Re: About the article

Subject: Re: About the article
From: Adam Di Carlo <adam@xxxxxxxxxxx>
Date: 01 May 1999 14:31:45 -0400
"Didier PH Martin" <martind@xxxxxxxxxxxxx> writes:
> As we learned 22 years ago with GUI research, spatial relationship is important for pattern recognition.

Actually, any standards in how DSSSL is represented were set in the
60s as LISP indentation patterns were set.  It has nothing to do in
particular to do with DSSSL.

> Now you know, why all my DSSSL script are more like:
> (bloc
>    (bloc
>      att:val
>    )
> ) 
> than like the traditional (the more containment relationship you have the worse the ")))))" is
> (bloc
>   (bloc
>   att:val))

Actually, standard Emacs scheme mode would be:


You mentioned ease of debugging and finding closing paras.  I find
this to not follow.  My editor is clever enough to always conveniently
indicate what opening paren a closing paren matches (whether on screen
or not).

My intention here is not to start a flame-war on indentation styles,
simply to inform Didier why things are the way they are.  Most people
program from within Emacs -- if you've never seen Emacs scheme mode in
action, then you might have difficulty comprehending some fundamentals
of most peoples' programming environment.

Personally, I think style is style, and up to personal taste.  If you
have a set of people working on one big piece of software, or if you
are patching other software, you should follow whatever conventions
the rest of the source has.  Otherwise, use whatever works for you.
Any decent structured editor will be flexible enough to implement any
style you care to come up with.

> Why I am doing this? because I want DSSSL to have the place it deserve in our world.

Then why do you only write tools that run on non-free platforms?  I've
seen several announcements of your stuff, and it looks interesting,
but I can't run it.

.....Adam Di Carlo....adam@xxxxxxxxxxxxxxxx<URL:>

 DSSSList info and archive:

Current Thread