Re: VC++ Net 2002 - 2003 upgrade

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance

From: Hendrik Schober (SpamTrap_at_gmx.de)
Date: 02/05/04 

Date: Fri, 6 Feb 2004 00:19:02 +0100

Jonathan Turkanis <technews@kangaroologic.com> wrote:
> [...]
> Okay, I'm not going to argue about the definition of 'first-rate'. ;-)

  <g>

> Let me put it this way. I haven't yet written any tricky
> template-based code which I haven't been able to make work correctly
> on VC7.1.

  Great. But what does that by itself mean?
  I can say, I have written no template code
  that I was not able to get work correctly
  on VC6. The questions remain: _Why_ didn't
  I write it? -- because it wouldn't work --
  and How long did it take? -- too long for
  too much of the code. [...]

  If I'm cornered with different errors spit
  out by VC, CW and GCC I usually cook up a
  sample and play with Comeau until I got it
  done after the book. _Then_ I try to make
  the other compilers accept the code. Now,
  the interesting thing about this is that,
  most of the time, once I got the code the
  way it used to be, it is a lot easier to
  make the rest of the compilers accept it.
  The biggest problem is to get there when
  all you have is broken compilers that ICE
  on simple syntax errors.
  This is where Comeau shines. It is (almost)
  100% conforming (so much that I have yet to
  find a compiler vendor that wouldn't accept
  a bug report if you argue with "but Comeau
  does/doesn't accept the code") and it does
  produce great error messages.
  While you can get your work done with just
  about any other compiler, it does take
  longer. And it hurts more.

> I'm aware of problems with VC7.1's conformance, but so far they
> haven't tripped me up.

  Missing two-pahes lookup is a PITA if you
  write library code. It is stupid to have
  your users find silly syntax errors in your
  code. I always check my code with CW before
  I release it, even if it is meant to be used
  with VC only, since CW does check non-
  dependend names.

> Jonathan

  Schobi 

-- 
SpamTrap@gmx.de is never read
I'm Schobi at suespammers dot org
"Sometimes compilers are so much more reasonable than people."
  Scott Meyers

Relevant Pages

  • Re: What a translation unit is.
    ... That's basically what the Standard says, what Borland docs say, ... >> Comeau docs say, what Greg Comeau himself says, what everyone ... I acknowledge a compiled translation unit. ... World Class Compilers: Breathtaking C++, Amazing C99, Fabulous C90. ...
    (alt.comp.lang.learn.c-cpp)
  • Re: Teaching new tricks to an old dog (C++ -->Ada)
    ... >> How many current compiler pass current ACATS test, ... Off the top of my head, Ada compilers are available ... >portions of the language standard are optional. ... Greg Comeau / Comeau for the Mac? ...
    (comp.lang.ada)
  • Re: Teaching new tricks to an old dog (C++ -->Ada)
    ... >> How many current compiler pass current ACATS test, ... Off the top of my head, Ada compilers are available ... >portions of the language standard are optional. ... Greg Comeau / Comeau for the Mac? ...
    (comp.lang.cpp)
  • Re: how to use pure c to write OO programs
    ... >> can some one tell that how to use pure c to write OO programs? ... Holding your head down, run towards the wall. ... We get can compilers going rapidly, but as above, not sure these ... Greg Comeau / Celebrating 20 years of Comeauity! ...
    (comp.lang.c)
  • Re: How did C++ beat the competition?
    ... By then, as Walter pointer out, we had Zortech, ... >compilers at the time were incredibly expensive. ... >the introduction of Windows 3 that Microsoft started to dominate. ... Greg Comeau / Comeau C++ 4.3.3, ...
    (comp.lang.cpp)