Re: help finding a file

[Thierry Thomas]

Le jeu. 20 mai 21 à  3:00:19 +0200, paul beard 
 écrivait :

> ldd /usr/local/bin/smbtorture | grep found
> 
> libpytalloc-util.cpython-38.so.2 => not found (0)
> 
> libpytalloc-util.cpython-38.so.2 => not found (0)
> 
> libpytalloc-util.cpython-38.so.2 => not found (0)
> 
> 
> find . -name pkg-plist -exec grep libpytalloc-util.cpython-38.so {} \;

$ pkg which /usr/local/lib/libpytalloc-util.cpython-38.so.2
/usr/local/lib/libpytalloc-util.cpython-38.so.2 was installed by package 
talloc-2.3.1

Unfortunately, many Python ports have a dynamic plist, and you cannot
grep them.
-- 
Th. Thomas.


signature.asc
Description: PGP signature

Re: Conflicts between py37-pygments and ipython5

[Thierry Thomas]

Le mar.  6 oct. 20 à  2:36:27 +0200, Danilo G. Baio 
 écrivait :

> Hi.

Hello,

> Could you test math/sage with this patch?
> https://people.freebsd.org/~dbaio/ipython5-pygments.patch
> 
> Looking the commit history of devel/ipython5, the requirement
> 'pygments<2.6' is only necessary to solve an issue with Python 2.7, and
> devel/ipython5 is now python3.5+.
> 
> https://github.com/ipython/ipython/pull/12174
> https://github.com/ipython/ipython/issues/12173

Great! With you patch applied to devel/ipython5 math/sage builds again!

I´m closing my PRs 250074 and 250144, they are no more needed.

Best regards.
-- 
Th. Thomas.


signature.asc
Description: PGP signature

Re: Conflicts between py37-pygments and ipython5

[Thierry Thomas]

Le jeu. 24 sept. 20 à 20:54:57 +0200, Steve Wills 
 écrivait :

Hello,

> I saw a similar thing with devel/py-jira with the SHELL option enabled. 
> I disabled it locally, but I think perhaps devel/nbconvert should use 
> textproc/pygments-25?

Thanks, this is a good idea, and it would fix the problem ATM.

But meanwhile I´m working on Sage´s next upgrade (ATM this is
sage-9.2.beta9), and this one depends on recent versions (devel/ipython,
etc.), and making devel/py-nbconvert would break Sage 9.2, to be
released RSN.

Is there a way to make these ports cohabit without conflicts?
-- 
Th. Thomas.
___
freebsd-python@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-python
To unsubscribe, send any mail to "freebsd-python-unsubscr...@freebsd.org"

Conflicts between py37-pygments and ipython5

[Thierry Thomas]

Hello,

math/sage is now BROKEN, caused by PR 249131 / r549076:

===>   sage-math-9.1_5 depends on package: py37-nbconvert>0 - not found
===>   Installing existing package /packages/All/py37-nbconvert-5.6.1_1.txz
[121amd64-default-build-as-user-job-04] Installing py37-nbconvert-5.6.1_1...
[121amd64-default-build-as-user-job-04] `-- Installing py37-nbformat-5.0.7...
[121amd64-default-build-as-user-job-04] `-- Extracting py37-nbformat-5.0.7:
.. done
[121amd64-default-build-as-user-job-04] `-- Installing
py37-pandocfilters-1.4.2...
[121amd64-default-build-as-user-job-04] `-- Extracting
py37-pandocfilters-1.4.2: .. done
[121amd64-default-build-as-user-job-04] `-- Installing py37-pygments-2.7.1...
pkg-static: py37-pygments-2.7.1 conflicts with py37-pygments-25-2.5.2
(installs files into the same place).  Problematic file:
/usr/local/bin/pygmentize-3.7

Failed to install the following 1 package(s):
/packages/All/py37-nbconvert-5.6.1_1.txz
*** Error code 70

math/sage does not directly depend on textproc/py-pygments, but it
depends on devel/py-nbconvert which does.

And it also depend on devel/ipython5, which requires py37-pygments-25.

Is there a way to solve this conflict?
-- 
Th. Thomas.


signature.asc
Description: PGP signature

Re: egg-info / dist-info

[Thierry Thomas]

Le lun. 29 juin 20 à 23:49:32 +0200, Thierry Thomas 
 écrivait :

> > I don´t care about egg-info or dist-info, but I´d like to keep the same
> > namings on each platform; is it possible to enforce that?
> 
> Well, it works, whatever the installation method, so I´ve just adapted
> the plist according to $OSREL in r540835.
> 
> But if someone has an explanation, I´ll be happy to know it!

Well, it does not only depend on $OSREL, another (yet unknown factor)
is also involved. I tried to fix it in r542957 just by checking which
method has been used.
-- 
Th. Thomas.


signature.asc
Description: PGP signature

Re: egg-info / dist-info

[Thierry Thomas]

Le ven. 26 juin 20 à 10:02:34 +0200, Thierry Thomas 
 écrivait :

> I don´t care about egg-info or dist-info, but I´d like to keep the same
> namings on each platform; is it possible to enforce that?

Well, it works, whatever the installation method, so I´ve just adapted
the plist according to $OSREL in r540835.

But if someone has an explanation, I´ll be happy to know it!
-- 
Th. Thomas.


signature.asc
Description: PGP signature

Re: egg-info / dist-info

[Thierry Thomas]

Le ven. 26 juin 20 à 11:04:50 +0200, Kubilay Kocak 
 écrivait :

> Hi Thierry,

Hello koobs!

> Would need to spend some time isolating/understanding the behaviour 
> delta in detail to be able to recommend the best course of action, 
> modulo trying to use USE_PYTHON=distutils which automates most/all of 
> this work, as it uses setuptools --record output to produce the pkg-plist

As I wrote, this is not a Python port, and it cannot use
USE_PYTHON=distutils.

It uses a Python virtual env and runs pip inside. What could be the
differences between FreeBSD-12.1 and -CURRENT?

Note: it does not use pip from the ports tree, but a pip bundled with
math/sage.

> We can help you figure some of the details out in #freebsd-python on 
> freenode IRC

OK - see you later there!
-- 
Th. Thomas.
___
freebsd-python@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-python
To unsubscribe, send any mail to "freebsd-python-unsubscr...@freebsd.org"

egg-info / dist-info

[Thierry Thomas]

Hello,

I´m maintaining the port math/sage. This is not a Python port, but it
builds and installs several Python libraries.

Untill now, it did´nt build on -CURRENT, because some dependencies were
missing, but they have been fixed.

And now I´m discovering that all the files such as '*.egg-info*' become
something like '*.dist-info*' on -CURRENT. And of course, the plist is
wrong!

I don´t care about egg-info or dist-info, but I´d like to keep the same
namings on each platform; is it possible to enforce that?

Regards.
-- 
Th. Thomas.


signature.asc
Description: PGP signature

py-wxPython30 and python 3.5?

[Thierry Thomas]

Hello,

I'm the new maintainer of the port textproc/py-enchant, and I'm working
on a new port which depends both on py-enchant and Python-3.5. It's why
I need to create a py35-enchant. But py-enchant depends on
x11-toolkits/py-wxPython30, and I need also a port of wxPython30 for
Python 3.5...

Is anyone working on something like a x11-toolkits/py35-wxPython30 ?
-- 
Th. Thomas.


signature.asc
Description: PGP signature

Re: svn commit: r369447 - head/math/py-numpy

[Thierry Thomas]

Hello,

Le sam  4 oct 14 à 16:36:40 +0200, Dmitry Sivachenko 
 écrivait :

> If there are no more similar build logs, may be we fix these by adding 
> LDFLAGS=-shared?

I tried to analyze the problem - without success ATM.

Antoine is right: this is rather strange, but I have not noticed any
problem with this solution.

Best regards,
-- 
Th. Thomas.
___
freebsd-python@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-python
To unsubscribe, send any mail to "freebsd-python-unsubscr...@freebsd.org"

Re: svn commit: r369447 - head/math/py-numpy

[Thierry Thomas]

Le mer  1 oct 14 à 22:53:45 +0200, Antoine Brodin 
 écrivait :

> And those 2:
> 
> http://package20.nyi.freebsd.org/data/91amd64-default-baseline/369680/logs/errors/wspr-3.00.r2511_8.log
> http://package20.nyi.freebsd.org/data/91amd64-default-baseline/369680/logs/errors/wsjt-9.1.r2511_5.log
> 
> Are you sure it's no a bug in numpy?
> Manually adding -shared to all those ports doesn't look too right.

I'm wondering too... This is rather strange, and I cannot explain why
this failure appears now.

For sure numpy was broken before my patch: it was impossible to use it
without setting LD_LIBRARY_PATH to the libfortran directory. But I don't
understand how it could be related to this missing '-shared'.

In the meantime we also upgraded Gcc from 4.7 to 4.8, and this produces
also new failures, e.g. with cad/kicad: it was OK with 4.7 but does not
compile anymore with 4.8.

Any ideas?
-- 
Th. Thomas.
___
freebsd-python@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-python
To unsubscribe, send any mail to "freebsd-python-unsubscr...@freebsd.org"

Re: svn commit: r369447 - head/math/py-numpy

[Thierry Thomas]

Le mer  1 oct 14 à 19:55:09 +0200, Antoine Brodin 
 écrivait :

> Hi,

Hello,

> 2 others:
> 
> http://package20.nyi.freebsd.org/data/91amd64-default-baseline/369680/logs/errors/py27-obspy-0.9.2_2.log
> http://package20.nyi.freebsd.org/data/91amd64-default-baseline/369680/logs/errors/py27-pymc-2.3_2,1.log

Same cause, same remedy!

I have tested the two attached patches, and they fix the issue.

Best regards,
-- 
Th. Thomas.


pgpvdTETGiXL1.pgp
Description: PGP signature

Re: ports/188327: [numpy] math/py-numpy: Numpy is broken

[Thierry Thomas]

The following reply was made to PR ports/188327; it has been noted by GNATS.

From: Thierry Thomas 
To: nemysis 
Cc: freebsd-gnats-sub...@freebsd.org
Subject: Re: ports/188327: [numpy] math/py-numpy: Numpy is broken
Date: Thu, 17 Apr 2014 23:03:05 +0200

 On sun  6 apr 14 at 22:28:54 +0200, nemysis 
  wrote:
 
 > >Number: 188327
 > >Category:   ports
 > >Synopsis:   [numpy] math/py-numpy: Numpy is broken
 
 The proposed work-around with LD_LIBRARY_PATH has been used to fix the
 port french/eficas: it works!
 -- 
 Th. Thomas.
___
freebsd-python@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-python
To unsubscribe, send any mail to "freebsd-python-unsubscr...@freebsd.org"

Re: FreeBSD python patch for science/mei

[Thierry Thomas]

Le lun 25 nov 13 à 19:34:01 +0100, Marcus von Appen 
 écrivait :

> Your port science/mei uses the `python` command directly at build
> and/or installation time. In order to avoid this, we created
> a patch to let the port use a specific python version instead.
> 
> http://people.freebsd.org/~mva/python/mei.diff

Committed, thanks!
-- 
Th. Thomas.


pgpJhWOQTuTnv.pgp
Description: PGP signature

Re: Expertise needed: setup.py problem.

[Thierry Thomas]

Le Dim 18 nov 07 à 14:51:25 +0100, Thierry Thomas <[EMAIL PROTECTED]>
 écrivait :

> I builds and installs fine, and runs OK on some machines but fails
> (Python core dumps) on some others.

Problem solved! After some aparently unrelated changes, it runs OK, and
I'll commit it after the ports thaw. No idea about the cause of the
failure...
-- 
Th. Thomas.


pgpaVRWeI1E7l.pgp
Description: PGP signature

Expertise needed: setup.py problem.

[Thierry Thomas]

Hello,

I have been trying to port Miro (formerly Democracy) for some time. This
application is very interesting: e.g. it lets you watch videos from
Youtube or Dailymotion without a flash plug-in.

My port is available from
.

I builds and installs fine, and runs OK on some machines but fails
(Python core dumps) on some others.

I have tried to find differences between a working machine and a failing
one, without success!

However, I've noticed that build scripts are different, even if both of
them produce the same files:

- when the resulting installation is OK, the build begins with
Compiling 'playlisttab' template to 
/usr/ports/multimedia/miro/work/Miro-1.0/resources/../portable/compiled_templates/playlisttab.py

- when it is not, the build begins with
Compiling 'new' template to 
/usr/ports/multimedia/miro/work/Miro-1.0/resources/../portable/compiled_templates/new.py

but I don't understand what causes this difference!

N.B.:
- there is no impact from py25-setuptools;
- tried with -vv, but that did not give me any clue;
- building in a tinderbox produces a failing application, then I guess
  that some dependency or some environment variable is missing;
- if you want to give it a try: it is quick to build, but requires many
  dependences (add them as packages if needed).

Thanks!
-- 
Th. Thomas.


pgpKKu5hmFdiC.pgp
Description: PGP signature