of attachment
counting change. If that makes it more appealing for CVS by addressing
more performance concerns, I'll do it.
--
-D. dgc@xxxxxxxxxxxx NSIT::ENSS
"So now, less than five years later, you can go up on a steep hill...
and with the right kind of eyes you can almost see the high-water
mark -- the place where the wave finally broke and rolled back." -HST