Re: [boost] Re: Rationale for a #define

From: Mark Rodgers (mark.rodgers_at_[hidden])
Date: 2000-10-06 03:34:25

> Can anyone confirm that the problem is gone in BCC 5.5.1?  

It seems to be.

> If so, then is BCC 5.5.1 a free, easily-available upgrade from BCC 5.5?

It is.

> If that's the case, then we could drop BCC 5.5 support

I agree.  I think you should just remove it.  I notice Beman is using
5.5.1 in his regression tests.

In general, I think we should assume all patches are applied to
any version we support, so ideally I'd like also to see Beman using 
SP4 for VC++, and also STLport 4.0.

Beman, I also think you could drop BCC 5.4 from the table since 
5.5.1 has fixed the annoying bugs in 5.5, so I don't think there is
any reason BCB 4 still needs to be used.

We can't really support every combination of versions, and if 
"upgrade" is an applicable answer to a problem, then I think it 
is appropriate.

Mark