David, You have brought up good use cases for using hCards to identify people, and this is definitely worth further exploring and defining.
On 12/22/05 1:58 PM, "David Osolkowski" <[EMAIL PROTECTED]> wrote: > To identify the source, > a page with XFN links should include an <address> element, which identifies > the author of the page. Yes. And you could just shorten that to: "a page should include an <address> element, which identifies the author of the page" This is true for all pages, not just pages with XFN. This is orthogonal to XFN. Using <address> accordingly is simply a good semantic XHTML authoring practice. > The <address> element should contain an hCard, > which--to my knowledge--is the best method we have for representing a person > in (X)HTML. Strongly agreed. This is IMHO a very good "hCard practice" and documented in the hCard FAQ: http://microformats.org/wiki/hcard-faq In addition, you've inspired me to write it up in the hcard-examples page as well: http://microformats.org/wiki/hcard-examples#Authors_of_Pages_and_Posts > Identifying the target is a little more complicated. Actually, I'm not sure that it is that complicated at all. First of all, there are two key cases here to satisfy illustrated by the scenario you mentioned (when author A uses XFN to indicate a relationship to another author B). 1. Author A asserts what Author B's name is. This is the common/80% case as in blogrolls, as those links nearly always say what the person's name is that they are linking to. This is easily done by taking a "typical" XFN link from a blogroll: <a href="http://jeff.example.org"; rel="friend">Jeff Xmpl</a> And marking it up with hCard: <span class="vcard"> <a class="fn url" href="http://jeff.example.org"; rel="friend">Jeff Xmpl</a> </span> In fact, I'll just go do this on my blogroll right now... ...that was fairly straight forward. I think this might be another improvement on the markup pattern for blogrolls. 2. Author B asserts what Author B's name is. Clearly this is the better of the two, as you can more often trust a person to spell their own name than others to spell their name for them (perhaps I'm biased from personal experience ;). This is the case you address (so to speak) with your example: > Generally, we should > check the linked page for an <address> element, Again, this is just a good idea in general, whether using XFN or not. > and assume that linking to a > page means relating to the author of that page. Yes, that's one of the established behaviors that XFN is built on > Basically, I'm proposing some best practices for using XFN with hCards that > seem to improve the semantics, without needing to invent anything new. Yes, and your methodology, "improve the semantics, without needing to invent anything new" is excellent. > If we agree on these practices, they should be explained on the XFN website. Agreed. There is a page on the XFN website that explains how XFN is used with other services and technologies: http://gmpg.org/xfn/and/ but it hasn't been updated in quite some time (it predates hCard ;). Let's start with flipping this around, and document on the microformats.org wiki the best practices of using hCards with XFN. http://microformats.org/wiki/hcard-examples#hCard_and_XFN Thanks, Tantek _______________________________________________ microformats-discuss mailing list [email protected] http://microformats.org/mailman/listinfo/microformats-discuss
