Re: Capitalizing on HTML (was Re: equivalent power in SGML and XML)
I realize this thread has gone pretty cold. Just wanted to add a couple
At 10:26 AM 9/19/96 -0400, Paul Prescod wrote:
>If it were not for the HTML experience, I might not think that Jane Notepad
>is worth supporting. But vi/emacs/notepad/hotdog/bbedit is still the editor
>of choice for HTML and if we want to win over that crowd, XML should be at
>least as easy to write by hand (with a simple DTD).
>If the target user community had demonstrably given up hand-coding, we could
>write a language that makes that process arduous. But right now, I think that
>doing so would delay XML's acceptance by end users and vendors. If XML code
>looks too verbose, the major browser vendors may even shun it on the
>presumption that the users would not embrace it. In other words, XML might
>be declared DOA (by the Vendors that Be) before the editing tools are
>That doesn't mean that we have to support markup minimization options. I
>think that getting rid of them makes hand-coding easier and less
>error-prone. On the other hand, getting rid of mixed content would make
>hand-coding more difficult and probably more error prone than a language
>that allows mixed content but has simple RS/RE rules.
> Paul Prescod
I agree that markup minimization offers dubious benefits for most
hand-coding, and that we should avoid egregious pilings-on of required
markup for simple cases. Our principles support this nicely.
I do think that your note about "(with a simple DTD)" is telling. In
the scenarios we're considering, the relative efficiency of hand-coding
has a lot more to do with DTD design than with any other factors, so it's
not worth spending much effort to optimize hand-coding in generic XML
anyway. HTML seems to have been designed (and I use this word loosely)
with hand-coding as one of its principles; other SGML/XML DTDs have
other primary purposes.