Right now, the multiprocessing code is littered with statements like:

    assert self.\_popen is None, 'cannot start a process twice'
    assert self.\_parent_pid == os.getpid(), \\
           'can only start a process object created by current 

process'
assert not _current_process._daemonic, \
'daemonic processes are not allowed to have children'

These will obviously be stripped out if running in optimized mode -
however its not cool to rely on these anyway, the code should be
refactored to proper checks, e.g.:

    if not hasattr(lock, 'acquire'):
        raise AttributeError("'%r' has no method 'acquire'" % lock)

@jesse: could you provide a patch that addresses this issue?

@mark Yeah - I'm the current multiprocessing maintainer. If I fix it, I'll just commit it :) I filed this as a to do against myself.

I attached a patch which replaces all asserts with checks that raise exceptions. I used my judgement in determining exception types but I might have been off in some places. Also, this patch replaces ALL asserts. It is possible that some of the internal functions should be kept using asserts but I am not familiar enough with the module to say which. I figured I'd better do the whole thing than reviewers can say where lines should be reverted to asserts.

Thank you. I've attached some comments, click on the "review" link to read them.

Thanks for the quick review! I attached second iteration addressing feedback + changed all occurrences of checks like "type(x) is y" to "isinstance(x, y)".

I would appreciate a second look because this patch has many small changes and even though I ran full test suit which passed, I'm afraid I made a typo somewhere :)

Thanks for the patches, vladris!

I've reviewed the latest version, and it addresses all of Antoine's review feedback. Ezio left some additional feedback (http://bugs.python.org/review/5001/#ps3407) which still needs to be addressed.

@drallensmith, the patch you commented on is very old and probably doesn't apply anymore. It would be useful to make an updated patch (or rather PR, as we are using Github for development now: see https://devguide.python.org/pullrequest/).

Patches (for 2.7 and 3.5; the code is identical, only the line numbers changed) for the portion I commented on can be found in bpo-31169.

New changeset 48d9823 by Antoine Pitrou (Allen W. Smith, Ph.D) in branch 'master':
bpo-5001, bpo-31169: Fix two uninformative asserts in multiprocessing/managers.py (bpo-3078)
48d9823

Pull request submitted (4 days ago) for managers.py, queue.py; just added pool.py changes to those.

No discussion yet on pull request (5 days); just submitted fixes for util.py.

Please allow us a bit of time :-) We are all volunteers here.

Sorry! I was starting to wonder if the PR had been overlooked somehow...

I've updated the PR to include all of the non-Windows-specific asserts; I am not sufficiently familiar with Windows multiprocessing to feel confident writing informative error messages.

New changeset bd73e72 by Antoine Pitrou (Allen W. Smith, Ph.D) in branch 'master':
bpo-5001: More-informative multiprocessing error messages (bpo-3079)
bd73e72

I suppose this can be closed now.