Subject: Re: time spent per package
To: James K. Lowden <>
From: None <>
List: tech-pkg
Date: 03/14/2003 13:19:22
The idea of posing this as a challenge for which people could receive
prizes is an interesting one (though I hope the reference to the
Longitude Prize is not entirely apt).  Thanks.

James K. Lowden writes:
 > Definition of project scope is notoriously slippery.  It's important that
 > the terms of the deal be clear, else its motivational force will be nil.  

So how could a packaging task be appropriately defined?  Elements that
must go into the definition must include:

- acceptance by NetBSD into pkgsrc (though this might give people with
  commit privileges an unfair(?) advantage; is that a problem?)

- functionally useful (though this is hard to define if there are bugs
  in the original software, how much debugging is appropriate during
  packaging, and who is the judge?)

What else should be included, if anything?

Given that packages may take anywhere from 10 minutes to a day or
more, what would be an appropriate prize?  Should it vary with each
package?  Should there be "easy" and "difficult" categories with
different prizes?

Sorry to prolong the discussion, but since this group is likely the
main target audience for such a prize offering, to make it effective I
would really like your input.

Thanks again for your ideas.