Re: NUL-terminated strings and eof-object-terminated generators Marc Nieper-WiÃkirchen 22 Feb 2022 08:13 UTC
Am Mo., 21. Feb. 2022 um 18:54 Uhr schrieb John Cowan <xxxxxx@ccil.org>: > On Sun, Feb 20, 2022 at 3:25 AM Daphne Preston-Kendal <xxxxxx@nonceword.org> wrote: [...] >> Anyone who is seriously maintaining an implementation which isn’t willing to make major changes for the sake of Large support was alienated, at the latest, on Tuesday this week by the adoption of syntax-case. (I don’t see Chicken or MIT adopting more of Large now, for example.) > That doesn't worry me. There is going to be a section that explains variances from Small (e.g. the numeric tower), and one of the things it will say is that a partial implementation of Large is a Better Thing than a non-implementation. While something instead of nothing is universally usually better, I fear that such an attitude is detrimental to the R7RS-large project. So far (and probably also finally), R7RS-large is basically a huge collection of SRFIs. Now, if we endorse that implementations take their pick of the parts of R7RS-large they intend to support (for whatever reasons), we are basically in a pre-R7RS-large situation: A huge corpus of SRFIs exists and portable code is possible under the assumption that the target system implements the relevant SRFIs. If this is going to be the de facto status of the implementation landscape, the efforts we make here are out of proportion to what R7RS-large would finally mean. Note that we are not talking about shipping or not shipping some non-central library like the Posix library SRFI 170 (which may not be implementable in some systems for good reasons) but we are talking about basic foundations like how the syntax expander works.