From: Joaquín Mª López Muñoz (joaquin_at_[hidden])
Date: 2004-07-27 10:53:51
Stefan Slapeta ha escrito:
> Joaquín Mª López Muñoz wrote:
> > Could some kind soul with access to MSVC 8.0 try TEST 0 and TEST 1 for
> > this compiler and report the results? If TEST 0 passes and TEST 1 fails, IMHO
> > BOOST_NO_MEMBER_TEMPLATE_FRIENDS should be set on for this
> > compiler, or at least a weak variant, something like
> > BOOST_NO_MEMBER_TEMPLATES_ACROSS_NAMESPACES
> Once again: VC 8.0 behaves _exactly_ like VC 7.1! (therefore, test 0
> succeeds and test 1 fails)
Oh, excuse my misunderstanding. I took this previous statement of yours:
[Michael Stevens] The error seems to indicate a namespace problem rather then
something specific to member template friends?
[Stefan Slapeta] Yes. Both 7.1 and 8 show exactly the same behaviour [...]
as meaning that you endorsed the pov that the problem was not template-related.
(Later on on the paragraph, your position is made clearer, though.)
Anyway, thank you for the further explanation and for trying
the tests. I'll update my lib accordingly to this info.
> Thus, as BOOST_NO_MEMBER_TEMPLATE_FRIENDS is defined in visualc.hpp for
> this compiler...
> #if _MSC_VER <= 1310
> # define BOOST_NO_MEMBER_TEMPLATE_FRIENDS
> ...there should be the same define for VC 8.
> Whether this is too strict or not and whether there should be a "weaker"
> macro (just for friends in other namespaces) is a completely different
> discussion and the outcome is equally valid for both VC 7.1 and VC 8!
I guess we'd need more direct exposure to the compiler to determine
this. Peter seems to point out that MSVC 8.0 has made some actual
progress in supporting member *class* template friends, but I don't really
Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk