`dill` and `cloudpickle` don't work with `require`

[basnijholt]

from ipyparallel import Client, require
client = Client()
lview = client.load_balanced_view()
dview = client[:]
dview.use_cloudpickle()

@require('types')
def func(x):
    return types.SimpleNamespace(n=x)

res = lview.map(func, range(10))

res.result()

gives

---------------------------------------------------------------------------NameError                                 Traceback (most recent call last)<string> in <module>()
/home/bnijholt/anaconda3/lib/python3.5/site-packages/ipyparallel/client/remotefunction.py in <lambda>(f, *sequences)
    248             if _mapping:
    249                 if sys.version_info[0] >= 3:
--> 250                     f = lambda f, *sequences: list(map(f, *sequences))
    251                 else:
    252                     f = map
/home/bnijholt/anaconda3/lib/python3.5/site-packages/ipyparallel/controller/dependency.py in __call__(self, *args, **kwargs)
     71 
     72     def __call__(self, *args, **kwargs):
---> 73         return self.f(*args, **kwargs)
     74 
     75     if not py3compat.PY3:
<ipython-input-8-293a3326374e> in func(x)
NameError: name 'types' is not defined

[minrk]

These decorators shouldn't be necessary if you use cloudpickle, because they should detect imported names correctly, unlike the default serialization. That doesn't address the concern of writing code that works in both cases, but for snippets like this, several of these serialization decorators change behavior with dill/cloudpickle vs default.