[sc34wg3] Comments on latest TMRM draft

Lars Marius Garshol sc34wg3@isotopicmaps.org
Fri, 15 Jul 2005 17:32:14 +0200

* Patrick Durusau
| Rather than simply "mak[ing] the defnitions/formalism easier to
| read" I see the prose as a context of usage that makes the
| definitions meaningful. Admittedly a definition should make a term
| "meaningful" but that succeeds only to a degree.

I guess we can stop this discussion here and just note that we have
different preferences. I guess we should turn to others and see what
they think.

| [the "property" terminology]
| I don't know that it would be any more representative that what you
| and I have said as individuals but perhaps other members of the
| committee would like to voice their preferences?

I agree that you and I have discussed this as far as we can.
| I really see it as a preference issue and not something fundamental.


* Lars Marius Garshol
| [are curlies set or proxy delimiters]
| But the examples are examples of sets, so what other notation could
| possibly have been used?

* Patrick Durusau
| Err, sorry, I was not reading the examples as using sets but simply
| as a notation of proxies since each example (as I read them) as
| single proxies.

In that case this is another case of the draft not agreeing with you.
The definition of a proxy says it's a set, but I realize you may
disagree with that.
* Lars Marius Garshol
| The definition of proxy says it's a set of (key, value) pairs, and so
| there is no identifier in the proxy.

* Patrick Durusau
| And the prose, sorry!, says all proxies have identifiers, unique
| ones at that.

Correct, but it's quite clear that the identifier is not inside the
proxy. The identifier is produced by a function, and there's a
function that given the identifier can give you the proxy. That's
something else entirely.

So as it stands the draft says the proxy is a set of (key, value)
pairs, but, again, you may disagree with that and want proxies to be
something else entirely.
| OK to 2 and 3, I think we need to discuss in Montreal #1.


| [the values(x) function]
| Hmmm, looks like an error in changing from T+'s saying that there
| could be more than one instance of a key. I think the reasoning for
| a bag was that if there was more than one occurrence of a key in a
| proxy that the multiple occurrences of the value for the mulitple
| occurrences of the key should not be treated as a set. Under the
| same reasoning, the keys were treated as a bag.

Yeah, I sort of guessed that, but it seems that this is another case
of discrepancy between the actual draft text and the intention behind

Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >