Meta-Orders (Spunoff: comments on 'border-radius')

> 
> Allan Sandfeld Jensen wrote:
> 
>> Maybe a way to use ECMAscript functions in generated content could be 
>> used, or we should just let the authors let scripts generated the 
>> entire list, if they are too lazy to do enumeration manually.
> 
> Philip TAYLOR wrote:
> "Too lazy" ? That seems a very odd accusation.  Surely an author
> who asks HTMML/CSS to enumerate his lists, rather than just typing
> in 1), 2), 3) ..., is exactly ths opposite of "lazy".  He is doing
> the right thing.  Whether or not it is feasible (or even sensible)
> for HTML+CSS to attempt to be able to correctly enumerate all the
> world's languages is an entirely different matter ...
> 

PT... I don't know about that.  I'm no expert, and not even in the 
ballpark... but in a way, the numerics that order a piece of data to 
show up in a certain place amongst its group... is metadata.  Then 
again, when has metadata ever been allowed to affect presentation OR 
structure?  Or does it all the time?

I've heard the word "assertion" used in RDF land a few times, but I'm 
not absolutely sure how it applies to RDF, or this conversation.  Isn't 
an OL element doing an assertion?  It seems to know three things.... and 
from those three, asserts a presentational flow.  The three things the 
OL knows... maybe are...

1. what LANG namespace its operating within
2. the ACTUAL order of its childnodes
3. the author wanted these number-ordered during display

Each childnode, standing ALONE in the metadata world, has no use for a 
numeric order... unless it knows (carries) its PLACEMENT within a group 
of other data... and maybe its PLACEMENT within a namespace realm. 
Should a wayfaring element be able to report...

  "I am the third of seven in an ordered list id'd as 'mylist' whose 
home is within namespace 'myweb.com'"?

The element is not part of that group now... or is it?  SOME say web 
node-cullers SHOULD be able to grab most-any node/element out of 
context, and use it in any way they'd like.  Authors often feel that 
this is wrong, and would, if possible, copyright-kybosh a few of these 
cullers for using THEIR nodes out of context.  They won't allow the node 
to be transcluded, locking it up at the content management layer.  ie. 
Thoust shall NOT take, nor display MY element, without taking it 
CORRECTLY from my nodeserver. (power trip thing) The author or 
content-manager may FORCE the parental OL to go along, or NOBODY IS GOING.

There is the possibility that the containing OL of the wandering LI 
might hold the author-preferred style FOR all the LI's beneath, as well. 
  The LI's author doesn't want his LI to be culled-away without taking 
the stylings with it.  Its the reason I still fight for the continuence 
of the style attribute, but that's beside the point.

So, is it important that elements be able to report that they are part 
of an ordered list in a certain namespace... when they are out traveling 
on their own?  Probably.  But me, being a nodeslinger from WAY back, 
want every element on the planet to carry a 10kb DublinCore-6 metadata 
ID card with it, and that's just not feasible, plausible, sensible, or 
any other ible.

I'm not even sure what point I'm making here.  I don't think I have one. 
  Your conversation just prompted my comment.  It might be something 
'meta'physical. :o

Bestest wishes!
Wingnut

Received on Sunday, 27 February 2005 16:13:58 UTC