[Catalyst] RE: Configuration Approach

[Christopher Hicks]

On Wed, 26 Oct 2005, Dami Laurent (PJ) wrote:
> Actually the places where I used AppConfig are not (yet!) on Catalyst. 
> However what I will most probably do when I port them to Catalyst is to 
> call AppConfig from MyApp->setup(), and insert the results into 
> $c->config(...) so that config data becomes available to all subsequent 
> requests.

AppConfig is really slick and we've used it in a number of CGI's we've 
shipped to customers before we started using Catalyst extensively.  I 
haven't used AppConfig with Catalyst either.  Some of that is because of 
frustration with trying to help Andy move things along.  He seemed too 
busy to try to deal with somebody "helping out" when I offered to submit 
patches to resolve some of the inconsitancies in AppConfig.  A rewritten 
AppConfig that supported the full body of the AppConfig syntax across all 
the file formats with cleaner Catalyst integratin would be great.  (If 
this is going over your head, AppConfig provides a few forms of 
substitution in the config files, but that substitution doesn't have any 
effect on the command line arguments which would often be quite handy. 
The failure here is that the individual drivers for the pulling stuff in 
from files, etc. are each built with these substitutions embedded in them 
instead of refactoring that to a higher level.)

Even in its current form though its one of the handier Perl modules. 
Integrating the concept of command line arguments, config files, and 
internal program defaults into a signgle framework has been attempted on 
many occasions, but usually the results are much uglier than AppConfig.

-- 
</chris>

"There are four boxes to be used in defense of liberty:
  soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)