In article <2rakm3$h66@paperboy.gsfc.nasa.gov> jf@cen.com (John Farley) writes:
>maxtal@physics.su.OZ.AU (John Max Skaller) writes:
>
>>         Very well. I will propose that templates and exceptions
>> and namespaces be removed from the Standard.
>
>I consider templates and exceptions to be part of the CURRENT feature set. They
>have been around for quite awhile. These are exactly the sorts of features that
>desperately need to be standardized. Templates are useable these days, although
>the implimentations differ from compiler to compiler. You can't possibly
>consider template code to be portable! IMHO, exceptions are currently unusable.
>Even though I would love to use them, I wouldn't put them in production code.
>Namespaces solve a problem common to all C++ programmers.

 So do all the other extensions I've seen -- both ones accepted
and rejected solved (or tried to) common problems.
>
>These are the types of problems that the Standards Committee should be
>solving for the users.  The point that I was trying to make is that the
>Standards Committee should NOT be trying to solve the "C++ compiles are slow"
>problem.

 Well I agree. But we must be sure when extending the
language that efficient implementations fo the extensions are possible.
For example, the order of initialisation problem requires some sort
of topological sort of dependencies at link time.

>Most of the language issues that have been resolved are quite good
>(although the RTTI still affects me like fingernails on a blackboard). What is
>needed now is to get the Standard approved and to get the features into the
>compilers as soon as possible.

 Get what 'Standard' approved? Its one thing to extend the
language, its another to describe it. Its not the extensions that
are causing problems but description of the most basic features.
"Simple things" like name lookup are harder to specify than you think.


--
        JOHN (MAX) SKALLER,         INTERNET:maxtal@suphys.physics.su.oz.au
 Maxtal Pty Ltd,      CSERVE:10236.1703
        6 MacKay St ASHFIELD,     Mem: SA IT/9/22,SC22/WG21
        NSW 2131, AUSTRALIA

In article <2r09d3$324@paperboy.gsfc.nasa.gov> jf@cen.com (John Farley) writes:
>Sorry, I dropped a line on the previous post:
>
>> As a full-full-time user of C++, I would rather have the compiler writers
>> of the world concentrating on getting the CURRENT feature set to work
>> correctly, than to spend their time changing the the compilers to meet new
>> language features that solve yesterdays problems! I especially don't want
>
>the standardization process to be slowed down.
>

 Very well. I will propose that templates and exceptions
and namespaces be removed from the Standard.

 Then I shall consider whether to buy and Eiffel or Ada
compiler, because C++ will be clearly inadequate for my needs.


--
        JOHN (MAX) SKALLER,         INTERNET:maxtal@suphys.physics.su.oz.au
 Maxtal Pty Ltd,      CSERVE:10236.1703
        6 MacKay St ASHFIELD,     Mem: SA IT/9/22,SC22/WG21
        NSW 2131, AUSTRALIA

maxtal@physics.su.OZ.AU (John Max Skaller) writes:

> In article <2r09d3$34@paperboy.gsfc.nasa.gov> jf@cen.com (John Farley) writes:
> ...
> >> As a full-full-time user of C++, I would rather have the compiler writers
> >> of the world concentrating on getting the CURRENT feature set to work
> >> correctly, than to spend their time changing the the compilers to meet new
> >> language features that solve yesterdays problems! I especially don't want
> >
> >the standardization process to be slowed down.
>
>
>         Very well. I will propose that templates and exceptions
> and namespaces be removed from the Standard.

I consider templates and exceptions to be part of the CURRENT feature set. They
have been around for quite awhile. These are exactly the sorts of features that
desperately need to be standardized. Templates are useable these days, although
the implimentations differ from compiler to compiler. You can't possibly
consider template code to be portable! IMHO, exceptions are currently unusable.
Even though I would love to use them, I wouldn't put them in production code.
Namespaces solve a problem common to all C++ programmers.

These are the types of problems that the Standards Committee should be
solving for the users.  The point that I was trying to make is that the
Standards Committee should NOT be trying to solve the "C++ compiles are slow"
problem. Most of the language issues that have been resolved are quite good
(although the RTTI still affects me like fingernails on a blackboard). What is
needed now is to get the Standard approved and to get the features into the
compilers as soon as possible.

Sorry, I dropped a line on the previous post:


> As a full-full-time user of C++, I would rather have the compiler writers
> of the world concentrating on getting the CURRENT feature set to work
> correctly, than to spend their time changing the the compilers to meet new
> language features that solve yesterdays problems! I especially don't want

the standardization process to be slowed down.