Recent Releases
4.3.2021062005 Jul 2021 06:05
minor feature:
Infrastructure:
Add SHA3-256 to release hash
Python-3.9.5, Python-3.10beta3
Python-3.7.10PyPy 64 bit variant (belief: PyPy3 problems of now are the same as cPython-3.11 speed problems)
Jupyterlab-3.0.16.
.
Add SHA3-256 to release hash.
Python-3.9.5, Python-3.10beta3.
Python-3.7.10PyPy 64 bit variant (belief: PyPy3 problems of now are the same as cPython-3.11 speed problems).
Jupyterlab-3.0.16.
New Packages:
Fastai
Sqlite_utils, csvs_to_sqlite, python_picard
.
Fastai.
Sqlite_utils, csvs_to_sqlite, python_picard.
Upgraded Packages:
Numpy-1.21.0+mkl, scipy-1.7.0, scikit_learn-0.24.2j
Jupyterlab-3.0.16, Spyder-5.0.4, torch-1.9.0
.
Numpy-1.21.0+mkl, scipy-1.7.0, scikit_learn-0.24.2j.
Jupyterlab-3.0.16, Spyder-5.0.4, torch-1.9.0.
Variation per Version:
Pytorch in Winpython64-3.9.5.0.exe only
PyPy branch (the 'JIT' Python):
Doesn't include a three eco-systems: Qt, numba, Torch, mypy, (70 packages, 14 of usual packages)
Has some huge speed-up, but also some slowness moments not always related to 'C' interfaces,
Will improve significantly with python-3.8 compatibility, HPy interface development, anaconda Community interest
.
Pytorch in Winpython64-3.9.5.0.exe only.
PyPy branch (the 'JIT' Python):
Doesn't include a three eco-systems: Qt, numba, Torch, mypy, (70 packages, 14 of usual packages).
Has some huge speed-up, but also some slowness moments not always related to 'C' interfaces.
Will improve significantly with python-3.8 compatibility, HPy interface development, anaconda Community interest.
4.1.2021041722 Apr 2021 03:25
minor feature:
Infrastructure:
Python-3.8.9 and Python-3.9.4
pip-21.1.dev0-2021-04-12
Jupyterlab-3.0.14, replacing back Jupyter-nbclassic per Jupyter-notebook
Python-3.9 is now feature-complete, so becomes the default download suggestion
VScode-1.55.2.0, nodejs-14.16.1
.
Python-3.8.9 and Python-3.9.4.
pip-21.1.dev0-2021-04-12.
Jupyterlab-3.0.14, replacing back Jupyter-nbclassic per Jupyter-notebook.
Python-3.9 is now feature-complete, so becomes the default download suggestion.
VScode-1.55.2.0, nodejs-14.16.1.
New Packages:
nbdime-3.0.0, jupyterlab_git,
voila_gridstack-0.2.0,
qpsolvers, quadprog, qdldl,
.
nbdime-3.0.0, jupyterlab_git.
voila_gridstack-0.2.0.
qpsolvers, quadprog, qdldl.
Upgraded Packages:
Spyder-5.0.1, Matplotlib-3.4.1, sympy-1.8
Panel-0.11.3, Voila-0.2.9, holoviews-1.14.3
numpy-1.20.2+mkl, SciPy-1.6.2, Pandas-1.2.4
Numba-0.53.1,Torch-1.8.1+cpu,
.
Spyder-5.0.1, Matplotlib-3.4.1, sympy-1.8.
Panel-0.11.3, Voila-0.2.9, holoviews-1.14.3.
numpy-1.20.2+mkl, SciPy-1.6.2, Pandas-1.2.4.
Numba-0.53.1,Torch-1.8.1+cpu.
variation per version:
Pytorch in standard branch only
portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other)
.
Pytorch in standard branch only.
portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other).
4.0.2021030711 Mar 2021 11:25
minor feature:
Infrastructure:
Python-3.8.8, Python-3.9.2
Jupyterlab-3.0.9, replacing Jupyter-notebook per Jupyter-nbclassic (icon name unchanged, but jupyterlab plumbing below)
Python-3.9 is now feature-complete, so becomes the default download suggestion
VScode-1.54.1.0, Nodejs-14.16.0
.
Python-3.8.8, Python-3.9.2.
Jupyterlab-3.0.9, replacing Jupyter-notebook per Jupyter-nbclassic (icon name unchanged, but jupyterlab plumbing below).
Python-3.9 is now feature-complete, so becomes the default download suggestion.
VScode-1.54.1.0, Nodejs-14.16.0.
New Packages:
Great_expectations, openpyxl, ipycanvas, nbclassic
Wasmer_compiler_cranelift, wasmer_compiler_singlepass, zstandard
.
Great_expectations, openpyxl, ipycanvas, nbclassic.
Wasmer_compiler_cranelift, wasmer_compiler_singlepass, zstandard.
Upgraded Packages:
Spyder-4.2.3,
Scikit_learn-0.24.1, sympy-1.7.1
Seaborn-0.11.1, dask_labextension-5.0.1, Torch-1.8.0+cpu
Numpy-1.20.1+mkl, SciPy-1.6.1, Pandas-1.2.3
Matplotlib-3.3.4, holoviews-1.14.2,
Numba-0.53.0rc3
.
Spyder-4.2.3.
Scikit_learn-0.24.1, sympy-1.7.1.
Seaborn-0.11.1, dask_labextension-5.0.1, Torch-1.8.0+cpu.
Numpy-1.20.1+mkl, SciPy-1.6.1, Pandas-1.2.3.
Matplotlib-3.3.4, holoviews-1.14.2.
Numba-0.53.0rc3.
Variation per version:
Pytorch in standard branch only
Portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other)
.
Pytorch in standard branch only.
Portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other).
3.4.2021021315 Feb 2021 03:16
minor feature:
Infrastructure:
Python-3.8.7, Python-3.9.1
Pip-20.3.3 with the new "Resolver":
Now pip wants to install maintain a "coherent" set of wheels, like in a Distribution,
WinPython has been tweaked to it, but there will remain an with Spyder if you auto-update it : it will downgrade PyQt5,
Jupyterlab-2.2.9, nodejs-14.15.0.
VScode-1.52.1
.
Python-3.8.7, Python-3.9.1.
Pip-20.3.3 with the new "Resolver":
Now pip wants to install maintain a "coherent" set of wheels, like in a Distribution,
WinPython has been tweaked to it, but there will remain an with Spyder if you auto-update it : it will downgrade PyQt5,
.
Now pip wants to install maintain a "coherent" set of wheels, like in a Distribution.
WinPython has been tweaked to it, but there will remain an with Spyder if you auto-update it : it will downgrade PyQt5.
Jupyterlab-2.2.9, nodejs-14.15.0.
VScode-1.52.1.
New Packages:
Testing packages: coverage, nbval
Sklearn_contrib_lightning, onnxruntime, anyio
.
Testing packages: coverage, nbval.
Sklearn_contrib_lightning, onnxruntime, anyio.
Upgraded Packages:
Spyder-4.2.1,
Scikit_learn-0.24.0, scikit_image-0.18.1, joblib-1.0.0, sympy-1.7.1
Seaborn-0.11.1, dask_labextension-4.0.1
Numpy-1.19.4, SciPy-1.5.4, Pandas-1.1.5, statsmodels-0.12.1
Matplotlib-3.3.3, holoviews-1.14.0
.
Spyder-4.2.1.
Scikit_learn-0.24.0, scikit_image-0.18.1, joblib-1.0.0, sympy-1.7.1.
Seaborn-0.11.1, dask_labextension-4.0.1.
Numpy-1.19.4, SciPy-1.5.4, Pandas-1.1.5, statsmodels-0.12.1.
Matplotlib-3.3.3, holoviews-1.14.0.
Variation per version:
Pytorch in standard branch only
in Python-3.9 branch, but with no official support yet: PyTorch, Numba
Portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other)
.
Pytorch in standard branch only.
in Python-3.9 branch, but with no official support yet: PyTorch, Numba.
Portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other).
3.0.20201121929 Dec 2020 13:45
minor feature:
Infrastructure:
Python-3.8.7, Python-3.9.1
Pip-20.3.3 with the new "Resolver":
Now pip wants to install maintain a "coherent" set of wheels, like in a Distribution,
WinPython has been tweaked to it, but there will remain an with Spyder if you auto-update it : it will downgrade PyQt5,
Jupyterlab-2.2.9, nodejs-14.15.0.
VScode-1.52.1
.
Python-3.8.7, Python-3.9.1.
Pip-20.3.3 with the new "Resolver":
Now pip wants to install maintain a "coherent" set of wheels, like in a Distribution,
WinPython has been tweaked to it, but there will remain an with Spyder if you auto-update it : it will downgrade PyQt5,
.
Now pip wants to install maintain a "coherent" set of wheels, like in a Distribution.
WinPython has been tweaked to it, but there will remain an with Spyder if you auto-update it : it will downgrade PyQt5.
Jupyterlab-2.2.9, nodejs-14.15.0.
VScode-1.52.1.
New Packages:
Testing packages: coverage, nbval
Sklearn_contrib_lightning, onnxruntime, anyio
.
Testing packages: coverage, nbval.
Sklearn_contrib_lightning, onnxruntime, anyio.
Upgraded Packages:
Spyder-4.2.1,
Scikit_learn-0.24.0, scikit_image-0.18.1, joblib-1.0.0, sympy-1.7.1
Seaborn-0.11.1, dask_labextension-4.0.1
Numpy-1.19.4, SciPy-1.5.4, Pandas-1.1.5, statsmodels-0.12.1
Matplotlib-3.3.3, holoviews-1.14.0
.
Spyder-4.2.1.
Scikit_learn-0.24.0, scikit_image-0.18.1, joblib-1.0.0, sympy-1.7.1.
Seaborn-0.11.1, dask_labextension-4.0.1.
Numpy-1.19.4, SciPy-1.5.4, Pandas-1.1.5, statsmodels-0.12.1.
Matplotlib-3.3.3, holoviews-1.14.0.
Variation per version:
Pytorch in standard branch only
in Python-3.9 branch, but with no official support yet: PyTorch, Numba
Portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other)
.
Pytorch in standard branch only.
in Python-3.9 branch, but with no official support yet: PyTorch, Numba.
Portable VSCode in "cod" branch only. (but you can copy the " t cod" directory from one to the other).
3.0.2020102802 Nov 2020 07:05
minor feature:
Infrastructure:
Python-3.9.0, Python3.8.6,
nodejs-14.14.0, VSCode-1.50.0, Python-VSCode-2020.9.114305
Qt Stack Qt-5.15.1 is now in WinPython-3.9
.
Python-3.9.0, Python3.8.6.
nodejs-14.14.0, VSCode-1.50.0, Python-VSCode-2020.9.114305.
Qt Stack Qt-5.15.1 is now in WinPython-3.9.
New Packages:
PyQt5-Tools (QtDesigner for PyQt5)
FastApi (see https://talkpython.fm/episodes/show/284/modern-and-fast-apis-with-fastapi)
.
PyQt5-Tools (QtDesigner for PyQt5).
FastApi (see https://talkpython.fm/episodes/show/284/modern-and-fast-apis-with-fastapi).
variation per version:
Pytorch is in standard 3.8 branch only,
portable VSCode is in "cod" branch only. (but you can copy the " t cod" directory from one WinPython to any other)
.
Pytorch is in standard 3.8 branch only.
portable VSCode is in "cod" branch only. (but you can copy the " t cod" directory from one WinPython to any other).
3.0.2020080823 Aug 2020 21:45
minor feature:
Infrastructure:
Simplification: use Christoph Cgohlke wheel msvc_runtime to ship msvc dlls
Python-3.8.5, Python-3.9.0rc1
.
Simplification: use Christoph Cgohlke wheel msvc_runtime to ship msvc dlls.
Python-3.8.5, Python-3.9.0rc1.
New external tool:
Pandoc-2.3 (is back) to support nbconvert
.
Pandoc-2.3 (is back) to support nbconvert.
New Packages:
Dask_labextension, dask_ml,
Jupyter_bokeh
Wasmer
on Python-3.9 : python-Levenshtein, astropy
.
Dask_labextension, dask_ml.
Jupyter_bokeh.
Wasmer.
on Python-3.9 : python-Levenshtein, astropy.
Upgraded Packages:
Jupyterlab-2.2.4, Qt-5.15.0, VSCode-1.48,+ python-vscode-2020.8.101144, Spyder-4.1.4
Numpy-1.19.1, SciPy-1.5.2, scikit_learn-0.23.2, pandas-1.1.0
Pyarrow-1.0.0, bokeh-2.1.1, Matplotlib-3.3.1
Torch-1.6.0+cpu
.
Jupyterlab-2.2.4, Qt-5.15.0, VSCode-1.48,+ python-vscode-2020.8.101144, Spyder-4.1.4.
Numpy-1.19.1, SciPy-1.5.2, scikit_learn-0.23.2, pandas-1.1.0.
Pyarrow-1.0.0, bokeh-2.1.1, Matplotlib-3.3.1.
Torch-1.6.0+cpu.
Dropped Packages:
Tensorflow-cpu + Keras : too big to ship
.
Tensorflow-cpu + Keras : too big to ship.
Variation per version:
Torch is not in "cod" branch
Portable VSCode is in "cod" branch only. (but you can copy the " t cod" directory from one to the other)
.
Torch is not in "cod" branch.
Portable VSCode is in "cod" branch only. (but you can copy the " t cod" directory from one to the other).
2.3.2020053001 Jun 2020 12:05
minor feature:
Infrastructure:
on wppm, packages labels, when not present in internal databases packages.ini,are taken in summary information of the installed package,
Automatic update of missing entries in packages.ini during build
Missing dll 'vcruntime140_1.dll' is now included (on 64 bit)
Python-3.8.3 final.
WARNING for WinPython-3.8 and later:
Since Python-3.8.2, the default settings is no more looked in " HOME " but in " USERPROFILE "
That is disrupting the old way of WinPython to keep users things in a directory.
in comments in "env.bat", a "set USERPROFILE= HOME " is created to keep user settings in winpython directory for Python-3.8.2+ and Python-3.9:
it's due to the above change in cpython-3.8.2+,
Effect is a way too big to put it by default, but it's also possible via script "make_working_directory_and_userprofile_be_winpython.bat"
.
on wppm, packages labels, when not present in internal databases packages.ini,are taken in summary information of the installed package,
Automatic update of missing entries in packages.ini during build
Missing dll 'vcruntime140_1.dll' is now included (on 64 bit)
.
Automatic update of missing entries in packages.ini during build.
Missing dll 'vcruntime140_1.dll' is now included (on 64 bit).
Python-3.8.3 final.
WARNING for WinPython-3.8 and later:
Since Python-3.8.2, the default settings is no more looked in " HOME " but in " USERPROFILE "
That is disrupting the old way of WinPython to keep users things in a directory.
in comments in "env.bat", a "set USERPROFILE= HOME " is created to keep user settings in winpython directory for Python-3.8.2+ and Python-3.9:
it's due to the above change in cpython-3.8.2+,
Effect is a way too big to put it by default, but it's also possible via script "make_working_directory_and_userprofile_be_winpython.bat"
.
Since Python-3.8.2, the default settings is no more looked in " HOME " but in " USERPROFILE ".
That is disrupting the old way of WinPython to keep users things in a directory.
in comments in
2.2.2019102225 Oct 2019 16:45
minor feature:
no Tensorflow nor PyTorch package yet.
no Cartopy/Geoviews/Geopandas/rpy2 package yet.
Early addition of Quart/Hypercorn/Uvicorn/fastAPI/python_language_server.
1.7.2016110113 Nov 2016 01:05
minor feature:
Highlights : pandas 0.19.1, scikit-learn-0.18.0, WinPython-3.6.0.0 (python-3.6.0beta3 quality / no Spyder nor other Qt package).
User facing : spyder 3.1.0dev0-20161104, matplotlib1.5.3.
Computation: numba-0.29, llvmlite-0.14, cython-0.25.1.
New packages:
Clrmagic.
Scikit-optimize-0.2
Jupyterlab-0.8 (only in WinPython-3.6, as still pre-alpha)
.
Clrmagic.
Scikit-optimize-0.2.
Jupyterlab-0.8 (only in WinPython-3.6, as still pre-alpha).
Adjustments for Python3.6.0beta3 (new python._pth method to isolate environment).
This time: no more WinPython-3.5 PyQt4 32bit.
MD5.
SHA-1
SHA-256
Binary
Size
Fd74fa93926fca050aa132c3456046e3
6d292d2cc1966f20ea08d136a4123c7506998bfc
4c925715733923e9f4fc8b1a44ad82e8398029a0a84ddd7d48a1fd689d7c7159
WinPython-64bit-3.6.0.0Zero_preview.exe
24 075 947 Bytes
1ace445989983c81110051c1cfa07e5b
C8d7480381e74eb9241ba2f370894769bc854555
Cbb1a934c83730a09dcad5811d94ec9210b77c0e18c542d62530f855d07933a8
WinPython-32bit-3.6.0.0Zero_preview.exe
23 203 484 Bytes
49eabdcc104772d458d272a6e797156d
500a9e8f5c500060e136e3c3efb826517cd15d16
E6e24a4deae4ae67388ff5b743fb3c27104b06921079a73f2af68d056cd6338b
WinPython-64bit-2.7.12.4Zero.exe
22 270 791 Bytes
Cbb5ed5d1d8fccae4adf79d1a625955f
1babd7b9eddbbfecd8c384af04a2b9bf94fcf362
8a0fd17529ad5a3a61a2026833806125803a37ea75ad3758d9ec71a460ae737b
WinPython-32bit-2.7.12.4Zero.exe
21 545 631 Bytes
Fc44358714dc30a1a85d0e7520ec936f
9cee6042b8120b89bf29beacacc46dcc6361bbad
E6db29f4c6a06b831a146d936c397ea90ba8ae4fb72cde76de7d8c29ebad8268
WinPython-32bit-3.4.4.5Zero.exe
24 489 833 Bytes
9dca9184b78c7728e4a8cc9d3367ee8c
Ae8ecdf59c6f87b5b0eb9098e02797b94aa5fd7a
D560821e88bc6520c4c501e0dcbff2399cd7a06d5540ecdd66b8dde49fb518f6
WinPython-64bit-3.4.4.5Zero.exe
25 217 133 Bytes
Ddbf14301965f3f9e8a2e5d5136f4c3f
8c016ce7b8986f9b63c353baaac3771843cbc364
Bf9332092f8c513a5c3efe799bfc827a15ce70f4053583687ad74e052007de64
WinPython-64bit-3.5.2.3Zero.exe
23 642 895 Bytes
25e5e62668d339719b468d912e72cb7e
Dd71a73e15be8b55a43a6c00aa502d
1.6.2016082829 Aug 2016 09:05
minor feature:
User facing : spyder3.0b4+, Bokeh 0.12.1.
Computation: dask/distributed/xarray updates.
Infrastructure: PyQt5.7.0 for Python3.5.
Altair/Vega.
Lot of packages, pandoc-1.17.1.
no more binary ".exe"' packages.
MD5.
SHA-1
SHA-256
Binary
Size
64ace04d35e8d2eaa4c67469152e61fd
563b1f2e0e01d6ef60c4f0c6c1ce6448a94c69ff
0664b241e5dd8ef02d18f9ef1c07ef3860d605e164c50564dec5e078779fc1fd
WinPython-32bit-3.4.4.4Zero.exe
23 881 113 Bytes
67c6ecb9e215a83551743426d7a08015
Bcfb98567bfd0c68f4488c93f89c3e0a6e252f12
Abecc24189db6ae5b10426c0eaba5c7f819a9d5362942808d0d960311a6480fe
WinPython-64bit-3.4.4.4Zero.exe
24 600 054 Bytes
31cc8cbd1c94f69332b137cfa4f12c5f
Bde744a8b442108bf200eb37799e3bacd6501f21
Eebb24cd2a0996853818d8bf87c19f8411de94c5b2d335fc7c978cbe8287bb37
WinPython-64bit-3.5.2.2Zero.exe
23 032 756 Bytes
6e738e456196b71b2b3c3d847d916f60
Aecf18ba1fbcd955ddd40d962bd5b94ce6bc4418
1f6fa6457c5ab41f4eaaab5cd443c20a9fbfb0a3caef80ad89c4d92582d216b5
WinPython-32bit-3.5.2.2Zero.exe
22 280 272 Bytes
0574aa934268fa6b4fcd450373ccf28e
7077e5185917d320084f8d63a13f524c2900a877
280a877d452e05f71bd6a72685b99b99afb257863b2b5a8e174007eb58288225
WinPython-64bit-2.7.12.3Zero.exe
22 221 744 Bytes
Fa4d385602f9e200b27970e5d863e656
1e0b915bb312a67df3bab39ba7078fc571f234be
032e367187eb85945c01e1711950eed65f077bfc254c0dea5279cba580bd45e7
WinPython-32bit-2.7.12.3Zero.exe
21 498 414 Bytes
F682dbc3eeb99843c8b848ade3b220cb
C39e10adc8f2f6c42c06ef538edb4779bcd85f34
7b38419233b775354ae09d53f253b7fa5407cf99ad15c20ea3646c1e51759c57
WinPython-64bit-3.4.4.4Qt5.exe
301 448 368 Bytes
D372671a80bad0bd2e652e569c45e4ae
Ead43341e0ae01860a90e6847333f5f6fe15a14c
B65c453fc09a5b5f585fd25aef23bb668ace6533e9f409e3d94125b4791c7d7c
WinPython-32bit-3.4.4.4Qt5.exe
256 752 166 Bytes
5a96da96e8a38700076b3200c6e60dfe
280cce807857e84b550246b155592456899364da
Ae9dd71ef002b3cf85dc9c05cfebd78acc714cfe1a5c1e69cbdf78f4558559ba
WinPython-64bit-3.5.2.2Qt5.exe
271 660 710 Bytes
E4e10b6ae96bf11874cad56bdd302990
Da6c45b49bf197fdc871188c7b6066
1.6.2016072224 Jul 2016 20:45
minor feature:
Highlight:
User facing: IPython 5.0, Matplotlib 1.5.2, Spyder 3.0.0beta 4.
Computation: numba 0.27, dask 0.10.1, numpy 1.11.1, scipy 0.18.0rc2, statsmodels 0.8.0rc1
Infrastructure: Qt5.6.0 for Python3.5 (beta quality)
.
User facing: IPython 5.0, Matplotlib 1.5.2, Spyder 3.0.0beta 4.
Computation: numba 0.27, dask 0.10.1, numpy 1.11.1, scipy 0.18.0rc2, statsmodels 0.8.0rc1.
Infrastructure: Qt5.6.0 for Python3.5 (beta quality).
New packages: distributed, ipyleaflet.
Infrastructure:
Switch default download to PyQt5 64bit, from PyQt4 64bit.
a WinPython2.7Zero (only for python-legacy constrained users)
.
Switch default download to PyQt5 64bit, from PyQt4 64bit.
a WinPython2.7Zero (only for python-legacy constrained users).
Cleanups:
Remove un-needed NSIS scripts still in projects.
.
Remove un-needed NSIS scripts still in projects.
Removals:
Drop of WinPython3.4 PyQt4 32bit branch.
Drop of pyreadline
Drop of QtDemo icon
Specific Notes:
First release built under Windows 10,
PyQt5 builds are suggested for new users.
MD5
SHA-1
SHA-256
Binary
Size
5634e7f090a9580b74ccc6d743403e62
Ccc0f215d30585982378840aa63976e60f5bb8e2
515476e01e97790d3baece0e6aa1e426011d8968b7800d6e212f80d6841e447e
WinPython-64bit-2.7.12.2Zero.exe
22 337 907 Bytes
A49915a70d7ae1cdb07348fb781e066d
3d2178bc8d1e18adae021c3a1be8aea9817b9668
01f1184eb352b1af8f39a7df19dff51c58b4b982880e52027df720804974800d
WinPython-32bit-2.7.12.2Zero.exe
21 614 863 Bytes
E651763a0c3c5e6173fe01463692f7b3
29372261d86a6b7d32350b361c54bac31edb6262
64e7d47d9a60194822b522febd382efa5e816c2459ba8414d236b3026e70eaaf
WinPython-64bit-3.5.2.1Zero.exe
23 152 399 Bytes
Baa0f3a4a0d5774a5ab02d1a9858e5ec
8730bee0c34fb840303fcde1eb0053b3a4fe39af
F6dba5ec1efa7eada5befd8dd38cf6d0e1dcacb9a27822e416fbb81eea66493d
WinPython-32bit-3.5.2.1Zero.exe
22 413 213 Bytes
Fa8d2888e4add0c0d987c50677220166
Cd6455cf0a7ff66d9850ea12c0dcdcee63a989e7
Adcc6c0897b4020f1536d163992be2d2f14dcea1708f0b860391081ca1db00c1
WinPython-32bit-3.4.4.3Zero.exe
23 996 736 B
1.3.2016020910 Feb 2016 05:45
minor feature:
Highlight:
Jupyter Notebook 4.1 / Ipython 4.1.1 ('Search replace' and 'run_all' options), bokeh 0.11.0.
Numpy 1.10.4 and numba 0.23.1 now allows usage of '@' multiplicator for WinPython 3.5
Two icons change:
"Spyder (reset)" replaces "Spyder (light)".
"IDLEX" superseds "IDLE".
.
Jupyter Notebook 4.1 / Ipython 4.1.1 ('Search replace' and 'run_all' options), bokeh 0.11.0.
Numpy 1.10.4 and numba 0.23.1 now allows usage of '@' multiplicator for WinPython 3.5.
Two icons change:
"Spyder (reset)" replaces "Spyder (light)".
"IDLEX" superseds "IDLE".
.
"Spyder (reset)" replaces "Spyder (light)".
"IDLEX" superseds "IDLE".
New packages:
Data visualization: ptpython, IDLEX.
Data acquisition: pyreadline
.
Data visualization: ptpython, IDLEX.
Data acquisition: pyreadline.
Upgrades:
Data visualization: matplotlib 1.5.1, Holoviews 1.4.1+.
Data crunching: scipy 0.17.0, Pulp for WinPython 3.5
Data acquisition: pep 8.0.2 / setuptools 19.6.2
.
Data visualization: matplotlib 1.5.1, Holoviews 1.4.1+.
Data crunching: scipy 0.17.0, Pulp for WinPython 3.5.
Data acquisition: pep 8.0.2 / setuptools 19.6.2.
Cleanups:
Reorganization of documentation Notebooks: be more "run all" and "beginner" friendly.
Improved pip compatibility
.
Reorganization of documentation Notebooks: be more "run all" and "beginner" friendly.
Improved pip compatibility.
Removals:
Scite.
no more mid-size WinPython "slim" ("uninstall" from WinPython full size is the suggested workaround)
.
Scite.
no more mid-size WinPython "slim" ("uninstall" from WinPython full size is the suggested workaround).
Specific Notes:
Users wanting 'R' in their WinPython can look at the included R_installation notebook.
Still no mingwpy for python 3.5, but ongoing progress: https://mingwpy.github.io
Still older numba 0.20.0 in WinPython 3.4, to spare users the "vs 2015 runtime".
.
Users wanting 'R' in their WinPython can look at the included R_installation notebook.
Still no mingwpy for python 3.5, but ongoing progress
1.2.2015102930 Oct 2015 06:45
minor feature:
Changelog.
re-add flavor option using the DOS build chain.
Prepare for next release cycle.
Shifting to Jupyter World.
of redundant definition.
Tweaks.
More tweaks.
Add package.
Qtconsole-4.0.1 removes this workaround need.
do changelogs via pip freeze.
do changelogs via pip freeze, part 2.
Step toward movable / non-movable option.
.
Add package.
re-arrange build process for requirements.txt use.
Initial handling of python 3.4.
Cleanups.
Ease wppm usage.
.
QT_API setting is now dynamic.
Move standard patchs in wppm.
Update version and remove python2.6.
in case of build.
Cleanup.
Workaround bad installers inside wppm.
Workaround bad installers inside wppm.
Cleanup around launchers scripts.
Cleanup wppm launcher.
New packages.
of.
Adapt wppm for pip requirements.
Switch to 96 build by requirements.
New package.
Matplotlib1.5 claims this package.
Twine claims that package.
Qt5 script tweak.
New packages.
Qt '5 otherwise 4' launchers.
Packages names addition.
More package name.
More package.
More packages.
Add scripts to make winpython movable or.
New package name.
New package names.
Keep nbextensions in Winpython directory.
Miss in previous commit.
Pyuic.
New packages (hdf4 change).
1.2.20150801Qt505 Aug 2015 03:16
minor feature:
third build with PyQt5, using PyQt5.5.0.
no more Qt4 packages.
PyQtGraph and Spyder are 'fresh from Github' versions (to get more Qt5 fixes).
one know bug: if you install or uninstall only one package at a time with WinPython Control Panel:
the package is installed (or uninstalled).
then a Window open to show a 'installing'/'uninstalling the package' message
== You have to click the "cancel" button of this useless window to make it go away.
.
the package is installed (or uninstalled).
then a Window open to show a 'installing'/'uninstalling the package' message
== You have to click the "cancel" button of this useless window to make it go away.
1.2.20150731Qt501 Aug 2015 14:05
minor feature:
Spyder 3.0.0dev of july30th, (contain some Qt5 fixes).
pyqtgraph from git master.
pyuic should be fixed (was a PyQt5 change to incorporate in Winpython).
PyQt5.4.1 instead of PyQt5.5.0.
remove Vitables (not PyQt5 compatible).
1.2.2015072528 Jul 2015 05:45
minor feature:
273 MB
WinPython-64bit-3.4.3.5_build4.exe.
.
Source code (zip).
.
Source code (tar.gz).
1.2.20150723qt524 Jul 2015 09:25
minor feature:
277 MB
WinPython-64bit-3.4.3.5_build3Qt5.exe.
.
Source code (zip).
.
Source code (tar.gz).
1.2.20150628b29 Jun 2015 06:25
minor feature:
shift to a ".bat" method to create builds.
adapt "spyder" and "spyder (light)" to change in spyder 2.3.5.
3.4.2.1 3.3.5.2 2.7.8.221 Oct 2014 19:39
minor feature:
New packages MinGW32 4.8.1, PyYAML 3.11, XlsxWriter 0.5.9,
beautifulsoup4 4.3.2, decorator 3.4.0, husl 4.0.0, llvmpy 0.12.7, lmfit 0.8.0, lxml 3.4.0, nltk 3.0.0, numba 0.15.1, prettytable 0.7.2, requests 2.4.3, scilab2py 0.5, sqlparse 0.1.13, wheel 0.24.0, Cython 0.21.1, Pillow 2.6.1, Python 3.4.2, SQLAlchemy 0.9.8, Sphinx 1.2.3, astroid 1.2.1, brewer2mpl 1.4.1, colorama 0.3.2, ipython 2.3.0, joblib 0.8.3, julia 0.1.1.4, matplotlib 1.4.1, networkx 1.9.1, nose 1.3.4, pandas 0.15.0, pg8000 1.10.1, pylint 1.3.1, pyparsing 2.0.3, pytz 2014.7, rpy2 2.4.4, seaborn 0.4.0, setuptools 5.8, simplejson 3.6.4, six 1.8.0, spyder 2.3.1, sqlite_bro 0.8.7.4, statsmodels 0.6.0, tornado 4.0.2, PyQtdoc 4.8.4, Scidoc 1.8.0, libpython 1.0.