[Rpm-devel] HELP: Worst ever RPM spec files needed

[Tom 'spot' Callaway]

On Wed, 2006-04-26 at 12:02 -0400, Jeff Johnson wrote:

> Illustrating bad examples might give you an entertaining, but perhaps  
> contentless, presentation.

I certainly do not plan to make the entire presentation that of
illustrating bad examples, merely one minor portion of that
presentation.

> The goal of a build recipe is to produce packages that install. Period.

The goal of C code is to produce programs that build and execute, but
none of the examples provided by the IOCCC (http://www.ioccc.org/) would
be considered good form. 

> Making sure a build recipe will pass muster when reviewed by various  
> committee's with an agenda
> is a political, not an engineering, task. If the package installs,  
> and functions sufficiently well,
> who are you to say what is "RPM best practices".

The goal of having "best practices" for anything is to assist people in
designing clean, maintainable, functional, and repeatable things.
By no means is my way the only way, best is certainly a loaded word, but
"best practices" is a recognizable phrase in the computing industry, and
it is something that I am repeatedly asked for in my job.

If it helps you to sleep better, you may insert "in the humble opinion
of Tom Callaway" after every reference to "best practices".

~spot
-- 
Tom "spot" Callaway: Red Hat Senior Sales Engineer || GPG ID: 93054260
Fedora Extras Steering Committee Member (RPM Standards and Practices)
Aurora Linux Project Leader: http://auroralinux.org
Lemurs, llamas, and sparcs, oh my!