comp.lang.ada
 help / color / mirror / Atom feed
From: Marin David Condic <mcondic.auntie.spam@acm.org>
Subject: Re: The Ada experience.
Date: Sat, 07 Sep 2002 09:55:28 -0400
Date: 2002-09-07T13:55:54+00:00	[thread overview]
Message-ID: <3D7A054F.25E84ECF@acm.org> (raw)
In-Reply-To: 4519e058.0209062259.1ab868de@posting.google.com

Your complaints about VisualC++ are fair, but there are two perspectives to look at. MSVC++ is very feature-rich and well
integrated. Gnat as a compiler is a command-line oriented thing with some add-on tools that might make it a bit more of
an IDE. (Waiting to see what GPS provides.) In this sense, I can understand the complaint about it looking like something
from the '60s in comparison to MSVC++.

The other perspective is that even though MSVC++ is a feature-rich, well integrated set of tools, the execution thereof
can be quite poor. Throw on top of it the naturally difficult syntax/semantics of C++ and all the traps that implies plus
the awkward and butt-ugly cripcrap that MSVC++ sticks into the code so it can connect up to the GUI building thing and
you've got a royal mess. It can make you yearn for the '60s where all you wanted to do was develop for a command line.
:-)

MDC
--
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jast.mil/

Send Replies To: m c o n d i c @ a c m . o r g

    "I'd trade it all for just a little more"
        --  Charles Montgomery Burns, [4F10]
======================================================================



Ted Dennison wrote:

>
>
> I'm guessing you are comparing it to VisualC++, no?
>
> I suppose it depends on your perspective. For me, developing using
> VisualC++ (version 6.0) feels like playing with a toy compared to the
> Gnat toolkit. The editor is nice looking, but not nearly functional
> enough. The compiler doesn't implement the language properly (or even
> very well). Template support is a particular disaster. It has
> gawd-aful error messages, which causes the user way more work that
> should be nessecary to figure out what they did wrong (its not
> uncommon for something simple like a missing semicolon to pull an
> error somewhere deep in a standard library header file). It comes with
> a semi-functional non-standard version of the STL. To top it off, its
> chock full of code-generation bugs, which act as little day-wasting
> land mines for you to periodicly step on. And good luck trying to get
> Bill G. to fix them for you, or getting the sources to fix them
> yourself.
>
> The code builder is too moronic figure out recompilation dependancies
> without manual user intervention (which is just begging for errors).
> Maintaining all those dependancies manually in a large project is
> simply impossible.
>
> But I'll grant you, its a nice *shiny* toy. :-)




  reply	other threads:[~2002-09-07 13:55 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-09-01 21:41 The Ada experience henderson was no auteur
2002-09-02 22:19 ` jim hopper
2002-09-04 10:40 ` Preben Randhol
2002-09-07  6:59 ` Ted Dennison
2002-09-07 13:55   ` Marin David Condic [this message]
2002-09-09 11:41     ` Marc A. Criley
2002-09-09 13:51       ` SteveD
2002-09-09 17:36       ` tmoran
  -- strict thread matches above, loose matches on Subject: below --
1996-07-23  0:00 Kenneth Mays
replies disabled

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox