CTM - any users left?

[Stephen Montgomery-Smith]

Mark Murray wrote:
> Stephen Montgomery-Smith writes:
> 
>
> 
> CTM was quite a clever hack at the time it was conceived, but it never
> really made it out of "hack" status. The CTM builder is as fragile as
> hell (I did the job myself a few years ago), and is incredibly resource
> hungry. It buys its owner very little, and its returns are really only
> in the goodwill department. If it is really useful, it really needs
> someone to own it from the builder downwards, and it needs a strong
> update to bring it into the 21st century.
> 


I don't think ctm is so fragile.  I have been maintaining it now for a few 
years, and it has survived a couple of disk crashes and such like.  One place it 
might be considered fragile from the end user point of view is that if he/she 
makes any change to the source code, that future CTM applications will fail when 
they try to update in the same place.

I will agree that it is resource hungry.  It might be worth rethinking how mkctm 
works.  It first uses cvsup and cvs to update its own copies of the sources, so 
maybe it could cleverly use the outputs from these programs to limit which parts 
of the sources need updating.  (The other possibility is to wait another five 
years, then the computers will handle it in no time at all.)

Anyway, right now I don't really have the time to invest in making these fixes.

-- 
Stephen Montgomery-Smith
stephen at math.missouri.edu
http://www.math.missouri.edu/~stephen