Re: [xdi] Metagraph symbols ==> Metagraph $words

[Drummond Reed <drummond.reed@xdi.org>]

Neither spaces nor empty strings are allowed in an XRI (or a URI). And underscore is strongly discouraged from use in URIs by W3C and is not a reserved character, so it's not an XRI delimiter.

So, as the representation for the concept of subcontext/supercontext, () is the best fit.

=Drummond

On Tue, Jun 21, 2011 at 1:15 PM, Joseph Boyle <joseph@josephboyle.com> wrote:

Are space or empty string acceptable in that position? If they're allowed but not currently used, could they be used to indicate subcontext instead of ()?

On Jun 21, 2011, at 11:57 AM, Drummond Reed wrote:

I have done the first part of the action item I had from the last telecon to analyze moving back from metagraph symbols to semantically-meaningful English $words.

As I expected, for three of the four, the translation is very easy

Metagraph symbol  ==> $word 
$  ==>  $is  (inverse is the same)
*  ==>  $has  (inverse is $is$has)
!  ==>  $a  (inverse is $is$a)

These all would be used just as we have used them for the past several years.

The only metagraph symbol that does not have a good English equivalent is () -- the metagraph symbol for subcontext. After thinking about this over the weekend, I realized there is a reason there is no good English word for expressing a pure subcontext relationship the way we use () in XDI statements, e.g.:

+snow/()/+ski
+water/()/+ski
@Rossignol/()/+ski

The reason is that when you translate these XDI statements into English, the semantics of the () context symbol are expressed by...WHITESPACE!

snow ski
water ski
Rossignol ski

It is literally the left-to-right order of the English words, delimited by a space, that expresses the supercontext/subcontext relationship.

No wonder we could never find the right word for it! And also a good case for why (), the metagraph symbol for subcontext, should remain a symbol. Expressing the inverse, supercontext, will be $is(), following the same rule as all of the other metagraph $words above.

So, the proposal is to move back to the $words $is, $has, and $a instead of the metagraph symbols for $, *, and ! when they are used as predicates in XDI statements. If everyone is in agreement about this, I will draft a new wiki page called MetagraphWords to replace

  http://wiki.oasis-open.org/xdi/MetagraphSymbols

Please post if you don't agree.

=Drummond