RE: table-cell border precedence

table-cell border precedenceI have pointed out these and other
inconsistencies already. My primary concern at this point is that the XSL FO
group has apparently punted on supporting the CSS2 semantics for collapsing
borders which they normatively reference. In particular, the XSL FO group
apparently believes that the border precedence can be computed on cells
during XSLT processing. However, I have shown definitively that this is
impossible to do when the border widths are based on relative units that can
only be computed during the formatting process.

I have urged that a simple solution be added to resolve this problem;
namely, than an "auto" value be added for the border precedence properties
which would have a meaning of "determine precedence base on CSS2 algorithm".

So far, however, the XSL FO group seems to have ignored my suggestion.

Regards,
Glenn Adams
  -----Original Message-----
  From: w3c-xsl-fo-sg-request@w3.org [mailto:w3c-xsl-fo-sg-request@w3.org]On
Behalf Of Roger Neate
  Sent: Thursday, July 26, 2001 7:39 PM
  To: xsl-editors@w3.org
  Cc: alpha-dog@arfarfarf.com
  Subject: table-cell border precedence


  W3C CR-XSL-20001121, Section 6.7.10. fo:table-cell, states: "If the value
of the border-collapse trait is 'collapse' the border for each side of the
cell is determined by, for each segment of a border, selecting, from all
border specifications for that segment, the border that has the highest
precedence. It is an error if there are two such borders that have the same
precedence but are not identitical."

  In addition, Sections 6.7.4. fo:table-column and 6.7.9. fo:table-row
state, "If the value of border-collapse is "collapse" for the table the
border properties also apply."

  These two quotes suggest (although it is not explicit) that the border
specifications of cells, rows, and columns must be considered when selecting
a cell-border in the collapsing border model. But some of these formatting
objects do not have border-precedence properties listed as applying to them.
How is this to be interpreted? If the borders of rows and columns are to be
considered, what is the appropriate precedence? If the borders of rows and
columns are not to be considered, under what circumstances are these borders
used?

  Also, Sections 6.7.6 fo:table-header, 6.7.7 fo:table-footer, and 6.7.8
fo:table-body all indicate that border and padding properties apply to these
formatting objects. It is not stated that they only apply when the value of
border-collapse is "collapse." How is this to be interpreted? Are these
borders always used and, if so, how exactly are they positioned with respect
to the borders of cells and of the table? Or are they used to select cell
borders in the collapsing border model and, if so, what is their precedence?
How is the padding used?

  I would note that all of this seems inconsistent with the CSS2
specification which indicates that the borders of column-groups, columns,
row-groups, and rows are only used if the border-collapse trait is
"collapse" and that only cells have padding (this latter point is made clear
in an erratum).