[Flac-dev] the road to 1.0...

[Josh Coalson]

On my lists of things to do for 1.0 were 1) improve seeking; and 2)
speed up both encoding and decoding.  Seeking seems better now (I
added the SEEKTABLE and tweaked the search algorithm).

On the way, one of my encoding experiments worked.  By taking
advantage of a relatively unused area in the Rice parameter
space, I added an escape code for switching to flat encoding
within a partition.  Technically this is a format change
though you might never notice since it is very unlikely
for the Rice parameter that is now the escape code to have
been used before.   It gives a slight but almost free extra
bit of compression.

With that out of the way, I am concentrating on speed.  There is
not a whole lot more I can do in the encoder in C, so I am thinking
about some hand-coded assembly in a few places.  The decoder still
has room for improvement with just C.

Given the nature of the changes I may have one more beta (0.10)
unless enough people are compiling out of CVS to give me some
confidence about the results.

Josh


__________________________________________________
Do You Yahoo!?
Get email at your own domain with Yahoo! Mail. 
http://personal.mail.yahoo.com/