Re: [boost] [parameter] Redefining BOOST_PARAMETER_MAX_ARITY

Subject: Re: [boost] [parameter] Redefining BOOST_PARAMETER_MAX_ARITY
From: Ryan Gallagher (ryan.gallagher_at_[hidden])
Date: 2008-10-14 23:05:06

Jamie Allsop <ja11sop <at> yahoo.co.uk> writes:
> David Abrahams wrote:
> > on Mon Oct 13 2008, Jamie Allsop <ja11sop-AT-yahoo.co.uk> wrote:
> >> Should I expect to be able to redefine BOOST_PARAMETER_MAX_ARITY 
> >> and be able to compile?
[snip]
> 
> yes it does (see my previous post) apologies for the noise

You mention in another message that this was due to defining it in 
two headers to different values.  The approach I took to deal with 
this in my usages is to use a central wrapper header for this library
and define the value there.  I left a comment that this value should
only be increased.  Don't know that there's a better way.

> >> Also I wonder if there is some theoretical or practical upper limit?
> > 
> > Surely there is a practical upper limit, but it depends on your 
> > compiler and I don't know what it is.
> > 
> 
> gcc 4.2 and as it happens I was using 15 parameters (yes a lot, time 
> pressures et al. caused both my hasty post and lack of simplification). 

Just for additional data points, I'm using 14 on msvc 8 sp1 and gcc 4.1.

> One thing I did not mention was that this is a very cool library.

Agreed. =)
Still trying to convince some co-workers it's okay to use though. ;-)