Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
I think it would be useful to have some way of marking which sections of the n-lab are original research, which exposition, and which something else. Something a little finer than on a per-page basis, as well.
Then it would be nice to have some way for people to say that they've read the OR bits and that they:
(Note that I don't put "Have verified it" in my list).
Any suggestions? Is this worthwhile?
(Hmm, something here is killing lists ... have to track that down and kill it instead)
I'm not so sure about enclosing whole sections in boxes, that could get a little wearing on the eyes.
And now I think of it, there's an accessibility issue: colours that I can see are not necessarily the same for everyone so lots of colour changes increases the chances that someone can't read the page. One could, of course, control all this by CSS.
I was thinking more of some sort of icon in the margin, textbooks often have this to denote a section that is harder than the rest.
The list of names would have to be hideable, otherwise it could get very long indeed. But I do think that such a list is a good thing. Just knowing it's been checked isn't really enough - you have to be able to assign a sort of Bayesian belief to the standard of the checker (and that would vary from topic to topic). I also think that one should have levels of checking to make it easier to distinguish rigorous checking from "seems alright to me" (the latter can be valuable, though, as it indicates that there's no obvious counterexample).
I like the idea of icons in the margin, like the "dangerous bend" icons in the TeXbook. That wouldn't need any CSS, just someone to find or make some good images and put them on the nLab where they can be included into other pages at the appropriate point.
Regarding hiding names, it seems that perhaps more generally a query-box-like syntax for a "click-to-display" box would be useful. For instance, answers to "exercises" such as the one here could be put in such boxes. Would that require javascript in addition to CSS?
Well, if we want to hide lists of names because they could get very long, we'll want something more than white-on-white, since that still takes up space.
I think that a full implementation of something like this will be a big thing and so needs more discussion to see what exactly we want. However, we won't know what we want until we have a workable system so here's a proposal for an interim method.
Original work should be signposted (anyone who feels like designing a logo should stick it on the SVG Sandbox - and yes, it should be an SVG) and signed. The "signature" should be a link to the n-lab page of the author (that also makes it clear who the original author(s) is/are and distinguishes them from the last person to edit the page). On the author's/s' page(s) (gosh, this single/plural thing is getting irritating) then the statement of the theorem should be included in a special section, say "Original Research on the n-Lab" with a link at least to the page containing the theorem, if not to the theorem itself. Then anyone who wishes to declare an opinion can sign the entry on the author's page, not the original entry.
Of course, this would work for things other than the traditional "theorem/proof" result. The author's page should contain enough to identify the work but not the whole proof (so Todd's characterisation of morphisms of manifolds could be something like "A recasting of the definition of a morphism of manifolds into the language of spans" or whatever it was). That way one isn't constantly trying to sync two pages.
Also, it's then up to the author what kind of feedback they want. They could leave it completely open or they could put up a table with entries a little like what I put in the first post and then checkers put their name in the box they feel happiest with.
With a system like this in place we could get the ball rolling without any changes to the infrastructure. That will give us a chance to experiment and see what exactly we do want (since this time we don't have the Wikipedia example to draw on! What a relief!).
1 to 8 of 8