[XSL-LIST Mailing List Archive Home]
Re: do you use pi's?
Subject: Re: do you use pi's?|
From: "John E. Simpson" <simpson@xxxxxxxxxxx>
Date: Sat, 03 Apr 1999 14:49:52 -0500
At 10:07 AM 4/3/99 -0800, Mark D. Anderson wrote:
>This *really* felt like something that might be done with a PI, but
>i've never used one before. But is there any point?
>I find it really hard to arrive at a definitive criterion for what
>should be in the instance, and what a PI (the example above is deliberately
>sharp; others are harder). And using a PI means that the processor
>(or processors, since there may be a stack) has to start treating
>multiple types of nodes.
>Comments? Does XML need PI's?
Not sure what you mean by "the processor... has to start treating multiple
types of nodes."
But in general I'd say no: XML *per se* doesn't need PIs, any more than it
needs comments. My reading of the purpose of PIs is that they signal to
some specific downstream application that it must do something in
particular, behave in some particular way, enter into some particular state
-- in short must *know* something meaningful to it, but not to an
interpreter (human or machine) of the document's actual content.
In particular, note that a PI's pseudo-attributes (like real ones) are
incapable of expressing any sort of structure: you can't nest attributes.
To answer the question asked in the Subject: line, I myself use only the
"standard" PIs, like <?xsl-stylesheet...>. I'm not a developer of
XSL-processing software; if I were, I think they'd assume much greater
significance to me.
John E. Simpson | The secret of eternal youth
simpson@xxxxxxxxxxx | is arrested development.
http://www.flixml.org | -- Alice Roosevelt Longworth
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list