Peter Lebbing

You can rest easy, that still fails in the exact same way :-)

Bisecting this is no fun at all; I think just analysing the code is more productive. But I did find that it was already broken in the same way in...

I like the elegant and principled approach here! But I do have one main criticism. But first, some simple observations. You write ```haskell module Lib where foozle = ... (foozleTop,...

That's because the name is `Just "topEntity_probe___VOID_TDECL_NOOP__"`, not just the suffix.

This is related to [my comment](https://github.com/clash-lang/clash-compiler/issues/2655#issuecomment-1925272120) in #2655: as soon as Clash needs to deduplicate, as here, it will prefix the name with a prefix based on name of the...

> Another option is to not make `KnownDomain dom` not part of the `HiddenClock/Reset/Enable` constraint synonym. Ah but I recently wrote [some documentation](https://hackage.haskell.org/package/clash-prelude-1.8.1/docs/Clash-Explicit-Signal.html#g:conveniencetypes) recommending people construct such tuples themselves...

I think the error message is misleading. I think the problem is that `PortProduct` only works on types with more than one "field". So when it does ```haskell PortProduct "A"...

Could you please edit the cover letter to mention that the two "implementations" are identical except regarding signal names? Right now, one wonders whether that is the case or not.

I think we should just add a prim for `unpack @Int` (but don't forget about `-fclash-intwidth`)

"Is the data a counter?" That is underspecified. Is the data a counter, and does it only increment or decrement by a maximum of one count each cycle? The Gray...