Eric Biggers wrote:
> Hi,
>
> I did some more testing with zlib-like and multiplicative hashing.  The
> differences for the files I was testing were quite small.  However, the 
> proposed
> hash function, as well as a slightly modified version of it, did come out
> slightly ahead.  So it might as well be used.
>
> I also did a few quick tests with greedy parsing.  In general, it seemed to
> improve performance by 10-20% and increase the size of the compressed files by
> 1-2%.  If the performance improvement is considered more desirable, then I can
> change the patch to use greedy parsing.  For now it's using lazy parsing, like
> it was before but more optimized.

I tend to be more interested in compression rates than
in timings. I mostly use compression for directories which
are rarely updated.

Jean-Pierre



------------------------------------------------------------------------------
_______________________________________________
ntfs-3g-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ntfs-3g-devel

Reply via email to