<div>====================</div><div>PyPy 1.9 - Yard Wolf</div><div>====================</div><div><br></div><div>We're pleased to announce the 1.9 release of PyPy. This release brings mostly</div><div>bugfixes, performance improvements, other small improvements and overall</div>
<div>progress on the `numpypy`_ effort.</div><div>It also brings an improved situation on Windows and OS X.</div><div><br></div><div>You can download the PyPy 1.9 release here:</div><div><br></div><div> <a href="http://pypy.org/download.html">http://pypy.org/download.html</a> </div>
<div><br></div><div>.. _`numpypy`: <a href="http://pypy.org/numpydonate.html">http://pypy.org/numpydonate.html</a></div><div><br></div><div><br></div><div>What is PyPy?</div><div>=============</div><div><br></div><div>PyPy is a very compliant Python interpreter, almost a drop-in replacement for</div>
<div>CPython 2.7. It's fast (`pypy 1.9 and cpython 2.7.2`_ performance comparison)</div><div>due to its integrated tracing JIT compiler.</div><div><br></div><div>This release supports x86 machines running Linux 32/64, Mac OS X 64 or</div>
<div>Windows 32. Windows 64 work is still stalling, we would welcome a volunteer</div><div>to handle that.</div><div><br></div><div>.. _`pypy 1.9 and cpython 2.7.2`: <a href="http://speed.pypy.org">http://speed.pypy.org</a></div>
<div><br></div><div><br></div><div>Thanks to our donors</div><div>====================</div><div><br></div><div>But first of all, we would like to say thank you to all people who</div><div>donated some money to one of our four calls:</div>
<div><br></div><div> * `NumPy in PyPy`_ (got so far 44502ドル out of 60000,ドル 74%)</div><div><br></div><div> * `Py3k (Python 3)`_ (got so far 43563ドル out of 105000,ドル 41%)</div><div><br></div><div> * `Software Transactional Memory`_ (got so far 21791ドル of 50400,ドル 43%)</div>
<div><br></div><div> * as well as our general PyPy pot.</div><div><br></div><div>Thank you all for proving that it is indeed possible for a small team of</div><div>programmers to get funded like that, at least for some</div>
<div>time. We want to include this thank you in the present release</div><div>announcement even though most of the work is not finished yet. More</div><div>precisely, neither Py3k nor STM are ready to make it in an official release</div>
<div>yet: people interested in them need to grab and (attempt to) translate</div><div>PyPy from the corresponding branches (respectively ``py3k`` and</div><div>``stm-thread``).</div><div><br></div><div>.. _`NumPy in PyPy`: <a href="http://pypy.org/numpydonate.html">http://pypy.org/numpydonate.html</a></div>
<div><br></div><div>Highlights</div><div>==========</div><div><br></div><div>* This release still implements Python 2.7.2.</div><div><br></div><div>* Many bugs were corrected for Windows 32 bit. This includes new</div><div>
functionality to test the validity of file descriptors; and</div><div> correct handling of the calling convensions for ctypes. (Still not</div><div> much progress on Win64.) A lot of work on this has been done by Matti Picus</div>
<div> and Amaury Forgeot d'Arc.</div><div><br></div><div>* Improvements in ``cpyext``, our emulator for CPython C extension modules.</div><div> For example PyOpenSSL should now work. We thank various people for help.</div>
<div><br></div><div>* Sets now have strategies just like dictionaries. This means for example</div><div> that a set containing only ints will be more compact (and faster).</div><div><br></div><div>* A lot of progress on various aspects of ``numpypy``. See the `numpy-status`_</div>
<div> page for the automatic report.</div><div><br></div><div>* It is now possible to create and manipulate C-like structures using the</div><div> PyPy-only ``_ffi`` module. The advantage over using e.g. ``ctypes`` is that</div>
<div> ``_ffi`` is very JIT-friendly, and getting/setting of fields is translated</div><div> to few assembler instructions by the JIT. However, this is mostly intended</div><div> as a low-level backend to be used by more user-friendly FFI packages, and</div>
<div> the API might change in the future. Use it at your own risk.</div><div><br></div><div>* The non-x86 backends for the JIT are progressing but are still not</div><div> merged (ARMv7 and PPC64).</div><div><br></div>
<div>
* JIT hooks for inspecting the created assembler code have been improved.</div><div> See `JIT hooks documentation`_ for details.</div><div><br></div><div>* ``select.kqueue`` has been added (BSD).</div><div><br></div><div>
* Handling of keyword arguments has been drastically improved in the best-case</div><div> scenario: proxy functions which simply forwards ``*args`` and ``**kwargs``</div><div> to another function now performs much better with the JIT.</div>
<div><br></div><div>* List comprehension has been improved.</div><div><br></div><div>.. _`numpy-status`: <a href="http://buildbot.pypy.org/numpy-status/latest.html">http://buildbot.pypy.org/numpy-status/latest.html</a></div>
<div>.. _`JIT hooks documentation`: <a href="http://doc.pypy.org/en/latest/jit-hooks.html">http://doc.pypy.org/en/latest/jit-hooks.html</a></div><div><br></div><div>JitViewer</div><div>=========</div><div><br></div><div>
There will be a corresponding 1.9 release of JitViewer which is guaranteed</div>
<div>to work with PyPy 1.9. See the `JitViewer docs`_ for details.</div><div><br></div><div>.. _`JitViewer docs`: <a href="http://bitbucket.org/pypy/jitviewer">http://bitbucket.org/pypy/jitviewer</a></div><div><br></div>
<div>
Cheers,</div><div>The PyPy Team</div>