[mythtv-users] MythTV and Python 3

[Gary Buhrmaster]

On Tue, Oct 14, 2014 at 1:26 AM, Stephen Worthington
<stephen_agent at jsw.gen.nz> wrote:
> Mythbuntu 14.04, MythTV 0.27+fixes
>
> I was just looking at using Python for another MythTV script and I
> realised I should be writing it in Python 3 now that Python 2 is
> deprecated.

There has been no official deprecation.  Python 2.7
is just "legacy".

There is certainly some in the python community
that like using the tactic of saying deprecation since
they no longer want to work on things called legacy
(support and fixing things is no fun; writing new code
is fun).

Certainly if someone is writing (entirely) new
development of (entirely) new applications with
(entirely) new support modules, one should look
towards python 3 (I know I do).  But python 2.7
will be supported for a very long time.  I expect
much longer than the official EOL 2020 date
because enterprise linux distributions have
committed support dates far longer than that
(and rumor has it that some major corporations
have substantial internal commitments to 2.7
that is not going away any time soon).

And in any event, 98% of the interesting stuff is
not the language itself (2 vs 3), but the surrounding
eco-system libraries/modules, and that continues
to thrive in both versions.

Python 2 is dead, long live python 2.

> Are the MythTV bindings Python 3 compatible?

They are currently written for Python 2.  Python 3
is not 100% language compatible (although for
some simple cases, one can do various translations
and/or compatibility functions (see six)).

> Is anyone working on
> moving MythTV scripts to Python 3?

I believe it is on the (very large) to-do list.  Given
that python 2.7 has a long life ahead of it I
personally see no particular rush to expect a
dev to rearrange their priorities.

<conjecture>
And with 0.28 having an even more complete
services API, there is an opportunity to rework
the entire bindings structure (no more direct
internal DB access, for example, since that,
too, is "legacy" and targeted to be dropped at
some unspecified future point in time).  But
that is even more work than some simple
2 to 3 translation.
</conjecture>

If I was writing a python 3 function today
to interface with MythTV, I would likely look
to using the services API.  Python has lots
of good modules for accessing such services.