Re: How about using bzip2 as the standard *.deb compression format?

[Paul Slootman <paul@wau.mis.ah.nl>]

On Mon 05 Oct 1998, Paul Slootman wrote:
> On Sun 04 Oct 1998, James Troup wrote:
> > Joseph Carter <knghtbrd@debian.org> writes:
> > 
> > > Old/slow/lomem machines can't properly compile X or Mozilla anyway.
> > 
> > Bzzt.  I've compiled xfree86 for Debian/m68k on a 386/25 equivalent
> > with only 14Mb (don't ask) of memory several times.  Took 5 days,
> 
> 14MB isn't that lomem...

BTW, I just had a look at the new bzip2 version. This are the relevant
lines from top while running 'bz2cat <linux-2.1.124.tar.bz2 | bzip2 >x':

  PID USER     PRI  NI  SIZE  RSS SHARE STAT  LIB %CPU %MEM   TIME COMMAND
30413 paul      20   0  6820 6820   288 R       0 72.0 10.7   3:55 bzip2
30412 paul       0   0  3928 3928   288 S       0 23.5  6.2   0:48 bz2cat

Decompressing doesn't take that much time nor memory, if I compare it
for example with my X server:

  265 root       0   0 15028  11M  1004 S       0  0.5 19.0 542:35 XF86_SVGA

Of course, 4MB is still quite a lot, but I guess that should be doable
for just about everyone. Alternatively, from the manpage:

       Compression  and decompression requirements, in bytes, can
       be estimated as:

             Compression:   400k + ( 7 x block size )

             Decompression: 100k + ( 4 x block size ), or
                            100k + ( 2.5 x block size )

and

       For files compressed with the  default  900k  block  size,
       bunzip2  will require about 3700 kbytes to decompress.  To
       support decompression of any file on a 4 megabyte machine,
       bunzip2  has  an  option to decompress using approximately
       half this amount of memory, about 2300 kbytes.  Decompres­
       sion  speed  is also halved, so you should use this option
       only where necessary.  The relevant flag is -s.

So, I think that some experimentation of what block sizes and flags to
use may be in order.  Besides, as decompression is done internally by
dpkg (right?), dpkg could check the memory available on the machine
and decide which decompression algorithm to use.

In short, I don't really think that there are compelling arguments
_not_ to consider bzip2.

And yes, x ended up identical to linux-2.1.124.tar.bz2 in case you're
wondering :-)


Paul Slootman
-- 
home: paul@wurtel.demon.nl | work: paul@murphy.nl | debian: paul@debian.org
http://www.wurtel.demon.nl | Murphy Software,   Enschede,   the Netherlands