[r6rs-discuss] [Formal] Allow compilers to reject obvious violations

[Anton van Straaten]

[I am posting this as an individual member of the Scheme community.  I 
am not speaking for the R6RS editors.]

William D Clinger wrote:
> By my accounting, no one has yet objected to giving
> implementations discretion with respect to warnings.
> 
> By my accounting, the following messages object to
> giving implementations discretion to reject programs
> before running them:

Thanks for the accounting, that's very helpful.

I should clarify the intent behind my comment: much of the discussion in 
the subthread in which I responded -- which would be tedious to define 
precisely, but it excludes the Clinger/Felleisen discussion -- seemed to 
be taking positions on when certain kinds of errors should or even must 
be reported, without making it completely clear (to me at least) exactly 
what was being expected of R6RS in that regard, if anything.

While comments have been clear about whether the author would find it 
acceptable for R6RS to either require or forbid a particular error 
response scenario, it has not always been as clear to me what the same 
author believes about the relevance to the case for implementation 
discretion.

Anton