Greetings! I see we've been on parallel paths.
I've committed a lisp-array-alike1 to gcl_cleanup for review. With this
and a gcl/unicode version of gobble-whitespace also committed to this
branch, there are no unexpected test failures with maxima/GCL 2.6.14.
I'll update the testsuite file for the GCL expectations, and review the
remaining expected failures.
To my understanding, I also 'own' the #-=kcl conditional. I'll review
and clean these as well if there are no objections.
Do we really support kcl at all anymore? I think kcl has been mostly replaced by gcl.
I think all the kcl conditionals should be removed. If that's the case, it's best to do that on a separate branch so that they all get removed together without other changes getting mixed in. It makes figuring out what happened easier and git bisect easier too.
I say your very helpful bug report at savannah -- thanks! They indeed
seem to have changed the web interface so that no one can close bugs.
I'll correspond with them on this.
Take care,
Robert Dodier <robert.dodier@gmail.com> writes:
> On Fri, Jan 20, 2023 at 4:56 PM Raymond Toy <toy.raymond@gmail.com> wrote:
>
>> On the other hand, I see:
>>
>> > (describe (format nil "~r" 8))
>>
>> "eight" - string
>> dimension: 64
>> fill pointer: 5
>>
>> so perhaps using array-dimensions and array-total-size are not the right things to use in this case. AFAICT, format returns a string and a string with a fill-pointer is a subtypep of string.
>
> Yeah, that's right, the strings in question have a fill pointer, and
> the fill pointer is valid, so there's no bug. I'm fixing up ALIKE1 to
> handle vectors which have a fill pointer. Sorry for the noise.
>
> best,
>
> Robert
>
>
>
--
Camm Maguire camm@maguirefamily.org
==========================================================================
"The earth is but one country, and mankind its citizens." -- Baha'u'llah