Release

[Waldek Hebisch]

Attached are proposed release notes for FriCAS 1.3.9.

--
Waldek Hebisch

[Grégory Vanuxem]

I would use "tagged" instead of taged.

https://www.spellmentor.com/t/ta/tagged/taged-or-tagged

--
You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/ZJ9hfRyW9KgmWuyL%40fricas.math.uni.wroc.pl.

[Ralf Hemmecke]

On 01.07.23 15:11, Grégory Vanuxem wrote:
> I would use "tagged" instead of taged.
> https://www.spellmentor.com/t/ta/tagged/taged-or-tagged

Yes, yes.

I would even put in a bit more... proposal at:

https://github.com/fricas/fricas/compare/master...hemmecke:fricas:proposed-release-notes

Ralf

[Waldek Hebisch]

Given URL above Github says:

: This comparison is taking too long to generate.
:
: Unfortunately it looks like we can't render this comparison for you
: right now. It might be too big, or there might be something weird with
: your repository.

Only after chasing a few links I got what looks like the proposal.

Some remarks about principles I use when preparing notes. I try
to keep them short and skip "less interesting" changes (where
"interesting" is from user point of view). In particular internal
changes normally are deemed "not interesing". That is why I skipped
"Simplify hashtables" and "Generalize new univariate factorizer".
Hashable change is mostly internal, but has potentially significant
impact on user code, so it is mentioned.

The other ommisions are borderline cases, I skipped them scanning
list of changes, but have no strong feeling one way or the other.

--
Waldek Hebisch

[Ralf Hemmecke]

> Some remarks about principles I use when preparing notes. I try
> to keep them short and skip "less interesting" changes (where
> "interesting" is from user point of view). In particular internal
> changes normally are deemed "not interesing". That is why I skipped
> "Simplify hashtables" and "Generalize new univariate factorizer".
> Hashable change is mostly internal, but has potentially significant
> impact on user code, so it is mentioned.

Fine with me. But you should correct "taged" to tagged.

I just went through the commits and added things that I think would be
relevant. Of course, a personal view in the things.

For example changing escape() to underscore() is an API change and
should be mentioned. Similar for introduction of
LinearSystemMatrixPackage2 and ParametricComplexCurve, even if they are
just small additions.

Decide what you take from my list. It is just a proposal, after all.

Ralf

[Tim Daly]

As a native american english speaker the use of 'taged' isn't "correct"

but it is clear in context. It falls into the category of things often written

by people who are not native speakers. I'm impressed at how well

people can write in more than one language. My German often

suffers from "The umlaut is silent" (e.g in Groebner/Grobner).

I'm surprised it was even worth mentioning.

Tim

[Ralf Hemmecke]

> My German often suffers from "The umlaut is silent"

What exactly should that mean?

As a native German I can tell you that o and ö are not the same letter.
And they are pronounced differently. So I don't understand your "silent".

> (e.g in Groebner/Grobner).

Well, if you write/say Flote or Krote instead of Flöte (flute) or Kröte
(toad), then a German would probably either understand that you have
problems with umlauts of (if the German does not have an understanding
for other languages) even look at you and hardly understand you at all.

People write Floete or Kroete on keyboards that do not have the umlaut
letters on them and this is usually commonly understood.

Ralf

[Tim Daly]

It is a normal in the U.S. to have silent letters that other

languages pronounce. The obvious one is the "H" which

the English pronounce and the American does not, such

as in the word "Herb" (a kind of plant). The English say

"Herb", the American say "erb".

Using the phrase "the (letter) is silent" is cheap U.S. comedy.

Being a poor comedian but aware that the unlaut is a different German

letter ("o" vs "ö", since Gröbner is the correct spelling) I tried to

make a joke that "the umlaut is silent". Jokes always loose

(the second "o" is silent) in translation. Sigh.

Anyway, this is off topic but I intended to defend Waldek's spelling mistake.

Having studied and forgotten both German and French

I am honestly impressed with all the group member's command of English.

Tim

--
You received this message because you are subscribed to a topic in the Google Groups "FriCAS - computer algebra system" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fricas-devel/0C7xQIeV0lA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fricas-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/34ee2858-c803-f411-1765-eea6434ce85d%40hemmecke.org.

[Waldek Hebisch]

I have now put preliminary source release tarball at:

http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9-pre.tar.bz2

If there are no significant problems it will be promoted to
release.

In about 2-3 days I plan to create binary release.

--
Waldek Hebisch

[Ralf Hemmecke]

I've just created a pull request #131 to fix some minor issues in
install.rst.

I actually do not quite understand why you insist on

Build *extra* documentation

I actually think that it is THE documentation, nothing extra.

On 02.07.23 20:08, Waldek Hebisch wrote:
> I have now put preliminary source release tarball at:
>
> http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9-pre.tar.bz2
>
> If there are no significant problems it will be promoted to
> release.

I'll look at it tomorrow night.

Ralf

[Nasser M. Abbasi]

No problems building and running this tar file on Linux kernel 6.1.19-1  Manjaro. No problems using it with sagemath 10.0

>./configure --with-lisp=sbcl

make

sudo make install

     FriCAS Computer Algebra System
                            Version: FriCAS 1.3.9
                 Timestamp: Sun Jul  2 01:31:24 PM CDT 2023

fricas --version
FriCAS 1.3.9
based on sbcl 2.3.0
>which fricas
/usr/local/bin/fricas
>


>sage
┌────────────────────────────────────────────────────────────────────┐
│ SageMath version 10.0, Release Date: 2023-05-20                    │
│ Using Python 3.10.9. Type "help()" for help.                       │
└────────────────────────────────────────────────────────────────────┘
sage: print(fricas.eval(")lisp |$build_version|"))

Value = "FriCAS 1.3.9"
sage:

Build first 12 files (1,892 integrals) of CAS integrations tests with it. No new problems.

I've put file #10 report PDF and the build log in this folder if needed

On these first 12 files Fricas 1.3.9 scored 95.51% compared to Fricas 1.3.8 score of 95.24%

 --Nasser

[Waldek Hebisch]

On Mon, Jul 03, 2023 at 12:04:36AM +0200, Ralf Hemmecke wrote:
> I've just created a pull request #131 to fix some minor issues in
> install.rst.

I have noticed a little thing the we probably should change: AFAICS
Sphinx is only needed to build .html pages. Building book.pdf works
without Sphinx. So more precise statement
should be

The ``.html`` documentation is built via Sphinx.

Ignoring that OK. Note that we now have release branch (r1.3.8), so
this should be commited twice, first to the trunk and then to release
branch.

> I actually do not quite understand why you insist on
>
> Build *extra* documentation
>
> I actually think that it is THE documentation, nothing extra.

For better clarity. Normal build from release tarball or
with access to X builds Hyperdoc documentation. Saying
"the documentation" would be confusing, implying either
that Hyperdoc pages are build together with book and
.html pages or that that Hyperdoc pages are not there at all.

--
Waldek Hebisch

[Grégory Vanuxem]

Hello  Waldek, *,

Just built using SBCL-2.3.6 (dynamic-space-size=4Gb) with GMP and Aldor support.

$make check

=> No unexpected failure.

┌──(greg㉿ellipse)-[~/Tmp]
└─$ fricas -nox
Checking for foreign routines
FRICAS="/usr/local/lib/fricas/target/x86_64-linux-gnu"
spad-lib="/usr/local/lib/fricas/target/x86_64-linux-gnu/lib/libspad.so"
foreign routines found
openServer result 0

                       FriCAS Computer Algebra System
                            Version: FriCAS 1.3.9

                 Timestamp: lun. 03 juil. 2023 08:20:03 CEST
-----------------------------------------------------------------------------
   Issue )copyright to view copyright notices.
   Issue )summary for a summary of useful system commands.
   Issue )quit to leave FriCAS and return to shell.
-----------------------------------------------------------------------------


(1) -> )co sieve
   Compiling FriCAS source code from file /home/greg/Tmp/sieve.as using
      Aldor compiler and options
-O -Fasy -Fao -Flsp -lfricas -Mno-ALDOR_W_WillObsolete -DFriCAS -Y $FRICAS/algebra -I $FRICAS/algebra
      Use the system command )set compiler args to change these
      options.
   Compiling Lisp source code from file ./sieve.lsp
   Issuing )library command for sieve
   Reading /home/greg/Tmp/sieve.asy
(1) -> sieve 100

   (1)  25

So all is right.

It is just a pity from my point of view there is no support of Hunchentoot in it.

I mean having an option for it. I know you do not want a  required Internet

access to build FriCAS and I understand that.

But if you can load, (require ...), Hunchentoot, even without Internet access,

that would make sense. Using Roswell I have Hunchentoot for Lisp

implementations supported by Roswell and that supports hunchentoot.

It is just a matter of:

$ros install hunchentoot

and later (require ...) it.

Later I can build FrICAS with Hunchentoot in it if I want, and if it is not

available it will not be in FRICASsys, it doesn't really matter, this is just an

option.

Personally I use in my FriCAS copy the following (borrowed from Ralf and Kurt work):

configure.ac:

This need to be adapted to add

SBCL_HOME and WEBSPAD_REQUIREMENTS

=================================================

+AC_ARG_WITH([quicklisp],
+    [AS_HELP_STRING([--with-quicklisp=PATH],
+        [use quicklisp installation path (default "~/quicklisp")])],
+    [USE_QUICKLISP=yes
+     AS_IF([test -d $withval],
+       [QUICKLISPPATH=$withval],
+       [AS_IF([test -d ~/quicklisp],
+          [QUICKLISPPATH=~/quicklisp],
+          [AC_MSG_ERROR([quicklisp path not available])])])],
+    [USE_QUICKLISP=no])
+
+INCLUDE_WEBSPAD=no
+AC_ARG_ENABLE(webspad,
+  [AS_HELP_STRING([--enable-webspad],
+                  [include a webserver into FriCASsys, (yes/no)])],
+  [AS_CASE([$enableval],
+    [yes], [AS_IF([test "$fricas_lisp_flavor" = "sbcl" \
+     || test "$fricas_lisp_flavor" = "openmcl" \
+ || test "$fricas_lisp_flavor" = "ecl"],
+       [INCLUDE_WEBSPAD=yes],
+       [INCLUDE_WEBSPAD=no
+        AC_MSG_ERROR([webspad only available for SBCL or Clozure CL])])],
+    [no], [INCLUDE_WEBSPAD=no],
+    [AC_MSG_ERROR([only yes/no allowed as value for --enable-webspad])]
+  )]
+)
+
+# Now Check that Hunchentoot is available.
+WEBSPAD_REQUIREMENTS=
+AS_IF([test x"$USE_QUICKLISP" = xyes],
+       [WEBSPAD_REQUIREMENTS="(load \"$QUICKLISPPATH/setup.lisp\")(require :hunchentoot)"],
+       [WEBSPAD_REQUIREMENTS="(require :asdf)(require :hunchentoot)"])
+
+AS_IF([test x"$INCLUDE_WEBSPAD" = xyes],
+  [AC_MSG_CHECKING([for asdf and hunchentoot])
+  AS_IF([$fricas_lisp --non-interactive --eval "(progn $WEBSPAD_REQUIREMENTS)"],
+    [AC_MSG_RESULT([yes])],
+    [AC_MSG_RESULT([no])
+     AC_MSG_ERROR([asdf and hunchentoot are required for --enable-webspad])
+     INCLUDE_WEBSPAD=no])]
+)
+
+AS_IF([test x"$INCLUDE_WEBSPAD" = xyes],
+  [AC_MSG_NOTICE([Webserver hunchentoot will be included into FRICASsys.])],
+  [AC_MSG_WARN([No webserver will be included into FRICASsys.])
+   WEBSPAD_REQUIREMENTS=""]
+)
+
+AC_ARG_WITH([sbcl-home],
+    [AS_HELP_STRING([--with-sbcl-home=PATH],
+                    [Value for SBCL_HOME.])],
+    # Action if --with-sbcl-home is given.
+    [AS_IF([test -d "$withval"],
+           [SBCL_HOME=$withval],
+           [AC_MSG_ERROR([--with-sbcl-home directory does not exist.])])],
+
+    [lisp_prog=`echo $fricas_lisp | sed 's/ .*//'` # remove parameters
+     lisp_path=`which $lisp_prog` # cannot be empty
+     SBCL_HOME="${lisp_path%sbcl}../lib/sbcl"]
+)
+
+#AC_MSG_NOTICE([SBCL_HOME=$SBCL_HOME])
+AC_SUBST(SBCL_HOME)
+AC_SUBST(WEBSPAD_REQUIREMENTS)

====================================================

I removed the documentation here.

And in src/interp/Makefile.in, something like:

=======================================================

+ WEBSPAD_REQUIREMENTS = @WEBSPAD_REQUIREMENTS@+
+ SBCL_HOME = @SBCL_HOME@
+ export SBCL_HOME

=======================================================

${FRICASSYS}: ../etc/stamp-databases
echo '(defparameter FRICAS-LISP::*building-fricassys* t)' \
  '(load "makeint.lisp") #-:ecl(BOOT::reclaim)' \
  '#+:cmu (setf (ext:search-list "libspad:")' \
  '(list "${FRICAS}/lib/" "${libdir}/fricas/target/${target}/lib/"))' \
  '#+:cmu (setq ext:*top-level-auto-declare* t)' \
  '#+:cmu (setq *compile-verbose* nil)' \
  '#+:cmu (setq *compile-print* nil)' \
  '#+:cmu (declaim (optimize (ext:inhibit-warnings 3)))' \

+   '${WEBSPAD_REQUIREMENTS}' \
  '#+:ecl(FRICAS-LISP::make-program "$(BASE)$@" nil)' \
  '#-:ecl(BOOT::spad-save "$(BASE)$@" t)' \
             | DAASE="$(BASE)$(fricas_targetdir)" ${BOOTSYS}
@ echo 6a ${FRICASSYS} created

exposed.lsp: $(fricas_src_algdir)/exposed.lsp

=====================================================

It is just as simple as this.

Of course it is too late  to add something like this

but I think this could be added to GitHub.

For me, this adds simplicity without adding _necessary_ required things.

And hsbcl can also be used I guess.

Regards,

__

Greg

 

--

You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.

To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/ZKG9FS3xy4iICEaI%40fricas.math.uni.wroc.pl.

[Ralf Hemmecke]

On 03.07.23 09:21, Grégory Vanuxem wrote:
> Just built using SBCL-2.3.6 (dynamic-space-size=4Gb) with GMP and Aldor
> support.

> It is just a pity from my point of view there is no support of Hunchentoot
> in it.

> I mean having an option for it.

Well, there is no option for it in the FriCAS build machinery, because
the need of Hunchentoot is currently somewhat orthogonal to the main
FriCAS area, i.e., you can work perfectly fine with FriCAS even if you
do not have Hunchentoot.

Hunchentoot is only needed if one wants the jFriCAS interface. And for
that one just needs to provide and use HSBCL during the build of FriCAS.
So in some sense there is a configure option, but agreed, the burden to
build/fetch HBCL is completely put onto the user.

See (currently, till the official docs are build):

https://hemmecke.github.io/fricas/install.html#hunchentoot-optional

Until the official distribution of hsbcl-1.3.9.tar is put onto the
release page you can use this:


wget http://www.math.uni.wroc.pl/~hebisch/fricas/hsbcl.tar
tar -xf hsbcl.tar
cd hsbcl
time ./build_hsbcl > rap1 2>&1

as written here:

https://groups.google.com/g/fricas-devel/c/ixaDVZ6inxw/m/KY4WWapYAwAJ

I find hsbcl-1.3.9.tar a bit misleading, because the version number
refers to FriCAS and not SBCL. In fact, hsbcl-1.3.9.tar just contains
build instructions to build an SBCL image with Hunchentoot included.
What version of SBCL you use is completely up to what the user provides
on his/her system.

Can you try out whether the build instructions on

https://hemmecke.github.io/fricas/install.html

work for you to build FriCAS with the jFriCAS interface? Maybe there is
still something unclear that we could fix before release.

Thanks.
Ralf

[Grégory Vanuxem]

Le lun. 3 juil. 2023 à 09:42, Ralf Hemmecke <ra...@hemmecke.org> a écrit :
>
> On 03.07.23 09:21, Grégory Vanuxem wrote:
> > Just built using SBCL-2.3.6 (dynamic-space-size=4Gb) with GMP and Aldor
> > support.
>
> > It is just a pity from my point of view there is no support of Hunchentoot
> > in it.
>
> > I mean having an option for it.
>
> Well, there is no option for it in the FriCAS build machinery, because
> the need of Hunchentoot is currently somewhat orthogonal to the main
> FriCAS area, i.e., you can work perfectly fine with FriCAS even if you
> do not have Hunchentoot.

Yes, I know, I was not clear.

>
>
> Hunchentoot is only needed if one wants the jFriCAS interface. And for
> that one just needs to provide and use HSBCL during the build of FriCAS.
> So in some sense there is a configure option, but agreed, the burden to
> build/fetch HBCL is completely put onto the user.

I think it does not need to be absolutely necessary if you want
to use FriCAS Jupyter's notebook. I agree that with Debian like systems
for example it is not easy to _install_ hunchentoot support if necessary,
so hsbcl is of help. And I do not suggest to delete the HSBCL possibility
to build FriCAS. As I said, it cans help (a lot). But this is a strong
requirement,
It requires a custom SBCL image. So I suggest to add an option if you already
have the hunchentoot package in your Lisp settings. You suggested that too
previously if memory does not deserve me.

Attached is a proof of concept patch that can be applied on fricas-1.3.9-pre.
I tried this with :
$./configure --enable-gmp --enable-aldor --with-lisp='ros run'
and
$./configure --enable-gmp --enable-aldor --enable-webspad --with-lisp='ros run'

'make check' gives no unexpected failure.
sieve.as can be compiled and returns 25 with 'sieve(100)'
and if --enable-webpad is added I can smoothly run the
FriCAS kernel with:
$jupyrer notebook

Even with ')set message time on' in my $HOME/.fricas.input.

'--enable-webspad' should be renamed though.

You can check that with:
(see https://github.com/roswell/roswell/releases/tag/v22.12.14.113)
Here it's for an AMD64 Debian like system.

$ wget https://github.com/roswell/roswell/releases/download/v22.12.14.113/roswell_22.12.14.113-1_amd64.deb
$ sudo dpkg -i roswell_22.12.14.113-1_amd64.deb
$ ros install hunchentoot

and in the patched source tree
$./configure --enable-gmp --enable-aldor --enable-webspad --with-lisp='ros run'
make and sudo make install as usual.

It's really easy to return to the previous Linux configuration with
'dpkg --purge roswell' and 'rm -Rf .$HOME/.roswell'

I have followed your discussion with Waldek on this, and
I agree, having Hunchentoot using Internet access etc.
is really not convenient. Here this is just another option,
simpler with correct Lisp settings if available.
Personally I would prefer to use, say
'sudo apt install cl-hunchentoot' but my Linux distribution
does not have this package for now.

> See (currently, till the official docs are build):
>
> https://hemmecke.github.io/fricas/install.html#hunchentoot-optional
>
> Until the official distribution of hsbcl-1.3.9.tar is put onto the
> release page you can use this:
>
>
> wget http://www.math.uni.wroc.pl/~hebisch/fricas/hsbcl.tar
> tar -xf hsbcl.tar
> cd hsbcl
> time ./build_hsbcl > rap1 2>&1
>
> as written here:
>
> https://groups.google.com/g/fricas-devel/c/ixaDVZ6inxw/m/KY4WWapYAwAJ

Yes, that's very handy. I didn't notice it in fact. Thanks.

> I find hsbcl-1.3.9.tar a bit misleading, because the version number
> refers to FriCAS and not SBCL. In fact, hsbcl-1.3.9.tar just contains
> build instructions to build an SBCL image with Hunchentoot included.
> What version of SBCL you use is completely up to what the user provides
> on his/her system.
>
> Can you try out whether the build instructions on
>
> https://hemmecke.github.io/fricas/install.html
>
> work for you to build FriCAS with the jFriCAS interface? Maybe there is
> still something unclear that we could fix before release.

For sure I will.
__
Greg

PS: The patch needs to be modified if one use '--enable-webspad' with an already
included Hunchentoot SBCL

[Ralf Hemmecke]

Hi Greg,

I understand your wish.

> $./configure --enable-gmp --enable-aldor --enable-webspad --with-lisp='ros run'

First of all, "--enable-webspad" has not made it into the current
official repo. And it probably never will.

After Waldek suggested the HSBCL idea, There is no need for --webspad
anymore. At the time I wrote these patches, I thought to include
webspad.lisp into the official fricas repo. But now essential parts of
it, namely interpret_block, the modified mathprintWithNumber and
*OBEY-STDOUT* are in official FriCAS, only hunchentoot was needed. So
the jFriCAS code is relatively orthogonal to FriCAS and can be
considered as just another interface similar to efricas and frimacs.

What actually could be done was a "service" option or script with the
meaning of downloading hsblc-1.3.9.tar, compiling it into a temporary
directory and then using it in the actual build of FriCAS.
To automate this, would certainly make the user experience better.
Whether this should be done at configure time or in a script earlier to
that is a matter of taste. Doable, but I am afraid it will not be done
in this release. We have the documentation that I mentioned in the last
mail and that must suffice for now. Maybe some more people "fight" for a
better solution, but I am happy that jFriCAS is not officially supported
so that with a "mild" user effort it can be installed.

Ralf

[Ralf Hemmecke]

On 02.07.23 20:08, Waldek Hebisch wrote:

> I have now put preliminary source release tarball at:
>
> http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9-pre.tar.bz2

I've created FriCAS from this release tarball and added jFriCAS and
jupytext. Things work smoothly when following the steps from the
installation guide.

https://hemmecke.github.io/fricas/install.html

The FriCAS-Aldor interface compiles fine from the git-repo, so I expect
no problems with your above release tarball.

I wanted to look at including the patches from Peter Broadbery, but ...
no time.

Ralf

[Grégory Vanuxem]

Yes, OK. It was a suggestion after all.

__
Greg

> --
> You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/b6816f9f-fab5-d859-4ed5-4337950122fc%40hemmecke.org.

[Waldek Hebisch]

Now at

http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9.full.tar.bz2
http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9.amd64.tar.bz2

you can find second preview of coming relase, this time source and
64-bit Intel/AMD binary tarball. ATM there is on 32-bit tarball,
to include hunchentoot I would have to create new 32-bit build
environment and it would take more time than 64-bit version.

ATM at Sourceforge in release area you can find hsbcl-1.3.9.tar
file which contains files that I used to include hunchentoot
in the binary. If there are no _serious_ problems in a day or
two the release will be at Github and Sourceforge.

--
Waldek Hebisch

[Waldek Hebisch]

On Wed, Jul 05, 2023 at 04:55:00PM +0200, Waldek Hebisch wrote:
> On Sun, Jul 02, 2023 at 08:08:21PM +0200, Waldek Hebisch wrote:
> > I have now put preliminary source release tarball at:
> >
> > http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9-pre.tar.bz2
> >
> > If there are no significant problems it will be promoted to
> > release.
> >
> > In about 2-3 days I plan to create binary release.
>
> Now at
>
> http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9.full.tar.bz2
> http://fricas-wiki.math.uni.wroc.pl/public/fricas-1.3.9.amd64.tar.bz2
>
> you can find second preview of coming relase, this time source and
> 64-bit Intel/AMD binary tarball. ATM there is on 32-bit tarball,

^^

I mean no.

> to include hunchentoot I would have to create new 32-bit build
> environment and it would take more time than 64-bit version.
>
> ATM at Sourceforge in release area you can find hsbcl-1.3.9.tar
> file which contains files that I used to include hunchentoot
> in the binary. If there are no _serious_ problems in a day or
> two the release will be at Github and Sourceforge.
>
> --
> Waldek Hebisch
>

> --
> You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/ZKWERKt3fC/NP0N8%40fricas.math.uni.wroc.pl.

--
Waldek Hebisch

[Nasser M. Abbasi]

I am getting strange error using Fricas 1.3.9 pre-release July 4, 2023 version (latest)

>fricas

                       FriCAS Computer Algebra System
                            Version: FriCAS 1.3.9

                 Timestamp: Wed Jul  5 01:45:22 PM CDT 2023

(1) -> integrate((b*x+a)^m*(d*x+c)^n,x)                    
 
   >> System error:
   The variable BREAK is unbound.

(1) -> integrate((b*x+a)^m*(d*x+c),x)  

   (1)
          2       2   2              2         2                 2
       ((b d m + b d)x  + ((a b d + b c)m + 2 b c)x + a b c m - a d + 2 a b c)
    *
         m log(b x + a)
       %e
  /
      2 2      2       2
     b m  + 3 b m + 2 b
                                         Type: Union(Expression(Integer),...)
(2) -> integrate((b*x+a)^m*(d*x+c)^n,x)
 
   >> System error:
   The variable BREAK is unbound.

(2) -> integrate((b*x+a)^(-4+n)/((d*x+c)^n),x)                    
 
   >> System error:
   The variable BREAK is unbound.

(2) -> integrate((b*x+a)^(-5-n)*(d*x+c)^n,x)                    
 
   >> System error:
   The variable BREAK is unbound.

These do not generate these error in 1.3.8.

>fricas --version
FriCAS 1.3.9
based on sbcl 2.3.0

--Nasser

[Nasser M. Abbasi]

Fyi, I see 1.3.9 have been officially released. So I download it and build it. This error is still there.

I see dozens such exceptions generated in a test I did.

./configure --with-lisp=sbcl
make
sudo make install

>which fricas
/usr/local/bin/fricas

>fricas --version
FriCAS 1.3.9
based on sbcl 2.3.0
>

>fricas
FRICAS="/usr/local/lib/fricas/target/x86_64-linux-gnu"
spad-lib="/usr/local/lib/fricas/target/x86_64-linux-gnu/lib/libspad.so"

FriCAS Computer Algebra System
Version: FriCAS 1.3.9

Timestamp: Fri Jul 7 03:57:05 AM CDT 2023

(1) -> integrate((b*x+a)^m*(d*x+c)^n,x)

>> System error:
The variable BREAK is unbound.

--- ON 1.3.8, no error --------------------

>which fricas
/usr/local/bin/fricas
>fricas --version
FriCAS 1.3.8
based on sbcl 2.2.5
>fricas
FRICAS="/usr/local/lib/fricas/target/x86_64-linux-gnu"
spad-lib="/usr/local/lib/fricas/target/x86_64-linux-gnu/lib/libspad.so"
FriCAS Computer Algebra System
Version: FriCAS 1.3.8
Timestamp: Wed Jul 6 05:49:26 PM CDT 2022
(5) -> ii:=integrate((b*x+a)^(-5-n)*(d*x+c)^n,x);

Type: Union(Expression(Integer),...)
(6) -> unparse(ii::InputForm)

(6)
"((6*b^4*d^4*x^5+((6*a*b^3*d^4+(-6)*b^4*c*d^3)*n+30*a*b^3*d^4)*x^4+((3*a^2*b^
2*d^4+(-6)*a*b^3*c*d^3+3*b^4*c^2*d^2)*n^2+(27*a^2*b^2*d^4+(-30)*a*b^3*c*d^3+3
*b^4*c^2*d^2)*n+60*a^2*b^2*d^4)*x^3+((a^3*b*d^4+(-3)*a^2*b^2*c*d^3+3*a*b^3*c^
2*d^2+(-1)*b^4*c^3*d)*n^3+(12*a^3*b*d^4+(-27)*a^2*b^2*c*d^3+18*a*b^3*c^2*d^2+
(-3)*b^4*c^3*d)*n^2+(47*a^3*b*d^4+(-60)*a^2*b^2*c*d^3+15*a*b^3*c^2*d^2+(-2)*b
^4*c^3*d)*n+60*a^3*b*d^4)*x^2+((a^4*d^4+(-2)*a^3*b*c*d^3+2*a*b^3*c^3*d+(-1)*b
^4*c^4)*n^3+(9*a^4*d^4+(-12)*a^3*b*c*d^3+(-9)*a^2*b^2*c^2*d^2+18*a*b^3*c^3*d+
(-6)*b^4*c^4)*n^2+(26*a^4*d^4+(-10)*a^3*b*c*d^3+(-45)*a^2*b^2*c^2*d^2+40*a*b^
3*c^3*d+(-11)*b^4*c^4)*n+(24*a^4*d^4+24*a^3*b*c*d^3+(-36)*a^2*b^2*c^2*d^2+24*
a*b^3*c^3*d+(-6)*b^4*c^4))*x+((a^4*c*d^3+(-3)*a^3*b*c^2*d^2+3*a^2*b^2*c^3*d+(
-1)*a*b^3*c^4)*n^3+(9*a^4*c*d^3+(-24)*a^3*b*c^2*d^2+21*a^2*b^2*c^3*d+(-6)*a*b
^3*c^4)*n^2+(26*a^4*c*d^3+(-57)*a^3*b*c^2*d^2+42*a^2*b^2*c^3*d+(-11)*a*b^3*c^
4)*n+(24*a^4*c*d^3+(-36)*a^3*b*c^2*d^2+24*a^2*b^2*c^3*d+(-6)*a*b^3*c^4)))*exp
(((-1)*n+(-5))*log(b*x+a))*exp(n*log(d*x+c)))/((a^4*d^4+(-4)*a^3*b*c*d^3+6*a^
2*b^2*c^2*d^2+(-4)*a*b^3*c^3*d+b^4*c^4)*n^4+(10*a^4*d^4+(-40)*a^3*b*c*d^3+60*
a^2*b^2*c^2*d^2+(-40)*a*b^3*c^3*d+10*b^4*c^4)*n^3+(35*a^4*d^4+(-140)*a^3*b*c*
d^3+210*a^2*b^2*c^2*d^2+(-140)*a*b^3*c^3*d+35*b^4*c^4)*n^2+(50*a^4*d^4+(-200)
*a^3*b*c*d^3+300*a^2*b^2*c^2*d^2+(-200)*a*b^3*c^3*d+50*b^4*c^4)*n+(24*a^4*d^4
+(-96)*a^3*b*c*d^3+144*a^2*b^2*c^2*d^2+(-96)*a*b^3*c^3*d+24*b^4*c^4))"
Type: String
(7) ->

[Ralf Hemmecke]

On 07.07.23 11:08, 'Nasser M. Abbasi' via FriCAS - computer algebra

system wrote:
> Fyi, I see 1.3.9 have been officially released.

I saw no message here at fricas-devel, so I wouldn't say that it is
released officially.

Also https://github.com/fricas/fricas/releases/ currently still shows 1.3.8.

However, THANK YOU for your feedback.
I think your issues must be fixed before the release.

I can confirm this issue on my installations.

Ralf

[Nasser M. Abbasi]

Ralf;

I just assumed it is officially released because I saw it here

https://sourceforge.net/projects/fricas/

It says

" Released /fricas/1.3.9/Notes-1.3.9.txt"

With 12 hrs ago timestamp. 

Thanks,

--Nasser

[Waldek Hebisch]

On Fri, Jul 07, 2023 at 02:24:24AM -0700, 'Nasser M. Abbasi' via FriCAS - computer algebra system wrote:
> Ralf;
>
> I just assumed it is officially released because I saw it here
>
> https://sourceforge.net/projects/fricas/
>
> It says
>
> " Released /fricas/1.3.9/Notes-1.3.9.txt

> <https://sourceforge.net/projects/fricas/files/fricas/1.3.9/Notes-1.3.9.txt/download>

> "
>
> With 12 hrs ago timestamp.

It is not offcially released, release it still in making, in stage of uploading
files. I am affraid we will need to restart.

And thanks for information.

--
Waldek Hebisch

[Nasser M. Abbasi]

" It is not offcially released, release it still in making, in stage of uploading
files."

I see. I did not know that as I normally go to sourceforge to download latest release.

If you need additional examples,  Here are two links showing all the 

cases that generated exception for Fricas from two test files I just finished.

Few of these exceptions as valid and known internal exception other than 

this new one "The variable BREAK is unbound." error

This page  at the bottom gives list of all integrals with F(-2) error (i.e. exception)

also this second page  gives additional list of integrals with F(-2).

Clicking on any link from the list shown above, opens a new page that shows 

the actual integral for Fricas down the page with the exception. Any one that says

Exception raised: NotImplementedError >> the translation of the FriCAS object sage2 to sage is not yet impleme

nted

will be such one. (the above is sagemath translation of the exception)

Maybe this will help in your investigation for the cause.

--Nasser

[Waldek Hebisch]

On Fri, Jul 07, 2023 at 08:33:19AM -0700, 'Nasser M. Abbasi' via FriCAS - computer algebra system wrote:
> " It is not offcially released, release it still in making, in stage of
> uploading
> files."
>
> I see. I did not know that as I normally go to sourceforge to download
> latest release.

Yes, there could be some confusion. But the main page at Sourceforge
listed 1.3.8 and I did not start updating Github pages (which would
tie release to specific revision of source code).

> If you need additional examples, Here are two links showing all the
> cases that generated exception for Fricas from two test files I just
> finished.

Thanks. I have now commited a workaround. This is actually very
old problem and had a workaround. I thought that problem is gone
and deactivated the workaround. But we still need it.


--
Waldek Hebisch