[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: XCatalog

• From: "Rick Jelliffe" <ricko@a...>
• To: "XML Dev" <xml-dev@i...>
• Date: Thu, 18 Mar 1999 11:35:12 +1100

From: David Megginson <david@m...>

>That same issue might be what keeps DTD syntax alive for simple uses
>-- sure, DTDs aren't all that powerful compared to most of the new
>proposals out there, but it's compact and easy to parse and it doesn't
>require you to process yet another XML document (it's also supported
>by lots of existing software and is backed up by approved ISO and W3C
>specs).

Furthermore, I think there is human factors element involved: when there
is a change in domain, (at least some) people expect or need this to be
flagged by using a difference syntax. When the data is highly cohesive,
it is natural to couple it syntactically to distinguish it from the
markup in which it is embedded.

I think it is a really important design principle, and too easily
dismissed. It helps explain
* why are URLs not factored out into attributes? &
* why are scripting languages not in instance notation? &
* why are the patterns in XSL not in instance notation? &
* why don't people like LISP syntax (i.e., does the unified syntax
actually cause reading panic in newcomers: it has been widely commented
that computer languages with different syntaxes each for assignment,
declarations, infix maths, and prefix functions, such as C and ALGOL
family languages have been much more successful)?

Apart from that, there is an issue of the perceived cost-benefit of
metadata (i.e. schemas) compared to data. If I have a 1K Docbook
instance and a 600K Docbook Schema in instance syntax, it is a strong
dissuader against using that schema. So I don't think that terseness is
of minimal importance for schemas: appropraite balance to the size of
the instance is, however.

There is also clearly some contradictory factors at work too: some
people feel saturated with too many little language and love a unified
syntax. You can see their personality, in this regard, reflected in the
DTDs they write.

Rick Jelliffe



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@i...
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo@i... the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo@i... the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@i...)

• Follow-Ups:
  • Re: XCatalog
    • From: Lars Marius Garshol <larsga@i...>
  • RE: XCatalog
    • From: "Jeffrey E. Sussna" <jes@k...>

• Prev by Date: Re: Need XML documents located on web servers for testing
• Next by Date: Re: Why is this JC test not-wf?
• Previous by thread: Re: XCatalog
• Next by thread: RE: XCatalog
• Index(es):
  • Date
  • Thread