[patch] buffered seeking

[Adam D. Moss]

Since the mailing list is active, this is the latest call
for any response whatsoever to my buffered-seek patch which
I've been repeatedly trying to draw attention to for over eight
months now.  Thanks!
--Adam
==============================================================
Hello!

The attached patch avoids throwing away our precious buffered
data (and saves an actual abstracted seek() and subsequent re-
buffering) when the user does a seek (forward or backwards) that
is short enough to lay within the currently-buffered region.  The
cost is an abstracted tell() per seek on a buffered handle (very
cheap in real life, I hope).

The patch isn't against HEAD, but it should apply okay.

Regards,
--Adam
-- 
Adam D. Moss   . ,,^^   adam at gimp.org   http://www.foxbox.org/   co:3
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: physfs-bufferseek.patch
URL: <http://icculus.org/pipermail/physfs/attachments/20040822/8d0423e7/attachment.diff>