Merge lp://qastaging/~coreygoldberg/selenium-simple-test/fix-set-window into lp://qastaging/selenium-simple-test

9 + 'Checking attribute %r of %r' % (attribute, element_to_string(elem)))

This is another example showing that we lack coverage. I think it's time to start addressing those by writing failing tests for them as we encounter them.

24 + def _was_resized(width, height):

This function is used only once and will work equally well without the parameters (which are just hiding the set_window_size ones).

17 + # Do not actually resize if the specified size is same as original.
18 if (orig_width == width) and (orig_height == height):
19 return (width, height)

I don't think the optimization is worth it there, presumably test writers will use set_window_size() when they intend the size to change so this code path is rarely reached and one can even argue that it may come in the way:

I can imagine cases where I want to shuffle the browser a bit and ensure *something* reaches it when I use set_window_size. Bypassing my request then comes in my way.

34 return (width, height)

If there is no way to return values different from the ones I passed, why should I care about what is returned ?

If that was available, I would rather get the *previous* values. But that would require querying the browser with get_window_size() which I can equally well call myself.

So, I'd rather return something else, including nothing ;)

In the same spirit, I think you can further simplify the tests while making them clearer (their intent is almost clear but comments won't hurt).

Basically, testing {get|set}_window_size requires to start with a known heigth/width and making sure set_window_size change it or errors out.

fixed based on comments.. see updated branch.

Looks simpler ;)