Re: [boost] Algorithms and Formal Reviews

From: Jeremy Siek (jsiek_at_[hidden])
Date: 2001-01-11 10:11:16

I'm also in favor of a Boost algorithms library (or libraries).

We'll probably end up with several categories of algorithms, so I suggest
we do the following. Have a <boost/algorithm.hpp> header that always
includes *all* the boost algorithms. As the number of algorithms grow, and
we figure out what categories are worth having, we create other headers,
e.g., <boost/algorithm/text.hpp>, and move algorithms from the main header
to the new header, and then include the new header in the main header.

Cheers,
Jeremy

On Wed, 10 Jan 2001, Beman Dawes wrote:

beman> I've been thinking about sequence algorithms because of several recent 
beman> boost postings, and string algorithms because I've seen several nice onces 
beman> recently.   See below for a string algorithm example (by Jens Maurer).
beman> 
beman> The Boost Formal Review procedure works pretty well for whole libraries, 
beman> and would presumably work well if someone submitted a whole library of 
beman> algorithms.
beman> 
beman> But what happens if someone just has one or two neat algorithms to 
beman> submit?  Neither the submitter nor the boost membership is going to want to 
beman> go through a whole Formal Review just for one or two algorithms, 
beman> particularly if they are only five or ten lines of code each.
beman> 
beman> One way to smooth the process would be to have a Boost Algorithms Policy 
beman> which defines what algorithms are acceptable and what aren't.  Appoint a 
beman> couple of people as moderators of the Boost Algorithms Library (with CVS 
beman> write access).  They will act as permanent Review Managers + library 
beman> maintainers.  They would check that submissions meet the Algorithms Policy, 
beman> and then (?) to ensure sufficient peer-review.  Once the algorithms 
beman> moderators agree the algorithm is accepted, they can move it into the 
beman> library with a minimum of fuss.
beman> 
beman> (?) might be something like stick the algorithm in a queue to give people 
beman> time to look at it and comment.
beman> 
beman> Comments?
beman> 
beman> --Beman
beman> 
beman> // replace the first instance of name in s with value
beman> void replace(std::string & s,
beman> 	     const std::string & name, const std::string & value)
beman> {
beman>    std::string::size_type p = s.find(name);
beman>    if(p != std::string::npos)
beman>      s.replace(p, name.length(), value);
beman> }
beman> 
beman> 
beman> 
beman> 
beman> 

----------------------------------------------------------------------
 Jeremy Siek                        www: http://www.lsc.nd.edu/~jsiek/
 Ph.D. Candidate                  email: jsiek_at_[hidden]
 Univ. of Notre Dame         work phone: [phone]
----------------------------------------------------------------------