Re: Support for undo-amalgamate in a version of the atomic-change-group macro (with patch)

[Stefan Monnier]

> In my use case this is an important guarantee since undo/redoing an
> action needs to ensure the action was fully undone before performing
> the new action. Failure to do so will cause bugs that could be
> difficult to track down.

Fair enough.  I think the better way to document it is that we want to
mimic the behavior we'd get if the undo-boundaries were never added in
the first place.

> Another note on naming, this could be called `with-undo-amalgamate',
> not sure if the "-change-group" suffix is worth including (I did this
> to fit in with existing names, although it seems a bit verbose).

The less verbose name is better, indeed.
Could even be `with-single-undo`.


        Stefan