Re: Guile versioning

[Thien-Thi Nguyen]

   From: address@hidden (Thomas Bushnell, BSG)
   Date: 25 Jun 2001 21:53:22 -0700

   I'm using Guile 1.4 because that's the *released* version.  Why is
   the released version nearly a year out of date?

because the feedback is likewise?  :->

from my perspective, it seemed the guile community was for a long time
fractured between users and developers (who may also be users), so both
groups went sort of open-loop implementing their own hacks (developers)
and workarounds (users) w/o consulting each other.  it is only of middle
2000 that i feel there is more communication betwixt.  probably w/ 1.5.0
release and active solicitation of user feedback this gap can be reduced
further.

   The following things suck in Guile 1.4, and I suspect they may be
   better in the current CVS source:

   There is no good emacs interface
   The debugger can't single step
   The debugger has no good breakpoints
   The debugger doesn't interact with emacs nicely

   What are guile-core's plans in this area?

it's difficult to answer for everyone; there is no gestalt "guile-core".
(personally, i think the "steering committee" idea is flawed, especially
if those named to such a position are absent from this mailing list.  i
became aware of a guile-devel-internal mailing list through a google
search which depressed me for a few days, not because i'm not
subscribed, but because its existence is a sign of weakness in our
development model.  those few days turned into weeks due to other
reasons, but i digress...)

re debugging, most of the pieces are lying around ready to be cohesed
and integrated.  sounds like Neil Jerram is looking into making the
current debugger more full-featured standalone.  perhaps the result can
be hooked into Keisuke Nishida's channels (which i confess to not fully
understanding yet) for full emacs puissance.

i am wondering if fsf would hire me (or someone capable :-) to integrate
guile/gdb/bash...

thi