There was a time when we were used to "broken links" when surfing the Web.
Webs became essentially dynamic and then broken links disspeared.
But the semantic web, reviving that era, shows us plenty of "broken links".
Look for a random ontology, find any supposedly triplified data set, try a SPARQL endpoint: you have high chances of finding the dreaded "broken link".
Common law of SemanticWeb projects: not profitable projects, only supported by public funding, which are abandoned as soon as the payment has been executed (and the taxpayers sacked for nothing).
I don't even bother in putting examples. Google yourself.
