It has been pointed out to me that I spend a lot of time criticizing (some
might say "bashing") C++.

In order to correct any misimpression I might have created, please allow
me to just suggest that people should think about why I bother.

The reason is that I actually think that C++ has an awful lot of good
points, and that it has an awful lot going for it.  If I didn't believe
that, I'd just ignore the language completely and be done with it.

But I don't ignore it.  Rather, I care (sometimes passionately) about it.
I believe it to be fundamentally a good language, but I'd like to see it
be better... much better.

The problems with the language (which I and others have noted) are (in no
particular order):

 a)  It isn't fully defined yet (because there is no standard yet).

 b)  The implementations are not all consistantly robust yet.

 c)  The complexity of the language is relatively high.

 d)  Some aspects of the language are perhaps less than perfect.

Points (a) and (b) are, of course, being worked on as we speak.  Point (c)
is (to some degree) an inevitable part of ANY sophisticated programming
tool.  Point (d) involves things which have been (and shall be) debated
both here and elsewhere for a long time to come (just as for all other
programming languages I know of).

The point I wish to make is that notwithstanding the above issues, I still
feel quite strongly that C++ brings a lot to the party in terms of new
capabilities to write efficient, maintainable, adaptable, and reliable
software.  I wish that the language was better than it is, and I shall
continue to note points about it where it is not as good as (in my opinion)
it could be, but I do so only because I feel that the language can and
should be used (and even improved, where possible).

--

-- Ronald F. Guilmette ------------------------------------------------------
------ domain address: rfg@netcom.com ---------------------------------------
------ uucp address: ...!uunet!netcom.com!rfg -------------------------------

In article <rfgCEA93t.9vo@netcom.com>, rfg@netcom.com (Ronald F. Guilmette) writes:
|> It has been pointed out to me that I spend a lot of time criticizing (some
|> might say "bashing") C++.
|>
|> In order to correct any misimpression I might have created, please allow
|> me to just suggest that people should think about why I bother.
|>
|> The reason is that I actually think that C++ has an awful lot of good
|> points, and that it has an awful lot going for it.  If I didn't believe
|> that, I'd just ignore the language completely and be done with it.
|>
|> But I don't ignore it.  Rather, I care (sometimes passionately) about it.
|> I believe it to be fundamentally a good language, but I'd like to see it
|> be better... much better.
|>
|> The problems with the language (which I and others have noted) are (in no
|> particular order):
|>
|>  a)  It isn't fully defined yet (because there is no standard yet).
|>
|>  b)  The implementations are not all consistantly robust yet.
|>
|>  c)  The complexity of the language is relatively high.
|>
|>  d)  Some aspects of the language are perhaps less than perfect.
|>
|> Points (a) and (b) are, of course, being worked on as we speak.  Point (c)
|> is (to some degree) an inevitable part of ANY sophisticated programming
|> tool.  Point (d) involves things which have been (and shall be) debated
|> both here and elsewhere for a long time to come (just as for all other
|> programming languages I know of).
|>
|> The point I wish to make is that notwithstanding the above issues, I still
|> feel quite strongly that C++ brings a lot to the party in terms of new
|> capabilities to write efficient, maintainable, adaptable, and reliable
|> software.  I wish that the language was better than it is, and I shall
|> continue to note points about it where it is not as good as (in my opinion)
|> it could be, but I do so only because I feel that the language can and
|> should be used (and even improved, where possible).
|>
|> --
|>
|> -- Ronald F. Guilmette ------------------------------------------------------
|> ------ domain address: rfg@netcom.com ---------------------------------------
|> ------ uucp address: ...!uunet!netcom.com!rfg -------------------------------

Dear Ronald,

I am completely agree to criticize the language, standarts are useful
to get human communication platforms, however when a human knows some technique
he observes always that he may do better things!

observation 1: The world is dynamic

observation 2: The human has limited understanding resources
  (resources are always expensive)

observation 3: Any language must be a communication tool (for humans)

Next, do you know any computer language which is "robuuuust"!

Further, complexity can be handle by us but every body should do it
at his own speed.

At last, our world is not perfect why do you expect a language is perfect?
I think the notion of perfection is "opportunistic". When we solve a problem
we have always the best reason to do it. Most of the reasons are around us.

Now on the technical viewpoint, C++ is one of the best language we have (at the moment)
and I will not change my mind so easily, however I have some smart critics:

a) How to use nice features provided in high level languages

 - pattern matching like in SML should be interesting
  (bad think C++ is an imperative language)
 - unification of Prolog should be interesting.

b) Which language is object oriented? No one!!!

 - What about active message passing?
 - What about getting persistence integrated ?
 - Why not a flexible garbage collector?
 - We have still no good templates.
 - What about other inheritance mechanisms?
 - Why numbers are not objects?
 - Why not agent concepts which require at least the ADA task capabilities?
 - ....

I think the language is not complex enough, but I like it.

Sincerely,
  Bernard De Cuyper
  bmcuyper@etro.vub.ac.be