[uf-discuss] Re: hCalendar for events in a table format

[Tantek Ç elik]

On 3/5/08 5:02 AM, "Toby A Inkster" <mail at tobyinkster.co.uk> wrote:

> However, implied headers like this while lowering the barrier to entry for
> authors, would considerably raise the barrier for parsers -- mostly because
> of colspan and rowspan, which would be an absolute pain to handle.

Speaking from the experience of working on a browser rendering engine which
*did* have to handle colspan and rowspan, I can certainly state that
requiring a microformats parser to perform a table layout (effectively what
it takes to support colspan and rowspan) would *drastically* raise the
effort necessary and would introduce numerous opportunities for subtle bugs
and incompatibilities.

I think it would be reasonable to adopt a design principle of *not*
requiring microformat parsers to perform a table layout, even if it can be
used to make inferring semantics easier.


> Although microformats' general principle is to place the burden of effort
> onto parsers, implied headers via the scope attribute may shift the effort
> *too* far in that direction. What do others think?

Yes, very much so too far in that direction.

Thanks,

Tantek