← Back to team overview

pynie-dev team mailing list archive

vision for Pynie

 

I talked with Guido at PyCon last week about the direction for Pynie, particularly whether it should target 2.x or 3.x. He says 3.x, and it makes a lot of sense. Pynie is still quite young, and by the time it is mature, Python 3.x adoption will have started to pick up. It also gives us an opportunity to work with the CPython developers over the next few years as they polish 3.0, 3.1, 3.2.

So, I've started shifting the implementation to 3.x, one feature at a time. The transition should be relatively quick, as Pynie is still small.

My work on Pynie over the next few months will be focused on getting Pynie to pass the Python 3.0 core test suite. To do this, I'm breaking down the syntax elements of the test suite into bootstrapping tests in Lib/test/parrot/. These tests are run using the SubUnit testing framework, which decouples the framework from the tests, so the testing framework can run on CPython, while the individual tests run on Pynie. This means Pynie doesn't have to be fully-featured in order to start passing tests, which is a definite plus for the early stages of implementation. There's plenty for contributors to do there, writing tests, adding builtins (which are small functions in Pynie), or implementing syntax.

Pynie also needs some work on the configuration system. This will probably use autoconf.

I'd like to see how far we can push Pynie by the Parrot 1.4 release in July. It seems likely that we could at least have the full core syntax for Python 3.0 working by then.

Allison