Am Fr., 11. Sept. 2020 um 19:55 Uhr schrieb Lassi Kortela <xxxxxx@lassi.io>: > > > Scheme and Common Lisp deal with > > almost the same concepts. A library interface rarely needs to be > > concerned with the parts where they diverge > > > > The differences in macros are very important too. Few would want to > > rewrite their syntax-rules or syntax-case macros in define-macro. > > Macros are mostly an internals issue, not an interface issue. > > In the rare somewhat case that a library exports a macro, equivalent > syntax-rules and defmacro definitions can be given. It doesn't take that > many minutes to write a macro. Even SRFI 177 wasn't a big macro-writing > effort by any means, we spent all our time discussing the design. True, but a major reason why we couldn't find a convincing design was the macro system or, rather, the lowest common denominator. When we have made our decision about the macro system of R7RS (large), which extends the syntax-rules of R7RS (small), we should give SRFI 177 another try. I know that John had reasons to do it the other way round, but I still think that fixing the unportable stuff (reader syntax, macro system, extensions to the library system) first (i.e. in the red edition) would have made a lot more sense. > Ideally most programming language features should be McDonalds instead > of fine dining; people shouldn't have to do anything fancy when they are > not solving fancy problems. Consistency and good defaults matter a lot. > Scheme is a pleasure to use in part because things are named so well. Luckily, there are alternatives besides fine dining and McDonald's! :) > If only there was a way to convince languages to use the same names for > the same things. Perhaps it could be done with shims if there's enough I think it is dangerous. Language semantics differ (good so because otherwise there's no reason for another language!) and making people believe things are the same because they look the same is not a good thing.
