Traits and fmt brought up to the compiler of 21 October

- Where an implementation is written: with its trait or its type, a
  newtype for anything else, and traits, types and impls usable above
  where they are written.
- One implementation that fits decides the type of a call.
- A function is written as #<procedure> wherever it is; Procedure.
- A (dyn T) implements the traits T requires.
- Formatting: the lambda that returns a vec of mixed pieces, in the
  table of errors; the short fmt tour in Language basics in the style of
  a string being a block.

🤖 Generated with [ECA](https://eca.dev)

Co-Authored-By: eca-agent <git@eca.dev>

_02_.Writing-programs.sz

@@ -932,8 +932,11 @@ The messages for the common mistakes, and what to write instead:
 | Block: ... found a boolean, or found a bool                            | @code{(if b "yes" "no")}                            |
 | Block: ... found a (Vec string)                                        | @code{(vec-map BStr xs)}, or @code{(: xs Block)} above @code{(def xs ...)} |
 | Block: ... found a (Vec (Vec ...)), at a lambda that returns a vec    | @code{(each [...])} around the vec the lambda returns |
+| these types do not match, naming two of int, string and char, at a lambda that returns a vec | @code{(each [...])} around the vec the lambda returns |
 }
 
+The last two are one mistake, a lambda given to a function that does not say it returns a block. Its vec is a vec of strings if all its pieces are strings, and then @code{show} gets a vec of vecs. If the pieces are of different kinds, such as a name and a count, the vec has no type at all, and the compiler says so inside the lambda.
+
 @subsubsection{A block of your own}
 
 @code{captured} renders content to a string and gives it to a function, which returns the block to write in its place. It is how a block that needs the whole of its content is made, and how padding and trimming are written: