[boost] Re: ETI in VC71?

Arkadiy Vertleyb writes:

I got the impression that the ETI bug has made its way into VC71 (I just got my entire encoded type full of integers, and worked it around in the usual way -- by providing meaningless specialization for it).

Is this true?

Not to my best knowledge. There is always a chance that you've came across some corner case that nobody hit before, though.

Are the conditions under which it shows up somehow formalized?

Only to the extent that is documented here -- http://www.boost.org/libs/mpl/doc/tutorial/eti.html I'd suggest you to post a query to MSVC-specific newsgroup, as Microsoft compiler engineers are obviously in a better position to answer your question. HTH, -- Aleksey Gurtovoy MetaCommunications Engineering