[Sorry for this resent. I just dumped crappy Postbox mail client, and went back to Thunderbird. Let's see if that does not screw up plain text that contains HTML-like constructs...] Roman Łapacz wrote:
While updated the UML schema, I noted we are missing a few group-to-element relations:
<BidirectionalPort> ????? <Port> <BidirectionalLink> ????? <Link> <Topology|Node> ????? <Link|LinkGroup> [...] <PortGroup> ????? <Port> <LinkGroup> ????? <Link>
Can we use hasPort/hasLink for the 5 missing relations?
<NetworkObject> hasPort<Port> <NetworkObject> hasLink<Port>
In my opinoin we don't have to use the relation element for mentioned cases. Simple inclusion would be enough.
I should have mentioned: I'm proposing this in the context of RDF, which does require explicit names. We indeed agreed on simple inclusion in in XML, and I don't think we should change that. So is the above fine in RDF or do we like something else, e.g. <Group> contains <NetworkObject>. Freek