fixes #770, date keyword from json

[schferbe]

No description provided.

[wasade]

Thanks, @schferbe. It looks like there are real failures here, can that be fixed?

[coveralls]

Coverage increased (+0.009%) to 86.447% when pulling bca1e02 on schferbe:master into 619e591 on biocore:master.

[coveralls]

Coverage remained the same at ?% when pulling a9cadeb on schferbe:master into 43b4217 on biocore:master.

[schferbe]

@wasade I think the failures I caused should be fixed now...

[wasade]

Thanks. It looks like the current failures are due to flux in flake8 and sphinx which are out of context for this PR. If you're able to fix those (either by restricting the versions, or addressing the errors) that would be fantastic, but I understand if you'd prefer not too. I cannot merge until they are addressed and unfortunately not something like to get to today.

Actually, one more request, would it be possible to add a mention into the changelog?

[schferbe]

It wasn't my intention to get that deeply involved into your project. I'd prefer you fixing the remaining issues. Thank you.

I will add something to the changelog...

[wasade]

Hi @schferbe, sorry for the delay on this. The dateutil library as specified isn't compatible with the (wide) range of numpy versions we support. Is it possible to bound the version, or preferably to perform the parsing using the standard library to avoid adding a dependency?

[wasade]

Closing as there doesn't seem to have been progress finalizing this PR