Re: [racket-users] Re: mixfix and racket 2

2021-03-24 Thread Jay McCarthy
I agree with Roger that the zero-precedence `mixfix` system is a really elegant middle-ground between Honu and a whole big language. Something I really like about Pyret is that it doesn't have precedence and requires parens for legibility. -- Jay McCarthy Associate Professor @ CS @ UMass Lowell ht

Re: [racket-users] Re: mixfix and racket 2

2021-03-24 Thread Roger Keays
Looks promising. The extra enforestation phase does add another layer of complexity though. Also, operators don't use pattern matching, so I presume to support Sorawee's example below, you'd have to parse the syntax after the ? yourself. I think there is some value to considering zero-precenden

[racket-users] Re: mixfix and racket 2

2021-03-23 Thread jackh...@gmail.com
Honu's enforestation is approximately this. See https://www.cs.utah.edu/plt/publications/gpce12-rf.pdf. There's been some discussion about whether to base rhombus on a similar approach. On Tuesday, March 23, 2021 at 8:59:50 AM UTC-7 Roger Keays wrote: > From racket-news [1]: > > mixfix (pkg/src)