TOO MUCH VOLUME -- a proposal

Fare Rideau rideau@ens.fr
Mon, 12 May 1997 21:11:28 +0200 (MET DST)


Dear LispOS subscribees,
   there is definitely TOO MUCH volume on the LispOS mailing list,
most of which is flames due to people not knowing/understanding
each other's point.
   The result is that there is a high e-mail volume,
with a highly redundant content.
Plus the ephemeral nature of e-mail discussions (even when archived)
hides the meaningful information in the volume,
instead of making it persistent.
   This is just not bearable. We're going nowhere that way.
	"Les paroles s'envolent, les \'ecrits restent".

What I propose is that discussion be moved *OUT* of the list.
into each of the subscribees' WWW page.
   If you wish to discuss a topic, expose the topic on the list,
and post pointers to relevant URLs that describe your opinion.
Similarly, when your position evolves significantly,
post a update announce on the list.
If you want to reply to someone else's ideas,
expand you page quoting his where needed,
and send the result to him, and an URLed announce to the list.
   For those without WWW publishing capabilities,
a server (CVS?) could be setup for remote publishing
(please someone with webmaster priviledge?).


On another meta-topic, it seems that we have lots of divergent ideas.
Well, as has already been pointed out,
we're NOT forced to all program the very same thing;
when two people have opposite opinions in such a *free* project,
there's no argument to force one's views against the other one's anyway.
   Happily, there's enough room and enthousiasm
for several rival Lisp-based OS projects; as long as everything is free,
we'll only gain, not lose, from fair competition.
And if the interfaces are high-level enough, Unification can come later.
   So after we know what each of us wants, we can group according
to our mutual affinities, instead of flaming each other according to
our reciprocal dissensions.

It seems to me that *candidate* projects for grouping would include:
various-Applications-on-top-of-CL,
CMUCL-extensions-toward-an-OS,
deeply-reflective-system-with-a-basic-modular-lisp-under-CL,
Persistent-Store-for-some-of-the-above,
some-of-the-above-with-Scheme-(EuLisp?)-instead-of-CL,
hacking-with-Flux-instead-of-Unix-below,
etc.
   Please no flames, but explanations and argued advocacy
on your page instead. Cross references to each-other's
good and bad arguments welcome.


To apply my proposed methodology, my personal tastes would go
towards the deeply-reflective-system-with-a-basic-modular-lisp
and a Persistent-Store for it.
My personal position is extensively described in the Tunes project:
http://www.eleves.ens.fr:8080/home/rideau/Tunes/
Words I use try to not buzz, but refer to my Glossary instead:
http://www.eleves.ens.fr:8080/home/rideau/Tunes/Glossary.html

As food for thought, I have lots of pointers to material relevant
to computer system design and implementation:
http://www.eleves.ens.fr:8080/home/rideau/Tunes/Review/

Contributions welcome.

Be constructive!

#f