[Freeciv-Dev] Re: Documentation, Usability and Development

[Raimar Falke <hawk@xxxxxxxxxxxxxxxxxxxxxxx>]

On Thu, Nov 29, 2001 at 01:20:23PM -0800, Kevin Brown wrote:
> But if you accept the patch, there is a chance (perhaps a good one,
> even) that it won't need to be reverted, and the program will be
> improved as a result.

Aeehmmm we don't play roulette here. We develop software. This should
be deterministic. Every patch which gets applied should be understood
by the maintainer who applies the patch. If this isn't the case the
patch could introduce Trojans for example. Smaller patches are easier
to understand.

> So in the case where you accept the patch, there is some nonzero
> chance that it'll pass muster, and the program will improve as a
> result.  In the case where you reject the patch, there is no
> improvement.
> 
> Sounds to me like accepting the patch is the way to go.

> And one other thing: you can't always get there from here via small
> patches.  

> Some things require large patches and there's no other way around
> it.

Ack.

> Furthermore, a set of small patches isn't necessarily going to be
> higher quality as a whole than a single large patch.  

> Oftentimes, quite the opposite will be true: the large patch will
> have more cohesion because it's written by a single person with a
> very clear idea of what's going on.

Ack. The maintainer should act here as a architect (he has to take a
broader view than the individual patch author).

        Raimar

-- 
 email: rf13@xxxxxxxxxxxxxxxxx
 "Despite all the medical advances of the 20th century, the mortality 
  rate remains unchanged at 1 death per person."