> It's in atlsync.inl for VC7.1. I just searched for
> "CCriticalSection::CCriticalSection" to find it, BTW.
Bizarre... I honestly ca't find it here, unless XP's Search feature is lying
to me.
> You seem to have big problems with C++ templates' syntax. Is
> there anything
> you'd like us to explain about them, which makes you particularly
> uncomfortable with them?
My problem is simply that I feel they reduce the readability of the code a
little because they introduce more non-alphanumerics - it's purely a problem
of perception. Simple template syntax isn't too bad, but once you start
nesting them, they become unwieldy, purely in my personal opinion.
I'm also not keen to mix "technologies" any more than I have to - I would
want to use ATL throughout, or MFC throughout, or STL throughout, etc. Even
though they are all legit C++, I still consider things like std::xxx to be a
different technology from MFC. That's a weaker reason, I know... but it's
just me {:v)
And in all honesty (and I apologise for this), it bugged me that I asked for
MFC and you gave me something clearly NOT MFC... almost as if you have a big
problem sticking within the boundaries people outline {:v( *ducks*
(This does not mean that I don't appreciate your contribution to MSVC in any
way)
> > Hmmm... the structure it was in WAS static... but it still seemed to
> > fail at Lock().
>
> Please take a look at the code again: it's a function having a
> local static
> variable. This way, the variable gets constructed the first time the
> function is entered. This is different with a static member of a
> *structure*, which gets constructed at program's startup.
Ah, sorry - I didn't recognise that properly for what it was. Interesting
idea...
--
Jason Teagle
[EMAIL PROTECTED]
