holtzermann17
commented
11 years ago This is relevant to outreach to the mainstream (more to mainstream computer science and publisher people than to mathematicians, but nevertheless...) Cf. https://github.com/holtzermann17/planetmath-docs/issues/21
Note that for now, we use a mixture of MySQL to deal with things like "which corrections are connected to which articles". We may want to add a "Virtuoso API" or "SPARQL API" so that we can send representations of that data to Virtuoso too.
This is presumably especially useful for "downstream" consumers of semantic data. See this PDF for some more information about the sorts of triples we would like to store (most article-level metadata already being available).
Other triples, like "A is a fork of B", "A is following B" or "A abandoned B", "A suggested B" have yet to appear! Similarly, issues around who is in what editing group, etc., are (so far) handled by Organic Groups only.
Note, again, that "activity data" ends up being handled by the userpoints module (so long as some score is associated with a given activity, it will be logged). This "works", but if we ever want to do more (e.g. with the ActivityStreams protocol or some Linked Data version of the same), that will be another layer.
Note that the semantics of "attachments" for example would be handled via "pm:parent", with the meaning