I do not want to make Scheme unimplementable efficiently, which is why I would oppose removing all instances of UB from Large (and clearly that would not go over well with our users and implementer...
I would like to see as many UB references as possible be replaced by `SHOULD do xxx'. That at least encourages implementers; and if we can find N≥3 implementations that already do xxx, we have...
This is why I hope we can put the correct behaviour in our Report, albeit with a SHOULD. This in no way invalidates any previously-valid, but it replaces UB with a desired behaviour that some...
I'm inclined to replace my earlier SHOULD with MUST, and recognize that some/many/most implementations will not meet this criterion. (I also support implementations having a safety parameter or...
I regard each occurrence of "UB" in our final Report as a black mark against us. I would prefer a statement such as "It is an error to mutate an immutable value, and the non-continuable...
especially code that nobody actually runs
set!-values doesn't match with its spec.
That is absolutely correct. Please substitute "informative" for "normative" in my previous comment.
set!-values doesn't match with its spec.
I strongly support sample implementations appearing in the Report, for two reasons. First, they do provide an operational semantics for the items being implemented; and, second, they are examples...
I have already volunteered to take responsibility for turning the DocBook version into LaTeX, so that the final PDF will be typographically compatible with the existing Reports. The goal is a...
I am willing to take care of setting up a workflow to produce LaTeX from the DocBook originals,
+1 in principle for adopting a possibly-modified SRFI231. I'd oppose a lexical syntax except that serialization might demand it. In that case, something very simple might have to be adopted. For...
(scheme base) and (rnrs base (6)) procedures out of the Foundations into the Batteries
A general comment on this thread: I'd prefer we define library contents by a coherent API rather than by what needs what. @readevalprintloop's comment must absolutely be followed, but it would be...
I would like to leave #/... as available for an implementation extension allowing regular expression literals (that might be compiled at compile time in some implementation), but #"/"blah blah /...
(scheme base) and (rnrs base (6)) procedures out of the Foundations into the Batteries
If we decide to do things like this, we should document our reasoning.
I will also add that software emulation of decimal floats is available "for free" in the International Components for Unicode package, and in [Mike Cowlishaw's decimal arithmetic package](http:...
I mostly have my teaching hat on with this. If a naive programmer sees Your current balance is 292ドル.77777777777777., it at least matches the result of hand calculation;l they then can ask how to...