Re: Request for comments: CONS specification

[F. Schoenahl]

On Mon, 31 May 2004 03:39:10 +0200
Pierre THIERRY <address@hidden> wrote:

> > Developing a new extensible version of cons is a good idea.

> I agree. ;-)

Not me! :)

Hi everybody 
I am not a contributor to any of the (s)cons projects, and had no time to help 
on other buildsystems, so you might don't care about my opinion - take this as 
a - user - opinion.

I have been playing around with cons, scons and aap to make my projects easier 
to develop. I see that scons and aap, and now cons, converge to the same point. 
I would say if you want to play around and do a new cons for your pleasure, I 
understand, but I really have the feeling you reinvent the wheel, and your 
analysis misses one fundamental point : what is missing in other 
implementations that you will bring with a new cons? I have read these pages 

  http://www.a-a-p.org/tools_script.html
  http://www.a-a-p.org/tools_build.html
  etc... 

Thus I don't know what your aims are, but I think most of the job is done 
already. 

As user, I will first see that 
- perl is not as easy to have running on windows as python, 
- perl is not easy to learn compared to python : the scripting of a   build 
system serves programmer aims only, and python is more familiar   to 
programmers than perl
- scons and aap are already modular - how will you improve the concepts?

So, I would recommend to focus on some brand new direction for a new cons, else 
it may abort very soon!

Good luck,
fred