# Start a new discussion

## Not signed in

Want to take part in these discussions? Sign in if you have an account, or apply for one below

## Site Tag Cloud

Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.

• CommentRowNumber1.
• CommentAuthorRodMcGuire
• CommentTimeApr 9th 2018
• (edited Apr 9th 2018)

test reference for inclusion test

this reference can be included using [[!include REF First Order Categorical Logic]].

See inclusion working SandboxSandbox/1477

other than the actual contents of this entry, discussion of this possible approach should happen in nForum: liberated books and REF inclusions.

1. Not sure what your intention here are/were, Rod, but your experiment got me thinking that it might be really useful if the references on the nLab had their own semantics, and were not just plain text. One can imagine for example something like a /reference endpoint, like the one I have recently added for /author. There is by now a huge bibliography spread out over the nLab, and I think there would be great value just in such a bibliography itself: even simple things like being able to search references, or being able to see upon which pages a given reference has been used, would be very useful. One could also add things like comments on a ’reference’ page, which could point out errors, or give a ’review’; over time, this might grow into a very useful resource.

(PS - No doubt you and the other readers have already realised it, but for the moment including links in the way you do in the experiment is probably not a good idea, because someone might change the page REP First Order Categorical Logic without realising it is being used a reference on a different page, which would cause complications on the other page. But always good to experiment with things like this!)

• CommentRowNumber3.
• CommentAuthorMike Shulman
• CommentTimeApr 9th 2018

This is a good idea, but it’s something that bears serious thinking about all the angles of, before doing anything substantial like updating all the references on the lab. In addition to the issue Richard mentions of someone unknowingly changing the reference page, for instance how will this interact with #anchor links to references? How about the “commented bibliographies” that appear on many nLab pages?

One thought is that rather than labeling such pages with a prefix like REF, we could put them in a separate “ref” web. and then include them from there. Then the page names would be prettier, one could search for a reference by simply searching the ’ref’ web, and we could put a header on the ’ref’ web warning people that its content is being included elsewhere.

Even better would be an “includeonly/noinclude” feature like wikipedia transclusion, so that comments and reviews (and annotated/linked tables of contents, like we have at some pages such as Elephant and CWM) could exist on the reference page, but only the citation itself would be included. Perhaps with an automatic link back to the reference page for more details? Just brainstorming here…

• CommentRowNumber4.
• CommentAuthorspitters
• CommentTimeApr 9th 2018

Earlier discussion.

I seem to recall there where more, but I cannot currently find them.

• CommentRowNumber5.
• CommentAuthorMike Shulman
• CommentTimeApr 9th 2018

Thanks for finding that, Bas.

• CommentRowNumber6.
• CommentAuthorRichard Williamson
• CommentTimeApr 10th 2018
• (edited Apr 10th 2018)

Funny coincidence that the issue with . mentioned in the discussion Bas links to reappeared here just a few days ago!

I can try something simple out here when I get the chance. My initial suggestion would be a /ref endpoint like I have done for author (so not a separate web, but a new kind of resource) for defining the reference, and to have a new syntax for including references in the main nLab pages, without reusing include. We’ll need to think about how to design the page creation/editing, though, because I think it important that we have some semantics around references: one should be able to identify author, journal, title, etc.

I think, to begin with at least, it would be best to just add references defined in the new way/edit existing references to use the new way manually and gradually, rather than do any kind of search and replace.

Even better would be an “includeonly/noinclude” feature like wikipedia transclusion, so that comments and reviews (and annotated/linked tables of contents, like we have at some pages such as Elephant and CWM) could exist on the reference page, but only the citation itself would be included.

Indeed, I think this is a good approach.