One issue surfaced mid-week; some Win32 compilers (VC++, possibly others) are silently failing to compile certain source lines terminated with a CR only. (The traditional Apple Mac line termination is a CR only.)
Gennadiy should get some kind of prize for figuring this out; I had failed several times to understand what was happening.
In the past we talked about restricting line terminations to either CR/NL or NL as a convenience to the bulk of users, but hadn't done anything about it since it seemed low priority. But now that we are seeing CR terminations actually causing silent miscompilations, I'd like to fix this ASAP.
I don't know how widespread the problem is; I'll look at it this morning and fix if possible.
In any case, branch-for-release should happen sometime today (Saturday, US Eastern time).
--Beman
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
