Pet Name Markup Language

Chip Morningstar
Sun, 7 May 2000 13:40:28 -0700 (PDT)

This is a very interesting proposal. I think it addresses the underlying
technical problem well, but I think it has some weaknesses from a UI
perspective. Using a markup language notation for encoding references as keys
when a message is transmitted from one person to another seems reasonable,
especially given the growing propensity for people to send email containing
HTML (as much as I detest that trend). However, I'm inclined to disbelieve that
people will be prepared to adopt the convention of typing <pn>...</pn> around
people's names as they are composing messages.

I suspect a lighter-weight syntax is possible. Given that the text is going to
be passed through a filter before and after transmission anyway, there is no
particular reason to slavishly follow the horrible markup language lexical
model anyway. Eudora, for example, when displaying a mail message,
automatically recognizes http://... as a URL and displays it as a clickable
link accordingly, even if there is no markup.

People are already used to using email addresses, even though these limit you
to a more constrained notation than human names in general do. One idea is to
simply adopt the convention that any alphanumeric token in the outgoing message
text that matches an entry in your pet name dictionary is a pet name. This is
certainly has problems (i.e., what about if I am asking a question about rocks
of my geologist friend whose pet name is "rock"; is an does "rock" in the text
efer to him or is it just the word "rock"?), but I like it better than using
markup notation.