Re: Missing typecasters [was: Decimal]

On 14/04/2010 09:59, Daniele Varrazzo wrote:
> On Wed, Apr 14, 2010 at 8:01 AM, Federico Di Gregorio <fog-NGVKUo/i/[email protected]> wrote:
> 
>> > I also merged everything else from the fix22 branch (the code seemed
>> > fine to me) but I didn't check the tests and now one test fails in a
>> > curious way; are you sure you committed everything? :)
>> >
>> > ERROR: test_flush_on_write (test_async.AsyncTests)
> This test depends on one of the commits Jan has pushed in his
> async_fixes branch. I added it to show a bug (flushing large queries)
> he has fixed later too, as described in this last weekend's emails.

Yep, I fixed it while merging Jan's and yours fixes.

I merged everything from you and almost everything from Jan (last two
changes require a little more review and I don't have that time today).
You both did an impressive work, thank you again.

Before the release I'd like to have a couple more changes and a final
decision on poll/fileno on the cursor. The changes are trivial and I'll
do them when I'll find 1h of free hacking time:

executing() -> isexecuting()
issync() -> async attribute (it really just expose the value of the
async parameter given at __init__ time.)

About poll/fileno, now that the semantics are exactly the same I'd like
to remove them from the cursor and have the client code access them
directly on the connection or via cursor.connection and for two reasons:

1) Smaller API is easier to maintain.

2)If in the future we discover we NEED poll() on the cursor (different
behaviour for named cursors?) adding a method won't break the API while
changing the semantics of an already existing one will.

federico

-- 
Federico Di Gregorio                                       fog-NGVKUo/i/[email protected]
 If we are going to teach "creation science" as an alternative to
  evolution, then we should also teach the stork theory as an
  alternative to biological reproduction.                -- Judith Hayes

_______________________________________________
Psycopg mailing list
Psycopg-IAPFreCvJWPBWskQ1e/[email protected]
http://lists.initd.org/mailman/listinfo/psycopg

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkvFeJ8ACgkQvcCgrgZGjetPyACcDt/r5NEM97SJo8RzjqvuyHZ2
tXIAn3fI/XrCDWyL+o/lMxoQ0D7FL0/E
=xnYA
-----END PGP SIGNATURE-----