From: Paul Moore (gustav_at_[hidden])
Date: 1999-12-19 13:25:21
From: jsiek_at_[hidden] [mailto:jsiek_at_[hidden]]
> Paul Moore writes:
> > Grrr. I think I'd blame them. If they define _MSC_VER, they should be
> > compatible. Bugs and all. After all, most people only use _MSC_VER
> > as a way of protecting bug-workarounds...
> The thing to do is define another macro, something like MSVCPP,
> which you define only when it is really VC++, and then use this
> macro to protect the workarounds.
But then the *user* would need to define it in his builds. No - the compiler
defines a macro which identifies itself. Other compilers shouldn't define it as
well in the name of "compatibility". On this issue I side with MSVC. If you
define _MSC_VER, you should implement *all* of the features of that version of
MSC - even the bugs.
BTW, what value does _MSC_VER have on other compilers - which version of MSVC do
they masquerade as? If they use a non-MS version number, we may have a chance...
But as I said, I'm for removing abs() and avoiding the whole problem...
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk