Re: GCL Version_2_6_13pre125

[Camm Maguire]

Greetings!

"Chun Tian (binghe)" <binghe.lisp@gmail.com> writes:

> Greetings,
>
> (yes, the Solaris feature request is still there. But take your time
> please.)
>
> I've learnt a lot from your Debian patch
> (axiom_20170501-6.debian.tar.xz) [1]. In particular, I need to take your
> patch which builds Axiom from system-installed GCL, instead of the one
> from "zips/" (although I have manually updated the GCL tarballs).  It
> seems that, with your patches now there's no need to modify GCL binary
> at all.  Perhaps it's those (obsoleted) patches causing the issues I met.
>
> Let's temporarily put my current issue aside.  There is one thing I know
> how to fix. Now I'm on Debian 11 amd64 with "axiom" package installed,
> if I start it and type just "0" and RET, the screen shows like this:
>
> $ axiom
>                         AXIOM Computer Algebra System
>                           Version: Axiom (May 2016)
>                Timestamp: Monday August 24, 2020 at 04:42:40
> -----------------------------------------------------------------------------
>    Issue )copyright to view copyright notices.
>    Issue )summary for a summary of useful system commands.
>    Issue )quit to leave AXIOM and return to shell.
>    Visit http://axiom-developer.org for more information
> -----------------------------------------------------------------------------
>
> (1) ->
> (1) -> 0
> (1) ->
>    (1)  0
>                                                      Type:
> NonNegativeInteger
> (2) ->
>
> Have you ever considered why there are two "(1) ->" prompts showing
> together, and whenever the user types something, the input string is
> also repeated?
>
> The answer is that, recent GCL versions have supported libreadline but
> Axiom assumes GCL doesn't have and in fact Axiom has its own
> command-line editing facility called "clef".
>
> I don't know how to disable GCL's readline feature after it has already
> been built with "--enable-readline" (by default), but if Axiom still
> builds its own GCL at beginning, the solution to the above issue is
> simply putting "--disable-readline" into GCLOPTS.
>

I think you are looking for:

)lisp (si::readline-off)

Thank you for the reminder -- I need to make sure the axiom package does
the same at startup.

Take care,


> Regards,
>
> Chun Tian
>
> [1] https://packages.debian.org/sid/axiom
>
> On 2022/11/11 18:11, Camm Maguire wrote:
>> Greetings, and thanks for your feedback.  If I recall, we had some
>> solaris correspondence outstanding, no?
>> 
>> GCL supports two applications which are incompatible with the ANSI
>> standard and require CLtL1 semantics, AXIOM and hol88.  Please make sure
>> your build is not ansi, adding --disable-ansi to configure if needed.
>> The key :ANSI-CL should be missing from *features*.
>> 
>> The debian GCL package builds four images, ANSI and non-ANSI, profiling
>> and non-profiling.  These are governed by system default settings, and
>> then by environment variables GCL_ANSI and GCL_PROF.
>> 
>> I've completed a successful build of debian's axiom package using latest
>> gcl in bookworm.  There is a small patch required in that latest latex
>> requires -interaction=nonstopmode to build the table of contents of the
>> book, toc.tex.
>> 
>> There are trace and debugging facilities that can be employed in the
>> axiom image to chase down your error, but I am confident a correct
>> build will remove it.  Please let me know if problems persist.
>> 
>> Take care,
>> 
>> "Chun Tian (binghe)" <binghe.lisp@gmail.com> writes:
>> 
>>> Hi Camm,
>>>
>>> I try to build AXIOM using this GCL version (actually it's
>>> Version_2_6_13pre126 as I found on Git). The building process didn't
>>> report any error (which is already quite amazing), but when I start
>>> Axiom I got this:
>>>
>>>                         AXIOM Computer Algebra System
>>>                           Version: Axiom (May 2017)
>>>               Timestamp: Friday November 11, 2022 at 10:54:16
>>> -----------------------------------------------------------------------------
>>>    Issue )copyright to view copyright notices.
>>>    Issue )summary for a summary of useful system commands.
>>>    Issue )quit to leave AXIOM and return to shell.
>>>    Visit http://axiom-developer.org for more information
>>> -----------------------------------------------------------------------------
>>>
>>>    >> System error:
>>>
>>>    >> System error:
>>>
>>> ...
>>>
>>>
>>> Unrecoverable error: invocation history stack overflow.
>>> Abort trap: 6
>>>
>>> This is on macOS 11 (Big Sur) where GCL 2.6.12 cannot even be built.
>>> But I do can build Axiom with GCL 2.6.12 on older macOS versions (and
>>> then install/use on macOS 11).
>>>
>>> Unfortunately I don't know how to show more detailed error messages
>>> beside "System error:".  Assuming there's no need to modify any Lisp
>>> code (they are working with GCL 2.6.12), do you think the above errors
>>> can be fixed by setting "--ihsize" when building GCL? If so, what's the
>>> suggested value to begin with?
>>>
>>> Regards,
>>>
>>> Chun Tian
>>>
>>> On 2022/11/8 21:58, Camm Maguire wrote:
>>>> Greetings!
>>>>
>>>> Fresh tag, git banner reporting, equalp hashtables, fixed only bogus nan
>>>> comparison I could find, fixed mod 2 bug reported in maxima, large
>>>> memory model support on amd64, ansi in-package support.
>>>>
>>>> This is all I know right now that should go in for the tarball.  Please
>>>> play with it if desired and let me know if there are any showstoppers.
>>>> If anyone has a macosx or cygwin environment for a build, please let me
>>>> know.  Thanks for the helpful suggestion about a vmware image -- my
>>>> current understanding is that one still needs to purchase the media for
>>>> the closed source system, but please correct me if I am wrong.
>>>>
>>>> Robert Dodier <robert.dodier@gmail.com> writes:
>>>>
>>>>> On Fri, Nov 4, 2022 at 10:32 AM Camm Maguire <camm@maguirefamily.org> 
>>>>> wrote:
>>>>>
>>>>> I reviewed the bug reports and found a few more. Here is a link to see
>>>>> bug reports which have been tagged with "gcl":
>>>>> https://sourceforge.net/p/maxima/bugs/search/?q=labels:%22gcl%22
>>>>>
>>>>> Just to be clear, these are bug reports which have been explicitly
>>>>> tagged "gcl". There are other bug reports which mention GCL, but,
>>>>> after having reviewed them, I don't believe they are specifically
>>>>> about GCL.
>>>>>
>>>>> I think these bug reports are all medium to low priority, although I
>>>>> would certainly be glad if they get resolved. The one which seems most
>>>>> pressing is:
>>>>> https://sourceforge.net/p/maxima/bugs/1234/
>>>>> which is about non-ASCII characters in a folder name on Windows; I
>>>>> seem to recall more than one user has bumped into that. I don't know
>>>>> the status of that at this point.
>>>>
>>>> Looking these over too.  I could really use some reproducible input
>>>> (which I noticed you requested but apparently have not received) for the
>>>> pathname special character report, and the SIGPIPE report (I requested
>>>> on savannah with no reply).  The NaN report gives false now as you note,
>>>> 2.0^1024.1 still gives the inf for testing purposes.  I think the false
>>>> is in simp-expt in maxima, but haven't got further.  Most of his results
>>>> had been fixed some time ago.  Found one remaining in interpreted code
>>>> only which should now be closed, see transcript below.  The dribble
>>>> error only omits error header output (which is sent to a different file
>>>> descriptor) -- all newlines are identical addressing the original
>>>> complaint.  I can get the error header in there too if this is
>>>> important.
>>>>
>>>>>
>>>>>>> The problems around GCL + Maxima stem from the capabilities (ASDF,
>>>>>>> Unicode, etc) which are present in other implementations and not GCL.
>>>>>>> It's not much about conditionalizations.
>>>>>>
>>>>>> OK, thanks for that.  I'll put this on the back burner then.  Might I
>>>>>> create a git branch for this purpose?
>>>>>
>>>>> I think one of the anonymous ones was reported by me, not that it
>>>>> matters too much. Also #47404 is a bona fide bug; all comparisons with
>>>>> NaN are false, except NaN != NaN, which is true, according to IEEE
>>>>> 754. The spec itself is proprietary and not available to casual
>>>>> readers, to the best of my knowledge, but summaries are all over the
>>>>> web, including many statements about comparisons with NaN.
>>>>>
>>>>> best,
>>>>>
>>>>
>>>> =============================================================================
>>>>> (setq si::*print-nans* t)
>>>>
>>>> T
>>>>
>>>>> (setq a (exp 10000.0))
>>>>
>>>> #<inf>
>>>>
>>>>> (setq b (- a) c (- a a))
>>>>
>>>> #<-nan>
>>>>
>>>>> (<= c c)
>>>>
>>>> NIL
>>>>
>>>>> (>= c c)
>>>>
>>>> NIL
>>>>
>>>>> (> c c)
>>>>
>>>> NIL
>>>>
>>>>> (< c c)
>>>>
>>>> NIL
>>>>
>>>>> (= c c)
>>>>
>>>> NIL
>>>>
>>>>> (/= c c)
>>>>
>>>> T
>>>>
>>>>> (dolist (l '(< <= = >= > /=)) (print (list (funcall l c c) (funcall 
>>>>> (compile nil `(lambda (x y) (declare (long-float x y)) (,l x y))) c c))))
>>>>
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x61f100 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x5badc0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x554a40 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x5cfe50 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f2f8c0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f412a0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (T T) 
>>>> NIL
>>>>
>>>>> (setq c a)
>>>>
>>>> #<inf>
>>>>
>>>>> (dolist (l '(< <= = >= > /=)) (print (list (funcall l c c) (funcall 
>>>>> (compile nil `(lambda (x y) (declare (long-float x y)) (,l x y))) c c))))
>>>>
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f52c90 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f64670 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (T T) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f76050 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (T T) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f87a40 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (T T) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2f99420 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2faae00 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> NIL
>>>>
>>>>> (setq c (- a a))
>>>>
>>>> #<-nan>
>>>>
>>>>> (setq cf (float c 0.0s0))
>>>>
>>>> #<-nan>
>>>>
>>>>> (type-of cf)
>>>>
>>>> SHORT-FLOAT
>>>>
>>>>> (dolist (l '(< <= = >= > /=)) (print (list (funcall l cf cf) (funcall 
>>>>> (compile nil `(lambda (x y) (declare (short-float x y)) (,l x y))) cf 
>>>>> cf))))
>>>>
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2fbc7f0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2fce1d0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2fdfbb0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x2ff15a0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x3002f80 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x3014960 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (T T) 
>>>> NIL
>>>>
>>>>> (dolist (l '(< <= = >= > /=)) (print (list l (funcall l cf cf) (funcall 
>>>>> (compile nil `(lambda (x y) (declare (short-float x y)) (,l x y))) cf 
>>>>> cf))))
>>>>
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x7e9e3d0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (< NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x7eafd60 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (<= NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x7ec1740 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (= NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x7ed3130 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (>= NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x7ee4b10 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (> NIL NIL) 
>>>> Compiling /tmp/gazonk_94832_0.lsp.
>>>> End of Pass 1.  
>>>> End of Pass 2.  
>>>> OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3
>>>> Finished compiling /tmp/gazonk_94832_0.o.
>>>> ;; Loading #p"/tmp/gazonk_94832_0.o"
>>>> start address -T 0x7ef64f0 ;; Finished loading #p"/tmp/gazonk_94832_0.o"
>>>>
>>>> (/= T T) 
>>>> NIL
>>>>
>>>>>
>>>> =============================================================================
>>>>
>>>>
>>>>> Robert
>>>>>
>>>>>
>>>>>
>>>>
>>>
>> 
>

-- 
Camm Maguire                                        camm@maguirefamily.org
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah