Re: NUL-terminated strings and eof-object-terminated generators Marc Nieper-WiÃkirchen 25 Feb 2022 17:23 UTC
(Added WG2.) Am Fr., 25. Feb. 2022 um 18:09 Uhr schrieb John Cowan <xxxxxx@ccil.org>: > > > > On Fri, Feb 25, 2022 at 2:29 AM Marc Nieper-Wißkirchen <xxxxxx@gmail.com> wrote: > >> >> If it emerges that essential compatibility with R6RS is not possible >> on technical grounds, this is one thing, but forbidding R6RS >> compatibility on political grounds > > > I didn't say they objected on political grounds; in fact, I don't know their reasons. I said that if you wanted that judgment changed, you'd have to use political means. >> >> is en clair just stupid and >> petty-minded. > > > Now that's a political judgement. I don't disagree. :) Actually, I chose my words deliberately, so leaving open on purpose whether the actual decision was based on political grounds or not. But if one of the requirements R7RS-large has to fulfill is "no full R6 compatibility" apparently for its own sake, I have a hard time imagining non-political grounds. If I am the only one who thinks that R7RS-large could or even should be a vehicle to unite the Scheme worlds of R7RS-small and R6RS again, asking the SC to revert their decision wouldn't be worth it. If not, I would be happy if others spoke up as well. >> The only practical solution that will also yield the best for the >> community is in my opinion not to modify an existing implementation to >> create an R7RS-large fork, but to persuade the maintainers of such an >> existing high-quality implementation to steer their implementation >> towards R7RS-large compatibility. > > > If you want that to happen, you need to put the least burden on the implementers that is possible. And one has to get the implementers on board. > >> vary a lot with respect to how companies influence them. >> >> >> In any >> >> case, R7RS-large has to offer more than existing standards and (!) >> >> tools. > > > Small offers less than almost any Scheme implementation, but many implementors support it anyway. R7RS-small was a step forward in the landscape of all R5RS+-implementations, unifying things like the library system or record systems. R7RS-large has a different goal. Moreover, by its sheer size, much fewer implementors could support it.