Re: [boost] RE: Standardization of Boost libraries

From: Beman Dawes (bdawes_at_[hidden])
Date: 2002-03-10 17:19:06

At 06:16 AM 3/7/2002, Moore, Paul wrote:

 >From: Beman Dawes <bdawes_at_[hidden]>
 >> 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.
 >
 >This raises a question. As the author of the rational number library, I
 >didn't particularly do anything, or get involved, regarding the inclusion 

 >of
 >the library in this list. Don't get me wrong - I don't have a particular
 >issue with this, inasmuch as I don't see that there's anything much I can
 >contribute that others aren't able to handle better on my behalf. But I
 >don't recall any particular feedback on the library - certainly nothing
 >direct to me (I may have missed something in the list, as traffic these
 >days is far too high for me to get through everything).

Sorry.  If you didn't get public or private feedback, that's my fault.  I 
thought I'd covered each library in one post or another.

 > Does this imply that
 >there was no comment? Or that the committee accepted the library as it
 >stands with wholehearted enthusiasm :-)? Or simply that other things took
 >precedence, and they'll get to it in time?

Matt Austern has already posted comments on the discussion (thanks 
Matt!).  The October meeting was the first time the committee's Library 
Working Group (LWG) actually considered whether specific libraries might be 
within the scope of the TR.  I picked libraries to present partially to get 
an idea just what that scope would be.

To me, it seemed that LWG members often zeroed-in on the question "How 
useful is that library likely to be to how broad a range of users?"

With rational numbers, some LWG members were unsure.  "I'm on the fence on 
this one" was Bill Plauger's final comment IIRC.

Thus if you (or anyone else) wants to push rational numbers, building a 
list of real-world uses would help convince the LWG that rational numbers 
are in scope.

 >I do have some fairly strong views on the issue of keeping the library
 >simple, for example. If the committee feels that some form of baroque
 >template/specialisation/policy/traits mechanism is appropriate (to allow
 >user-level tuning of overflow behaviour vs efficiency, for example) do I
 >have any voice (beyond that of any random C++ user) to say "heck, no,
 >that's not what I intended"?

While in theory the LWG can make any change they want in a library, in 
practice that won't happen.  The person who does the work of writing the 
formal proposal (and that is almost always the developer) makes most of the 
choices.  The changes the LWG has made in the past are things like name 
changes, deciding what header declarations go in, and adding or removing 
signatures.  Unless the proposal author volunteers, it is unlikely that 
someone else would be willing to make major design changes in a library.

--Beman