Re: Psycopg 2.5 released

Hi all,

On Sun, Apr 7, 2013 at 10:07 PM, Daniele Varrazzo <
daniele(dot)varrazzo(at)gmail(dot)com> wrote:

> On Sun, Apr 7, 2013 at 8:19 PM, Gavin M. Roy <gmr(at)meetme(dot)com> wrote:
> > Awesome! I appreciate all of the work that has gone into psycopg and your
> > ongoing work on it. Any chance you'll add what's needed for native PyPy
> > support any time soon?
>
> Hi Gavin,
>
> I'm a total supporter of psycopg for PyPy, but not being a PyPy
> developer I can go up to a certain limit. I've helped developing
> psycopg2-ctypes 2.4.4
> (https://github.com/mvantellingen/psycopg2-ctypes), which passed the
> entire psycopg 2.4.4 test suite so had our "stamp of approval". More
> precisely, I've never run it on PyPy: Michael did; I run it with
> regular cpython (psycopg2-ctypes is not specific for PyPy: it just
> happens to be pure python so it can run either on cpython, on PyPy and
> whatever flavour has a ctypes implementation)
>
> Then some PyPy-related bugs started coming in, then there was a -cffi
> implementation which is supposed to be better than -ctypes but again
> I'm not a good judge for it. I was hoping that somebody caring for
> PyPy fixed them but nothing happened. I've avoid adding further
> features to the -ctypes project under the principle of not adding
> features where there are bugs.
>
> If I see activity again on either -ctypes and -cffi and there is some
> maintainer fixing PyPy issues I'd be happy to give my part to the
> project, which could be implementing the missing part in pure python,
> maintaining the test suite etc. I'd be also happy to have the project
> under <https://github.com/psycopg/>. But I have to see it maintained,
> and now either is not the case or the people hacking on it didn't get
> in touch with us.
>

My goal when porting psycopg2 was both as an exercise as well as seeing how
fast it could be with pypy.
In the end the performance was okayish, for some workloads it was quite
fast but I never actually used it in production. The cffi port is in that
aspect much more promising from what I've read. Never tried it myself
though.

Regarding further updates to psycopg2-ctypes I'm afraid I just can't find
the time for it. So if anyone is interested in taking over the development
the please let me know!

Regards,
Michael