Re: [docbook-apps] Re: Simpler XHTML output

["John L. Clark" <jlc6@po.cwru.edu>]

On Tue, Apr 26, 2005 at 11:36:17AM -0400, Norman Walsh wrote:
> I'm not sure what the right answer is...

> I'd be reluctant to make large changes to the current HTML output
> as there are *a lot* of pages out there that may rely on the current
> functionality.

I think this is quite reasonable...

> I'd be reluctant to fork the XHTML stylesheets and maintain them
> separately from the HTML stylesheets.

... but this I don't understand.  There are clearly different profiles
of HTML; the existence of xhtml is merely one dimension of that
profiling.  There are also certainly a lot of options for controlling
output from the current (x)html stylesheets.  I think that even with all
of these options, though, the current stylesheets really only output one
profile, which is "in spirit" similar to xhtml-transitional.  I think
forking the current HTML sources to create a new profile (along the
lines of xhtml-strict) would address some of the desire for
structurally-rich, CSS-controlled xhtml output.

Here's the directory structure I imagine:

  ...
  html                  Currently supported html output
  xhtml-transitional    Automatically generated from html; corresponds
                        to the current 'xhtml' directory
  xhtml-strict          Completely different profile supporting
                        "structural" xhtml with hooks for CSS

I think that this last profile should be treated as an independent
module of the stylesheets because it has such different design
considerations.  It will probably share some core functionality (and
hence, code) in common with the current HTML output stylesheets.
Perhaps that is the reason for the hesitation in forking, but I would
hope that such common code could be independently managed (e.g. in an
xhtml-common directory that could be automatically generated from a
subset of the html code).

Take care,

    John L. Clark

PGP signature