Attendees:
Hans Boehm
Victor Luchangco
Jens Maurer
Michael Scott
Michael Spear
Michael Wong

Some discussion of std::shared_ptr.
Support is implementation defined in P2066.
That's correct, since we could otherwise detect in another thread that
we're in the middle of a transaction. Both weak_pointer and arguably
use_count allow us to observe intermediate state without introducing a data
race.

std::future is disallowed together with other thread support.

There may be other library facilities that unavoidably use shared_ptr
internally. The TS will be a good way to flush those out.

Brief discussion of directory enumeration. But since that involves IO, it's
out anyway.

I suggest we return to our normal schedule, and meet again in 2 weeks,
since that will be shortly after the EWG discussion. Please let me know if
that doesn't work. This makes our next meeting:

May 4
8am Pacific, 11a.m Eastern, 17:00 Central Europe
meet.google.com/sbj-cvgh-vnd
‪(US) +1 208-925-0196‬ PIN: ‪255 542‬#

The agenda will be discussion of any EWG feedback. The meeting may again be
brief if they're mostly happy with the current proposal.