I'm editing Is there a canonical map from the cohomology of orbifold Chiral de Rham on an orbifold to the cohomology of Chiral de Rham on a crepant resolution? as there was a broken arXiv Front link (and I took the opportunity to add some paper titles in). But I cannot submit the edit as the question title is 'Too long by 1 character'. I have a vague memory of this issue coming up before, but I cannot find it here on meta. Obviously I could shorten the title (say by making it as "A canonical map....resolution?"), but this does change the sense slightly.
Here's a screenshot:
What should be our practice here? And I'm leaning towards calling this slightly buggy behaviour hanging over from the transition from SE1.0 to SE2.0, though I know it's operating as intended. Presumably this came up long ago on other sites, if it's a SE1.0 -> 2.0 issue?
I have saved a local copy of the new source text of the question, but have left it open for editing, in case there's a quick answer. If not, I'll cancel the edit, and paste in the updated source later, once there's consensus on how to proceed.
Edit: title has been tweaked, but I think discussion needs to be had around this, or else the old discussion found and linked to.