alfredo at uch
Sep 5, 2003, 3:10 AM
Post #4 of 8
(919 views)
Permalink
|
El vie, 05-09-2003 a las 06:49, Antony Stone escribió:
> On Friday 05 September 2003 10:34 am, Adrian von Bidder wrote:
>
> > On Friday 05 September 2003 00:04, Alfredo Rezinovsky wrote:
> > > The virus definition database update downloads a big file (ok, itīs no so
> > > big) to update a file that is almost the same. I think it will be good to
> > > have a non compressed virus definition database accesible via rsync
> > > protocol.
> >
> > Just FYI (and I have no idea how good or whether at all it works): gzip
> > supports a an output format that should work well with rsync.
>
> rsync contains its own compression algorithm which does an excellent job on
> its own.
>
rsync transfers only the differences between two files compressing them
(optional)
(file2-file2).gz will be small
If the database are already compressed there are very low probabilities
on data coincidence, thus the differences will be almost the whole file.
file2.gz-file1.gz will be BIG
If you look at the kernel archive, for example, patch you will allways
see a compressed diff file, not a diff of the compressed files.
The only need is to have clamav to support non compressed virus
definition databse. And avaiability of rsync server with those files.
--
Alfrenovsky
|
