On Sun, 2009-02-22 at 11:28 -0500, Mitchell Wand wrote:
> IIRC, we are in the waning hours of the voting period for the Steering
> Committee who will have the responsibility of guiding Scheme into the
> next stage of its existence.
> 
> And all anybody wants to talk about is case-sensitivity?!?


There are several things going on there.

First, no - there has been conversation about
topics more central to SC duties, such as voting
rules for making changes.   If you look at those,
I think you'll see that they mostly failed to 
shed much light.   For example, there was chat
about super-majority voting rules.  Do they 
protect minority opinions?  Do they lead to
hijacking by minority groups?

Only, look at how ungrounded that chat is:
*what* minority groups are we talking about
here?  And, in the end, what difference do the rules
make to someone who wants to know if their favorite
feature or change or whatever has a chance of
making it into R7?   We have no good basis to 
answer questions like that - we're stuck with
abstractions about voting system in general.

Whatever it is that we're talking about in those
threads it's not clear that we're talking about 
the future of Scheme.   But, wait - isn't the 
"future of Scheme" pretty much the overarching
concern here?

Those conversations lack lasting appeal
because it's not clear they are about anything
meaningful and "on topic" if the topic is the
future of Scheme.

-------------------------

Second: Case (in)sensitivity at least is a concrete
topic - grounded in the technical details of
the language.   And it's an interesting concrete
topic because of the historic change.   So, it's
a topic where people get to talk about things like
the importance of upwards compatibility, about 
the impact of a design detail on implementation
complexity, about the importance of supporting
non-english users and processing non-english data,
about the design process of R6, and about the possible
design process going forward.

-------------------------

Third: I think you ought to "take away"
from the discussion about case is that, evidently,
a great deal of the community interest in
direct participation in the deliberative 
process that produces new or revised language
in the standard.

Dillinger said it well a few messages
back:  Rationales were obscured in R6
and the thing was plunked down for basically
an all-or-nothing, thumbs-up-or-down.  That's
a formula for generating resentment.

Oh, sure, there was a public discussion period.
My sense of it was that it was very effective
at cleaning up a lot of nits and typos but...

On the topic of any of the larger, controversial
changes, the discussion period was not very effective.
It felt to me like a lot of cases of strong arguments
being offered against some of the R6 ways only to 
get back responses that said little more than "We don't
think that's important," or "We don't see it that way,
sorry, not gonna happen."   In some cases, my sense
was that the message from the editors was really
"Perhaps there is a point there or not but the fact
is we worked really hard on this and we just want to 
finish up now."

That is why I suggested in an earlier (ignored)
message a radical reinvention of the R7 process in
which, before we try to figure out even how to 
ratify a draft, we first figure out how better to facilitate
a meaningful, more formal design conversation with
more of the community members.

-----------------------------

Finally, the revised report series is an economically
significant object.   For people with their names on it,
it's a valuable addition to a CV.   For a larger audience,
it's a touchstone against implementations and a standard 
reference item to list on the "skills" section of a 
resume.   It's also a document that can be referred to
in conversations with other language design efforts.

Well, the Scheme community has an economic crisis, if
you ask me.   It comes down to it being an institutional
and economic privilege of a very few to determine 
the contents of the Report, even though economic interest
in the the report and its economic impact is much larger
than that.   This is taxation with representation.
It's tyranny, of a sort.

The result is less a "Report" and more a "Proclamation".

--------------------------------------

The original choice of title - "Report" - is suggestive.

It suggests that a lot of related work was going on 
in various places, by various people, all of whom recognized
one another as all working "on Scheme".

There were things they shared in common, areas where
their interests diverged.   It was believed to be valuable
between them and as their collective face to the world to 
find (or make up) a set of essential elements upon which
they could agree.   The statement of such agreement is
an account of what the community was doing - a "Report" - 
not a sanctified "Standard" (until IEEE happened).  Not
a "Proclamation" by an Authority.

The community is quite a bit larger now and comes from
a vastly more diverse economic and institutional place.
Yet the "Report" does not reflect this diversity and
work on the report (by anyone) tends to accrue benefits
mainly for a few.

As things stand right now - looking at various implementations,
and diverging opinions, and all of that - an honest R7
*could* very well be one that winds up being subtitled:

"Five similar languages, all called Scheme."

(With some suitable constant substituted for "Five".)

Perhaps it would have an appendix on "Best Portable
Practices".

Or, perhaps if we start off on a process that *can*
have that outcome, we'll actually discover consensus 
and have a single-language R7 after all - but one more
people are happier about.

And, again: money is money.  Looking at myself and looking
around the virtual room here I keep thinking that we
would see better work, better focus, and better outcome if 
more of us were getting paid to play.   The office of
the Steering Committee seems to me the proper place in the
existing order to begin to think about that and do something
about it.

-t





_______________________________________________
r6rs-discuss mailing list
[email protected]
http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss

Reply via email to