Re: C Language

[Simon Coulter <shc@xxxxxxxxxxxxxxxxx>]

On 08/09/2004, at 7:48 AM, Dave Odom wrote:

> However, I'm not sure either of them have the power of Assembler or
> "Assembler in Cursive": C/C++ as only they can address memory and you'd
> better know what you're doing as it won't give you Mickysoft "are you
> sure you want to do that" types of messages.   It ASSUMES you are a
> professional and know what your doing.   Whether that's a good
> assumption is another story.

Any language that supports pointers let's you address memory thus 
memory access is not a prerogative of C (nor its incremental 
derivative). Even COBOL and RPG can provide arbitrary access to memory.

C's major problem is that it makes no distinction between pointers, 
arrays, or strings. That, along with no idea about fixed-length 
character variables, no idea of true varying-length character 
variables, no decimal data type, excessive data typing, the treatment 
of character as a numeric data type, and a number of other things makes 
it unsuitable for anything except truly low-level code where an 
assembler would be ideal but you'd like a bit more function than an 
assembler generally provides.

C is OK (not good) for low-level stuff but crap for anything higher. 
The incremental derivative of C is no better at high-level code because 
it has its own collection of unexpected behaviours, traps, and flaws. 
The cost in time of developing in C is higher than almost any other 
language.

There is a deal of difference between expecting the user of a 
programming tool to be a professional and providing absolutely no 
support from the compiler. Even seasoned C professionals get bitten by 
the if (x=y) rubbish and any number of other stupid things that C 
compiles--hence the proliferation of Lint-like tools to catch the sort 
of stupidities that a decent compiler could be expected to reject.

PL/1 will do everything C can do but still manages to provide 
protection to the programmer. If the programmer chooses to use pointers 
to directly access storage in PL/1 they can (and are subject to the 
same caveats as doing this in C) but it is a conscious choice of the 
programmer rather than the compiler forcing that level of care on the 
programmer because the original C compiler writers wanted to make life 
easy for themselves.

Regards,
Simon Coulter.
--------------------------------------------------------------------
   FlyByNight Software         AS/400 Technical Specialists

   http://www.flybynight.com.au/
   Phone: +61 3 9419 0175   Mobile: +61 0411 091 400        /"\
   Fax:   +61 3 9419 0175                                   \ /
                                                             X
                 ASCII Ribbon campaign against HTML E-Mail  / \
--------------------------------------------------------------------