* Lint: Disable a couple potentially problematic/unneeded hooks
* Lint: Refine regex for existing Content-Type and PEP references checks
* Lint: Refine regex for Python-Version check and fix a couple deviations
* Lint: Refine regex for Created check and fix a couple deviations
* Lint: Refine regex for Resolution check and fix a few deviations
* Lint: Add new check for Post-History header
* Lint: Add check for Discussions-To header link
* Lint: Add basic check for PEP title presense & excessive length
* Lint: Add check for PEP/BDFL-Delegate header
* Lint: Add check for Sponsor header
* Lint: Add check for legacy Author field format
* Lint: Add check for RST Author field
* Lint: Use more helpful, user-oriented descriptions for header checks
* PEP 245: Add historical note and archive links to wiki page/email lsit
* PEP 628: Add direct link to resolution in acceptance message
* Lint: Explictly allow all valid RFC 2822 line continuations in headers
* Lint: Tweak Post-History and Resolution checks to accept fragments
* Builtin that determines an object's size in bytes. This is much more
intractable than it would seem. Allocating a single object may result
in a large block being requested from the O/S and never being returned
until shutdown.
* Test suite. This was a useless entry and more applicable before huge
efforts were expended to improve testing coverage.
* Four-space tabs. This was fixed-up a long time ago. Since then,
a whitespace czar routinely comes along correct subsequent transgressions.
* Generalized eval. This was done for Py2.4.
* Generalizing builtins to accept UserDict, UserList, sequences etc.
This is somewhat out-of-date with the introduction of subclassable
builtin types and the iterator interface.
it easier to improve the cmath algorithms, but no one had shown an
interest. The new plan is to await implementation of a full blown
complex type in C99.