[Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Victor Stinner
Hi,

I created the issue "Add sys.debug_build public variable to check if
Python was compiled in debug mode": http://bugs.python.org/issue25256

I would like to add an obvious way to check if Python was compiled in
debug mode, instead of having hacks/tips to check it. On the Internet,
I found various recipes to check if Python is compiled is debug mode.
Sadly, some of them are not portable. For example, 3 different checks
are proposed on StackOverflow but 2 of them are specific to Windows:
http://stackoverflow.com/questions/646518/python-how-to-detect-debug-interpreter

 Even if the exact impact of a debug build depends on the Python
implementation and the Python version, we can use it to have the same
behaviour on all Python implementations. For example, the warnings
module shows warnings by default if Python is compiled in debug mode:
Extract of my patch:

-if hasattr(sys, 'gettotalrefcount'):
+if sys.debug_build:
 resource_action = "always"
 else:
 resource_action = "ignore"

Alternative: Add a new sys.implementation.debug_build flag. Problem:
extending sys.implementation requires a new PEP, and I don't think
that debug_build fits into this object.

Berker Peksag likes the idea.

Serhiy Storchaka dislike the new flag: "I don't like this. The sys
module is one of most used module, but it has too many members, and
adding yet one makes the situation worse." (sys has 81 symbols)
"Checking for debug mode is not often needed, and mainly in tests.
Current way ``hasattr(sys, 'gettotalrefcount')`` works good. You also
can check ``'d' in sys.abiflags`` if it looks cleaner to you. Or add a
member to test.support."

The name "debug_build" comes from the existing
sysconfig.is_python_build() function. There is a sys.flags.debug flag,
so "sys.debug" can be confusing. I prefer to attach the "build"
suffix.

First I proposed a function sys.is_debug_build(), but a flag is
simpler than a function. There is not need to compute a version it's
known at build time.

What do you think? Should we add sys.debug_build?

Victor
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Nick Coghlan
On 2 October 2015 at 17:18, Victor Stinner  wrote:
> What do you think? Should we add sys.debug_build?

Spell it as "sys.implementation.debug_build" and I'm in favour.

Cheers,
Nick.

-- 
Nick Coghlan   |   [email protected]   |   Brisbane, Australia
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Victor Stinner
2015-10-02 9:37 GMT+02:00 Nick Coghlan :
> Spell it as "sys.implementation.debug_build" and I'm in favour.

Oh, in fact, I don't have no preference between sys.debug_flag and
sys.implementation.debug_flag. If I understood correctly, Serhiy would
prefer sys.implementation.debug_flag because he doesn't want to add
yet another symbol to the sys namespace.

But Berker Peksag wrote:

"According to the sys.implementation documentation and PEP 421, we can
only add a private attribute without writing a PEP. But I find
sys.implementation._debug_build too long and ``from sys import
implementation; implementation._debug_build``(or ``from sys import
implementation as i; i._debug_build``) is also not easy to write. So
I'm +1 to sys.debug_build."

Should I write a PEP for a new field in sys.implementation?

Victor
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Nir Soffer
Whats wrong with:

>>> sysconfig.get_config_var('Py_DEBUG')
0

Nir

On Fri, Oct 2, 2015 at 10:18 AM, Victor Stinner 
wrote:

> Hi,
>
> I created the issue "Add sys.debug_build public variable to check if
> Python was compiled in debug mode": http://bugs.python.org/issue25256
>
> I would like to add an obvious way to check if Python was compiled in
> debug mode, instead of having hacks/tips to check it. On the Internet,
> I found various recipes to check if Python is compiled is debug mode.
> Sadly, some of them are not portable. For example, 3 different checks
> are proposed on StackOverflow but 2 of them are specific to Windows:
>
> http://stackoverflow.com/questions/646518/python-how-to-detect-debug-interpreter
>
>  Even if the exact impact of a debug build depends on the Python
> implementation and the Python version, we can use it to have the same
> behaviour on all Python implementations. For example, the warnings
> module shows warnings by default if Python is compiled in debug mode:
> Extract of my patch:
>
> -if hasattr(sys, 'gettotalrefcount'):
> +if sys.debug_build:
>  resource_action = "always"
>  else:
>  resource_action = "ignore"
>
> Alternative: Add a new sys.implementation.debug_build flag. Problem:
> extending sys.implementation requires a new PEP, and I don't think
> that debug_build fits into this object.
>
> Berker Peksag likes the idea.
>
> Serhiy Storchaka dislike the new flag: "I don't like this. The sys
> module is one of most used module, but it has too many members, and
> adding yet one makes the situation worse." (sys has 81 symbols)
> "Checking for debug mode is not often needed, and mainly in tests.
> Current way ``hasattr(sys, 'gettotalrefcount')`` works good. You also
> can check ``'d' in sys.abiflags`` if it looks cleaner to you. Or add a
> member to test.support."
>
> The name "debug_build" comes from the existing
> sysconfig.is_python_build() function. There is a sys.flags.debug flag,
> so "sys.debug" can be confusing. I prefer to attach the "build"
> suffix.
>
> First I proposed a function sys.is_debug_build(), but a flag is
> simpler than a function. There is not need to compute a version it's
> known at build time.
>
> What do you think? Should we add sys.debug_build?
>
> Victor
> ___
> Python-Dev mailing list
> [email protected]
> https://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe:
> https://mail.python.org/mailman/options/python-dev/nirsof%40gmail.com
>
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Victor Stinner
2015-10-02 13:16 GMT+02:00 Nir Soffer :
> Whats wrong with:
>
 sysconfig.get_config_var('Py_DEBUG')
> 0

Again, refer to my first message "On the Internet, I found various
recipes to check if Python is compiled is debug mode. Sadly, some of
them are not portable."

I don't think that sysconfig.get_config_var('Py_DEBUG') will work on
other Python implementations.

On Windows, there is no such file like "Makefile" used to fill
syscofig.get_config_vars() :-( sysconfig._init_non_posix() only fills
a few variables like BINDIR or INCLUDEPY, but not Py_DEBUG.

Victor
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Maciej Fijalkowski
Speaking of other python implementations - why would you even care?
(the pypy debug build has very different properties and does very
different stuff for example). I would be very happy to have this
clearly marked as implementation-dependent and that's why it would be
cool to not be in sys (there are already 5 symbols there for this
reason, so hasattr totalrefcount is cool enough)

On Fri, Oct 2, 2015 at 2:19 PM, Victor Stinner  wrote:
> 2015-10-02 13:16 GMT+02:00 Nir Soffer :
>> Whats wrong with:
>>
> sysconfig.get_config_var('Py_DEBUG')
>> 0
>
> Again, refer to my first message "On the Internet, I found various
> recipes to check if Python is compiled is debug mode. Sadly, some of
> them are not portable."
>
> I don't think that sysconfig.get_config_var('Py_DEBUG') will work on
> other Python implementations.
>
> On Windows, there is no such file like "Makefile" used to fill
> syscofig.get_config_vars() :-( sysconfig._init_non_posix() only fills
> a few variables like BINDIR or INCLUDEPY, but not Py_DEBUG.
>
> Victor
> ___
> Python-Dev mailing list
> [email protected]
> https://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe: 
> https://mail.python.org/mailman/options/python-dev/fijall%40gmail.com
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Barry Warsaw
On Oct 02, 2015, at 11:46 AM, Victor Stinner wrote:

>Should I write a PEP for a new field in sys.implementation?

Specifically PEP 421 says that a PEP is needed if the new sys.implementation
attribute is required to be defined in all implementations, i.e. it's a new
required attribute.  Will debug_build be required of all implementations?

The PEP can be short.

https://www.python.org/dev/peps/pep-0421/#id30

If it's only relevant for CPython then an underscore-prefix symbol in
sys.implementation is the right place for it, and no PEP is needed.  Just open
an issue on the tracker.

Cheers,
-Barry
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Issue #25256: Add sys.debug_build?

2015-10-02 Thread Serhiy Storchaka

On 02.10.15 10:18, Victor Stinner wrote:

I would like to add an obvious way to check if Python was compiled in
debug mode, instead of having hacks/tips to check it. On the Internet,
I found various recipes to check if Python is compiled is debug mode.
Sadly, some of them are not portable.


I agree with Maciej. Why do you need to check if Python is compiled in 
debug mode? Because either you need to use some feature that is 
available only in debug mode (such as sys.gettotalrefcount), or to 
distinguish debug and non-debug binaries (in sysconfig and distutils), 
or make the behaviour different in debug mode (in warnings), or handle 
other behaviour differences (such as additional asserts or different 
stack consumption). In the first case you should just explicitly check 
the existence of related function. In the second case I suggest to use 
sys.abiflags as a suffix (likely you want to distinguish pymalloc from 
non-pymalloc build), or at least make a suffix depended on sys.abiflags. 
In the third case perhaps we have to set default warning level depending 
on sys.flags.debug, or sys.flags.verbose, or other dynamic flags. In the 
fourth case there is no good solution, but in any case this behaviour is 
implementation specific, and other implementation can have different 
consistency checks and different limits.




___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


[Python-Dev] Summary of Python tracker Issues

2015-10-02 Thread Python tracker

ACTIVITY SUMMARY (2015-09-25 - 2015-10-02)
Python tracker at http://bugs.python.org/

To view or respond to any of the issues listed below, click on the issue.
Do NOT respond to this message.

Issues counts and deltas:
  open5155 (+13)
  closed 31906 (+57)
  total  37061 (+70)

Open issues with patches: 2286 


Issues opened (48)
==

#25235: EmailMessage.add_attachment() creates parts with spurious MIME
http://bugs.python.org/issue25235  opened by groner

#25237: Add doc for tkinter commondialog.Dialog and subclasses
http://bugs.python.org/issue25237  opened by terry.reedy

#25238: Version added of context parameter for xmlrpc.client.ServerPro
http://bugs.python.org/issue25238  opened by desbma

#25239: HTMLParser handle_starttag replaces entity references in attri
http://bugs.python.org/issue25239  opened by frogcoder

#25240: Stack overflow in reprlib causes a core dump
http://bugs.python.org/issue25240  opened by ceridwen

#25242: Failed tests for Python 3.5.0 on shared virtual host
http://bugs.python.org/issue25242  opened by Open Genomes

#25243: decouple string-to-boolean logic from ConfigParser.getboolean 
http://bugs.python.org/issue25243  opened by jab

#25244: Idle: refine right-click behavior
http://bugs.python.org/issue25244  opened by terry.reedy

#25246: Alternative algorithm for deque_remove()
http://bugs.python.org/issue25246  opened by rhettinger

#25247: Tkinter modules built successfully but removed because they co
http://bugs.python.org/issue25247  opened by abrantesasf

#25251: Unknown MS Compiler version 1900
http://bugs.python.org/issue25251  opened by Matt.Hickford

#25252: Hard-coded line ending in asyncio.streams.StreamReader.readlin
http://bugs.python.org/issue25252  opened by eric.smith

#25254: Idle: debugger source line highlighting fails again
http://bugs.python.org/issue25254  opened by terry.reedy

#25256: Add sys.debug_build public variable to check if Python was com
http://bugs.python.org/issue25256  opened by haypo

#25257: In subject line email library inserts unwanted space after a t
http://bugs.python.org/issue25257  opened by SegundoBob

#25258: HtmlParser doesn't handle void element tags correctly
http://bugs.python.org/issue25258  opened by Chenyun Yang

#25259: readline macros can segfault Python
http://bugs.python.org/issue25259  opened by gumnos

#25263: test_tkinter fails randomly on the buildbots "AMD64 Windows10"
http://bugs.python.org/issue25263  opened by haypo

#25264: test_marshal always crashs on "AMD64 Windows10 2.7" buildbot
http://bugs.python.org/issue25264  opened by haypo

#25266: mako benchmark not working in Python 3.6
http://bugs.python.org/issue25266  opened by florin.papa

#25268: Support pointing frozen modules to the corresponding source fi
http://bugs.python.org/issue25268  opened by eric.snow

#25269: Add method to detect if a string contains surrogates
http://bugs.python.org/issue25269  opened by r.david.murray

#25270: codecs.escape_encode systemerror on empty byte string
http://bugs.python.org/issue25270  opened by reaperhulk

#25272: asyncio tests are getting noisy
http://bugs.python.org/issue25272  opened by gvanrossum

#25274: sys.setrecursionlimit() must fail if the current recursion dep
http://bugs.python.org/issue25274  opened by haypo

#25275: Documentation v/s behaviour mismatch wrt integer literals cont
http://bugs.python.org/issue25275  opened by shreevatsa

#25276: Intermittent segfaults on PPC64 AIX 3.x
http://bugs.python.org/issue25276  opened by haypo

#25277: test_eintr hangs on randomly on "AMD64 FreeBSD 9.x 3.x"
http://bugs.python.org/issue25277  opened by haypo

#25278: Unexpected socket exception on SFTP 'STOR' command
http://bugs.python.org/issue25278  opened by blanquier

#25282: regex: Support for recursive patterns
http://bugs.python.org/issue25282  opened by Sworddragon

#25283: Make tm_gmtoff and tm_zone available on all platforms
http://bugs.python.org/issue25283  opened by belopolsky

#25285: regrtest: run tests in subprocesses with -j1 on buildbots
http://bugs.python.org/issue25285  opened by haypo

#25286: views are not sequences
http://bugs.python.org/issue25286  opened by akira

#25287: test_crypt fails on OpenBSD
http://bugs.python.org/issue25287  opened by haypo

#25289: test_strptime hangs sometimes on AMD64 Windows7 SP1 3.x buildb
http://bugs.python.org/issue25289  opened by haypo

#25290: csv.reader: minor docstring typo
http://bugs.python.org/issue25290  opened by wasserverein

#25291: better Exception message for certain task termination scenario
http://bugs.python.org/issue25291  opened by ovex

#25292: ssl socket gets into broken state when client exits during han
http://bugs.python.org/issue25292  opened by ovex

#25293: Hooking Thread/Process instantiation in concurrent.futures.
http://bugs.python.org/issue25293  opened by Antony.Lee

#25294: Absolute imports fail in some cases where relative imports wou
http://bugs.python.org/issue25294  opened by Patrick Maupin

#25295: fun

[Python-Dev] Migrating to Python 3: the python 3 install issue

2015-10-02 Thread Terry Reedy

On python-list, Chris Warrick reported (thread title):
"The Nikola project is deprecating Python 2.7 (+2.x/3.x user survey 
results)"  This is for the November release, with 2.7 dropped in the 
next version next year. (Nikola is a cross-platform unicode-based app 
for building static websites and blogs from user-written templates and 
(marked-up) text files.   https://getnikola.com/ )


Since users do not write code to use Nikola, the survey was about 
installation of Python 3.  At present, 1/2 have 3.x only, 1/3 2.x only, 
and 1/6 both.  (So much for 'nobody uses 3.x for real work'.)  Most of 
the 2.x only people are able and willing to install 3.x.

https://getnikola.com/blog/env-survey-results-and-the-future-of-python-27.html

When Stefan Behnel asked why they did not drop the hard-to-maintain 2.7 
version once they ported to 3.3, Chris answered


> We did it now because it all started with frustration with 2.7 [0].
> Also, doing it back in 2012/2013 would be problematic, because back
> then not all Linux distros had an easily installable Python 3 stack
> (and RHEL 7 still doesn’t have one in the default repos)
>
> [0]: http://ralsina.me/weblog/posts/floss-decision-making-in-action.html

--
Terry Jan Reedy


___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Migrating to Python 3: the python 3 install issue

2015-10-02 Thread Brett Cannon
Thanks for the info, Terry! Glad people are realizing that Python 3 is now
available widely enough that applications can seriously consider dropping
Python 2 support now. I still think 2016 is going to see this happen more
and more once the Linux distros make their switches to Python 3.

On Fri, 2 Oct 2015 at 15:16 Terry Reedy  wrote:

> On python-list, Chris Warrick reported (thread title):
> "The Nikola project is deprecating Python 2.7 (+2.x/3.x user survey
> results)"  This is for the November release, with 2.7 dropped in the
> next version next year. (Nikola is a cross-platform unicode-based app
> for building static websites and blogs from user-written templates and
> (marked-up) text files.   https://getnikola.com/ )
>
> Since users do not write code to use Nikola, the survey was about
> installation of Python 3.  At present, 1/2 have 3.x only, 1/3 2.x only,
> and 1/6 both.  (So much for 'nobody uses 3.x for real work'.)  Most of
> the 2.x only people are able and willing to install 3.x.
>
> https://getnikola.com/blog/env-survey-results-and-the-future-of-python-27.html
>
> When Stefan Behnel asked why they did not drop the hard-to-maintain 2.7
> version once they ported to 3.3, Chris answered
>
>  > We did it now because it all started with frustration with 2.7 [0].
>  > Also, doing it back in 2012/2013 would be problematic, because back
>  > then not all Linux distros had an easily installable Python 3 stack
>  > (and RHEL 7 still doesn’t have one in the default repos)
>  >
>  > [0]:
> http://ralsina.me/weblog/posts/floss-decision-making-in-action.html
>
> --
> Terry Jan Reedy
>
>
> ___
> Python-Dev mailing list
> [email protected]
> https://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe:
> https://mail.python.org/mailman/options/python-dev/brett%40python.org
>
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Migrating to Python 3: the python 3 install issue

2015-10-02 Thread Victor Stinner
Fedora 23 (scheduled for the end of this month) will only come with
python3 (/usr/bin/python3), no python2 (nor python), *in the base
system*. Obviously, it will be possible to install Python 2 to install
applications not compatible with Python 3 yet.


Note: the current development version is Fedora 23, Fedora 24 is the next one.

Ubuntu is still working on a similar change.

Victor

2015-10-03 0:55 GMT+02:00 Brett Cannon :
> Thanks for the info, Terry! Glad people are realizing that Python 3 is now
> available widely enough that applications can seriously consider dropping
> Python 2 support now. I still think 2016 is going to see this happen more
> and more once the Linux distros make their switches to Python 3.
>
> On Fri, 2 Oct 2015 at 15:16 Terry Reedy  wrote:
>>
>> On python-list, Chris Warrick reported (thread title):
>> "The Nikola project is deprecating Python 2.7 (+2.x/3.x user survey
>> results)"  This is for the November release, with 2.7 dropped in the
>> next version next year. (Nikola is a cross-platform unicode-based app
>> for building static websites and blogs from user-written templates and
>> (marked-up) text files.   https://getnikola.com/ )
>>
>> Since users do not write code to use Nikola, the survey was about
>> installation of Python 3.  At present, 1/2 have 3.x only, 1/3 2.x only,
>> and 1/6 both.  (So much for 'nobody uses 3.x for real work'.)  Most of
>> the 2.x only people are able and willing to install 3.x.
>>
>> https://getnikola.com/blog/env-survey-results-and-the-future-of-python-27.html
>>
>> When Stefan Behnel asked why they did not drop the hard-to-maintain 2.7
>> version once they ported to 3.3, Chris answered
>>
>>  > We did it now because it all started with frustration with 2.7 [0].
>>  > Also, doing it back in 2012/2013 would be problematic, because back
>>  > then not all Linux distros had an easily installable Python 3 stack
>>  > (and RHEL 7 still doesn’t have one in the default repos)
>>  >
>>  > [0]:
>> http://ralsina.me/weblog/posts/floss-decision-making-in-action.html
>>
>> --
>> Terry Jan Reedy
>>
>>
>> ___
>> Python-Dev mailing list
>> [email protected]
>> https://mail.python.org/mailman/listinfo/python-dev
>> Unsubscribe:
>> https://mail.python.org/mailman/options/python-dev/brett%40python.org
>
>
> ___
> Python-Dev mailing list
> [email protected]
> https://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe:
> https://mail.python.org/mailman/options/python-dev/victor.stinner%40gmail.com
>
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Migrating to Python 3: the python 3 install issue

2015-10-02 Thread Victor Stinner
(grr, again i sent a draft by mistake, sorry about that)

Fedora 23 (scheduled for the end of this month) will only come with
python3 (/usr/bin/python3), no python2 (nor python), *in the base
system*. Obviously, it will be possible to install Python 2 to install
applications not compatible with Python 3 yet.

https://fedoraproject.org/wiki/Releases/23/ChangeSet#Python_3_as_the_Default_Implementation
https://twitter.com/_solotraveller/status/645559393627435008

Ubuntu is also working on a similar change. I don't know when it will happen.

Victor

2015-10-03 0:55 GMT+02:00 Brett Cannon :
> Thanks for the info, Terry! Glad people are realizing that Python 3 is now
> available widely enough that applications can seriously consider dropping
> Python 2 support now. I still think 2016 is going to see this happen more
> and more once the Linux distros make their switches to Python 3.
>
> On Fri, 2 Oct 2015 at 15:16 Terry Reedy  wrote:
>>
>> On python-list, Chris Warrick reported (thread title):
>> "The Nikola project is deprecating Python 2.7 (+2.x/3.x user survey
>> results)"  This is for the November release, with 2.7 dropped in the
>> next version next year. (Nikola is a cross-platform unicode-based app
>> for building static websites and blogs from user-written templates and
>> (marked-up) text files.   https://getnikola.com/ )
>>
>> Since users do not write code to use Nikola, the survey was about
>> installation of Python 3.  At present, 1/2 have 3.x only, 1/3 2.x only,
>> and 1/6 both.  (So much for 'nobody uses 3.x for real work'.)  Most of
>> the 2.x only people are able and willing to install 3.x.
>>
>> https://getnikola.com/blog/env-survey-results-and-the-future-of-python-27.html
>>
>> When Stefan Behnel asked why they did not drop the hard-to-maintain 2.7
>> version once they ported to 3.3, Chris answered
>>
>>  > We did it now because it all started with frustration with 2.7 [0].
>>  > Also, doing it back in 2012/2013 would be problematic, because back
>>  > then not all Linux distros had an easily installable Python 3 stack
>>  > (and RHEL 7 still doesn’t have one in the default repos)
>>  >
>>  > [0]:
>> http://ralsina.me/weblog/posts/floss-decision-making-in-action.html
>>
>> --
>> Terry Jan Reedy
>>
>>
>> ___
>> Python-Dev mailing list
>> [email protected]
>> https://mail.python.org/mailman/listinfo/python-dev
>> Unsubscribe:
>> https://mail.python.org/mailman/options/python-dev/brett%40python.org
>
>
> ___
> Python-Dev mailing list
> [email protected]
> https://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe:
> https://mail.python.org/mailman/options/python-dev/victor.stinner%40gmail.com
>
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] VS 2010 compiler

2015-10-02 Thread Matthew Einhorn
On Wed, Sep 30, 2015 at 3:57 PM, Carl Kleffner  wrote:

> Concerning the claims that mingw is difficult:
>
> The mingwpy package is a sligthly modified mingw-w64 based gcc toolchain,
> that is in development. It is designed for simple use and for much better
> compatibility to the standard MSVC python builds. It should work out of the
> box, as long as the \Scripts folder is in the PATH.
>

Indeed, I tested it with 2.7.9 x86/x64, and wheres before I had to apply at
least http://bugs.python.org/issue4709, and
http://bugs.python.org/issue16472 and then generate the .a files to get it
to work, this worked out of the box! (except of course still needing to
change distutils.cfg).

I do wonder though, which may be slightly off topic here, whether putting
the mingw binary stub files (?) in python/Scripts, rather than directly
having python/shared/mingwpy/bin be on the path, is the best approach? I
suppose it's no worse than having the other stuff in python/scripts on the
path (although with e.g. pip you could do the python -m pip trick). Are
there guidelines about this (using python/scripts) for Windows? I didn't
see anything mentioned when python -m pip install ... vs. pip install is
discussed.

Although I hope a 7z version of that mingw will be separately available for
those that want to share a copy between versions of python, since the files
in share/mingwpy is presumably the same between python versions for a
specific bitness?

Matt
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com


Re: [Python-Dev] Migrating to Python 3: the python 3 install issue

2015-10-02 Thread Barry Warsaw
On Oct 03, 2015, at 01:05 AM, Victor Stinner wrote:

>Ubuntu is also working on a similar change. I don't know when it will happen.

For the desktop, we're aiming for 16.04 LTS.

Cheers,
-Barry
___
Python-Dev mailing list
[email protected]
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: 
https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com