Solving environment fails on windows

[arthur-parmentier]

Describe your Problem:

conda env create -f environment.yml fails on windows and results in several packages not being installed.

To Reproduce:

On windows 10 (lastest anaconda 1.10), conda env create -f environment.yml

Error report:
`(base) C:\Users\dell\Desktop\ChessVisionBot>conda env create -f environment.yml
Collecting package metadata (repodata.json): done
Solving environment: failed

ResolvePackageNotFound:

• libopencv==3.4.2=h7c891bd_1
• libedit==3.1.20181209=hb402a30_0
• libcxxabi==4.0.1=hcfea43d_1
• libpng==1.6.37=ha441bb4_0
• pixman==0.38.0=h1de35cc_0
• grpcio==1.16.1=py37h044775b_1
• libssh2==1.8.2=hcdc9a53_2
• pyobjc-core==5.2=py37_0
• yaml==0.1.7=hc338f04_2
• freetype==2.9.1=hb4e5f40_0
• jpeg==9b=he5867d9_2
• pyproj==1.9.6=py37h01d97ff_1002
• bcrypt==3.1.6=py37h01d97ff_1
• bzip2==1.0.8=h1de35cc_0
• pyobjc-framework-quartz==5.2=py37_0
• zlib==1.2.11=h1de35cc_3
• scipy==1.3.1=py37h1a1e112_0
• libnetcdf==4.6.1=h5e45101_3
• numpy-base==1.16.5=py37ha711998_0
• tornado==6.0.3=py37h01d97ff_0
• c-ares==1.15.0=h1de35cc_1001
• libopenblas==0.3.6=hdc02c5d_1
• matplotlib-base==3.0.2=py37hf043ca5_1002
• mkl-service==2.0.2=py37h1de35cc_0
• basemap==1.2.1=py37hd103db2_1
• proj4==5.2.0=h6de7cb9_1005
• tk==8.6.8=ha441bb4_0
• mkl==2019.4=233
• libvpx==1.7.0=h378b8a2_0
• xz==5.2.4=h1de35cc_4
• markupsafe==1.1.1=py37h1de35cc_0
• opencv==3.4.2=py37h6fd60c2_1
• libopus==1.3=h1de35cc_0
• cairo==1.14.12=hc4e6be7_4
• jasper==2.0.14=h636a363_1
• glib==2.56.2=hd9629dc_0
• libcurl==7.62.0=h051b688_0
• harfbuzz==1.8.8=hb8d4a28_0
• hdf5==1.10.2=hfa1e0ec_1
• numpy==1.16.5=py37h926163e_0
• gettext==0.19.8.1=h15daf44_3
• hdf4==4.2.13=0
• libtiff==4.0.10=hcb84e12_2
• sqlite==3.29.0=ha441bb4_0
• expat==2.2.6=h0a44026_0
• qt==5.6.3=h1d42b2f_0
• lxml==4.4.1=py37h08abf6f_0
• cryptography==2.7=py37h212c5bf_0
• graphite2==1.3.13=h2098e52_0
• libffi==3.2.1=h475c297_4
• libxslt==1.1.32=h1e79283_1003
• openssl==1.1.1d=h1de35cc_3
• fontconfig==2.13.0=h5d5b041_1
• libgfortran==3.0.1=h93005f0_2
• python==3.7.4=h359304d_1
• libxml2==2.9.9=hf6e021a_1
• libiconv==1.15=hdd342a3_7
• sip==4.18.1=py37h0a44026_1000
• curl==7.62.0=ha441bb4_0
• fribidi==1.0.5=h1de35cc_0
• cffi==1.12.3=py37hccf1714_0
• py-opencv==3.4.2=py37h7c891bd_1
• protobuf==3.8.0=py37h0a44026_0
• tensorflow-base==1.14.0=mkl_py37h5a24fda_0
• netcdf4==1.4.1=py37ha292673_200
• pillow==5.4.1=py37hb68e598_0
• pcre==8.43=h0a44026_0
• tensorboard==1.14.0=py37h80053f4_0
• libcxx==4.0.1=hcfea43d_1
• astropy==3.2.1=py37h01d97ff_0
• readline==7.0=h1de35cc_5
• graphviz==2.40.1=hefbbd9a_2
• zstd==1.3.7=h5bba6e5_0
• geos==3.7.2=h6de7cb9_1
• pyyaml==5.1.2=py37h1de35cc_0
• wrapt==1.11.2=py37h1de35cc_0
• libprotobuf==3.8.0=hd9629dc_0
• pyobjc-framework-cocoa==5.2=py37_0
• psutil==5.6.3=py37h01d97ff_0
• kiwisolver==1.1.0=py37h770b8ee_0
• h5py==2.8.0=py37h878fce3_3
• tensorflow==1.14.0=mkl_py37h085be34_0
• cftime==1.0.3.4=py37h1e5eb4f_1001
• nomkl==3.0=0
• intel-openmp==2019.4=233
• ncurses==6.1=h0a44026_1
• pynacl==1.3.0=py37h1de35cc_1000
• icu==58.2=h4b95b61_1
• pyqt==5.6.0=py37hc26a216_1008
• pango==1.42.4=h060686c_0
• ffmpeg==4.0=h01ea3c9_0`

Expected behavior:
I expected the env to solve and the package to install as usual

System:

• OS: Windows 10
• Interpreter: Python 3.9 (lastest)

Additional context:
The env file is made for OSX. Can that be the issue?

[arthur-parmentier]

The same error appears with >conda create --name chess --file requirements.txt:

Collecting package metadata (current_repodata.json): done
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed

PackagesNotFoundError: The following packages are not available from current channels:

• pango==1.42.4=h060686c_0
• python-chess==0.30.0=pypi_0
• cryptography==2.8=py37hafa8578_1
  ....

[dynamicdiscord]

Hiya, wanted to put my two cent in. Since OSX is unix, have you tried running any WSL versions to build this? Thanks!

[dynamicdiscord]

@kochsebastian -- tagging you to help this user.

[kochsebastian]

Yeah some of these packages are unix specific.
I wanted to make a slim requirements file for ages but I haven't got the time at the moment. I would advise to start installing the essential things like tensorflow, opencv, python-chess etc. And then try to run the program. See want decencies are missing and install them manually. Try again until it runs without an import error.

You could also look into this install-only-available-packages
With this you can at least install all available packages and then install the rest manually.

Sorry for the inconvenience but this is not one of my priorities at the moment

[dynamicdiscord]

Whenever I get a free moment, I'll look into creating a slim requirements file. Just found this again.