New Python block cipher API, comments wanted

[Paul Rubin]

Erik Max Francis <max at alcyone.com> writes:
> In traditional development cycles, new features get introduced long
> before alpha.  Alpha is when you have all the new features in place in
> at least some fashion, although they may not be working too well yet. 

Well, the idea is to have the features in place before the coming
alpha, though maybe not "long before".  There are also application
packages in Python that want to use the module and will use it if it's
available.

> Alpha is not when you introduce new features, particularly ones that
> are just reviewed APIs.

It sounds like you're thinking of this as some complicated subsystem
where a lot of API-level stuff needs to be ironed out that hasn't been
thought about yet.  It's not anything like that.  This functionality
has been done many times before and crypto application writers are
used to it.  

> I think you'd be much better off in this process aiming for (2.4 or
> whatever the next release will be) and starting with a PEP.

This is not a Python enhancement; it's just some library functions
that implement an existing, uncomplicated, well-accepted standard.  If
you look at the existing set of Python library functions, hardly any
of them have PEPs.  There's no PEP for the existing rotor module, for
example.  A PEP for this seems to me like overkill.

Anyway, so far at least, the actual crypto application developers who
have looked at this module have either said it looks fine or else have
suggested minor changes which are easy to incorporate.  So my attitude
is that a whole lot more process is not needed--this is either doable
properly in time for 2.3 or else it isn't, and the only way to find
out is to try to get it done and see what happens.  If it's not done
in time, that's ok, there's always the next release.  But such a delay
means that useful, real-world applications that need the module will
take longer to become widely deployable, etc.  I'd rather not make
them wait if they don't have to.