Pyallegro status (is it dead?). What about pygame.

[Grzegorz Adam Hankiewicz]

On 2005-02-27, Przemys?aw Ró?ycki <P.Rozycki at elka.pw.edu.pl> wrote:
> Has anyone used pyallegro? - I want to give allegro library a
> chance - but it seems that python bindings for this library are
> not maintained anymore.

They aren't if you look at the web page. If you look at the CVS
activity you will see that I'm doing work on it. But slowly, I'm
the only one interested on this right now.

> How does pygame compare to pyallegro - both in maturity and
> convenience of usage in python?

Pyallegro is itself incomplete. For example, there are serious string
conversion problems. As you probably know, Allegro supports Unicode,
but in a slightly broken way of having a global encoding. Pyallegro
simply ignores string encodings and passes happily raw Python string
(whatever their format) to Allegro, which then tends to choke on
8-bit characters.

Other issues (which AFAIK have not been dealt with other bindings)
are memory integration of both the C and Python sides. This is,
if you happen to create lots of python memory and then delete
references to it (implicitly marking them for garbage collection)
and try to create an Allegro bitmap, the C version will likely be
unaware of all that garbage collected memory and fail. IIRC this
could be solved by forcing the C library to use Python's malloc,
but I'm no Python expert, so could be talking rubish here.

Compared to pygame, pyallegro doesn't have anything. No
documentation, no support, no complete bindings, etc, etc. If you
want to do some serious work, ignore pyallegro for a few decades.

OTOH, if you like wasting time like me, you are welcome to join
me hacking the CVS. I wrote myself an image viewer program using
pyallegro and PIL for file format support. Works quite nicely,
in a perverted broken way, that is.

-- 
  Please don't send me private copies of your public answers. Thanks.