Matthew Dillon wrote: > :% ls -l > :248643584 Sep 17 00:03 5.0-CURRENT-20020917-JPSNAP.iso > :212988130 Oct 13 10:39 5.0-CURRENT-20020917-JPSNAP.iso.gz > : > :Compression gets rid of about 36MB. > : > :That's 3.4 hours saved on a 28.8K modem download time, overall... > :a 14% reduction in size. > > Well, ok, but on a percentage basis you don't get much out of it. > If someone is downloading via a modem they're probably doing it > overnight anyway.
And that's OK, because we all live in countries that don't charge time or message units for phone calls, right? 8-) 8-). > bzip2 does even worse then gz in this instance, so no magic > there either. Bzip sucks. It was invented to get out from under a patent that is now expired. > -rw-r--r-- 1 dillon wheel 179985801 Oct 13 15:00 bzip2.bz2 (bzip2 -9) > -rw-r--r-- 1 dillon wheel 178963831 Oct 13 14:56 gzip9.gz (gzip -9) > -rw-r--r-- 1 dillon wheel 187006976 Jun 8 00:04 miniinst-RC4-8Jun2002.iso The gzip number I gave was for the default (-6), not -9. The -9 only dropped 600K more out; every little bit helps, though. I gave the default number to make it repeatable, and to give a valid baseline vs. all the compressed data, which was uncompressible because it was otself "-6". A lot of things in the less minimal distributions are more compressible. The ISO for the "live FS image" is *immensely* compressible. -- Terry To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
