Equation bank removal and upgrade to current sphinx

[lheagy]

Upgrade sphinx version

• remove dependency on sphinx 1.4.8 and docutils 0.12 and depend on latest versions (closes #428)
• all development should now take place on the most recent version of sphinx. If you use conda, please run

  conda update sphinx

  and if you use pip

  pip install --upgrade sphinx

• deploy docs from python 3

Equation bank

• temporarily remove usage of equation bank: we are running into errors with multiple references to a single equation. In order to overcome this, we will need a custom solution to label the equations uniquely for each usage. This is outlined in issue #547
• fix equation labels. All equations should be labeled with :label: (not :name:)

Unused references

• currently quite a number of references are not actually referred to in the text. This is causing warnings in the build process. I have temporarily removed those references from the main bibliography and have put them in unused_references.bib. Presumably they were initially included because we intended to use them, and it might indicate that citations are not properly formatted in the site. I have created issue #548 to keep track of this.

[lheagy]

@dccowan: would you mind giving this a quick review? I am happy to answer any questions you have. This will hopefully resolve the issues you were running into the other day

[dccowan]

@dccowan: would you mind giving this a quick review? I am happy to answer any questions you have. This will hopefully resolve the issues you were running into the other day

@lheagy I upgraded to the latest requirements and was able to successfully build the website locally. It seems like some of the original citations for some of the case histories have been replaced by new ones which reference SEG owned materials. If we are confident these are the correct citations, I can remove/comment the redundant ones.

[lheagy]

sounds good, thanks @dccowan: lets get this merged first and then you can make changes on a new branch so it is a bit better scoped. If you are happy with this, can you please approve the pr?

https://github.com/geoscixyz/em/pull/546/files