Ross Gardler a écrit :
Nicola Ken Barozzi wrote:
Thorsten Scherler wrote:
...
Summary:
A forrest:hook is an abstract layer for *any* "graphical helper
container" (not only html), like an "interface". Each format can
implement special matcher and transformer which decide how and whether
forrest:hooks get transformed in the final output format.
Wait a second, I don't mean that writing <span> I want that outputted
directly.
What I really mean is:
<forrest:span forrest:class="testClass"> </forrest:span>
Using a default namespace, it becomes:
<span class="testClass"> </span>
I have to admit that my first response was to agree with Nicola, but
Thorsstens reply convinced me otherwise. The main reason I like
Thorstens approach is that it keeps the language small.
I think the argument is a little bogus (less HTML dependant, as Nicola
says it's just naming), but with Thorstens version we will not need to
invent new tags for new formats, whereas with Nicolas we will.
I also agree with Thorsten, the main thing is that this solution solve
the previous problem of tag closure.
I forgot the language independance that we want to keep.
I think that's because we talked about type="span" and type="div" which
are well known for XHTML writer.
We might have talked about type="inline" and type="block" but it would
have been a little bit <fo> oriented :-) .
Maybe we should find others words to clearly means that we are not
systematically going to produce XHTML or FO or IDon'tKnowWhat.
The concepts of div and span in XHTML can be respectively assimilated to
block and inline in <fo>.
Maybe something like type="layer" - for div, block and type="flow" for
span or inline ...
Regards,
Cyriaque,
Ross
