[EMAIL PROTECTED] wrote:
> 
> >While working on the "isaexec" patch I stumbled over the following
> >issue:
> >-- snip --
> >% cstyle -pP isaexec.c
> >isaexec.c: 67: line > 80 characters
> >-- snip --
> >I'd like to propose to relax this limit, making 132 characters the
> >maximum line length in OS/Net.
> 
> NO WAY.

Why ? I did not suggest to reformat all existing code - I am just
begging for the option that people can choose to use a little bit wider
width, maybe even optionally via setting a flag in the source code
header explicitly.

> >The 80 char limit is IMO outdated and it may be less annoying to use a
> >slightly larger boundary here (the isaexec patch itself is no good
> >example why the 80 char limit is bad, but there are other examples where
> >the readability can be seriously damaged by the enforcement of the
> >current length limitation).
> 
> Sorry; my xterms are still 80 characters and for good reason.
> 
> It allows me to have many more xterms open side-by-side so I can see much
> more stuff at the same time.
> 
> By relaxing the limit to 132 characters you are *forcing* everybody to
> switch to 132 wide xterms.

You are wrong.
1. Many of those tools support horizontal scrolling and continuation.
2. I did not say to reformat all existing code. I just asked for the
option to allow wider character widths on demand when it is usefull.
3. I do not expect that everyone will use all 52 (= 132 - 80) additional
characters to the full extend

> (Or to face code which is much less readable
> than code formatted to fit 80 columns; 132 wrapped to 80 is just not
> a pretty sight.

I agree. But if you look at the existing code it's usually only a few
extra characters which are needed to make the code MUCH more readable.

> And while you argue that "132 columns is more readable", the 80 column
> viewer is exposed to code which is much less readable.
> 
> I don't think "modern" comes into this in anyway; if your lines need
> to be 132 columns wide you may be nesting code too deeply, may be using
> identifiers which are too verbose.

The problems I've seen to far are string literals. Another bunch of good
examples could be found in mozilla.org's (see content/ and layout/ - an
enforcement of the ON cstyle would have devastating effects for the
maintability of the code) and KDE's codebase.

> I think you'll find that the C-style rules are not open for negotiation.

What do other Sun engineers think about this ?

> While I don't agree with some of them (such as the () required for return
> and the space after sizeof, in years I have learned to appreciate why
> the rules are there: they give consistent code; this code is consistently
> readable.  Variations in style hamper the accessibility and maintainability;
> the style was fixed in stone precisely because the maintainers vary over
> time.

Yes, I agree that using "cstyle" is a good idea and even that enforcing
cstyle's rules is a good thing. I did not argue that. My primary
complait is the fixed line length which is very ugly. We're not in the
days of Fortran 77 anymore... or are we ?

----

Bye,
Roland

-- 
  __ .  . __
 (o.\ \/ /.o) [EMAIL PROTECTED]
  \__\/\/__/  MPEG specialist, C&&JAVA&&Sun&&Unix programmer
  /O /==\ O\  TEL +49 641 7950090
 (;O/ \/ \O;)
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to