Error building mpc-1.1.0 [Sage 8.8, Linux 3.10 x86

Marco Castronovo

unread,

Sep 16, 2019, 11:12:51 AM9/16/19

to sage-devel

Dear all,

I tried to build Sage 8.8 from source on a Linux 3.10 x86_64 machine, but make fails while building mpc-1.1.0. The error I get is:

[mpc-1.1.0] Making all in doc

[mpc-1.1.0] make[6]: Entering directory `/cache/home/mc1903/sage-8.8/local/var/tmp/sage/build/mpc-1.1.0/src/doc'

[mpc-1.1.0] restore=: && backupdir=".am$$" && \

[mpc-1.1.0] am__cwd=`pwd` && CDPATH="${ZSH_VERSION+.}:" && cd . && \

[mpc-1.1.0] rm -rf $backupdir && mkdir $backupdir && \

[mpc-1.1.0] if (/usr/bin/bash /cache/home/mc1903/sage-8.8/local/var/tmp/sage/build/mpc-1.1.0/src/missing makeinfo --version) >/dev/null 2>&1; then \

[mpc-1.1.0] for f in mpc.info mpc.info-[0-9] mpc.info-[0-9][0-9] mpc.i[0-9] mpc.i[0-9][0-9]; do \

[mpc-1.1.0] if test -f $f; then mv $f $backupdir; restore=mv; else :; fi; \

[mpc-1.1.0] done; \

[mpc-1.1.0] else :; fi && \

[mpc-1.1.0] cd "$am__cwd"; \

[mpc-1.1.0] if /usr/bin/bash /cache/home/mc1903/sage-8.8/local/var/tmp/sage/build/mpc-1.1.0/src/missing makeinfo -I . \

[mpc-1.1.0] -o mpc.info mpc.texi; \

[mpc-1.1.0] then \

[mpc-1.1.0] rc=0; \

[mpc-1.1.0] CDPATH="${ZSH_VERSION+.}:" && cd .; \

[mpc-1.1.0] else \

[mpc-1.1.0] rc=$?; \

[mpc-1.1.0] CDPATH="${ZSH_VERSION+.}:" && cd . && \

[mpc-1.1.0] $restore $backupdir/* `echo "./mpc.info" | sed 's|[^/]*$||'`; \

[mpc-1.1.0] fi; \

[mpc-1.1.0] rm -rf $backupdir; exit $rc

[mpc-1.1.0] Can't locate Texinfo/ModulePath.pm in @INC (@INC contains: /home/mc1903/share/texinfo /cache/home/mc1903/sage-8.8/local/lib/perl5 /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5 .) at /home/mc1903/bin/makeinfo line 83.

[mpc-1.1.0] BEGIN failed--compilation aborted at /home/mc1903/bin/makeinfo line 86.

[mpc-1.1.0] make[6]: *** [mpc.info] Error 2

[mpc-1.1.0] make[6]: Leaving directory `/cache/home/mc1903/sage-8.8/local/var/tmp/sage/build/mpc-1.1.0/src/doc'

[mpc-1.1.0] make[5]: *** [all-recursive] Error 1

[mpc-1.1.0] make[5]: Leaving directory `/cache/home/mc1903/sage-8.8/local/var/tmp/sage/build/mpc-1.1.0/src'

[mpc-1.1.0] make[4]: *** [all] Error 2

[mpc-1.1.0] make[4]: Leaving directory `/cache/home/mc1903/sage-8.8/local/var/tmp/sage/build/mpc-1.1.0/src'

[mpc-1.1.0] ********************************************************************************

[mpc-1.1.0] Error building mpc-1.1.0

[mpc-1.1.0] ********************************************************************************

Please find attached a more detailed log.

Best regards,

Marco Castronovo

mpc-1.1.0.log

Dima Pasechnik

unread,

Sep 16, 2019, 11:35:00 AM9/16/19

to sage-devel

Hi,

It looks as if your /home/mc1903/bin/makeinfo is broken.

here is what happens: the configure of mpc finds makeinfo there, and
the mpc build tries
to use it, but to no avail.

Either fix it, or move it out of the way (for the building of Sage, at least)

HTH
Dima

PS. Normally, makeinfo is installed systemwide, and you might ask your
sysadmins to install it...

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/27c1661f-ca01-4f1d-acf8-de8ae122fd46%40googlegroups.com.

Marco Castronovo

unread,

Sep 16, 2019, 11:56:54 AM9/16/19

to sage-devel

Thank you, this fixed the problem. However, now make can't build gf2x-1.2.p0 , see attached log. It's complaining that GNU Autoconf is not installed. I'm pretty sure it is installed system-wide (I'm on a node of a computing cluster), so I guess the question is: what directory does make go to look for these utilities? Maybe I can modify this directory with some parameter passed to configure before running make?

Best,

Marco

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-...@googlegroups.com.

gf2x-1.2.p0.log

Dima Pasechnik

unread,

Sep 16, 2019, 12:03:22 PM9/16/19

to sage-devel

Are you building on a networked file system?
That's the only reason for the gf2x install to try running autotools,
that some timestamps are off a bit...
Sage normally can be built from source without autotools.

If it's possible please do building on a local filesystem
(usually referred to "scratch partition" or something like this).

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/eab0b97d-3dcd-4360-ae2a-901fd528883b%40googlegroups.com.

Marco Castronovo

unread,

Sep 16, 2019, 3:01:45 PM9/16/19

to sage-devel

Compiling in the scratch partition made the process go smoothly for a while, until it got stuck again on sqlite-3270100 ... I'm attaching the log for that. I don't understand if this is an issue of different nature.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/eab0b97d-3dcd-4360-ae2a-901fd528883b%40googlegroups.com.

sqlite-3270100.log

Dima Pasechnik

unread,

Sep 16, 2019, 3:19:06 PM9/16/19

to sage-devel

Are you by any chance running the build from zsh shell?
What is the output of

echo $SHELL

if you run it at the command prompt?
If this is the case, could you run the build from bash instead?

(this error message with sqlite looks quite similar
to the previous one, so it might actually have
nothing to do with network file systems...)

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/f066b4a6-cc19-41e1-a3c7-db5d01127cad%40googlegroups.com.

Marco Castronovo

unread,

Sep 16, 2019, 3:34:28 PM9/16/19

to sage-devel

I'm using a bash shell.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/f066b4a6-cc19-41e1-a3c7-db5d01127cad%40googlegroups.com.

Marco Castronovo

unread,

Sep 16, 2019, 3:50:07 PM9/16/19

to sage-devel

It seems that every time the compilation stops with an error, the error is fixed by running ./configure and then ./make again. I don't know how to interpret this, but it seems to work.

Marco

Dima Pasechnik

unread,

Sep 16, 2019, 3:52:48 PM9/16/19

to sage-devel

Do you have ZSH_VERSION set?
The lines in your log

make[4]: Entering directory
`/scratch/mc1903/sage-8.8/local/var/tmp/sage/build/sqlite-3270100/src'
CDPATH="${ZSH_VERSION+.}:" && cd . && /usr/bin/bash
/scratch/mc1903/sage-8.8/local/var/tmp/sage/build/sqlite-3270100/src/missing
aclocal-1.15

look a bit strange to me...

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/25d4b9a4-087d-41ec-bbc6-7b63f8d9f61a%40googlegroups.com.

Marco Castronovo

unread,

Sep 16, 2019, 3:57:43 PM9/16/19

to sage-devel

No, echo $ZSH_VERSION returns an empty line.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/25d4b9a4-087d-41ec-bbc6-7b63f8d9f61a%40googlegroups.com.

Dima Pasechnik

unread,

Sep 16, 2019, 4:07:03 PM9/16/19

to sage-devel

On Mon, Sep 16, 2019 at 8:50 PM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> It seems that every time the compilation stops with an error, the error is fixed by running ./configure and then ./make again. I don't know how to interpret this, but it seems to work.

And the error is always of this sort, an attempt to run autotools?

maybe a skew system clock, or some other joy coming
from a big cluster...

I guess this can be fought by using AM_MAINTAINER_MODE([disable])
or directly by passing

--disable-maintainer-mode

to the respective ./configure scripts.

E.g. with sqlite one can change the line

sdh_configure

in build/pkgs/sqlite/spkg-install to

sdh_configure --disable-maintainer-mode

to get rid of this spurious autotool rebuilds.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/fdc5c2c7-e918-4f34-8ca7-fa3b55f42ff1%40googlegroups.com.

Marco Castronovo

unread,

Sep 17, 2019, 10:39:18 AM9/17/19

to sage-devel

The option --disable-maintainer-code doesn't seem to help. Here is the log for the latest package (pynac) that fails to build. The error seems to be always related to an attempt to run autotools.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/fdc5c2c7-e918-4f34-8ca7-fa3b55f42ff1%40googlegroups.com.

pynac-0.7.24.p0.log

Dima Pasechnik

unread,

Sep 17, 2019, 12:22:33 PM9/17/19

to sage-devel

On Tue, Sep 17, 2019 at 3:39 PM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> The option --disable-maintainer-code doesn't seem to help. Here is the log for the latest package (pynac) that fails to build. The error seems to be always related to an attempt to run autotools.

pynac's configure does not support this option.

probably you also didn't pass it properly, as there should be a warning
in the log about unknown option, but it's not there; just in case, it
should go into the line

sdh_configure --disable-static --with-giac=no PYTHON=sage-python23

of build/pkgs/pynac/spkg-install

Would you like to try a version of pynac with this option supported?
I could easily make one for you.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/f2b3e832-bf9c-49a8-b7c7-630f12957e55%40googlegroups.com.

Marco Castronovo

unread,

Sep 17, 2019, 2:11:03 PM9/17/19

to sage-devel

I included the old log file, here is the new with the warning you were mentioning:

configure: WARNING: unrecognized options: --disable-maintainer-mode

Is there a way to consistently pass this option for all packages? Any help is very welcome.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/f2b3e832-bf9c-49a8-b7c7-630f12957e55%40googlegroups.com.

pynac-0.7.24.p0.log

Dima Pasechnik

unread,

Sep 17, 2019, 2:29:27 PM9/17/19

to sage-devel

On Tue, Sep 17, 2019 at 7:11 PM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> I included the old log file, here is the new with the warning you were mentioning:
>
> configure: WARNING: unrecognized options: --disable-maintainer-mode
>
> Is there a way to consistently pass this option for all packages? Any help is very welcome.

We're talking about this issue here:
https://trac.sagemath.org/ticket/28511
I'll post a patch there.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/37fc33b7-63b1-43bd-9af8-c6a3e4f52570%40googlegroups.com.

Dima Pasechnik

unread,

Sep 17, 2019, 2:46:55 PM9/17/19

to sage-devel

So you get try the branch on https://trac.sagemath.org/ticket/28511

Or simply replace build/bin/sage-dist-helpers with
https://git.sagemath.org/sage.git/plain/build/bin/sage-dist-helpers?id=b17854b509b5721830344766ba0bb61b9523c9f6

(I haven't tested it yet, but it's a very easy change...)

Marco Castronovo

unread,

Sep 17, 2019, 3:43:55 PM9/17/19

to sage-devel

There is no file named "sage-dist-helpers" in my directory sage-8.8/build/bin/ , do you mean that I should just copy the new file in there?

Marco

> > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/37fc33b7-63b1-43bd-9af8-c6a3e4f52570%40googlegroups.com.

Dima Pasechnik

unread,

Sep 17, 2019, 4:10:00 PM9/17/19

to sage-devel

On Tue, Sep 17, 2019 at 8:43 PM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> There is no file named "sage-dist-helpers" in my directory sage-8.8/build/bin/ , do you mean that I should just copy the new file in there?

Oh, right, it was moved after 8.8, in 8.8 it is
src/bin/sage-dist-helpers

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/e10c878b-20de-43c1-8f92-29b5b02fabcd%40googlegroups.com.

Marco Castronovo

unread,

Sep 17, 2019, 4:19:20 PM9/17/19

to sage-devel

I copied the file in sage-8.8/src/bin/sage-dist-helpers and ran chmod +x sage-dist-helpers, then ran make again. I seem to be getting the same error. The log is attached.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/e10c878b-20de-43c1-8f92-29b5b02fabcd%40googlegroups.com.

pynac-0.7.24.p0.log

Dima Pasechnik

unread,

Sep 17, 2019, 4:24:18 PM9/17/19

to sage-devel

yes, because, as I explained, this has no effect on pynac - you need
to change its source to make this option work, as I propose here:
https://github.com/pynac/pynac/pull/347

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/d62147bf-0adb-4bfb-9f4b-95ad234ed998%40googlegroups.com.

Marco Castronovo

unread,

Sep 17, 2019, 4:34:47 PM9/17/19

to sage-devel

The problem still persists with the file build/pkgs/pynac/spkg-install modified to contain the line:

sdh_configure --disable-maintainer-mode --disable-static --with-giac=no PYTHON=sage-python23

Is this what you meant? Sorry maybe I misunderstood.

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/d62147bf-0adb-4bfb-9f4b-95ad234ed998%40googlegroups.com.

pynac-0.7.24.p0.log

François Bissey

unread,

Sep 17, 2019, 4:49:10 PM9/17/19

to sage-...@googlegroups.com

Quick question Marco. What file system are you running on? It seems to me you have all
these build issues that could be related to timestamps being messed up on your file system.

François

Dima Pasechnik

unread,

Sep 17, 2019, 4:51:48 PM9/17/19

to sage-devel

On Tue, Sep 17, 2019 at 9:49 PM François Bissey <frp.b...@gmail.com> wrote:
>
> Quick question Marco. What file system are you running on? It seems to me you have all
> these build issues that could be related to timestamps being messed up on your file system.
>

OK, Francois knows much more about "interesting" HPC systems quirks than me. :-)

Apparently Marco is using a scratch partition - but it looks like it's
still network one, just like his home directiry

> François
>
> > On 18/09/2019, at 8:34 AM, Marco Castronovo <marco.ca...@gmail.com> wrote:
> >
> > The problem still persists with the file build/pkgs/pynac/spkg-install modified to contain the line:
> >
> > sdh_configure --disable-maintainer-mode --disable-static --with-giac=no PYTHON=sage-python23
> >
> >
> >
> > Is this what you meant? Sorry maybe I misunderstood.
> >
> >
> >
>

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/A3BA0831-3847-4B0D-A756-123726795D69%40gmail.com.

François Bissey

unread,

Sep 17, 2019, 4:53:47 PM9/17/19

to sage-...@googlegroups.com

> On 18/09/2019, at 8:51 AM, Dima Pasechnik <dim...@gmail.com> wrote:
>
> On Tue, Sep 17, 2019 at 9:49 PM François Bissey <frp.b...@gmail.com> wrote:
>>
>> Quick question Marco. What file system are you running on? It seems to me you have all
>> these build issues that could be related to timestamps being messed up on your file system.
>>
> OK, Francois knows much more about "interesting" HPC systems quirks than me. :-)
>
> Apparently Marco is using a scratch partition - but it looks like it's
> still network one, just like his home directiry

A network one is “vague”. nfs (3,4), GPFS, lustre….

Dima Pasechnik

unread,

Sep 17, 2019, 5:28:03 PM9/17/19

to sage-devel

On Tue, Sep 17, 2019 at 9:34 PM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> The problem still persists with the file build/pkgs/pynac/spkg-install modified to contain the line:
>
> sdh_configure --disable-maintainer-mode --disable-static --with-giac=no PYTHON=sage-python23
>
>
> Is this what you meant? Sorry maybe I misunderstood.

You did. One needs to update pynac package to include the change
https://github.com/pynac/pynac/pull/347

this means to prepare a new pynac tarball to be placed in upstream/
and few other things.
I asked whether you could try such a new tarball, which I am willing to prepare.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/b1963033-b5e9-40a6-9387-410ff118846a%40googlegroups.com.

Marco Castronovo

unread,

Sep 17, 2019, 8:07:19 PM9/17/19

to sage-devel

I'm on gpfs.

Marco

On Tuesday, September 17, 2019 at 4:49:10 PM UTC-4, François Bissey wrote:

Quick question Marco. What file system are you running on? It seems to me you have all
these build issues that could be related to timestamps being messed up on your file system.

François

Marco Castronovo

unread,

Sep 17, 2019, 8:08:29 PM9/17/19

to sage-devel

Sure I'm happy to try the new tarball. Thanks!

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/b1963033-b5e9-40a6-9387-410ff118846a%40googlegroups.com.

François Bissey

unread,

Sep 17, 2019, 8:15:01 PM9/17/19

to sage-...@googlegroups.com

Unless your GPFS cluster is seriously misconfigured it should be fine. However
the only way I can think of getting your systematic symptoms is if timestamps
of the file system either lack precision or are slightly in the future
(timezone problems between the GPFS cluster and your build node???).
That would be a problem for an administrator to look at.

François

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/a518a65d-b642-42fa-bf42-aa12308842f0%40googlegroups.com.

Dima Pasechnik

unread,

Sep 18, 2019, 5:26:55 AM9/18/19

to sage-devel

On Wed, Sep 18, 2019 at 1:08 AM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> Sure I'm happy to try the new tarball. Thanks!

Please follow the instructions on
https://github.com/dimpase/pynac/releases/tag/am_mainta

(and let us know whether this at least helped to fix the pynac build -
assuming you keep my previous hack from
https://trac.sagemath.org/ticket/28511, too)

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/66232ced-e3ce-4235-9e96-5e1cdc220d14%40googlegroups.com.

Marco Castronovo

unread,

Sep 18, 2019, 9:13:58 AM9/18/19

to sage-devel

I contacted the administrators of the cluster asking about possible issues with timestamps. The answer is:

Hello,

I am happy to report there are no timestamps issues on the Amarel cluster at this time.

Our nodes utilize a ntp client that maintains synchronization with a central time source. I can readily check the time on all nodes and verify all are as expected. When/if a time synchronization problem is experienced, other casualties will be noted such as inability to submit a batch job. Following is a sampling of compute node times.

[root@master ~]# date
Wed Sep 18 00:15:23 EDT 2019
[root@master ~]# pdsh -w hal[0001-0020] date | sort
hal0001: Wed Sep 18 00:15:29 EDT 2019
hal0002: Wed Sep 18 00:15:28 EDT 2019
hal0003: Wed Sep 18 00:15:28 EDT 2019
hal0004: Wed Sep 18 00:15:28 EDT 2019
hal0005: Wed Sep 18 00:15:28 EDT 2019
hal0006: Wed Sep 18 00:15:29 EDT 2019
hal0007: Wed Sep 18 00:15:29 EDT 2019
hal0008: Wed Sep 18 00:15:29 EDT 2019
hal0009: Wed Sep 18 00:15:28 EDT 2019
hal0010: Wed Sep 18 00:15:28 EDT 2019
hal0011: Wed Sep 18 00:15:28 EDT 2019
hal0012: Wed Sep 18 00:15:28 EDT 2019
hal0013: Wed Sep 18 00:15:29 EDT 2019
hal0014: Wed Sep 18 00:15:29 EDT 2019
hal0015: Wed Sep 18 00:15:29 EDT 2019
hal0016: Wed Sep 18 00:15:28 EDT 2019
hal0017: Wed Sep 18 00:15:28 EDT 2019
hal0018: Wed Sep 18 00:15:28 EDT 2019
hal0019: Wed Sep 18 00:15:28 EDT 2019
hal0020: Wed Sep 18 00:15:28 EDT 2019

On Tuesday, September 17, 2019 at 8:15:01 PM UTC-4, François Bissey wrote:

Unless your GPFS cluster is seriously misconfigured it should be fine. However
the only way I can think of getting your systematic symptoms is if timestamps
of the file system either lack precision or are slightly in the future
(timezone problems between the GPFS cluster and your build node???).
That would be a problem for an administrator to look at.

François

> On 18/09/2019, at 12:07 PM, Marco Castronovo <marco.c...@gmail.com> wrote:
>
> I'm on gpfs.
>
> Marco
>
> On Tuesday, September 17, 2019 at 4:49:10 PM UTC-4, François Bissey wrote:
> Quick question Marco. What file system are you running on? It seems to me you have all
> these build issues that could be related to timestamps being messed up on your file system.
>
> François
>
> > On 18/09/2019, at 8:34 AM, Marco Castronovo <marco.c...@gmail.com> wrote:
> >
> > The problem still persists with the file build/pkgs/pynac/spkg-install modified to contain the line:
> >
> > sdh_configure --disable-maintainer-mode --disable-static --with-giac=no PYTHON=sage-python23
> >
> >
> >
> > Is this what you meant? Sorry maybe I misunderstood.
> >
> >
> >
>
>
> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-...@googlegroups.com.

Dima Pasechnik

unread,

Sep 18, 2019, 9:17:57 AM9/18/19

to sage-devel

Is it really "expected" that clocks are off by a second?!

hal0001: Wed Sep 18 00:15:29 EDT 2019
hal0002: Wed Sep 18 00:15:28 EDT 2019

Perhaps Francois may tell...

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/54740e71-b6ad-421e-bef1-5b52b1be4fd2%40googlegroups.com.

Marco Castronovo

unread,

Sep 18, 2019, 9:25:54 AM9/18/19

to sage-devel

Dear Dima,

thank you for the help. I followed the instructions and I get:

make[3]: Entering directory `/scratch/mc1903/sage-8.8/build/make'

sage-logger -p 'sage-spkg pynac-0.7.24.p0' '/scratch/mc1903/sage-8.8/logs/pkgs/pynac-0.7.24.p0.log'

[pynac-0.7.24.p0] Attempting to download package pynac-0.7.24.p0

[pynac-0.7.24.p0] >>> Checking online list of optional packages.

[pynac-0.7.24.p0] >>> Checking online list of experimental packages.

[pynac-0.7.24.p0] >>> Checking online list of huge packages.

[pynac-0.7.24.p0] Error: could not find a package matching pynac-0.7.24.p0

[pynac-0.7.24.p0] Try 'sage --package list' to see the available packages

[pynac-0.7.24.p0] There is no package similar to pynac-0.7.24.p0

[pynac-0.7.24.p0] You can find further packages at http://files.sagemath.org/spkg/

make[3]: *** [/scratch/mc1903/sage-8.8/local/var/lib/sage/installed/pynac-0.7.24.p0] Error 1

make[3]: Leaving directory `/scratch/mc1903/sage-8.8/build/make'

make[2]: *** [all-start] Error 2

make[2]: Leaving directory `/scratch/mc1903/sage-8.8/build/make'

It seems that make is still trying to build pynac-0.7.24 , while the version that your test made me download is pynac-0.7.26 ?

Marco

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/66232ced-e3ce-4235-9e96-5e1cdc220d14%40googlegroups.com.

Dima Pasechnik

unread,

Sep 18, 2019, 9:31:17 AM9/18/19

to sage-devel

On Wed, Sep 18, 2019 at 2:25 PM Marco Castronovo
<marco.ca...@gmail.com> wrote:
>
> Dear Dima,
>
> thank you for the help. I followed the instructions and I get:
>
> make[3]: Entering directory `/scratch/mc1903/sage-8.8/build/make'
>
> sage-logger -p 'sage-spkg pynac-0.7.24.p0' '/scratch/mc1903/sage-8.8/logs/pkgs/pynac-0.7.24.p0.log'
>
> [pynac-0.7.24.p0] Attempting to download package pynac-0.7.24.p0
>
> [pynac-0.7.24.p0] >>> Checking online list of optional packages.
>
> [pynac-0.7.24.p0] >>> Checking online list of experimental packages.
>
> [pynac-0.7.24.p0] >>> Checking online list of huge packages.
>
> [pynac-0.7.24.p0] Error: could not find a package matching pynac-0.7.24.p0
>
> [pynac-0.7.24.p0] Try 'sage --package list' to see the available packages
>
> [pynac-0.7.24.p0] There is no package similar to pynac-0.7.24.p0
>
> [pynac-0.7.24.p0] You can find further packages at http://files.sagemath.org/spkg/
>
> make[3]: *** [/scratch/mc1903/sage-8.8/local/var/lib/sage/installed/pynac-0.7.24.p0] Error 1
>
> make[3]: Leaving directory `/scratch/mc1903/sage-8.8/build/make'
>
> make[2]: *** [all-start] Error 2
>
> make[2]: Leaving directory `/scratch/mc1903/sage-8.8/build/make'
>
>
> It seems that make is still trying to build pynac-0.7.24 , while the version that your test made me download is pynac-0.7.26 ?

Indeed. You need to sort out what happened to the files from
https://github.com/dimpase/pynac/releases/download/am_mainta/pynac-am_mainta.tar

that were supposed to overwrite the pynac version in step 3) on
https://github.com/dimpase/pynac/releases/tag/am_mainta

You probably did not run 3), or ran it in a wrong directory

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/2be00305-c7ac-49b0-b075-7347f6c2516e%40googlegroups.com.

Dima Pasechnik

unread,

Sep 18, 2019, 9:32:39 AM9/18/19

to sage-devel

what is the content of build/pkgs/pynac/package-version.txt ?
(i.e. SAGE_ROOT/build/pkgs/pynac/package-version.txt)

Marco Castronovo

unread,

Sep 18, 2019, 9:36:30 AM9/18/19

to sage-devel

It's 0.7.26.p1 .

> > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/2be00305-c7ac-49b0-b075-7347f6c2516e%40googlegroups.com.

Dima Pasechnik

unread,

Sep 18, 2019, 9:37:43 AM9/18/19

to sage-devel

hmm.

Try running

./sage -f pynac

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/11aec0b2-0d4f-426e-9f53-7428ea7694e9%40googlegroups.com.

Marco Castronovo

unread,

Sep 18, 2019, 9:43:12 AM9/18/19

to sage-devel

make[1]: Entering directory `/scratch/mc1903/sage-8.8/build/make'