Re: some statistics on using bz2 for packages
In article <19991030233207.I5999@lappy.djj.state.va.us> you wrote:
> bzip2 and libbz2 together are almost 400k (260k for the packages), so yes,
> this is a small issue (I don't see why bzip2 can't be essential anyway,
> since tar already requires it for some functionality).
Tar does not "depend" on either gzip or bzip2. Some completely optional
functionality ( the z and I options that fork gzip or bzip2 into the pipe,
respectively ) is nonsensical without them... but that does not constitute
grounds for declaring a "dependency". Tar is quite useful without either
gzip or bzip2...
Since gzip is essential, nothing needs to declare a dependency on it. Ditto
for tar. Since bzip2 is not essential, I suppose you could argue that the
tar package deserves a 'suggests bzip2', except that since tar is part of
base, this won't get noticed by many people...
There has been discussion "upstream" about gzip incorporating bzip2's
algorithm(s) at some point, but I have mixed feelings about it. If it
happened, it would make the whole question we're debating moot, but it would
also bloat gzip, which is currently in a fairly sweet spot on the "what it
does for what it costs" tradeoff curve.
I don't think allowing bzip2 for .deb's is any more useful or important than
doing a "pentium-optimized" distribution is... which is to say, not very.
Bdale
Reply to: