W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > May 1997

Re: PIFLE (was Re: Are PIs useful?)

From: Alex Milowski <lex@www.copsol.com>
Date: Thu, 15 May 1997 09:58:47 -0500 (CDT)
Message-Id: <199705151458.JAA00581@copsol.com>
To: w3c-sgml-wg@w3.org
Rick Jelliffe wrote:
> In any case, I think PIs should be used for more than just simple formatting.
> For example, centerpoint PIs embedded in text to allow smarter page-up/down
> scrolling. Or timeout and caching policy for external entities.
> Or to trigger external entity prefetching when in the locality of some link. These kind
> of things would allow much smoother operation to the user, rather than just 
> the extremes of pull and push.

I simply can't agree on this.  Such information does not belong in the source
document.  You can easily have ancillary document which describe how to use
a document or additional semantics.

In many of the systems we are developing in our company, we are creating 
systems that use *many* document to describe how to accomplish some task or
present some piece of information.  This allows us to create documents that
are reusable.

I can see how this use of processing instructions would be useful but, in my
opinion, it breaks encapsulation.  In the object-oriented programming world,
this usually means you are design one class for something that should be
many.  Applying this rule to the above, you could be trying to describe more
than one set of information in one document.

Now, CONCUR was suppose to be able to allow you to do this.  Unfortunately,
CONCUR can make life quite difficult--for everyone: parsers, authors, etc.  I
like the design pattern of "many documents make up the whole" much better than
compressing information into one.  It seems to work quite well for us.

...again, this is my opinion based on my experience.

I'd be all for formalizing the structure of a processing instruction if we have
to have them.  I'm just not for using them ever.  Hence, I don't want to be
forced to use them unless it is something like the RMD or VERSION declarations.

We should *not* force people to encode document semantics via processing
instructions in XML.

R. Alexander Milowski     http://www.copsol.com/   alex@copsol.com
Copernican Solutions Incorporated                  (612) 379 - 3608
Received on Thursday, 15 May 1997 11:00:22 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 10:04:31 EDT