Re: Standardization of Boost libraries

From: bill_kempf (williamkempf_at_[hidden])
Date: 2002-03-06 13:54:15

--- In boost_at_y..., Beman Dawes <bdawes_at_a...> wrote:
> Ouch!  This just arrived from Clark Nelson:
> 
>  >This is your reminder that the deadline for the pre-Curaçao 
mailing is 
> one
>  >week from today, 2002-03-12. Acceptable document formats include 
plain
>  >text, HTML and PDF. Please contact me with questions, and for 
numbers to
>  >apply to your documents.
> 
> That means it is time to think again about what additional Boost 
libraries 
> should be submitted to the committee for possible inclusion in the 
upcoming 
> TR. (Because they will be presented but not voted on, we don't 
actually 
> have to have stuff ready in a week.  But the notice is a wake up 
call to 
> start getting ready for April's committee meeting.)
> 
> The following Boost libraries are already on the committee's 
proposal list:
> 
> 1) Header <cstdint>.  (Tabled pending more comprehensive proposal 
from Bill 
> Plauger.)
> 2) Type Traits.
> 3) Regular Expressions.
> 4) Smart Pointers.
> 5) Random Numbers.
> 6) Rational Numbers.
> 7) Threads.
> 
> How about suggestions for perhaps six more Boost libraries to add 
to the 
> list?  Which of the remaining libraries are most "widely useful" 
for 
> example?

How about Boost.Function.  I think this is one of the most useful 
libraries that's included in Boost... though my opinion is clouded a 
bit by the fact that it's need by Boost.Threads.  I also think that 
Boost.Bind is something that should be addressed by the standard as 
well, and makes using Boost.Function much more bearable for a lot of 
applications (again, opinion clouded by Boost.Threads' needs).

A lambda library could replace these, but the concepts of lambda are 
foreign enough to most C++ users that I think the simpler 
Boost.Function and Boost.Bind libraries would still be helpful even 
if a lambda library were standardized.  So don't consider my 
suggestion for these two is meant to preclude lambda libraries for 
either Boost or the standard, but these two libraries are absolute 
must haves for familiar ways to simplify the usage of generic 
algorithms, and so fill a known hole.

Bill Kempf