Hacker News new | ask | show | jobs
by grumbel 310 days ago
Cyc[0] tried for 40 years to make handwritten semantic rules a thing, but still don't have much to show for. Humans just aren't good or fast at writing down the rather fuzzy semantics of the real world into a computer readable format.

With RDF specially, there was also the issue of "WTF is this good for?". Semantic Web sounds lofty in theory, but was there ever even a clear plan on how the UI would look like? How would I explore all that semantic data if it ever came into existence? How would it deal with link rot?

And much like with RSS, I think a big failure of RDF is that it's some weird thing outside the Web, instead of just some addition HTML tags to enrich existing documents. If there is a failure, it's that. Even today, a lot of basic semantic tags are missing from HTML, we finally got <date> in 2011, but we still have nothing for names, cities, units, books[1], movie, gps coordinates and a lot of other stuff we use daily.

Another big failure is that HTML has become a read-only format, the idea that one uses HTML as a source format to publish documents seems to have been completely abandoned. HTML is just a UI language for Web apps nowadays, Markdown, LaTeX or whatever is what one uses to write content.

0. https://en.wikipedia.org/wiki/Cyc

1. <a href="urn:isbn:..."> exists, but browsers don't support it natively