[Distutils] setuptools controversy (was Re: Prototype setuptools-specific PyPI index.)

[Phillip J. Eby]

At 05:56 PM 7/23/2007 -0400, Jim Fulton wrote:

>On Jul 23, 2007, at 5:08 PM, Fred Drake wrote:
>
>>On 7/23/07, Phillip J. Eby <pje at telecommunity.com> wrote:
>...
>>>Usually, people think they can get away with making "simpler"
>>>versions of Zope because they don't understand the full range of
>>>requirements which it is intended to meet.  The situation with
>>>setuptools is much the same, except that sometimes now it's the Zope
>>>folks making the accusations of too many features, excess complexity,
>>>evil, etc.  :)
>
>I can't find your message, so I'm responding via Fred's.
>
>I hope none of the "Zope folks" have said setuptools is evil.

I was referring to your comments regarding case-insensitivity being 
considered harmful.


>Speaking for myself, I think it would benefit from being simpler.

Good lord yes, and *I* of course would have benefited more than most, 
had that been possible.  (Since I would've needed to do a lot less work.)


>It is a lot more complex that I suspect you realize.

Right back at you.  :)  Except that in my case, the complexity I'm 
aware of is in the complexity of the requirements that had to be met 
to reach the goal of Total World Domination.

If those requirements hadn't been met at the start, it would never 
have gotten so far as to allow for the luxury of discussing whether 
perhaps some of those requirements could have been overlooked or 
handled differently.  (E.g., the recent progress on 
setuptools-specific indexing.)


>Of course, there are things I might do differently.

Again, right back at you.  :)  If I had the luxury of being able to 
build up some kind of community consensus to use whatever was "pure" 
(as opposed to practical), I most likely would've done a lot of 
things differently.

I certainly don't mind disagreement, in any case.  It's just that 
somebody's got to weigh the cost and benefits of a decision, viewed 
over the entire scope of setuptools' many audiences, which sometimes 
means I'm going to Pronounce against something even when I agree with 
your view (within the scope of your perspective).  Other times, 
simple backward-compatibility or other practical issues will rule out 
changing things that even I might myself consider to be a mistake in 
retrospect.