Re: [boost] Re: Named Pairs

From: Emily Winch (boost_at_[hidden])
Date: 2003-01-24 15:40:05

On Thu, 2003-01-23 at 20:30, Fredrik Blomqvist wrote:
> This is interesting stuff, but I think you should checkout Emily Winch's
> more general compile-time associative list implementation also.
> (http://groups.yahoo.com/group/boost/files/alist-feb14-02.zip seems to be
> the latest)
> 
> What's the status of the Emily's lib btw? Is an MPL/Boost integration under
> consideration?

It's still very much proof-of-concept, and it doesn't work under VC6:
the only compiler I have ever successfully compiled it on was Comeau.

I think that some kind of half compile-time/half run-time
associative_list container, plus a set of algorithms working on both the
associative_list and boost::tuple, would be a good thing to have in
Boost. Nevertheless, there are two things that I can see are necessary
before my associative_list could serve that purpose:

1. A good discussion/criticism of it. At the moment it is entirely
conceived and written by one person, and that doesn't make for
Boost-quality code. There is, for example, the alternative approach from
Mike Burton's paper to be considered (www.cs.ucsd.edu/~wgg/Statements/
mburton-ifip-gw-07-2002.pdf). 

If anyone would like to grab associative_list and post some comments or
suggestions, that would be a start in this direction! (Although, next
week I will be away in Hong Kong with uncertain internet access).

2. An implementation that works on more than one compiler, and
preferably most compilers. I've seen the amount of work that went into
the MPL to make it portable, and it's scary :) What would be really
useful would be a list of useful workarounds for broken compilers!

Emily