Will Robertson

I may have forgotten the discussion :-) Keeping them as "others" is totally fine -- I think I would agree we don't really want names like \foo_if_bar?:n and so on....

Good argument! I would be very happy for this to be the considered official stance on the matter.

What is the scope of this issue now? Should we consider making & an other? I'm of a mind to leave the others. Even & I would be worried about...

I guess we all agree we can't change any catcode settings as they currently are, to avoid spectacular breakage. I note that \ExplSyntaxOn ensures that catcodes of `&` and `|`...

I'm happy to take a look at this as long as it's not time-critical :)

Not at the “layer of the onion”. Both Frank and you (I think) have presented at different times the need for design-level interfaces, but the need here is literally just...

For what it's worth, I still very much like the "extended" abilities of `l3build manifest` that generates a table of files and their descriptions: This requires descriptions are embedded within...

My take on this is that we should simply add a customisation variable or two around `unpack`. The easiest way, I think, would be to allow something like unpackfiles =...

Of course I would see this to have some other indirect use cases when testing things like output routines and commands like `\clearpage` :)

I had this problem too. (Doubling \ does indeed work for me these days.) I believe the issue comes from curl rather than l3build. I guess we should be able...