Re: Guile is a great idea, but where's the community?

[Richard Todd]

Neil Jerram wrote:
>  1. It may involve some ideas previously discussed under the name of
>     "GUMM", but I shouldn't worry too much about that - what we really
>     need is for someone to just _do_ something, that can then be
>     incrementally refined with usage.

I agree, and have already started to a degree.  I've been making guile 
ports of modules from perl and python that I find useful.  I posted 
ansi-color last night in an attempt to show that it can be pretty easy 
to do.  In early January, I'll set up a project on Savannah for the project.

I'll also be contacting authors of modules I'd like to pull in over the 
next couple weeks, but there could be lots of good code that I don't 
know about.  So, the best thing people wanting to help in the near term 
can do is:
  1) point me to modules that you think should be incorporated
  2) implement modules that don't exist, but should be incorporated.
     (Many of these don't take much time; It's just work that has yet to
      be done.)

I'm just going to throw a start at it on the server, and get a mailing 
list started so we can hash out things like how the modules should be 
organized, how we want to manage releases and backwards compatibility, etc.

I'd like to have a 1.0-type release in 6 months (because June 1 is my 
birthday...I've worked in software too long to believe in estimating 
coding work!)  This would mean that we are stable with respect to how we 
want the project to run, and have some initial set of modules 
incorporated and documented.  After that it should be just add, add, 
add, release 1.2, add, add, etc.

Richard Todd