Re: [Maxima-discuss] Status of 5.48.1

[Jaime V. <vi...@fe...>]

Hello,

Now that I have also installed 5.47.0 in Windows, I've come to the 
conclusion that the problems I found already existed in that version, 
using Clisp. the difference is that we now have another lisp flavor 
(ccl64) that does not nave some of those problems but brings new ones. 
Sbcl seems fine.

Here is a more complete set of tests.

Notation:
v1= version 5.47.0, v2= 5.47post (last night master)

L1=Sbcl, L2=Clisp, L3= ccl64

tests:

T1= load ("diag")
T2= load ("orthopoly")
T3= load ("eigen")
T4= load ("dynamics")
T5= load ("plotdf")
T6= system ("echo Hello")
T7= plot2d (x, [x,0,2], [plot_format, xmaxima])
T8= plotdf ([y,x], [x,0,2], [y,0,2])

Running all those test in the console in Windows, the results are the 
following

Test   (v1, L1)   (v2, L2)   (v1, L2)  (v2, L2)   (v2, L3)
T1      pass        pass         pass    pass       pass
T2      pass        pass         fail    fail       pass
T3      pass        pass         pass    pass       pass
T4      pass        pass         fail    fail       pass
T5      pass        pass         fail    fail       pass
T6      pass        pass         pass    pass       fail
T7      pass        pass         pass    pass       fail
T8      pass        pass         fail    fail       fail


The packages that fail to be loaded, apparently is due to their path and 
not to anything inside the packages. The failure of tests T6, T7 and T8 
are in fact lisp programming problems. If T6 fails, T7 will always fail, 
and if either T5 or T6 fail, T8 will fail.

I can easily change plot2d and plotdf to make them work even if T6 
fails. But I cannot fix the failure of T8 when T5 fails. That would 
leave us with 0 failures for Sbcl, 1 failure for Ccl64 and 4 failures 
for Clisp.

Regards,

Jaime

On 03/08/25 19:07, Jaime Villate via Maxima-discuss wrote:
> On 03/08/25 17:23, Jaime Villate via Maxima-discuss wrote:
>> Now we must fix system("echo Hello") in Windows (defined in 
>> src/macsys.lisp).
>
> That works in some Lisp flavors and not in other. But I have noticed 
> an even more serious problem!
>