Meh.. I missed an autocorrect error ('y' -> 'why'), and I also
overlooked that there's several similar instances of that test in
gpoly which can fail.Hopefully, though, the issue is clear enough: J does not reliably pass this test. Thanks, -- Raul On Fri, Nov 18, 2022 at 3:55 PM Raul Miller <rauldmil...@gmail.com> wrote: > > This line in test/gpoly.ijs has been an ongoing source of woe for me: > > test j./_10+?2 11$25 > > Examples of why where it fails include: > _10j_9 5j14 _8j10 3j4 2j7 _3j1 13j1 9j_8 11j4 7j3 _9j1 > and > 11j_5 13j_2 0j12 7j_6 4j6 4j4 _1j10 9j_1 _9j14 14j9 0j1 > > This also fails in j903 and j807. > > I have not investigated if it had worked in some previous version of J. > > FYI, > > -- > Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
