status and priorities

[Loren James Rittle]

Hello fellow libstdc++-v3 maintainers and other readers,

> First, I'm pleased to report that many of these items have already been
> done! The progress has been fast and furious, and I think everybody can
> be truly pleased with the progress. A lot of hard work by many people
> has made a noticeable improvement. Some issues remain, however.

I too am extremely pleased with the rate of progress I've seen.  From
past history, I estimate that Benjamin will install the patch to build
the shared libstdc++ with IMI-enabled within 48-hours of the C++
compiler actually supporting it (granted, this will require us to
finally remove use of libtool).  And, Paolo is on course to have more
changelog entries than all others this year.  ;-)

> -  look ahead for unbuffered io, so know when multiple putc's can be
>   coalesced.

Does anyone know how to do this?  I think I'd know how this could be
done in general CS terms but not with a standard C++ compiler...

General example: If you see A then B back-to-back (or, more generally,
without C in between) then remove both A and B (if we are talking
about mutex locking analysis and optimization) or merge A and B by
using replacement text/call D (if we are talking about OS call
batching).  How do I write these rules in plain C++ and/or g++ with
extensions?  Or, are you talking about arranging to optimize:

  std::cout << "a=" << a << ',' << "b=" << b << std::end;

> -  generic named facets work (Loren were you working on this?)

I think I got it working well enough for FreeBSD 5 needs; then got
sidetracked with a semi-general framework for supporting
target-specific flags/macros and the FreeBSD-specific bits.

If I have any time, then I'd be more inclined to work on the
performance and interactive test suite framework that you (Benjamin)
started.  I have some patches in my tree but they aren't ready for
public posting.  I have no idea when they will be done.

Regards,
Loren