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.
• CommentAuthormaxsnew
• CommentTimeMay 12th 2018

Is there a reason the article on abstract stone duality uses the term $\infty$-calculus when linking to $\lambda$ calculus? Looks like it’s been there since the original page by David Corfield. I was going to just change it to $\lambda$, but it seems intentional.

• CommentRowNumber2.
• CommentAuthorDavid_Corfield
• CommentTimeMay 12th 2018
• (edited May 12th 2018)

No, there’s a bug involving unicode symbols. There’s a conversation about the problem in this thread. So please change any of these you find.

1. I will try to address these issues this evening.

• CommentRowNumber4.
• CommentAuthorRichard Williamson
• CommentTimeMay 13th 2018
• (edited May 13th 2018)

I have fixed a collection of similar unicode issues now. I have not yet fixed those referring to $\lambda$-calculus and $\beta$-reduction or $\eta$-reduction, though. The problem is that it is non-trivial to isolate them correctly. For $\infty$-calculus there are also things which should be $\pi$-calculus; and one needs context to understand whether $\beta$ or $\eta$-reduction is meant for $\infty$-reduction. Nightmare! I will do my best to fix it as soon as possible, but I will not have more time today.

• CommentRowNumber5.
• CommentAuthorMike Shulman
• CommentTimeMay 14th 2018

How did all these different unicode symbols get collapsed? Is there any help in the database history?

• CommentRowNumber6.
• CommentAuthorRichard Williamson
• CommentTimeMay 14th 2018
• (edited May 14th 2018)

Unfortunately the database history does not help (or, at least, not in a direct way; occasionally there is something in the database of relevance which helps).

You know, I’m not completely sure what happened in all cases. It’s a real mess. It seems that amongst some of the encodings backwards and forwards before I became involved, some unicode symbols got converted to question marks, for example $\Pi W$-pretopos to $?W$-pretopos, and an en dash in Homotopy Type Theory – Univalent Foundations to Homotopy Type Theory ? Univalent foundations. Others got converted to HTML code, which is problematic in page titles or symbols with a special meaning in Instiki.

When replacing an instance of the HTML code for an $\infty$-symbol with unicode in a page title a while ago, I compounded the problem somehow, meaning that I had to try to reverse replacement of the unicode symbol $\infty$ by ?. As you can imagine, there are rather a lot of pages on the nLab with $\infty$- occurring! There is no easy way to make only the correct replacements, so a few too many things got replaced (basically things with a unicode symbol before a hyphen).

It will be a long time before we clear up all of the mess. All we can do is clean it up as best we can when we come across it (I am also trying to clear it up in the history). I think the cases mentioned in #4 are the most serious on the main nLab that remain. There was a serious one with $\ell$-adic cohomology, which is hopefully now fixed.

I think the lesson here is to be very careful about altering content of pages programmatically.