Re: funny article

From: erk <>
Date: 9 Dec 2004 10:49:18 -0800
Message-ID: <>

[much snippage]

Gene wrote:

> Eric wrote:
>> I'm still unclear about how a "data exchange model" differs from a
>> model
> Dawn wrote:
>>'cause it doesn't ;-)
> Oh, but it does. A data model has semantics. Data exchange only

Cute, and true. I'm still toying with my idea of a particular "data exchange message", like a UI, being the result of overlaying a "path pattern" on top of proper relations. [Much fuzziness in my description, obviously.]

> Dawn:
> >data model (not from all perspectives) it seemed to work better than
> >relational structures -- why? I suspect that some XML research is
> > motivated.
> Adhockery.

I'm not criticizing research or researchers in general to suggest that much of it (in many fields, but especially CS) strikes me as apologetics: explaining and layering atop what's already there, without much attempt to reach any underlying truths. Of course, CS is harder because the underlying truths are, to a large degree, subject to definition; we can invent the logics we want.

> >an unearned status is not unlike a similar elevation of relational
theory a
> >couple of decades ago.
> No, I do not think it is connected at all. RM makes sense. XML
is an ugly,
> ugly kludge.

XML for marking-up a document makes some sense for data exchange, since it's "data-lite" (basically waving your hand at what, in a big long String, might be actual data and some hint at what sort of data it is). But I consider it merely a handy (?) front-end processor; any further extension is heinous.

XSLT hints at usefulness, but devolves into some of the most butt-ugly crap I've ever seen. The syntax utterly overwhelms its value, and the reason is that it's based on XML. An language to transform Data Format X into something else is best NOT expressed in Data Format X. That would be like using exclusively relations to express relational queries; it can be done, but it simply ugly.

> I would rather start with a foundation. I have solved some
> problem very quickly because I had the appropriate theory to apply.
> Those who did not floundered.

> Good theory can be *extremely* practical.

Agreed ^ 2.

  • erk
Received on Thu Dec 09 2004 - 19:49:18 CET

Original text of this message