Getting things rolling

["Scott E. Fahlman" <Fahlman@C.CS.CMU.EDU>]

I'm sending this to the steering committee because that mailing list has
the technical people on it as well.  And these issues of how to get
started involve all of us.

Here are some things that we need to do, in more or less the order in
which I think we need to deal with them.  The list is probably not
complete, so feel free to propose additions to it.  Entires are marked
with S (steering), T (technical), S&T (both) and S,T (each committee
separately).

Get the two committees organized.  Probably we don't need much
organization, but some sort of chairman or moderator is needed for each
-- someone who will feel responsible for goosing the agenda along. (S,T)

Make a list of formal things that the steering committee needs to do as
part of the ANSI and ISO formalities.  Is some sort of a kick-off
meeting necessary? (S)

Conduct a poll on the Common Lisp mailing list (and maybe by other
channels) to identify what companies have people participating in
this process, even if they are content merely to observe.  Prepare
a list of such participating companies, along with the name and address
of a contact person at each.  This will be used for formal X3J13
notifications, etc.  It will also make it clear to everyone that many
companies are participants in this process, even if they don't have
someone on the Technical committee.  (S)

Determine if there are any companies (or other implementation groups)
that need arpanet access but don't have it.  Try to get them accounts
somewhere. (S)

Try to establish reliable netmail contact with Japan.  Once this is in
place, select a Japanese member for the technical committee.  (Masayuki
Ida was discussed earlier, but we couldn't get mail to him, though mail
 from him has reached me.) (S)

Send an embassy to the Eulisp people and see if they have any interest
at all in participating in this process, given our unreasonable desire
to standardize something like the current Common Lisp and not start
over as they are doing.  If they want to participate, invite one or more
European members to join the committee.  Netmail seems not to be a
problem. (S)

Decide on how we are going to run the technical decision-making process,
how we are going to record and communicate the results, and what sort of
form the standard document will take.  (S&T).  I'll describe some
thoughts I have on this issue in a later message.

Decide what major areas we are going to try to fix and/or extend in this
first version of the standard.  Make for ourselves some guidelines on
how deeply we want to change things.  Set ourselves a realistic schedule
for completing the first spec.  (T)

Provide some guidance to the ISI folks about what services are most
needed, and what their priorities should be.  (S&T)

Create an inventory of issues that have been raised.  Guy made a start
at this, but it needs to be kept up to date.  Divide these issues into
three classes: issues where we think there is (or could easily be)
agreement on the right solution, isolated issues that are controversial,
and issues that are all tangled together in rotten areas of the current
spec (e.g. that part about what the compiler does). (T)

For the issues where there is or could be agreement, confirm this on the
mailing list and then record the decisions so that they don't unravel
again. (T)

Debate the isolated but controversial issues in public, then make the
decisions and record them. (T)

For areas where many interrelated things need work, choose someone in
each case to put together a comprehensive proposal.  Then debate the
result and nail it down. (T)

As coherent proposals for extensions appear (error, objects, whatever),
run these through the process.  If there are areas where progress is
needed but no proposals exist, find someone willing to take a crack at
it. (T)

Make a final pass over the completed specification document (T) and try
to get it through ISO and ANSI (S).

Get to work on Common Lisp 2001. (Whoever survives.)

-- Scott