NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c++-part2/16797

Internet Message Format | 1996-08-05 | 3.4 KB

Path: news.zocalo.net!news
From: Paul Hsieh <[e-mail]>
Newsgroups: comp.lang.c++,rec.games.programmer,alt.msdos.programmer,comp.programming
Subject: Re: Young programmers read me.
Date: Thu, 11 Apr 1996 20:55:30 -0700
Organization: Zocalo Engineering - Berkeley, California, USA
Message-ID: <[e-mail]>
References: <4kj3rp$[e-mail]> <[e-mail]>
NNTP-Posting-Host: paulh.chromatic.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.0 (Win95; I)

My personal feeling on coding style is that it should reflect what you
are doing.  To that end I try to use the latitude given to me by the
ability to have arbitrary white space in my code to identify "scope".
I think this was the entire intention of K&R when they designed C and
gave their samples.

> >: >if(something_happened)
> >: >{
> >: >     do_something_else();
> >: >}
> >: >is_n't_it();   :)

The code above is practically unreadable to me, because the lone "{"
confuses me about where the actual scope begins and ends.  Alternatively,

if(something) {
    do_something_special();    // The do_something_special function
} // endif

Is_not_what_was_typed_above();    // This is called regardless.

In this way, the beginning and end clause appear more symmetrical and
each line is self describing in what they do.  A single line that says
"{" doesn't describe anything.  It can't be commented beyond saying
"// beginning of scope" which would be really redunant to be looking
at sprinkled all over your code.

> >       Sure, I can understand that this might be a problem if you're
> >still using edlin. :)  This is really a personal preference issue and
> >nobody is going to change anyone else's mind.  My problem with the less
> >space argument is that it implies that you don't comment your code either.

I'll take the previous poster's personal preference over you *implication*
any day.  I'm a concise but descriptive camper myself.  Code is something
to be written and debugged.  The less time I spend hitting the down arrow
key or the pagedown key and the less time I spent staring at blank white
space that tells me nothing, the better.

> >Not to mention the fact that you might as well shove most of your code
> >onto the same lines, only line-breaking when absolutely needed.
>
> >void foo(int i, char *list){int t;for(t=0;t<i;t++){printf("#%i",t);
> >printf("%s\n",list[t]);}printf("%i total.\n",t);}

I would prefer this code to the code above if it were better commented and
used more descriptive tokens.  If you write code in this form though, you
have simply declared that the subtleties of how it works is unimportant and
that you are highly confident of its functionality; its only good for the
global symbol it defines.  Perfectly acceptable, just comment it.

> >       Wow!  Look at how much space that saved.  I could put all my code
> >in just two or three pages.  :)  I much prefer.
>
> There is a neat feature in Amos on the Amiga called "fold", well, really
> it is a feature of the Amos editor, to pack functions or expand them as
> you require.

There are utilities for the PC for obfusticating and de-obfusticating your
C-code as well.  AN obfusticator is sitting on the WATCO

--
Paul Hsieh
[e-mail]
Graphics Programmer
Chromatic Research

What I say and what my company says are not always the same thing