Re: lowercase requirement for py-ports (Re: svn commit: r405240 - in head/devel: . py-PyLD)

[Mikhail T.]

On 04.01.2016 11:38, Kubilay Kocak wrote:
> Agreed, its just not a 'current practice' thing that makes that relevant :)
True. But insisting on changes of current practice needs a good reason.
> How does the figure change only including py-* ports?
The share drops from 25% to 9.5%:

% awk -F\| '\!/^py/ {next} $1 ~ /.*[A-Z].*/ {upper++} {total++} END
{print upper " of " total " py-ports have capital letter in name - "
upper*100./total "%"}' INDEX-10
182 of 1906 py-ports have capital letter in name - 9,54879%

Yours,

-mi

___
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"

Re: numpy would not load: libgcc_s vs. libgfortran

[Mikhail T.]

On 04.01.2016 14:34, Olivier Duchateau wrote:
> Can you post the complete backtrace?
>
Well, here goes:

___ ERROR at setup of
TestAPI.test_can_post_image 
Traceback (most recent call last):
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/tests/conftest.py",
line 33, in test_app
return tools.get_app(request)
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/tests/tools.py",
line 132, in get_app
'config:' + new_paste_config)
  File "/opt/lib/python2.7/site-packages/paste/deploy/loadwsgi.py",
line 247, in loadapp
return loadobj(APP, uri, name=name, **kw)
  File "/opt/lib/python2.7/site-packages/paste/deploy/loadwsgi.py",
line 272, in loadobj
return context.create()
  File "/opt/lib/python2.7/site-packages/paste/deploy/loadwsgi.py",
line 710, in create
return self.object_type.invoke(self)
  File "/opt/lib/python2.7/site-packages/paste/deploy/loadwsgi.py",
line 146, in invoke
return fix_call(context.object, context.global_conf,
**context.local_conf)
  File "/opt/lib/python2.7/site-packages/paste/deploy/util.py", line
55, in fix_call
val = callable(*args, **kw)
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/app.py",
line 362, in paste_app_factory
mgoblin_app = MediaGoblinApp(mediagoblin_config)
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/app.py",
line 107, in __init__
setup_plugins()
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/init/plugins/__init__.py",
line 56, in setup_plugins
__import__(plugin_module)
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/media_types/audio/__init__.py",
line 18, in 
from mediagoblin.media_types.audio.processing import
AudioProcessingManager, \
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/media_types/audio/processing.py",
line 30, in 
from mediagoblin.media_types.audio.transcoders import (
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/media_types/audio/transcoders.py",
line 23, in 
from mediagoblin.media_types.audio import audioprocessing
  File

"/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin/media_types/audio/audioprocessing.py",
line 26, in 
import numpy
  File "/opt/lib/python2.7/site-packages/numpy/__init__.py", line
180, in 
from . import add_newdocs
  File "/opt/lib/python2.7/site-packages/numpy/add_newdocs.py", line
13, in 
from numpy.lib import add_newdoc
  File "/opt/lib/python2.7/site-packages/numpy/lib/__init__.py",
line 8, in 
from .type_check import *
  File "/opt/lib/python2.7/site-packages/numpy/lib/type_check.py",
line 11, in 
import numpy.core.numeric as _nx
  File "/opt/lib/python2.7/site-packages/numpy/core/__init__.py",
line 14, in 
from . import multiarray
ImportError: /lib/libgcc_s.so.1: version GCC_4.6.0 required by
/opt/lib/gcc48/libgfortran.so.3 not found

> I already encountered this error, and if often due to a wrong python import 
> (cycling import).

The script is from MediaGoblin's own self-tests. I doubt, they have a
cycling import -- supposedly, goblin-developers use them regularly...
Thanks! Yours,

-mi

___
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"

Re: numpy would not load: libgcc_s vs. libgfortran

[Mikhail T.]

On 04.01.2016 16:31, Olivier Duchateau wrote:
> After all it's perhaps Numpy's problem, but nobody reported this issue.
Well, I decided to try figuring this out myself and got the following
error from a very simply one line script consisting of "import numpy;":

Traceback (most recent call last):
  File "tn.py", line 1, in 
import numpy;
  File "/opt/lib/python2.7/site-packages/numpy/__init__.py", line
180, in 
from . import add_newdocs
  File "/opt/lib/python2.7/site-packages/numpy/add_newdocs.py", line
13, in 
from numpy.lib import add_newdoc
  File "/opt/lib/python2.7/site-packages/numpy/lib/__init__.py",
line 8, in 
from .type_check import *
  File "/opt/lib/python2.7/site-packages/numpy/lib/type_check.py",
line 11, in 
import numpy.core.numeric as _nx
  File "/opt/lib/python2.7/site-packages/numpy/core/__init__.py",
line 14, in 
from . import multiarray
ImportError:
/opt/lib/python2.7/site-packages/numpy/core/multiarray.so: Undefined
symbol "cblas_cdotc_sub"

Any ideas? Thanks!

-mi

___
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"

SOLVED: numpy would not load: libgcc_s vs. libgfortran

[Mikhail T.]

On 04.01.2016 16:29, Mikhail T. wrote:
> ImportError:
> /opt/lib/python2.7/site-packages/numpy/core/multiarray.so: Undefined
> symbol "cblas_cdotc_sub"
Ok, the above went away, when I rebuilt all of the Fortran-using
dependencies of numpy with gfortran5. Don't know, if they were built
incorrectly somehow, or if something is wrong with gfortran48.

My one-liner test script now "works", but the original problem is still
here -- only now it complains about gcc5's libgfortran:

  File "/opt/lib/python2.7/site-packages/numpy/core/__init__.py",
line 14, in 
from . import multiarray
ImportError: /lib/libgcc_s.so.1: version GCC_4.6.0 required by
/opt/lib/gcc5/libgfortran.so.3 not found

I wonder, if this has something to do with my setting PYTHONPATH -- I
need the not-yet-installed mediagoblin packages to be found in ${WRKSRC}
by the tests. What is the proper way of to do this?

Thank you!

-mi

___
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"

Re: SOLVED: numpy would not load: libgcc_s vs. libgfortran

[Mikhail T.]

On 04.01.2016 18:59, Olivier Duchateau wrote:
>> > I wonder, if this has something to do with my setting PYTHONPATH -- I
>> > need the not-yet-installed mediagoblin packages to be found in ${WRKSRC}
>> > by the tests. What is the proper way of to do this?
> You can add module (and submodules) with sys module.
>
> For example in mediagoblin/tests/tools.py before line beginning by 'from 
> mediagoblin ...'
>
> import sys
>
> # Full path of mediagoblin
> sys.append('/spare/usr/ports/www/py-mediagoblin/work/mediagoblin-0.8.1/mediagoblin')
Ouch! I was hoping, I could set an environment variable and be done with
it -- something like PATH or LD_LIBRARY_PATH or TCLPATH...

-mi

___
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"

Re: SOLVED: numpy would not load: libgcc_s vs. libgfortran

[Mikhail T.]

On 05.01.2016 05:41, Konstantin Belousov wrote:
> Some shared object in your process is linked to libgcc_s.so.1, and does not
> have the DT_RPATH set.  You might try to verify this by procstat -v and
> seeing /lib/libgcc_s.so.1 loaded.
...

Yes, I think so. But it is not a bug -- that (unknown yet) object is not
using any of the gcc-4.6 features and so is satisfied with the stock
/lib/libgcc_s.so.1

It is not its fault, that the scripts happen to load it before they try
to load numpy:

import foo-> needs libgcc_s.so.1, found in /lib
...
import numpy  -> needs gcc5/libgcc_s.so.1 but is fed the /lib's
version, because that is already loaded

One could change the order of the imports to get numpy loaded first, but
then the foo will also be using the gcc5 version.

The bottom line is that if /any/ shared library in a process needs
gcc5/libgcc_s.so.1, then /all/ of them have to use it. I believe, this
is  something lovingly referred to as "DLL hell"... :-)
> An easy way out of this issue is to set
>   LD_PRELOAD=/usr/local/lib/gcc/libgcc_s.so.1
> variable in your environment, but this could have some other consequences
> and leaves the cause undiagnosed.
I simply added the mapping:

libgcc_s.so.1   gcc5/libgcc_s.so.1

into my /etc/libmap.conf. Now my entire system uses gcc-version and I
can go back to porting mediagoblin :) The "cause" remains unresolved --
I do not know, which other library carelessly loads from /lib, but I
don't really care either because it is not wrong for it to do so, is it?

The real cause -- newer versions of GNU compilers requiring their own
libgcc_s.so.1 -- is something for their maintainer (CC-ed) to consider.
Maybe, the src/ version of libgcc_s needs to be updated to the latest
available -- clang ought to be able to build it. Or, if the license is
not acceptable, let's make it a port of its own -- and have the port
drop a mapping file into ${PREFIX}/etc/libmap.d .

The third option would be to rename one of the two libraries -- either
the src-provided one to libbsdgcc_s.so.1 or the port-installed -- into
libgccXX.so.1. Then both will be able to coexist in a process. But with
so much between them now /duplicated/, I like this option the least.

Yours,

-mi

___
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"

libreoffice-5.3.5.2 would not build for me with Python-2.7

[Mikhail T.]

I have python=3.6 in my DEFAULT_VERSIONS, but both 2.7 and 3.6 are 
installed.


While trying to build libreoffice, the build attempted to use 2.7 and I 
kept getting the compiler errors like this:


   pyuno_util.cxx:43:18: error: no matching function for call to
   'PyUnicodeUCS2_FromUnicode'

until I changed the USES line in editors/libreoffice/Makefile from 
simply "python" to "python:3".


I don't think, that should be necessary -- simply having "python" there 
should've picked up my desired default version, right? Not sure, if this 
is the problem of libreoffice or python... Yours,


   -mi

___
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"

Having trouble with USES=python

[Mikhail T.]

Hello!

I've created the below port "by the book". However, something is not 
right about the install -- some of the files duly installed into 
STAGEDIR make their way neither into package nor into PREFIX.


The check-plist complains about /some/ of the omissions:

   ===> Checking for items in STAGEDIR missing from pkg-plist
   Error: Orphaned:
   %%PYTHON_SITELIBDIR%%/scudcloud/resources/disable_snippets.js
   Error: Orphaned: %%PYTHON_SITELIBDIR%%/scudcloud/resources/leftpane.js
   Error: Orphaned: %%PYTHON_SITELIBDIR%%/scudcloud/resources/scudcloud.js
   ===> Checking for items in pkg-plist which are not in STAGEDIR

but not /all/ of them -- none of the icons, which are found under 
STAGEDIR, make their way into ${PREFIX}/share/icons/


What am I doing wrong? Thank you! Yours,

   -mi

# This is a shell archive.  Save it in a file, remove anything before
# this line, and then unpack it by entering "sh file".  Note, it may
# create directories; files and directories will be owned by you and
# have default permissions.
#
# This archive contains:
#
#   distinfo
#   Makefile
#   pkg-descr
#
echo x - distinfo
sed 's/^X//' >distinfo << '57677d168a5ec21bdf22c9501f075a8f'
XTIMESTAMP = 1505245632
XSHA256 
(raelgc-scudcloud-2017.09.12-3d689139a0fa65813caf54ea5457727c6b48b432_GH0.tar.gz)
 = 4d859527d27293a9461514382376ad4d5543842ba3f4c361d0c1314d2eab0aa4
XSIZE 
(raelgc-scudcloud-2017.09.12-3d689139a0fa65813caf54ea5457727c6b48b432_GH0.tar.gz)
 = 185690
57677d168a5ec21bdf22c9501f075a8f
echo x - Makefile
sed 's/^X//' >Makefile << 'b67911656ef5d18c4ae36cb6741b7965'
X# Created by: Mikhail T. 
X# $FreeBSD$
X
XPORTNAME=  scudcloud
XPORTVERSION=   2017.09.12
XCATEGORIES=net-im python
XPKGNAMEPREFIX= ${PYTHON_PKGNAMEPREFIX}
X
XMAINTAINER=m...@aldan.algebra.com
XCOMMENT=   Open source client for Slack messaging system
X
XLICENSE=   MIT
X
XLIB_DEPENDS=   libQupZilla.so:www/qupzilla-qt5
X
XUSES=  python pyqt:5
X
XUSE_GITHUB=yes
XGH_ACCOUNT=raelgc
XGH_TAGNAME=3d689139a0fa65813caf54ea5457727c6b48b432
XNO_ARCH=   yes
XUSE_PYTHON=autoplist distutils hunspell
XUSE_PYQT=  core_run dbus dbussupport gui_run network_run webkit_run 
webkitwidgets
X
X.include 
b67911656ef5d18c4ae36cb6741b7965
echo x - pkg-descr
sed 's/^X//' >pkg-descr << '7f9555c3c1089940396b503dd37f3979'
XScudCloud is a non official open-source desktop client app for Slack.
X
XScudCloud improves the Slack integration with Unix desktops featuring:
X
X   * multiple teams support
X   * native system notifications
X   * count of unread direct mentions at launcher/sytray icon
X   * alert/wobbling on new messages
X   * channels quicklist (Unity only)
X   * optional tray notifications and "Close to Tray"
X   * follow your desktop activity and will stay online while you're logged
X in (if correct packages are installed)
X
XWWW: https://github.com/raelgc/scudcloud
7f9555c3c1089940396b503dd37f3979
exit
___
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"

Re: Having trouble with USES=python

[Mikhail T.]

So, it being in STAGEDIR is not enough for it to be installed?..
-- 
Sent from mobile device, please, pardon shorthand.

12 вер. 2017 р. о 21:22 Shane Ambler  написав(ла):

> As you are using autoplist the file list is generated based on data in
> setup.py, like "package_data". You should patch setup.py so that it
> picks up the missing files and also adjust where it is installing data
> files such as icons, that may mean prefixing ${PREFIX} with ${STAGEDIR}.
> 
> You may find those missing icons in /${PREFIX}/share/icons/ rather than
> within the ports work dir.

___
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"

Re: [lini...@freebsd.org: ports/170946: [patch] mark certain ports broken on ARM]

[Mikhail T.]

On 23.08.2012 19:27, Mark Linimon wrote:

ge...@freebsd.org  devel/nspr
m...@aldan.algebra.com   devel/qmake
off...@freebsd.org devel/icu
po...@freebsd.org  lang/tcl86
I'd give a shot at repairing the above four, if I had access to an ARM system... 
None are present on the current list 
, however...


Or, if you shared the errorlogs, a simple fix might be obvious even without 
hands-on access to a build host.

+.if ${ARCH} == "arm"
+BROKEN=Does not compile on arm
+.endif

Would not the below be better?

   NOT_FOR_ARCHS=arm
   NOT_FOR_ARCHS_REASON_ARM=Does not compile on arm

The above method may spare you having to switch some of the affected ports from 
the simple bsd.port.mk to bsd.port.pre.mk.


Yours,

   -mi

___
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: [lini...@freebsd.org: ports/170946: [patch] mark certain ports broken on ARM]

[Mikhail T.]

On 23.08.2012 21:47, Eitan Adler wrote:

NOT and ONLY are meant for ports which by definition will never work
another arch (think x86info).  In this case the ports are just broken.
That's a fine distinction, that makes little sense to me -- the total 
number of ports, which can not ever work on some architecture is, 
probably, a couple dozen... It is also not mentioned in the Handbook 
 
at all:


 *

   If a port should be marked IGNORE only on certain architectures,
   there are two other convenience variables that will automatically
   set IGNORE for you: ONLY_FOR_ARCHS and NOT_FOR_ARCHS. Examples:

   ONLY_FOR_ARCHS= i386 amd64
   NOT_FOR_ARCHS= ia64 sparc64

   A custom IGNORE message can be set using ONLY_FOR_ARCHS_REASON and
   NOT_FOR_ARCHS_REASON. Per architecture entries are possible with
   ONLY_FOR_ARCHS_REASON_/ARCH/ and NOT_FOR_ARCHS_REASON_/ARCH/.

Thus, I still think, using these knobs is better -- especially, when 
doing otherwise requires splitting bsd.ports.mk into 
bsd.ports.{pre,post}.mk. Yours,


   -mi

___
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"

Question about a new Python port

[Mikhail T.]

Hello!

Attached is a shar for a new port -- of the bencode-module 
. Everything is fairly straightforward 
about it, except:


1. Just like devel/py-setuptools is creates/updates the
   site-packages/easy-install.pth. However, using the same method of cleaning
   up the file as the setuptools will not work, when more than one port updates
   the file. The .dist extension for the backed-up version should, probably, be
   replaced by a port-specific one -- so that each port creates its own backup.
   Or am I misreading it?
2. The new port also drops an egg -- without even the egg-info file -- maybe,
   all such ports should be forced (by bsd.python.mk) to properly unzip and
   install the contents? Instead of unzipping at run-time every time? It works
   either way, I'm sure, but it seems to me, the .egg concept was created for
   the benefit of people, who do not use anything like the Ports system.

Thanks for the feedback. Yours,

   -mi

# This is a shell archive.  Save it in a file, remove anything before
# this line, and then unpack it by entering "sh file".  Note, it may
# create directories; files and directories will be owned by you and
# have default permissions.
#
# This archive contains:
#
#   py-bencode
#   py-bencode/Makefile
#   py-bencode/distinfo
#   py-bencode/pkg-descr
#
echo c - py-bencode
mkdir -p py-bencode > /dev/null 2>&1
echo x - py-bencode/Makefile
sed 's/^X//' >py-bencode/Makefile << '8e9e08c27483a0042d47ca86ecca9e56'
X# Created by: Mikhail Teterin 
X# $FreeBSD$
X
XPORTNAME=  bencode
XPORTVERSION=   1.0
XCATEGORIES=devel python
XMASTER_SITES=  ${MASTER_SITE_CHEESESHOP}
XMASTER_SITE_SUBDIR=source/${PORTNAME:C/(.).*/\1/}/${PORTNAME}
XPKGNAMEPREFIX= ${PYTHON_PKGNAMEPREFIX}
X
XMAINTAINER=m...@aldan.algebra.com
XCOMMENT=   Standalone implementation of BitTorrent's encoding/decoding
X
XUSE_PYTHON=2.5+
XUSE_PYDISTUTILS=   yes
XPYDISTUTILS_NOEGGINFO= yes
XPLIST_FILES=   
%%PYTHON_SITELIBDIR%%/${PORTNAME}-${PORTVERSION}-py${PYTHON_VER}.egg
X
X.include 
8e9e08c27483a0042d47ca86ecca9e56
echo x - py-bencode/distinfo
sed 's/^X//' >py-bencode/distinfo << 'bf094c3eb541cf2d243da8ee22ffe064'
XSHA256 (bencode-1.0.tar.gz) = 
0301c2202233a3f274c940702bac7bd02d5d053b8bf9502b085156270e30be9b
XSIZE (bencode-1.0.tar.gz) = 2385
bf094c3eb541cf2d243da8ee22ffe064
echo x - py-bencode/pkg-descr
sed 's/^X//' >py-bencode/pkg-descr << '10e1a7f0605ca4b1e8dc0bdcb59c21bb'
XThe BitTorrent bencode module as light-weight, standalone package.
X
XThis package simply re-packages the existing bencoding and bdecoding
Ximplemention from the 'official' BitTorrent client as a separate,
Xleight-weight package for re-using them without having the entire
XBitTorrent software as a dependency.
X
XWWW: http://pypi.python.org/pypi/bencode/
10e1a7f0605ca4b1e8dc0bdcb59c21bb
exit
___
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"

lowercase requirement for py-ports (Re: svn commit: r405240 - in head/devel: . py-PyLD)

[Mikhail T.]

On 04.01.2016 10:23, Kubilay Kocak wrote:
> https://wiki.freebsd.org/Python/PortsPolicy [...] Feedback very welcome.
>
Well, you asked for it... I propose removing the lowercase-requirement.
Instead, the portname should match the capitalization found in PyPi.
This will better match the current practice and may even avoid some
future naming conflict.

Of course, it is likely, you've already discussed this -- if so, please,
send me a link to the discussion's archive. Thank you!

-mi

___
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"

Re: lowercase requirement for py-ports (Re: svn commit: r405240 - in head/devel: . py-PyLD)

[Mikhail T.]

On 04.01.2016 11:04, Kubilay Kocak wrote:
> No discussion archive
Well, if it is not a result of a discussion arriving to a consensus,
then, perhaps, the rule is not as hard, is it? What are you reasons
behind it?
> also PyPI is case insensitive

Ah, good to know. Perhaps, that's because Python has to work on OSes
with case-insensitive filesystems. But FreeBSD is not among those...

> FWIW, some of the current practice also includes arbitrarily changing
> the names/versions, missing prefixes/suffixed (causing known conflicts),
> among other non-standard things.
I agree, that /these/ things are best avoided. Mixed-case portnames,
however, are not so obviously wrong.
> My guess is also that multi-case ports are in the minority.
It is a /substantial/ minority. Analyzing the INDEX-10 here, I get:

% awk -F\| '$1 ~ /.*[A-Z].*/ {upper++} {total++} END {print upper "
of " total " ports have capital letter in name"}' INDEX-10
6312 of 25575 ports have capital letter in name

Yours,

-mi

___
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"

numpy would not load: libgcc_s vs. libgfortran

[Mikhail T.]

Hi!

I built py-numpy only a few weeks ago with default config options (I
think). I'm trying to port some numpy-using software now and its
self-tests fail with the following error:

...
import numpy
  File "/opt/lib/python2.7/site-packages/numpy/__init__.py", line
180, in 
from . import add_newdocs
  File "/opt/lib/python2.7/site-packages/numpy/add_newdocs.py", line
13, in 
from numpy.lib import add_newdoc
  File "/opt/lib/python2.7/site-packages/numpy/lib/__init__.py",
line 8, in 
from .type_check import *
  File "/opt/lib/python2.7/site-packages/numpy/lib/type_check.py",
line 11, in 
import numpy.core.numeric as _nx
  File "/opt/lib/python2.7/site-packages/numpy/core/__init__.py",
line 14, in 
from . import multiarray
*ImportError: /lib/libgcc_s.so.1: version GCC_4.6.0 required by
/opt/lib/gcc48/libgfortran.so.3 not found*

I checked all of the shared objects installed by numpy with ldd. All
report properly using the /opt/lib/gcc48/libgcc_s.so.1 (installed by the
gcc48 port) rather than the stock /lib/libgcc_s.so.1. For example:

/opt/lib/python2.7/site-packages/numpy/core/multiarray.so:
libblas.so.2 => /opt/lib/libblas.so.2 (0x80178)
libm.so.5 => /lib/libm.so.5 (0x8019d2000)
libpython2.7.so.1 => /opt/lib/libpython2.7.so.1 (0x801bfb000)
libc.so.7 => /lib/libc.so.7 (0x80082)
libgfortran.so.3 => /opt/lib/gcc48/libgfortran.so.3
(0x801fbc000)
*libgcc_s.so.1 => /opt/lib/gcc48/libgcc_s.so.1 (0x8022d2000)*
libquadmath.so.0 => /opt/lib/gcc48/libquadmath.so.0
(0x8024e8000)
libthr.so.3 => /lib/libthr.so.3 (0x802723000)
libintl.so.8 => /opt/lib/libintl.so.8 (0x802948000)
libutil.so.9 => /lib/libutil.so.9 (0x802b53000)

And yet, the error occurs...

Has anyone else seen this? Thanks! Yours,

-mi

___
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"

narrowed down troubles with devel/gobject-introspection

[Mikhail T.]

Ok, some Python script somewhere is trying to do chflags on a file under
~/.cache. On this system, my home directory is NFS-mounted and the
chflags fails. This should not be a fatal error, but it is... CC-ing
Python-maintainers... Yours,

-mi

Mikhail T. написав(ла):
> Building devel/gobject-introspection -- a requirement for various
> things, including gstreamer and thus KDE4 -- fails here:
>
> /bin/sh /opt/bin/libtool --tag=CC   --mode=link cc 
> -I/opt/include/glib-2.0 -I/opt/lib/glib-2.0/include   -D_REENTRANT
> -I/opt/include/glib-2.0 -I/opt/lib/glib-2.0/include   -Wsign-compare
> -Wcast-align -Wpointer-arith -Wnested-externs -Wmissing-prototypes
> -Wmissing-declarations -Wchar-subscripts -Wall -O2 -pipe
> -march=pentium4 -fno-strict-aliasing -avoid-version  -L/opt/lib -o
> libannotation.la -rpath /opt/unused annotation.lo  -L/opt/lib
> -lgio-2.0 -lgobject-2.0 -lgmodule-2.0 -lglib-2.0   -pthread
> -L/opt/lib -lgthread-2.0 -lglib-2.0 
> libtool: link: cc -shared  .libs/annotation.o   -Wl,-rpath
> -Wl,/opt/lib -Wl,-rpath -Wl,/opt/lib -L/opt/lib
> /opt/lib/libgio-2.0.so /opt/lib/libgobject-2.0.so
> /opt/lib/libgmodule-2.0.so /opt/lib/libgthread-2.0.so -pthread
> /opt/lib/libglib-2.0.so -licui18n /opt/lib/libintl.so
> /opt/lib/libiconv.so /opt/lib/libpcre.so  -march=pentium4 -pthread
> -pthread -pthread   -pthread -Wl,-soname -Wl,libannotation.so -o
> .libs/libannotation.so
> libtool: link: ar cru .libs/libannotation.a  annotation.o
> libtool: link: ranlib .libs/libannotation.a
> libtool: link: ( cd ".libs" && rm -f "libannotation.la" && ln -s
> "../libannotation.la" "libannotation.la" )
> env LPATH=.libs  env PYTHONPATH=../..:../..:YTHONPATH
> UNINSTALLED_INTROSPECTION_SRCDIR=../..
> UNINSTALLED_INTROSPECTION_BUILDDIR=../.. ../../tools/g-ir-scanner -v
> --add-include-path=../../gir --add-include-path=. -v
> --add-include-path=../../gir --add-include-path=.
> --namespace=annotation --nsversion=1.0 --libtool="/bin/sh
> /opt/bin/libtool"  --library=libannotation.la --pkg=gobject-2.0
> --include=GObject-2.0 --include=utility-1.0   ./annotation.c
> ./annotation.h --output annotation-1.0.gir
> Traceback (most recent call last):
>   File "../../tools/g-ir-scanner", line 38, in 
> sys.exit(scanner_main(sys.argv))
>   File
> 
> "/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7/giscanner/scannermain.py",
> line 277, in scanner_main
> transformer.register_include(include_obj)
>   File
> 
> "/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7/giscanner/transformer.py",
> line 112, in register_include
> self._parse_include(filename)
>   File
> 
> "/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7/giscanner/transformer.py",
> line 137, in _parse_include
> self._cachestore.store(filename, parser)
>   File
> 
> "/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7/giscanner/cachestore.py",
> line 112, in store
> shutil.move(tmp_filename, store_filename)
>   File "/opt/lib/python2.6/shutil.py", line 264, in move
> copy2(src, real_dst)
>   File "/opt/lib/python2.6/shutil.py", line 100, in copy2
> copystat(src, dst)
>   File "/opt/lib/python2.6/shutil.py", line 77, in copystat
> os.chflags(dst, st.st_flags)
> OSError: [Errno 45] Operation not supported:
> '/home/mi/.cache/g-ir-scanner/fc59b7fa095cd6403cb3917129208269cb202f6b'
> gmake[3]: *** [annotation-1.0.gir] Error 1
> gmake[3]: Leaving directory
> 
> `/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7/tests/scanner'
> gmake[2]: *** [all-recursive] Error 1
> gmake[2]: Leaving directory
> 
> `/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7/tests'
> gmake[1]: *** [all-recursive] Error 1
> gmake[1]: Leaving directory
> 
> `/home/usr/ports/devel/gobject-introspection/work/gobject-introspection-0.6.7'
> gmake: *** [all] Error 2
> *** Error code 1
>
> Stop in /usr/ports/devel/gobject-introspection.
> *** Error code 1
>
> Stop in /usr/ports/multimedia/gstreamer.
> *** Error code 1
>
> Stop in /usr/ports/multimedia/gstreamer-plugins.
> *** Error code 1
>
> Stop in /usr/ports/multimedia/qt4-phonon-gst.
> *** Error code 1
>
> Stop in /usr/ports/www/qt4-webkit.
> *** Error code 1
>
> Stop in /usr/ports/devel/qt4-assistant.
> *** Error code 1
>
> Stop in /usr/ports/x11/kdelibs4.
> *** Error code 1
>
> Stop in /usr/ports/x11/kdebase4.
___
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: narrowed down troubles with devel/gobject-introspection

[Mikhail T.]

On 07.03.2010 20:21, Joe Marcus Clarke wrote:

The recommended workaround for gobject-introspection is to symlink
~/.cache/g-ir-scanner to /dev/null, or some locally mounted directory.
   
Thanks, I built the port as `root', whose home is local. But, in 
general, a build, that uses anything outside the ${WRKDIR} is buggy... 
Perhaps, all that's needed is some env-variable? Yours,


   -mi

___
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"