RE: [boost] Signals Review

From: Gary Powell (powellg_at_[hidden])
Date: 2002-02-27 14:09:43

>>=============================

What would users prefer? The first option, that would guarantee O(1)
insertion time for unnamed slots, O(lg n) insertion time for named slots,
but
complicate (and somewhat slow down) the calling of slots? Or the second
option, that would guarantee O(lg n) insertion time for all slots but
provide
a simpler implementation that might be (slightly) more efficient when
calling
slots?
<<===============================
In general when I've used systems like this, the program didn't change the
insertion more than once. (On some part of initialization, not necessarily
on startup, but rather on startup of some portion of the program, say
entering a dungeon) And runtime calling was much more time sensitive (say
drawing to the screen).

------------------------------------------------------------------
Since today is the last day of the review, I'd just like to get my vote in.

I vote to accept this library. (with the minor stuff including documentation
fixes.)

On the other side, I haven't used it in production code and until there is
more use of it, I would regard the implementation details, and the interface
open to future change. (With Doug not having to support previous versions)
I'd like to see a phase 2, which had more control of the calling order and
the ability to group things by name. But the lack of this shouldn't prohibit
the current acceptance and use of this library. Plus there may be some other
things we haven't thought of that will influence phase 2.

Doug, you've done a nice job.

  Yours,
 -gary-