Bruno Le Floch

After reading the doc of the silence package, I think that `\WarningsOn/Off[`_list_of_packages_`]` would be a good to have in some form.

Relatedly, David Kastrup had an [old BachoTeX contribution](http://www.gust.org.pl/projects/pearls/2008p/david-kastrup/bachotex2008-david-kastrup-pearl1.pdf) in which he build an improved `\vtop` that detects whatsits at the start. I don't yet understand it because I'm not familiar...

One option in a given document would be to apply \smash to the section title (or only to the Ä?) It seems to give consistent spacing.

I vaguely remember this being by design. Let's see what Frank thinks.

Well, it seems we disagree somewhat: the original definition `\My:cn` seems perfectly fine to me: it is just simpler than having to do `\cs_new:Npn \My:Nn #1 #2 { \cs_new:Npn #1...

It may be unexpected, but `p` arguments are supported by the variant mechanism, so `\cs_generate_variant:Nn \cs_new:Npn { Npo }` works perfectly fine. Base functions should be documented as being some...

That's unfortunately a very good point. Basically `p` only makes sense before braced arguments, so `n`, `c`, `o`, `f`, `x`, `T`, `F`, and not before `N` or `V`. Clearly this...

How general should that be? In particular with respect to catcodes, newlines etc? Maybe say what you would use for tls?

I'm of two minds concerning this proposed addition. In such a non-expandable setting, just use ``` \exp_args:Nx \sean_subscript:n { \clist_use:nnnn {#2} {,} {,} {,\dots,} } ``` since `\clist_use:nnnn` wraps its...

Here is some possible documentation that I had in some file I'm cleaning up. (Obviously the added date would have to be changed to whichever date we end up implementing...