From: William E. Kempf (williamkempf_at_[hidden])
Date: 2002-08-09 08:56:48
----- Original Message -----
From: "Darryl Green" <green_at_[hidden]>
> > -----Original Message-----
> > From: William E. Kempf [mailto:williamkempf_at_[hidden]]
> > This doesn't work because users will want to use
> > Boost.Threads in their
> > libraries, where they have no control over the application
> > which implements
> > main(). It's a hack that may be acceptable in applications
> > development, but
> > I don't find it acceptable in library development.
> This implies that the library author cares about and needs to control the
> consistent treatment of all threads including "main" - is this really
> necessary - or even possible to control?
A library won't know what thread is the main thread. I can't predict what a
library is going to do, but yes, I can envision some cases where it may well
treat the main thread like any other thread and cause unexpected behavior
because the application didn't use the main() wrapper.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk