From: Vladimir Prus (ghost_at_[hidden])
Date: 2003-10-13 00:20:37
David Abrahams wrote:
> Vladimir Prus <ghost_at_[hidden]> writes:
> > There's one question: I would be interested to know if the problem that
> > Dave reported in email called "dependency analysis bug" is cleared or
> > not. But I don't think it's show-stopper.
> IIRC, you diagnosed my report as a non-bug (the #include of the
> dependent file was masked by PP directives). Then you found other
> problems, which I suppose you could check yourself(?)
I've originally diagnosed the problem as non-bug, that's true. However, later
I've found out a long dependency chain from is_class_test.cpp to is_class.hpp
via Boost.Test, which means that both is_class_test.cpp and Boost.Test should
be recompiled when you change is_class.hpp is touched. That happens to me,
however, I'm not sure if this happens to you.
IIRC, the symptoms of the original problems were somewhat different for me:
you had no recompiles at all, while I had only some files recompiled. That's
why there's a chance the bug still exists for you.
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk