Re: improving async support in psycopg

On 30/03/10 15:17, Federico Di Gregorio wrote:
> On 30/03/2010 00:17, Jan Urbański wrote:
>> The behaviour I proposed in a previous mail was that repeated execute()
>> attempts, commit() and rollback() were blocking, so app code would have
>> to make sure it's not calling and take care of serialising execute()
>
> That was also my initial idea. But more I think about it more I feel
> that async code should be _completely_ async and the client code should
> cope with it (and get loud exceptions in case of errors).
>
>> Maybe it could work like this:

 >> [things raise errors in async mode]

I've pushed these changes to my branch, along with a few bugfixes in the 
async execution/connection implementation.

> During the next few days I'll start pulling from your tree. I won't pull
> any change that does non-trivial changes to the current code base but
> everything that add a different code path is ok. I'll post on this list
> as I progress.

Thanks! If you'd like me to rework anything, just say so. I'm looking 
forward to having a fully operational async psycopg2 driver. On the PG 
Python dirver market psycopg2 is the only choice for me and I hope that 
it will eventually estabilish itself as the de facto standard for PG+Python.

Cheers,
Jan
_______________________________________________
Psycopg mailing list
[email protected]
http://lists.initd.org/mailman/listinfo/psycopg