iherman
commented
5 years ago I have made a small experiment, to see what it would mean to ‘manage’ older recs. The one I was most afraid of was the RDF1.1 spec, I was worried that this was done in a pre-respec era. It was also done on mercurial, not git(hub). But it is not that bad after all:
- W3C (thanks God!), although closed the mercurial service, it has archived (of course:-) the full mercurial content that I could retrieve. There are also tools to get the whole content into a github repository (which I did not do for now).
- RDF 1.1 was all done in respec. I took one of the files (the concepts’ one): it did not work out of the box with respec, because the bibliography file was old-skool, and it did not use the
<dfn>-<a>tricks the way respec does today. But those can be handled relatively easily. I would think that convert the concepts’ document into today’s respec would take 1-2 hours, probably less. There are some local scripts used for Turtle which may be a problem, because it uses jquery and today’s respec does not any more, so that is unknown. But I would expect that converting the full RDF set of documents into today’s respec could be done in 1-2 days.
Taking into account that we have LOT of documents, it is still a significant effort, but a bit better than I thought.
I cannot judge the effort it would take to adapt the SPARQL query algebra and SHACL to account for literals. For RDFa there is a processing step series that should be updated, probably not a huge deal (and the other documents are probably unchanged except for editorial errata). @gkellogg can probably handle the CSVW case easily, I do not think there are lots to do.
gkellogg
pchampin
The current charter proposal lists a HUGE number of documents that have to be updated (even if the changes are minimal or non-existent). That raises the issue whether this endeavor is feasible in the first place.