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

Re: Paranoia and Slippery Slopes (was Re: bohemians,

• To: <xml-dev@l...>
• Subject: Re: Paranoia and Slippery Slopes (was Re: bohemians, gentry)
• From: "Karl Waclawek" <karl@w...>
• Date: Thu, 5 Dec 2002 12:05:11 -0500
• References: <B885BEDCB3664E4AB1C72F1D85CB29F8040EF82E@R...>

> Thousands of developers using Microsoft technologies benefit from viewing XML as strongly typed
> data. Users of the .NET Framework DataSet, ASP.NET web services and SQLXML all utilize XML
> as strongly typed data. However these people are rarely markup weenies thus you won't find them on
> XML-DEV arguing the finer points of the PSVI.
>
> The question isn't whether there are scenarios where strongly typed uses of XML exist or are
beneficial
> because they do and are but whether users of XML in loosely typed  or untyped scenarios should be
> penalized for this.

Response from a developer:

I use "typed" processing all the time, but I would like to keep it modular.
No data types in XML, just don't see the necessity.

With most networked apps you can agree on the exact semantics of the content - and describe
it separately in whatever way suits the parties involved best. There is always something
a Schema language cannot describe, and conceivable there can always be a context where a
document's schema should be different or does not matter.

For general interoperability (if that can be achieved) you probably need
Artificial Intelligence, and then it would not even matter if you use XML or not.

Karl

• References:
  • RE: Paranoia and Slippery Slopes (was Re: bohemians, gentry)
    • From: "Dare Obasanjo" <dareo@m...>

• Prev by Date: Re: bohemians, gentry
• Next by Date: Re: Typing and paranoia
• Previous by thread: Re: Paranoia and Slippery Slopes (was Re: bohemians,gentry)
• Next by thread: RE: Paranoia and Slippery Slopes (was Re: bohemians, gentry)
• Index(es):
  • Date
  • Thread