Linear display not correct?



>>>>> "Dieter" == Dieter Kaiser <drdieterkaiser at web.de> writes:

    Dieter> Ray wrote:
    >> Neat.
    >> 
    >> I see that the symbol plists for $bessel_j and %bessel_j are different
    >> between the beginning and end of your tests.  In fact,
    >> grind('bessel_j(n,1)) causes the symbol plists to change.  The verb
    >> and noun properties are added.  And just doing 'bessel_j(n,1) by
    >> itself adds the noun and verb properties.  I guess the reader does
    >> that.

    >> That seems to hint we need to add those properties from the beginning.

    Dieter> To put the noun and verb properties on the property list don't work completely.
    Dieter> I have tried it. We get an additional ' in the output:

Yes.  As you mentioned in another e-mail, you need alias and
reversealias.

    Dieter> (%i15) display2d:false;
    Dieter> (%o15) false
    Dieter> (%i16) bessel_j(n,x);
    Dieter> (%o16) ?%bessel_j(n,x)
    Dieter> (%i17) :lisp ($nounify '$bessel_j)
    Dieter> %BESSEL_J
    Dieter> (%i17) bessel_j(n,x);
    Dieter> (%o17) 'bessel_j(n,x)   << output has additional '

    Dieter> I think there has something really changed in the handling of the noun and verb
    Dieter> forms for output. I have not Maxima 5.14.0 on my PC but it might be interessting

FWIW, the functions that deal with this issue with linear display are
msize, msize-atom, and msize-function in grind.lisp.

In msize-function, we see that if there's no reversealias, a quote is
prepended.  cvs annotate says that particular line hasn't changed
since the Great Downcasing back in 2004, which goes back to version
5.9.2 or earlier.

Ray