Re: [boost] Metrowerks 6.0 regression compile results

From: David Abrahams (abrahams_at_[hidden])
Date: 2000-10-14 12:55:54

If we have been making workarounds for CW bugs, have we also saved
reproducible cases for them? Their compiler guy is very responsive, and even
if Howard isn't paying close attention right now, I can send him something.

-Dave

----- Original Message -----
From: "Beman Dawes" <beman_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Saturday, October 14, 2000 10:46 AM
Subject: Re: [boost] Metrowerks 6.0 regression compile results

> At 07:59 AM 10/14/2000 -0400, John Maddock wrote:
>
>  >>These attempts to instantiate individual member functions may work in
> some
>  >>compilers, but they are non-standard.
>  >
>  >That does seem to be the cause of the problem: however it does seem to
be
>  >legal in the standard: see line 3 of the example in 14.7.2 para 2.
>  >
>  >Whatever I've disabled that for MW builds - so hopefully it should build
>  >now.
>
> Yes, thanks, John!
>
> Jens also updated random_test.cpp, so that's now working for all compilers
> tested.  Thanks, Jens!
>
> BCC 5.4 has been removed from the test set.
>
> regex/src/cregex.pp has been added to the regression test, and is working
> for all the Windows compilers tested.  [Jens: this is a compile only test;
> you may have to add -c or whatever to the Linux compiles.]
>
> cs-win32.html has been updated in the CVS.  It's looking much cleaner that
> the 1.18.0 version.
>
> I'll soon release 1.18.1.  But it is a gorgeous day here, so I'm headed
> outside.  Maybe tomorrow.
>
> --Beman
>
>
>
>