RE: [xml-dev] SGML complexity (was: RE: [xml-dev] Re: Recognizing...)

["Len Bullard" <cbullard@xxxxxxxxxx>]

On the other hand, the source for SGMLS was out there.  Would you write a
new compiler for Java just to use the bytecodes? Ah, well Microsoft did that
already and renamed it too.  Sound familiar?

No, I think it came down to politics and egos but history isn't a reason to
undo a good thing.  So far, XML is a Good Thing.  I just find all of the
'one little feature more' re-threads fun to watch because they so neatly
replicate the original SGML On The Web discussions in reverse.  There is
something Marxian about it but I'm not sure if that is Karl or Groucho.

Usually, more is more.

len


From: Tatu Saloranta [mailto:cowtowncoder@yahoo.com] 

> So the complaint is that SGML has a lot of features
> and buying a
> fully-conforming system is expensive?  I can't
> quarrel with that.  SGML
> systems were expensive and that was the real barrier
> to its adoption, not
> its complexity.  Does anyone really believe anymore
> that it was too complex
> or merely too complex for a Desperate Perl Hacker?

There is another side to the complexity, beyond the
question whether it's too complex to use. I don't
doubt that many DPHs (heck, perhaps even PHBs) would
be able to use SGML, as in manually write and debug
content marked up with it. Just like people write c++
code, despite complexity of the language.
But what I do doubt is that many developers would be
able to develop systems to process it. Like someone
already said it, writing an XML parser  is a
non-trivial task -- writing an SGML parser is a
10-year-backed-up-by-IBM kinda project, and ditto for
processing systems that make use of the features.
If I had to choose between task of writing an SGML
processor, or C++ compiler, I wouldn't know which one
would be less work.

Considering complexity of a processing system that
allows using the full power of SGML (which in the end
might give more power to the user), it's no wonder
systems were expensive. So thus perhaps SGML just
didn't and wouldn't offer that much more, to justify
the extra overhead all around?

Sometimes less is more,