HTML hyperlinks as they are - unidirectional - are a very long standing design decision the last futile revision of which was in 1099 or so with XLink. You prob

HTML hyperlinks as they are - unidirectional - are a very long standing design decision the last futile revision of which was in 1099 or so with XLink. You probably have it seen in action with SVG-in-HTML (and SVG fragment inclusion in SVG itself) where it kindof rears its ugly head in that, for the longest time, you had to use xlink:href (with xlink bound to the XLink XML namespace). XLink itself is based on SGML/HyTime concepts from 1990s. I believe Sun back then patent-encumbered XPointer (which is the part of XLink for addressing fragments), and especially claimed IP rights on its representation of range addressing, which didn't help to make it popular I guess. Before changing HTML link semantics though, HTML should really consider allowing placing href on any element rather than using the dedicated <a href=...> element though (with the expectation that this acts equivalently to an anchor link), like MathML does. In general, I'm not sure hosting discussions on third-party aggregators such as discus.com is really a win for privacy, though.