Re: [Bug-gnupedia] just HTML??

[Hook]

David Tanzer wrote:
> Tom Chance wrote:
>
> > I've been talking to RMS who agrees that it would be best to
> > keep it simple with HTML (not XML or other text
> > display technologies), not only because it will work
> > with any browser (even lynx!) but also because it will
> > be easy to change into other formats should we wish
> > to. Especially easy if we display with HTML
> > (compatible) and save data with a database, not a
> > large number of files.
>
> This is is just an argument for using HTML for the presentation
> format.  But it would be incredibly shortsighted to require that articles
> can only be submitted in HTML.  All that should be required for a
> submission format X is that there be a standardized way of converting X
> to HTML.
>
> For example, it would very desirable to accept submissions in
> Latex.  The latex source is much more intelligible than the HTML
> formatting commands into which it gets converted.  By posting
> that source as part of the encyclopedia, the reusability and
> modifiability of the document is greatly enhanced.
>
> Since XML is essentially HTML with some arbitrary restrictions
> removed--whereas HTML has a fixed set of tags, XML
> allows the author the freedom to use whatever tags are most
> descriptive--it is more natural and flexible for human _writers_.
> The tagging allows that writers to add descriptive metadata,
> and this makes the information reusable for new, unanticipated
> applications in the future.  I.e., it lays the foundation for
> the subsequent _querying_ of the encyclopedia.

And this very flexibility is one of the big issues with XML. I can easily
(and, in my own mind at least) justifiably invent all sorts of new tags for
my definitive article on Soviet nuclear powered satellites. Searching for
articles based on an open choice of tags is certainly no simpler than a
conventional free text search, and may end up more complex.

> So with XML open to the writers (which includes HTML as
> a subset), and HTML needed for the browsers, all that is needed
> is a conversion script.  If authors choose to use their own
> XML schema, we could ask that they send in the script that
> converts it to HTML.

Which realistically, you won't often get. An XML DTD with the tools to make
use of it is by far the easiest solution. The DTD I'm sure we can come up
with (recognising that it's going to evolve as we see the need), the tools
are a problem right now. I can't think of a single word processor which
outputs usable XML.

Paul