delta/python-packages/psycopg2.git/psycopg/adapter_binary.c, branch separate-binary github.com: psycopg/psycopg2.git Dropped ifdef guards against obsolete libpq versions 2016-08-15T01:55:44+00:00 Daniele Varrazzo daniele.varrazzo@gmail.com 2016-08-15T01:31:39+00:00 5ddc952dbb93167d19d84981debb741011958efc One of them was actually wrong: lobject_type.c wouldn't have compiled pre 8.3 (broken in 6e841a41, 2 years ago). 
One of them was actually wrong: lobject_type.c wouldn't have compiled
pre 8.3 (broken in 6e841a41, 2 years ago).
Dropped PG_VERSION_HEX constant 2015-06-02T09:54:08+00:00 Daniele Varrazzo daniele.varrazzo@gmail.com 2015-06-02T09:54:08+00:00 73d17e3c5e80c66738435cd6733e999cb9c21ca7 At PostgreSQL 10.0 it would have become awkward. 
At PostgreSQL 10.0 it would have become awkward.
Dropped simple type wrapper functions 2014-08-15T01:54:31+00:00 Daniele Varrazzo daniele.varrazzo@gmail.com 2014-08-15T01:37:37+00:00 a2b01cdf4221f4fad94143606c06141ce91ef7bd These functions don't need to exist: exposing the type in the module is enough. It is actually better as one may use isinstance and such. 
These functions don't need to exist: exposing the type in the module is
enough. It is actually better as one may use isinstance and such.
Dropped almost-no-op customized objects repr() 2014-08-15T01:54:10+00:00 Daniele Varrazzo daniele.varrazzo@gmail.com 2014-08-15T01:06:27+00:00 95165cef7d6157945802ffeb98d51c43fd4033e9 The default repr is enough: it prints <TypeName at 0xADDR> instead of <TypeName object at 0xADDR>. The only people being hurt by this change are the ones using doctests: they deserve it. 
The default repr is enough: it prints <TypeName at 0xADDR> instead of
<TypeName object at 0xADDR>.

The only people being hurt by this change are the ones using doctests:
they deserve it.
Name the types after the module they are exposed from 2014-08-15T00:47:19+00:00 Daniele Varrazzo daniele.varrazzo@gmail.com 2014-08-15T00:45:50+00:00 669e7879190c80d484bba5e969c834b0b3422331 






Use the Py_RETURN_NONE macro
2013-04-05T00:25:05+00:00

Daniele Varrazzo
daniele.varrazzo@gmail.com

2013-04-05T00:25:05+00:00

beea32ef03b6d0e599ae0f2b6b236da857e82077












Dropped GC support for several objects
2013-04-05T00:02:47+00:00

Daniele Varrazzo
daniele.varrazzo@gmail.com

2013-04-05T00:02:47+00:00

e14440d96d5aa70285e77bb26b979ea0dad84d6d

Non-containers don't need GC. It was half-baked anyway
as the tp_clear was often not set. Dropped tp_traverse too
for these objects as unused.





Non-containers don't need GC. It was half-baked anyway
as the tp_clear was often not set. Dropped tp_traverse too
for these objects as unused.







PyType_GenericAlloc is the default allocator: no need to specify
2013-03-20T23:53:20+00:00

Daniele Varrazzo
daniele.varrazzo@gmail.com

2013-03-20T22:27:10+00:00

d5316d7eb2e3cb75bbfdbd9b7a10f10240c4f462












Dropped "customized" pg_free functions
2013-03-20T01:48:21+00:00

Daniele Varrazzo
daniele.varrazzo@gmail.com

2013-03-20T01:34:50+00:00

5dfba462da063f8520ed55d193b929fa8257e3a8

The defaut is already to call PyObject_GC_Del.





The defaut is already to call PyObject_GC_Del.







Check if the object wrapped in binary is not None before trying the other types
2012-03-05T02:08:45+00:00

Daniele Varrazzo
daniele.varrazzo@gmail.com

2012-03-05T02:08:45+00:00

735d50c782b436d6807fd15e1b37e2231ce3e146

Otherwise it seems we clobber some result with NULL.





Otherwise it seems we clobber some result with NULL.