[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]


Jeffrey Watts wrote:
> On 15 Mar 2000, Jochem Huhmann wrote:
> > Debian is at least here in Europe quite common (although I personally
> > do know Redhat better, I've only recently started to use Debian more
> > often) and it is not enough just to have a "standard", it also has to
> > make sense.
> I agree, but a common package format is desirable, and RPM is adequate
> enough technically, and it has the dominant mindshare.  It makes an
> excellent choice for a standard.

I disagree.  Maybe writing to the RPM database should be considered for
standardization, but not endorsing RPMs.  Maybe working within RPM
development to make the accounting widely readable and manageable for
all packaging systems, would be good, but not endorsing RPM.
> > RPM imposes this /usr-bloat, since packages seldom make differences
> > between "system as delivered" and absolutely unimportant stuff.
> I think you are confusing "RPM" and "Red Hat". 

No, I'm not.  I make my own RPM's, I have made them for Red Hat systems,
Mandrake systems, and SuSE.  I know how they work.

RPM by default has one installation path, /usr/bin, that is the
problem.  Move that to /usr/local, and grow RPM itself from the inside
to recognize "base" packages and put them where they belong, that might
be ok. (actually, this can be done in the package .spec file itself,
which may be the ultimate solution, simple to do, and easy to move the
default for the rest of the packages, I may try that on a full install
if I have time someday).

> RPM doesn't force packages
> to be in /usr.   In fact, RPM supports relocatable packages, and a well
> written /etc/rpmrc can overcome many differences between distributions.

Exactly.  But, the issue remains when it comes to packaging that RPM
becoming the de facto standard is the "easy way out," when what really
should be done is probably a definition of what the package database
format should be.  If all packages put their "I installed this" list in
the same format, it would be a non-issue.
org:Hazard Evaluation Laboratories, Inc.;Applications and Installation Engineer
adr:;;1 Deer Park Dr., Suite L;Monmouth Junction;NJ;08852-1921;USA
fn:Robert W. Current

Reply to: