On 01/06/13 06:37, Olemis Lang wrote:
On 5/31/13, Gary Martin <[email protected]> wrote:
On 30/05/13 18:47, Anze Staric wrote:
[...]
In the
longer term, I would like to be removing the dutest dependency for these
(I may have already said this)
yes, I recall . It'd be interesting (for me) to know the reasons . If
it is related to the extremely harmful «bad doctest» meme then please
read this recent python-dev thread [1]_ . Even considering points of
view like Michael Foord's [2]_ there's actually no benefit in using
unittest over doctests once dutest is installed in place [3]_
.. [1] Purpose of Doctests [Was: Best practices for Enum]
(http://www.gossamer-threads.com/lists/python/dev/1060688?page=last)
.. [2] Michael Foord in Purpose of Doctests [Was: Best practices for Enum]
(http://www.gossamer-threads.com/lists/python/dev/1061085#1061085)
.. [3] /me in Purpose of Doctests [Was: Best practices for Enum]
(http://www.gossamer-threads.com/lists/python/dev/1061093#1061093)
My main objection is that, as far as I can tell, dutest is not a common
tool with a wide community interested in keeping it maintained.
Beyond that, I would not want to go down the road of suggesting that
doctests are bad, but I would like to appeal to some consistency in our
approach. As far as I am aware, the only doctests that we have are the
ones for the dashboard code. While this is limited evidence, it seems to
suggest that most of us will be happy with using the unittest module.
Cheers,
Gary
