Re: List archive developments?

[Earl Hood <earl@earlhood.com>]

(Note, I have directed this reply to the mhonarc-dev and mhonarc lists
 since mhonarc users may be interested, especially wrt the mhonarc
 plans mentoned below.  Original author will receive a separate copy
 of this reply.)

On June 3, 2002 at 10:41, someone wrote:

> What are your plans with MHonArc? I'd like to get all the folks 
> interested in improving the web archiving options for e-mail lists 
> together.  Who else is a leader in this area?
> 
> See all the junk below ...

There was alot of junk :-)

  [To summarize in a nutshell, there is a start of a new project,
   <http://sourceforge.net/projects/archetype>, related to web-based
   email archiving a web-based discussion forums.  Discussion
   started on do-code@yahoogroups.com list.]

Plans for MHonArc?

  First, I want to finalize the MHonArc 2.x codebase.  MHonArc has
  been around for years, but it is straining the initial design due
  to the limitations that Perl 4 imposed when MHonArc was initially
  written (note, MHonArc does use some Perl 5 features as it has been
  enhanced over the years).

  MHonArc 2.x appears to be quite usable for many, but dynamic-type
  capabilities (like integration with a database) it was not designed
  for.  So what I have been doing is modifying MHonArc 2.x to provide
  as much as it can to allow for custom hooks (like to a DB) by
  providing a public API for embedded usage (see MHonArc docs for
  some details).  Unfortunatly, some design limitation are very
  hard to overcome.

  MHonArc 3.x is planned to be a complete rewrite of MHonArc using
  an OO model, which can be done with Perl 5.  I have even thought
  of going to Java (for character encoding management), but would
  prefer to stay with Perl if it can support the new design.

  I actually made a start years ago, but stopped due to other
  priorities and interests.  I have been thinking of restarting
  the initiative since I now have the MHonArc source under CVS at
  savannah.gnu.org.  Plus, I currently have more time now to devote
  to such a effort.  With the repository at savannah, it also
  makes it easier to bring on some additional developers (with
  the proper skills of course :-)

  One nice side-effect of not doing the work years ago is that
  the feedback for MHonArc 2.x has provided more things to consider
  in the MHonArc 3 design that may have been overlooked.  I actually
  try to implement what I can in MHonArc 2.x and to use it as
  way to prototype some stuff.

  The primary goal of MHonArc 3 is to make it as modular as possible
  to allow for things like database hooks and custom filtering.  I.e.
  MHonArc 3 would primarily be a library with a default application
  that functions like MHonArc 2.x for those that do not care about
  a programmable library.

Initial thought on archetype project:

  archetype will probably have similiarities to MHonArc, but it
  appears to have overall different scope.  For example, I could see
  archetype using MHonArc as component of the entire system.

Other "leaders":

  The message you sent me appears to cover the major players.  The
  only project I can think of that was not mentioned is Hypernews,
  www.hypernews.org.


There is probably more that can be discussed, but I am not sure
what the entire scope of the project you are trying to organize and
how it will be managed.  I have a suspicion that what you hope to
achieve can be achieved by a collection of tools instead of single
monolithic system.

If the base requirements are written down and made available from the
archetype project page, it will help others determine how existing
projects fit, or don't, fit in.

--ewh

P.S. Have thought about what license you want to use for the project?
Your choice could impact on what existing software you can leverage,
unless you want to write everything from scratch.