On Tuesday 16 November 2004 08:48, Viktor T. Toth wrote:
> Hello everyone,
>
> First of all, Jim, thanks for the explanation regarding the switch to
> case-sensitive mode; I admit I did not quite understand where we were at
> prior to your e-mail.
>
> I just submitted a set of changes (sorry for the multiple CVS COMMITs) for
> share/tensor. The downcasing went fairly smoothly, with only minor
> problems. The only one I cannot fix at this time concerns the infix()
> function. It appears that it no longer accepts operators containing special
> symbols. For instance, the following excerpt worked previously, defining
> '@L' as an infix operator:
>
> (%i1) infix("xl");
>
> (%o1) "xl"
> (%i2) a xl b;
>
> (%o2) a xl b
> (%i3) infix("@l");
>
> (%o3) "@l"
> (%i4) a @l b;
> Incorrect syntax: @ is not an infix operator
> a @l
> ^
> (%i4)
> (%o4) b
>
>
> Other than this, the tensor packages should be functional, except when they
> depend on other, not yet downcased share packages (e.g., swartz.dem will
> fail, due to its dependence on simplification/facexp.mac, which has not yet
> been downcased.)
I confirm that everything works in share/tensor if to downcase all packages
that invoke from there. I have been experimenting with the changed defintion
for the Lie derivative from infix "@l" to Lie_deriv(forma,[vec, ind])).
With this change exterior algebra demos and the general demo 'ten_alg.dem'
again work well.
There is a small bug in function show in itensor.lisp. It show covariant
indices in upper case. Fortunately, it does not affect the mathematical
correctness of expression.
I'll check things further on my workable files.
>
> While I was at it, I also used this opportunity to change some file names,
> where the old names were obviously leftovers from when MAXIMA was being
> used on a file system limited to 6-character (or 8-character) file names.
> ctensr.mac is now ctensor.mac, and ctenso_n.dem (n=1..4) are now renamed to
> ctensor_n.dem.
Viktor, please let me know when I can start to change user's documetation
about tensors. I guess as a first step you could apply the proposed changes
to names of maxima tensors functions to make them in correspondence with
commercial macsyma. Then we will test and after this it is reasonably to
start working on documentation.
best regards,
valery