Use lpsum instead of plain addition for constraints

[meatballs]

Fixes #56

[meatballs]

Whoops. I committed a change to setup.cfg that I've been testing on this branch by accident.

[drvinceknight]

I've been down the road of getting pytest as a runner for the doctests: #28

I came to the conclusion that it wasn't the way to go:

• I could tell travis specifically (not appveyor) to run them so that made appveyor happy (it looks like the order is OS dependent)
• That doesn't work because travis doesn't get failures as a failed signal.

[meatballs]

Pretty much the conclusion I'm coming to - just didn't intend to do it on this branch!

[meatballs]

Oh good grief, it failed to commit the reversal!!! Agggghhhh......

[meatballs]

coveralls.io is playing up. It's showing that there's no change in coverage at https://coveralls.io/builds/11432442 but not reporting it here

[drvinceknight]

Yeah I've been waiting on it. This looks good to me, feel free to merge 👍

[coveralls]

Coverage remained the same at 100.0% when pulling 1cf6e71 on issue-56 into 4044583 on master.

[coveralls]

Coverage remained the same at 100.0% when pulling 2dff7b8 on issue-56 into 4044583 on master.

[coveralls]

Coverage remained the same at 100.0% when pulling 4fcb4a1 on issue-56 into 4044583 on master.

[coveralls]

Coverage remained the same at 100.0% when pulling 0354e62 on issue-56 into 4044583 on master.