Hmm... only if we're updating stdlib, supporting versions of J back to 4.0.1. A worthwhile effort, though, if someone was actually tackling those issues.
Otherwise, I think an idiom would be about as good as it gets. -- Raul On Thu, Dec 21, 2023 at 9:24 PM Ian Clark <earthspo...@gmail.com> wrote: > > For those of us digital archaeologists running back-levels of J, it would > be nice to have a robust test for a given version of J or later which > tolerates versions down to 4.01, say. > > > What represents best-practice? The best I can come up with is: > > jversion=: 3 : 0 > > ". 'j.' -.~ '/' taketo 9!:14'' > > ) > > > > which reliably returns a 3-digit integer, permitting (say): > > if. 900 <: jversion'' do. … > > > A candidate for stdlib? > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm