JOSS review - conda installation error

[lheagy]

Related to #2 and openjournals/joss-reviews#990, I was also unsuccessful in trying to install from conda (on my local machine: osx and on binder - e.g. try jupyter, which runs linux 64bit). Here is the error log I got:

Solving environment: failed

CondaHTTPError: HTTP 404 NOT FOUND for url <https://conda.anaconda.org/leampy/noarch/repodata.json>
Elapsed: 00:00.148486
CF-RAY: 4968e3cded3fc56a-ORD

The remote server could not find the noarch directory for the
requested channel with url: https://conda.anaconda.org/leampy

As of conda 4.3, a valid channel must contain a `noarch/repodata.json` and
associated `noarch/repodata.json.bz2` file, even if `noarch/repodata.json` is
empty. please request that the channel administrator create
`noarch/repodata.json` and associated `noarch/repodata.json.bz2` files.
$ mkdir noarch
$ echo '{}' > noarch/repodata.json
$ bzip2 -k noarch/repodata.json

You will need to adjust your conda configuration to proceed.
Use `conda config --show channels` to view your configuration's current state.
Further configuration help can be found at <https://conda.io/docs/config.html>.

For the JOSS submission, you do not necessarily have to have this resolved, but I would suggest that in the README, you

• [ ] describe the machine specs for which the conda install should work
• [x] provide alternate installation instructions for installing from source (as @kbevers described in #2):
  • git clone https://github.com/laempy/pyoints.git
  • pip install gdal # is there a reason this is not listed in your requirements.txt?
  • conda install gdal, pyproj, rtree
  • pip install -e .

[laempy]

Thank you @lheagy. The installation failed because of a typo in the installation instructions. Many thanks to @ReedAnders, who has noticed it. I have also added alternative installation instructions to the README file.

Gdal, pyproj, rtree and numpy are not included in the requirements file, because of version clashes during conda installation (pygdal does not recognize the available gdal versions).

Unfortunately I have no glue which machine specs are required. Packages for Windows (64 bit) and Linux (64 bit) have been prepared with conda. I assume, also 32 bit and Mac should work, but in the moment I have no chance to test it. Thus, also a conda recipe is provided to be able to compile a package on your own.

[lheagy]

Thanks for clarifying @laempy, this looks good.