> I know that there are many different ways to represent
> a graph, but your graph format *is just plain wrong.*

Yet when I offered to support an option to support the graph format you insist is uniquely "right", you poo-poo'ed the idea. So what is your real complaint? I don't know. Here you once again complain about the graph format, after rejecting an idea to address that directly.

Sorry, but you don't make sense to me. 

> if you're willing to rewrite all of the documentation you
> probably can explain this in a clear and simple way, and
> without breaking compatibility.

I don't see that happening. The docs are quite clear to me already, but I recognize that I'm too close to them to be objective about that. Instead I'm going on ample evidence from mailing lists, blogs, and StackOverflow that yours is the first complaint I've seen that the basic functionality here is any way unclear or ill-documented. People pick it up and use it productively at once. Nobody else has appeared to be confused.

Where things get unclear is limited to novel applications of multiprocessing. But people always stumble over multiprocessing, even for something as simple and bulletproof as a multiprocessing.Queue. Comes with the territory.

Finally, the distinction between "predecessors" and "dependencies" here strikes me as being uselessly pedantic. Nobody on the planet has an actual problem with recognizing that, e.g., "a" is predecessor of "n" in the string "pedantic". That's so in both technical language and plain English. Same here.

Still, if you want to rewrite the docs, go for out. I'll stay out of it completely.