On Sat, 3 Nov 2001, Karl J. Runge wrote:
> Why don't you guys try bzip2(1).

  In my experience, bzip2 is *much* slower than gzip.  In a test against a
16 MB segment of the kernel source tarchive, it yielded a compressed output
19% smaller than gzip, but took 73% more CPU time.  In cases that are
pathologically bad for bzip2 (lots of repeated bytes), the difference is
more like 10,000% times slower.  No, that is not a typo -- one hundred times
slower.  Now, granted, with today's CPU speeds, one may have the performance
to spare, but I am still worried about the tape drive stalling.

> The b stands for "block compression"

  It stands for "block-sorting", actually.  :-)

> bzip2recover evidently aids in this.

  From what I understand, bzip2recover simply reads the input, looks for
block boundaries, and creates output files for each recognizable block.  I
suspect that would not work well on a 300 gigabyte tape set.  :-)

-- 
Ben Scott <[EMAIL PROTECTED]>
| The opinions expressed in this message are those of the author and do not |
| necessarily represent the views or policy of any other person, entity or  |
| organization.  All information is provided without warranty of any kind.  |


*****************************************************************
To unsubscribe from this list, send mail to [EMAIL PROTECTED]
with the text 'unsubscribe gnhlug' in the message body.
*****************************************************************

Reply via email to