Re: FW: Hmmm.... future of cons?

[Steven Knight]

> Try swallowing a python, you might get swallowed by it. What you write make
> sense but I am not part of the Scons team. I'd rather add a perl (lua
> whatever) front end to Scons, that would fix my problem at least. I can get
> myself to do that if possible. I need to know more about SCons. This takes
> me to the next question. How stable, complete is SCons as of today?

Scons stability is quite good, IMHO.  The bug rate is pretty low (due
at least in part to the small install base of a new tool); we've been
fixing bugs quickly; we're disciplined about adding to the test base to
guard against regressions.

Completeness:  The biggest Cons things missing from SCons are Repository
support, UseCache support, and SIGNATURE configurability.  Options
for debugging your build are missing or sketchy.  Error messages are
insufficient.

SCons has a bunch of other stuff not in Cons, so I expect that SCons
will pull ahead once the above features are added, probably within
two-three months.

> Some times ago we had a similar discussion (in another context) where it
> seemed that sensitivity to perl-python, Scons not wanting to walk on cons
> toes, stopped the discussion. If I am to use SCons I'd like Scons
> developer(s) to speak out and ,please, convince me (I guess many would like
> that). I don't see why we couldn't discuss Scons (or moving to Scons)
> openly.

If people want to discuss this sort of thing, I'm okay with it.  Does
the above description of how "complete" SCons is provide enough
information?  What else do people need to feel like they're informed
enough for their purposes?

        --SK