Documents and meeting

[Jim Dehnert]

Hi, all,

I've updated the documents on the web site.  Since the meeting yesterday
was, uh, lightly attended, we didn't do much, and not much is reflected
from that.  The only real fallout is a proposed elaboration (or change)
in the template argument mangling, which we've been discussing on the
reflector.

I also reorganized the ABI document.  I removed the fiction that it is
just data layout, added chapter/section numbering (high-level, at
least), and reorganized in chapters that are (more or less)
introduction, data layout, calling conventions, and linkage.
I also changed the name from abi-layout.html to abi.html.  (I changed
the link in the summary, but it occurs to me that I have a bunch of
others to change -- you might, too.)  Let me know what you think.

I think we're close to resolving all of the important, tractable
issues.  I would like to push to clear them out over the next couple
of weeks.  So it would help if you would all take a hard look at the
issues on the most recent agenda, clarify them for yourselves,
suggest modifications/resolutions by email if appropriate, and
generally be prepared to push things through.  (Part of my motivation
is that I'll be traveling (business and vacation) for the five
Thursdays starting the 18th, so it would be nice to give you all
a vacation :-).

There is some feedback from the base ABI committee about priority
initialization.  They don't see a compelling need to do anything at
their level, and haven't been hearing demand from their C++ cohorts.
So they suggest that if we want something done, we should contact
our ABI counterparts within our companies and push them to do
something.  I think our current state is that IBM is pushing ahead
with their current implementation, gnu/Linux is doing the same, and
I'm not sure of other interest.  Unless something develops, I think
we'll have to drop this one (or perhaps agree on the source API and
leave it at that).

Jim
-		Jim Dehnert  x3-4272