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

RE: [SML] Numeric character references

• From: "Gavin Thomas Nicol" <gtn@e...>
• To: <xml-dev@i...>
• Date: Sun, 28 Nov 1999 16:28:26 -0500

> Minimalism *does* have long-term advantages to the end user
> because the tools they use will tend to be less buggy, delivered on time,
> etc.

There is probably some degree of truth to this, but I'm not sure
it would be orders of magnitude.

> More to the point, the higher-level abstractions and APIs built on
> top of SML will be easier to use because they don't have to hide
> as much underlying complexity as XML APIs do.

I really stronly disgree with this.

The DOM is a good example if a general-purpose/application independent
API. Very complex, with many special cases. In general though, an API,
by definition, is *application* specific. For a given application,
I cannot see why the API needs to reflect the underlying
implementation of the system. For example, I just implemented a
fairly complex security system that has calls like:

   Principal authenticate(String id, String key);

and it has 3 different implementations: one using XML, one using ODBC,
and another using LDAP. From an *application* perspective, these
differences are irrelevant.

You know that I was, and still am, in favor of having an API
other than the DOM which would allow *simpler* optimization for
performance. That doesn't imply the requirement for something like
SML to me though....



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 unsubscribe, mailto:majordomo@i... the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo@i... the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@i...)

• References:
  • Re: [SML] Numeric character references
    • From: "Michael Champion" <mike.champion@s...>

• Prev by Date: RE: Handheld computers
• Next by Date: RE: [SML] Whether to support Attribute or not?
• Previous by thread: Re: [SML] Numeric character references
• Next by thread: RE: Forget SML. We need RTSML.
• Index(es):
  • Date
  • Thread