Re: [Python-Dev] PEP 443 Accepted
On 5 June 2013 02:32, Guido van Rossum gu...@python.org wrote: Łukasz, Congratulations! I've accepted PEP 443. I've already marked it as Accepted in the repo. I've also applied some very minor edits in order to make the text flow a little better in a few places. I think this is a great PEP -- it's simple, doesn't overreach, and you've managed the bikeshedding admirably. Thank you for your great contribution to Python! Excellent news! Congratulations. Paul ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] PEP 443 Accepted
As somebody who missed the discussion about it and right now took a quick look at the PEP, i ask myself how subclasses are handled, as i don't see anything about it in the PEP, just support for ABCs. E.g if issubclass(Apple, Fruit) And i call a function which has registered an implementation for the Fruits type with an object of type Apple, is this implementation used? I assume so, but as said, i don't see it mentioned anywhere. -- Markus (from phone) Guido van Rossum gu...@python.org wrote: Łukasz, Congratulations! I've accepted PEP 443. I've already marked it as Accepted in the repo. I've also applied some very minor edits in order to make the text flow a little better in a few places. I think this is a great PEP -- it's simple, doesn't overreach, and you've managed the bikeshedding admirably. Thank you for your great contribution to Python! -- --Guido van Rossum (python.org/~guido) ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/markus%40unterwaditzer.net ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] PEP 443 Accepted
On Wed, Jun 5, 2013 at 5:29 PM, Markus Unterwaditzer mar...@unterwaditzer.net wrote: As somebody who missed the discussion about it and right now took a quick look at the PEP, i ask myself how subclasses are handled, as i don't see anything about it in the PEP, just support for ABCs. E.g if issubclass(Apple, Fruit) And i call a function which has registered an implementation for the Fruits type with an object of type Apple, is this implementation used? I assume so, but as said, i don't see it mentioned anywhere. That's covered by walking the MRO the same way method dispatch does it, so it didn't really get discussed much (the question never came up). The ABC handling is described explicitly as the generic dispatch implementation that the PEP was based on didn't handle them, and because it required a fair bit more thought than just walking the MRO. Cheers, Nick. -- Nick Coghlan | ncogh...@gmail.com | Brisbane, Australia ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] PEP 443 Accepted
On 5 cze 2013, at 09:29, Markus Unterwaditzer mar...@unterwaditzer.net wrote: As somebody who missed the discussion about it and right now took a quick look at the PEP, i ask myself how subclasses are handled, as i don't see anything about it in the PEP, just support for ABCs. E.g if issubclass(Apple, Fruit) And i call a function which has registered an implementation for the Fruits type with an object of type Apple, is this implementation used? I assume so, but as said, i don't see it mentioned anywhere. Yes, this is the supported behaviour. The PEP briefly explains that Where there is no registered implementation for a specific type, its method resolution order is used to find a more generic implementation. As Nick noted, the reason the description is so brief is that this works just like calling methods. -- Best regards, Łukasz Langa WWW: http://lukasz.langa.pl/ Twitter: @llanga IRC: ambv on #python-dev ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] New FreeBSD 10.0-CURRENT buildbot
On 2/06/2013 3:12 PM, Kubilay Kocak wrote: Afternoon (UTC+10), I'd like to request a new user/pass for a FreeBSD 10.0-CURRENT VM guest I've just setup as a dedicated buildbot slave to complement my other koobs-freebsd slaves. Also, with this and future additions to the FreeBSD buildslave set in mind, I think it might also be prudent for a rename to take place: koobs-freebsd9-amd64 koobs-freebsd9-amd64-clang (CC=clang) koobs-freebsd10-amd64 (clang is default here) Convention being: koobs-freebsdX[Y]-arch[-config] (Happy for feedback here) If there are any permutations or additions you'd like to specifically see for FreeBSD to increase coverage just let me know (Hint: I have a PandaBoard arm board here i'm getting up and running) I have ZFS DTrace to work with among other things, and the long term plan is to have FreeBSD buildbots running the full gamut of versions, from -RELEASE through -STABLE to HEAD branches. I'm on #python-dev IRC if someone needs to discuss. -- Regards, Koobs Ping :) ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] [Python-checkins] cpython: Add reference implementation for PEP 443
Dnia 5 cze 2013 o godz. 16:31 Brett Cannon br...@python.org napisał(a): Any chance you could move your definitions for generic function and single dispatch to the glossary and just link to them here? Sure thing. -- Ł ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
[Python-Dev] PEP 442 accepted
I (and Guido) are accepting PEP 442 (Safe object finalization) on the condition that finalizers are only ever called once globally. Congratulations to Antoine on writing yet another PEP that deeply touches the core language in a way that everyone can agree is an improvement.. I look forward to reviewing the code. -- Regards, Benjamin ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] PEP 442 accepted
Le 05/06/2013 18:10, Benjamin Peterson a écrit : I (and Guido) are accepting PEP 442 (Safe object finalization) on the condition that finalizers are only ever called once globally. Congratulations to Antoine on writing yet another PEP that deeply touches the core language in a way that everyone can agree is an improvement.. I look forward to reviewing the code. Thank you! Antoine (on holiday - is anyone I know in Hungary?). -- Regards, Benjamin ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
[Python-Dev] Compiling Python with Python
Hello all I'd like to start this email by saying this is not a proposal to change Python's build system. This is just the results of some experimentation you might be interested it. I have been working on a cross-platform build system called Meson, which is implemented in Python 3. For symmetry I wanted to see if it could be used to build Python itself. After about an evening worth of work, I got the basic C parts (that is, the build targets in the main Makefile such as core Python, pgen etc) built. Main highlights: - pyconfig.h generation in a fully cross-platform way without Autoconf (not tested with Visual Studio but should work as Meson has unit tests for this, tests for functions in header files and some others not done) - builds in a separate build directory, can have arbitrarily many build dirs with different configurations (for gcc/clang/static analysis/debug/release/etc) - configure time 5s, build time 8s on an i5 Macbook running Ubuntu (Autotool-configure takes 37 s but it's not directly comparable because it does a lot more) - the file describing the build is 433 lines, most of which look like this: if cc.has_header('io.h') pyconf.set('HAVE_IO_H', 1) endif - implementation of Meson is 100% Python 3, it does not have a dependency on the shell and in fact already works on Windows If you want to try it yourself, here are the steps (only 64 bit Linux tested thus far): - install python3-ply and Ninja (Ubuntu package ninja-build) - get Meson git head: https://sourceforge.net/p/meson/code/ - get Python3 trunk - download and extract the build files into trunk: https://dl.dropboxusercontent.com/u/37517477/python-meson.tar.gz - cd into Python trunk and do the following commands: mkdir build cd build path/to/meson.py .. ninja And there you have it. You can't do much with it, though (except run pgen to ensure that it actually did something ;) ). If you have any questions that are not directly related to Python, feel free to email me or the Meson mailing list. Enjoy, ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] Compiling Python with Python
On Thu, Jun 6, 2013 at 5:21 AM, Jussi Pakkanen jpakk...@gmail.com wrote: - implementation of Meson is 100% Python 3, it does not have a dependency on the shell and in fact already works on Windows Since you're talking about a bootstrap requirement here, the obvious question is: What version of Python 3 does it require? Will it be a lot of hassle to get hold of (say) Python 3.2, only to uninstall it when you have your 3.4 built? ChrisA ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] Compiling Python with Python
On Wed, Jun 5, 2013 at 10:37 PM, Chris Angelico ros...@gmail.com wrote: Since you're talking about a bootstrap requirement here, the obvious question is: What version of Python 3 does it require? Will it be a lot of hassle to get hold of (say) Python 3.2, only to uninstall it when you have your 3.4 built? The implementation does not use deep Python magic such as C extensions or the like so it should work with future releases of Python. The current version requires 3.3 or newer but only because it uses a couple of pythonlib functions that were added at 3.3. Changing it to work with 3.2 or earlier should not be a big task. ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] HAVE_FSTAT?
Hi. My 2 cents about this: (well I'm only a noob) I had this problem; I don't know about other people's environment, but my environment's problem was that it was actually not POSIX-compliant: it didn't have other file functions as well, but anyway the `fstat` error is the FIRST error you get when you compile in such environments, so people as unaware as me think the problem is with fstat only. Anyway I think if you are going to remove anything, you should think in terms of POSIX-compliancy of the target system. Removing HAVE_FSTAT might be fine (as user can easily write his own version of the function and have it included into the python's sources), but if you instead provide the user with the ability to use his custom functions when POSIX one's aren't available, it would help make python compile on even more platforms. Sorry if this last one was off-topic. Best regards. ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] Compiling Python with Python
What's the advantage in writing a new build tool? I'm asking this because I'm doing the same using scons: https://bitbucket.org/cavallo71/fatpython At the moment I'm very interested into this problem: the main advantages I see so far are (in scons) are node dependencies and the fact it is plain python syntax. Thanks On 5 Jun 2013, at 20:21, Jussi Pakkanen jpakk...@gmail.com wrote: Hello all I'd like to start this email by saying this is not a proposal to change Python's build system. This is just the results of some experimentation you might be interested it. I have been working on a cross-platform build system called Meson, which is implemented in Python 3. For symmetry I wanted to see if it could be used to build Python itself. After about an evening worth of work, I got the basic C parts (that is, the build targets in the main Makefile such as core Python, pgen etc) built. Main highlights: - pyconfig.h generation in a fully cross-platform way without Autoconf (not tested with Visual Studio but should work as Meson has unit tests for this, tests for functions in header files and some others not done) - builds in a separate build directory, can have arbitrarily many build dirs with different configurations (for gcc/clang/static analysis/debug/release/etc) - configure time 5s, build time 8s on an i5 Macbook running Ubuntu (Autotool-configure takes 37 s but it's not directly comparable because it does a lot more) - the file describing the build is 433 lines, most of which look like this: if cc.has_header('io.h') pyconf.set('HAVE_IO_H', 1) endif - implementation of Meson is 100% Python 3, it does not have a dependency on the shell and in fact already works on Windows If you want to try it yourself, here are the steps (only 64 bit Linux tested thus far): - install python3-ply and Ninja (Ubuntu package ninja-build) - get Meson git head: https://sourceforge.net/p/meson/code/ - get Python3 trunk - download and extract the build files into trunk: https://dl.dropboxusercontent.com/u/37517477/python-meson.tar.gz - cd into Python trunk and do the following commands: mkdir build cd build path/to/meson.py .. ninja And there you have it. You can't do much with it, though (except run pgen to ensure that it actually did something ;) ). If you have any questions that are not directly related to Python, feel free to email me or the Meson mailing list. Enjoy, ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/a.cavallo%40cavallinux.eu ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
Re: [Python-Dev] Compiling Python with Python
On Thu, Jun 6, 2013 at 6:00 AM, Jussi Pakkanen jpakk...@gmail.com wrote: On Wed, Jun 5, 2013 at 10:37 PM, Chris Angelico ros...@gmail.com wrote: Since you're talking about a bootstrap requirement here, the obvious question is: What version of Python 3 does it require? Will it be a lot of hassle to get hold of (say) Python 3.2, only to uninstall it when you have your 3.4 built? The implementation does not use deep Python magic such as C extensions or the like so it should work with future releases of Python. The current version requires 3.3 or newer but only because it uses a couple of pythonlib functions that were added at 3.3. Changing it to work with 3.2 or earlier should not be a big task. Newer versions shouldn't be a problem, older ones will. I'm mainly thinking about systems that can't just casually apt-get a Python 3.3. With Ubuntu, most people probably don't even need to worry about building from source, as there'll be a decently-recent Python in the repo; but what happens when you're on Debian Squeeze and the only Python 3 you can get is 3.1.3? Even Wheezy (the current stable Debian) comes with only 3.2. I do like the symmetry of using Python to build Python. But I also like using Python 3.3+ for everything, and not having to support the older Pythons. Unfortunately those two capacities clash, my lords, they clash! ChrisA ___ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com