On Thu, Dec 21, 2023 at 11:40 PM Jim Rees <[email protected]> wrote:
re: issue #2. > I agree with Mr. Feeley that the closure-api is very limiting and makes > finding guardian operations in code challenging because there are lots of > ways to create procedure calls, but api-names are easier to spot. > I agree, so I'll change the next version to use this style. For my own testing until these questions are resolved, I used: > > (guardian-register g obj [rep ob>]) > I agree that this makes sense. (guardian-next g [default #f]) <-- pop the next triggered rep, or return > default if none are available > As I noted earlier, I was careful not to imply that the set of guardians constitutes a sequence (following Chez). So I'd prefer something like guardian-retrieve. (guardian-all g) <-- pops the whole list directly from the tconc without > re-consing > It doesn't say what to do with "all"; we need a verb here, so "guardian-unregister-all". Because the internals need not be a list (see above), I'm not going to say "without re-consing".
