Frank Bemelman <f...@xs4all.invalid.nl> says...

>It takes more than intelligence, to pick a bunch of quotes making
>GWB look good.

The PR flacks that politicians hire do it all of the time, and 
I wouldn't call them "intelligent."  They do have a sort of low 
cunning...

Spehro Pefhany <s...@interlogdotyou.knowwhat> wrote:

> I think C89 is a subset of C++, but C99 adds new stuff. 

Still wrong.  C was never a subset of standard C++, ever since the
first standard definition of C++.  To give just a few examples:

* C has implicit conversion of all pointers to and from pointer-to-void.
  C++ explicitly disallows that, for reasons that are fundamentally 
  important in its view of the world, but counter-productive in C.

* sizeof('a')==sizeof(int) in C, but it's sizeof(char) in C++.  Which
  usually means they're different.

* C (at least as of C89) still allows "implicit int" declarations.
  C++ never did.

> Of course you can choose to write code that's compatible with both. 

You can, but it's a widely accepted opinion by the C gurus that you
shouldn't.  What point could there be to break a perfectly fine C
program, just to please a foreign tool that was never meant to be used
on C programs in the first place?

-- 
Hans-Bernhard Broeker (b...@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.

On 2004-08-06, CBFalconer <c...@yahoo.com> wrote:

> No.  The C const is peculiar, and effectively means write
> protect.

Why is that peculiar?  Would you prefer it was spelled
"readonly" or somesuch?  I find it extremely useful for placing
look-up tables and the like into ROM rather than RAM.

> The i can't be used in constant expressions, such as array
> sizes.

Ah. You think it should be a compile-time constant (like
Pascal's) rather than a run-time constant?  

> The point is C is not a subset of C++.  They are different
> languages with superficial similarities.

That's definitly true.

-- 
Grant Edwards                   grante             Yow!  I always have fun
                                  at               because I'm out of my
                               visi.com            mind!!!

On Thu, 5 Aug 2004 09:07:24 +0100, Chris Hills <c...@phaedsys.org>
wrote:

>In article <g...@4ax.com>, Anton Erasmus
><n...@nowhere.net> writes
>>Hi,
>>
>>I just need to get my facts straight. I have been given a spec, where
>>the use of C++ and a '51 core MCU is mandatory. AFAIC a C++ compiler
>>for the 51 does not exist. 
>
>
>DON'T DO IT

I am trying to find diplomaticaly acceptable reasons for why we should
not do this.

>The only C++ compiler system for the 8051 AFAIK is the Ceibo one that
>Keil can supply. However it is not very efficient and I can't see any
>reason for using it. 
>
>You C++ will be so restricted that is will be virtually C or it will be
>a VERY small program.
>
>Why is C++ Mandatory? This spec is from some one who clearly does not
>understand what they are asking for.

That is certainly true. The problem is, that it IS the client, and
hence must be kept happy.  I am gathering enough reasonable 
arguments for the changing of the spec.

>BTW which 51 is it?

The spec says any MCU with a '51 core. The existing design which we
are updating is based on a 4 clock/machine cycle 8032 with external
flash and SRAM together with a US$150 EPLD to provide enough I/O.

We want to change to either an AVR without the EPLD or a Philips or
Atmel ARM MCU that has enough capabilities without the need of the
EPLD and all the extra external components. 

Regards
   Anton Erasmus

On Thu, 05 Aug 2004 15:21:53 GMT, Bob Stephens
<s...@earthlink.net> wrote:

>On 5 Aug 2004 11:12:53 GMT, Hans-Bernhard Broeker wrote:
>
>> Anton Erasmus <n...@nowhere.net> wrote:
>>> Hi,
>> 
>>> I just need to get my facts straight. I have been given a spec, where
>>> the use of C++ and a '51 core MCU is mandatory. AFAIC a C++ compiler
>>> for the 51 does not exist. 
>> 
>> It does (see the other replies), but that's not really the answer to
>> the problem.  That combination is like asking an architect to build a
>> family home with a fully service-qualified road inside to go from the
>> couch to the toilet in an 40-ton 18-wheeler.  It's completely crazy
>> idea.
>> 
>> I'd interpret that spec as someone's sick try at a prank.
>
>Sounds more like some junior, pointy haired middle management type whose
>golfing buddy told him C++ was hip.
>

The spec has got all the right buzzwords etc. We have managed to stear
the client towards a more reasonable stance on many points, but where
we have unknowingly stepped on someone toes, it becomes much more
of a diplomatic game in stead of an engineering one.  It is much
easier to say that we believe we have found a typo, since no C++
compiler is available for the '51, than to in any way imply "What
idiot specced a '51 core and C++ as the language"

Regards
   Anton Erasmus

Grant Edwards wrote:
> On 2004-08-06, CBFalconer <c...@yahoo.com> wrote:
> 
>> No.  The C const is peculiar, and effectively means write
>> protect.
> 
> Why is that peculiar?  Would you prefer it was spelled
> "readonly" or somesuch?  I find it extremely useful for placing
> look-up tables and the like into ROM rather than RAM.
> 
>> The i can't be used in constant expressions, such as array
>> sizes.
> 
> Ah. You think it should be a compile-time constant (like
> Pascal's) rather than a run-time constant?

No, I don't think anything.  The point is that is not a real
compile time constant, unlike C++.  Read the C faq.

-- 
"The most amazing achievement of the computer software industry
 is its continuing cancellation of the steady and staggering
 gains made by the computer hardware industry..." - Petroski

martin griffith wrote:

>On Fri, 06 Aug 2004 04:06:54 GMT, in comp.arch.embedded you wrote:
>
>>nappy wrote:
>>> "CBFalconer" <c...@yahoo.com> wrote in message
>>>
>>> > --snip
>
>--
>"The most amazing achievement of the computer software industry
> is its continuing cancellation of the steady and staggering
> gains made by the computer hardware industry..." - Petroski

A more optimistic view is that an amazing achievement of the computer
software industry is its continuing use of the steady and staggering gains
made by the computer hardware industry to implement systems that computer
users desire.

>this intrigued me, so
>http://www.sysprog.net/quotdesi.html

There are some good ones there.  One that I found especially insightful:

Walking on water and developing software from a specification are easy if
both are frozen. (Edward V Berard)

-- 
========================================================================
          Michael Kesti            |  "And like, one and one don't make
                                   |   two, one and one make one."
          m...@gv.net            |          - The Who, Bargain

Anton Erasmus wrote:
> 
... snip ...
> 
> The spec has got all the right buzzwords etc. We have managed to
> stear the client towards a more reasonable stance on many points,
> but where we have unknowingly stepped on someone toes, it becomes
> much more of a diplomatic game in stead of an engineering one. 
> It is much easier to say that we believe we have found a typo,
> since no C++ compiler is available for the '51, than to in any
> way imply "What idiot specced a '51 core and C++ as the language"

Luverly.  By contrast this points out, once more, why I am no
diplomat and need buffering.  I wouldn't even think of the 'typo'
ploy.

Of course in certain circles this spec is an open ended license. 
It requires a compiler, which is not available, and therefore must
be developed (at X units of currency per hour until done).  After
this necessary preliminary we can quote the fixed price for the
actual job.  The compiler development might be inserted by making
the customer responsible for supplying the compiler, and if this
is awkward we will undertake .....  The system will increase the
critical buzzwords per page value.

With proper verbiage you could even supply the totally useless end
product in source form, after development on some fully featured
and standards compliant system.

I told you I was no diplomat. :-)

-- 
"The most amazing achievement of the computer software industry
 is its continuing cancellation of the steady and staggering
 gains made by the computer hardware industry..." - Petroski

On 2004-08-06, CBFalconer <c...@yahoo.com> wrote:

>>> No.  The C const is peculiar, and effectively means write
>>> protect.
>> 
>> Why is that peculiar?  Would you prefer it was spelled
>> "readonly" or somesuch?  I find it extremely useful for placing
>> look-up tables and the like into ROM rather than RAM.
>> 
>>> The i can't be used in constant expressions, such as array
>>> sizes.
>> 
>> Ah. You think it should be a compile-time constant (like
>> Pascal's) rather than a run-time constant?
>
> No, I don't think anything.  The point is that is not a real
> compile time constant, unlike C++.  Read the C faq.

I know the difference, I just don't understand why it's the C
implimentation of const that's "peculiar" instead of the C++
const being "peculair".  Using the term "peculiar" led me to
believe that you thought it somehow violated your (or the
world's) expectations and didn't do what it should.

-- 
Grant Edwards                   grante             Yow!  Now KEN and BARBIE
                                  at               are PERMANENTLY ADDICTED to
                               visi.com            MIND-ALTERING DRUGS...

I am sorry about entering in the middle of the discussion, but I could not
resist anymore.
CBFalconer wrote:
> [...]
>
> The point is C is not a subset of C++.  They are different
> languages with superficial similarities.

1. Though formally C is not a subset of C++, C++ was created with huge
     effort to make it compatible with C. Making C code correct C++ code
    usually is very easy - I did it many times and sometimes it helped to
find
    some unknown bugs - thanks to some more strict rules in C++.
2. C++ is a multy-style languages and one of its use is described as
    "better C".
- so calling their similarities superficial is not true. For more details I
would
    recommend to read:

1. The Design and Evolution of C++ by Bjarne Stroustrup

2. http://wwwold.dkuug.dk/jtc1/sc22/open/n3502.pdf

3. More on topic:
I always prefer to use C++ - even in embedded systems. It is not necessarily
increase
resources needed for executing the program code. Sometimes it even helps
make code smaller and more effective - of cause not directly because of
using C++.
but C++ helps make source code smaller and more maintainable which helps
to find and implement some solutions, which could be much harder with C.

Of cause some techniques that are popular in C++ often not appropriate in
embedded area.

Last time I developed 8051 software (~5 years ago) I could not find C++
compiler. It was ~5 projects and would use C++ for all but one (this one
had CPU with only 128 bytes RAM),

    Regards,
                      Michael Furman