Re: Resiliency To New Data Requirements

Bob Badour wrote:
> Keith H Duggar wrote:
> > It probably is off-topic, however, for general usage I'm
> > forced to partly agree with Dawn that the various *ML's do
> > have semi-structure. I think almost anyone would understand
> > what you meant to communicate if you said something like
> > "plain text is unstructured, relational data is structured,
> > and the stuff in between like HTML is semi-structured". Sure
> > the semi-structure sucks in major ways but the word semi-
> > structured communicates the concept just fine.
>
> I tend to disagree. The physical data structure of a web page is a
> sequential stream. HTML/XML/SGML provide means for creating "logical"
> hierarchies within sequential streams.

I'll agree with Bob - producing a parsable marked-up value from a text (string) value is basically applying substring/instr/indexOf operations to strings, and possibly checking validity through a boolean expressions involving regular expressions. Those are all functions over a string (possible subtype: XML) domain, which are the only way the hierarchies are "seen" by a user (developer).

Since relations offer far more useful structuring capabilities, there's no reason to elevate these hierarchies to "top-tier" structures. They are where they belong, in types, even if their use is often ill-advised. They're not structures, just values; they are "seen" as structures the same way any other type is "seen" as "composed" of more primitive values via the lens of functions over that type.

• erk