>Agreed that a table is (i.e., should be) "an ordered collection of
>information", it is often not _just_ that; sometimes it is an ordered
>collection of information plus some explicit formatting hints needed
>to represent that information in the way that I want to see it
>represented. I am thinking specifically of column widths. The
>algorithm does not exist that will allocate widths based on cell
>content as well as I can do it myself. (Since the definition of "as
>well as I can" is "in the way that best expresses what I wish to
>convey", I would maintain that such an algorithm never will exist,
>either.)
I would agree that the quality of a particular table rendering is very
subjective, and hence not formally definable. I also acknowledge that
certain constructs can lead to computationally expensive layout
being required.
However, I feel that it is certainly possible for *reasonable* output
to be produced quite effectively without hints, and point to LaTeX as
an example. So, while such hints might be desirable, I doubt they are
*required*.
This does not mean I think they should be ommitted. People would like
to have them, and part of our responsibility is to create something
people can use in a manner natural to them.
--- NOT speaking for EBT!