|
Boost : |
From: Lois Goldthwaite (loisg_at_[hidden])
Date: 2000-10-06 12:02:08
I will second the accolade for Borland C++, which was out much
earlier[1] than Microsoft's, and was always much better. You may
remember that the original Stepanov/Lee 1994 STL was developed using
Borland 4.5. This was at a time when VC++ was just barely starting to
recognize templates.[2]
VC++ didn't begin to catch up until Borland fell on hard times [3] and
laid off or antagonized its compiler developers into a mass exodus.
Lois
[1] And before Borland's came out, there was Zortech C++, and even
before that was Glockenspiel's, though I never used that one.
[2] IIRC Borland 3.1 (!) had a fair bit of support for templates.
[3] IMHO the problem there was that Borland got so carried away with
challenging Microsoft and Lotus and Ashton Tate -- all at once -- in the
corporate market that they abandoned the individual developers and
price- and quality-conscious software shops that were the backbone of
their market. So I'm inclined to think that Borland threw away its
advantages through bad decisions, rather than putting all the blame on
Microsoft. Novell and Lotus are more examples of companies who should
have paid attention to what they did well instead of anti-Microsoft
activities.
Greg Colvin wrote:
>
> From: William Kempf <sirwillard_at_[hidden]>
> > Actually, at one point in time, VC++ was one of the best C++
> > compilers available, especially on the Windows platforms.
> >
> > So, they didn't get a lock because of a monopoly, they got the lock
> > because they were the best.
>
> Actually, the Borland compiler was a much superior product. The real
> winner was the Microsoft Foundation Classes, which encapsulated the
> horrors of the Windows API just well enough that shops that couldn't
> handle Windows programming in C but were too proud to use Basic moved
> to C++. Borland had a competing product called OWL, and Microsoft
> refused to license MFC to Borland unless Borland dropped support for
> OWL. Borland refused to do that to their customers. Eventually
> Borland got an MFC license via a third party but by then it was too
> late.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk