[Competencies] corrected link

Tim Willett emailtotim at hotmail.com
Wed Sep 7 16:57:09 UTC 2011

Thanks for these thoughts, Simon.  I must say most are beyond my level of understanding of current practices.  See specific comments below in red.


Date: Wed, 7 Sep 2011 17:43:01 +0100
Subject: Re: [Competencies] corrected link
From: asimong at gmail.com
To: emailtotim at hotmail.com
CC: vsmothers at jhmi.edu; competencies at medbiq.org

Hi everyone

Just a couple of minor things I didn't want to raise in the call as we were short of time, and they may be questions rather for the TSC, though it seems right to raise them here first.

1. Relationships are just plain text from a vocabulary. Would it be better, from a spec point of view, if we made the relationship element content a URI (or other identifier) instead? That would seem to me to be an easy way towards being able to display different languages, among other things, and start off doing some reasoning with the relationships. Fits into the "linked data" pattern much better. Beyond me, but I wonder how the semantic web (e.g. OWL) handles this?

2. Sort of related, all the other identifiers are given using the "Catalog/Entry" approach. I recognise we may be caught in legacy here, but it would be nice to know whether we *really* believe that a Catalog/Entry approach is actually better (and in what way) to simply specifying a URI in all cases, or whether we are doing this simply to conform to old-established ways? No prizes for guessing that my preference would be for a mandatory URI, which again fits in very well with the Linked Data approach to the world.  Beyond me.

3. XtensibleInfo. Yes, OK, I know this has "always" been done before, and I know IMS have done it all the time, but what are people's actual experiences of using XtensibleInfo? My own are disappointing, suggesting that it is more useful in practice to have an open spec like Atom (and, by inheritance, Leap2A) where anything from a different namespace is allowed unless expressly forbidden. Beyond me.

If there are convincing reasons for these, fine, I'll be happy to learn what they are. If not, I hope we will move with the times rather than being stuck in practice that evolved maybe 10 years ago, and isn't in tune with current developments.

Thanks and best wishes

Simon Grant
+44 7710031657
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://medbiq.org/pipermail/competencies_medbiq.org/attachments/20110907/11ae1504/attachment-0002.html>

More information about the Competencies mailing list