Subject | Re: [ib-support] Re: Canceling Queries |
---|---|
Author | Ivan Prenosil |
Post date | 2002-02-27T15:01:57Z |
> From: Claudio Valderrama C.Below is another explanation why MS compiler was chosen.
> ""Artur Anjos"" <arsoft@...> wrote in message
> news:00c701c1be61$58dd7c60$0202a8c0@......
> >
> > ----- Original Message -----
> > From: "Leyne, Sean"
> >
> > > - a Microsoft C++ compiler (it's a loonngg story)
> >
> > I hate this lonnnnnnnnnnnggggggggggggg story! I'm stuck with this!
>
> When the IB team tried to compile IB with BC++ 5.02, they found the
> executable wasn't able to run without crashing. They found that the solution
> was to use the MSVC compiler. Besides that, MSVC generates more optimized
> code.
Ivan
http://www.volny.cz/iprenosil/interbase
====================
Date: Fri, 16 Apr 1999 23:54:33 -0400
From: Bill Karwin <bkarwin@...>
Subject: Re: Event problem over TCP/IP
wade wrote:
>This has been discussed from time to time. The answer is that in
> Of course the obvious followup question is why does Interbase not use
> Inprises own borland c++ compiler in the first place instead of the
> supposedly inferior MS compiler? Interesting to ponder.
1993/1994 when we were first doing the Win32 port, there existed a
possibility to add Windows NT on PowerPC and Alpha to the list of our
supported platforms. Borland C++ couldn't do that but Microsoft's
compiler could.
We even released an NT PowerPC port of InterBase 4.0, with somewhat
unfortunate timing because PowerPC suddenly withdrew support for Windows
NT.
In subsequent releases, we never had a compelling reason to adapt the
product to work with Borland's compiler. We were on the lookout for a
good business case to do a Win NT Alpha port (haven't seen sufficient
demand yet), and there always seemed to be more useful things to do in
engineering.