[rfc-i] Section 2.35 - style attribute on ol
ietf at augustcellars.com
Tue Nov 11 11:02:20 PST 2014
> Section 2.35 is the style attribute mandatory? It only says it can't be
empty if not, is there a default
[paul] "style" is optional.
I am happy to have style be optional, however this means that there needs to
be a default specified. The fun part of this is that the default is not a
single value but a behavior. This behavior needs to be documented based on
the principle of least surprise.
The current behavior - if you don't specify a style and you are at level 0,
then it uses empty. This is not possible for an ol list, so that would be
an error in this case not to specify one.
If you don't specify a style and you are not at level 0, then it uses the
same style as the previous one. However there is an interesting behavior
for letters where it will alternate between upper and lower case.
Not all of this can be kept with the current set of styles. I don't know
that inheritance makes any sense except for the case of numbers. If you are
doing letters or roman numerals then you generally do want to do alternation
Either there needs to be a defaulting behavior that can be implemented
consistently by all of the convertors or this should become a mandatory
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the rfc-interest