Re: comments on CSS1

On Tue, 5 Dec 1995, Scott E. Preece wrote:

>    From: David Seibert <seibert@hep.physics.mcgill.ca>
> 
> |   <h3>Tables</h3>
> |
> |   When conflicts arise while a UA is rendering a table, the most
> |   important thing is that the information present be displayed in a clear
> |   manner, so that it is useful to the reader.  For this reason, frame 
> |   conflicts within tables should be resolved in favor of the 
> |   choice which is likely to make the information clearest (e.g., leaves 
> |   the most space around the element, or provides a line separating it 
> |   from adjacent elements).  This is similar in spirit to the policy of 
> |   using the larger of the two frames when joining two adjacent table 
> |   elements.
> ---
> 
> I don't object to the rule (I think it's good to say what should happen
> when conflicts occur), but I strongly disagree with the statement of the
> rule: "frame conflicts within tables should be resolved in favor of the
> choice which is likely to make the information clearest".  The author
> presumably intended to make the data as clear as possible; if she has,
> in doing so, created a conflicct, there is no way for the browser to
> know which choice leads to more clarity.  In particular, extra space or
> extra frames may obscure relationships that the author meant to express
> (by, for instance, putting related data physically closer together and
> without separating lines).
> 
> So don't editorialize, just present the decision:
> 
>    When the styling information for a table is ambiguous, the browser
>    should resolve the ambiguity towards the default form for a
>    table: all cells separated by same-width rules, all data separated
>    from cell edges by the same amount of space, all data cells in the same
>    typographic style, etc.
> 
> Actually, I think you argue about as cogently that either towards or
> away from the default makes sense.  I don't think I care, so long as
> you say one or the other, to make it predictable.
> 
> scott
> 

Moving away from the default makes much more sense.  If an author puts an 
extra frame around a table element, he probably wants the frame there to 
emphasize that element.  I think that a UA should try to do what the 
author wants, and should not take out things that the author wants to add 
unless it cannot render them, since it can always make the document 
larger.  UAs should be smart, but they should not assume that they are 
smarter than authors.  

David

Work: seibert@hep.physics.mcgill.ca         Home: 6420 36th Ave.
Physics Department, McGill University       Montreal, PQ, H1T 2Z5 
3600 Univ. St., Mtl., PQ, H3A 2T8, Canada   Canada
(514) 398-6496; FAX: (514) 398-3733         (514) 255-5965

Received on Tuesday, 5 December 1995 11:56:26 UTC