Gnus development remains haphazard?

[Matt Swift <swift@alum.mit.edu>]

I just read a thread here in which the question, does Oort Gnus have a
set of particular goals?  Replies suggested about half a dozen
possible major goals.  I've been away from this group for a while, but
if this thread is indicative, Gnus is still being developed in a
haphazard way.  Perhaps before tackling those major new goals, it
would be best to make development more efficient in general.

A bug tracking system would be very helpful for all users, would 
avoid duplication of labor in analyzing and solving problems, and
would increase the number of people who make occasional contributions
of an analysis or fix.

My impression is that for the last five years, bugs that are reported
are not always analyzed, those that are analyzed are not always fixed,
and fixes are not always incorporated into releases.  One consequence
of the difficulty of bug squashing in Gnus is that development efforts
are often channeled instead toward new functionality, rather than
making core functions more stable, sensible, documented, and versatile
(not to mention accessible to new users, etc.).  

In addition, a bug tracking system is also an effective way to
organize, coordinate, and focus development of new functionality -- a
"wishlist" or "missing feature" bug.

I don't have the resources or knowledge to establish a bug tracking
system, but I can promise that I would use it, and I do believe that
having one would "pay for itself" rather quickly with improvements in
Gnus.