SQLite 3.8.4 fails test_sqlite on 2.7 and 3.4.

[ 296s] test test_sqlite failed -- Traceback (most recent call last):
[ 296s] File "/home/abuild/rpmbuild/BUILD/Python-2.7.6/Lib/sqlite3/test/hooks.py", line 165, in CheckOpcodeCount
[ 296s] self.assertTrue(first_count > second_count)
[ 296s] AssertionError: False is not true

This seems to be a change in behaviour callback of SQLite: (one "." per callback issued)
sqlite3_progress_handler(db, 1, progress_callback, NULL);
sqlite3_exec(db, "create table foo(a,b)", callback, 0, &zErrMsg);
sqlite3_progress_handler(db, 2, progress_callback, NULL);
sqlite3_exec(db, "create table bar(a,b)", callback, 0, &zErrMsg);

3.7.17> ./test
..................................................................
...........................
3.8.3.1> ./test
........
.......
3.8.4> ./test
.....
.....
3.8.4.1> ./test
.....
.....

This may be a side effect of optimization Documentation does day the callbacks are approximate. Can the test be adjusted to run more than CREATE TABLE so that a difference in the number of callbacks is tested for?

Also see openSUSE bug: https://bugzilla.novell.com/show_bug.cgi?id=867887

New changeset 1763e27a182d by Benjamin Peterson in branch '2.7':
weaken callback count inequality (closes #20901)
http://hg.python.org/cpython/rev/1763e27a182d

New changeset dbc9e3ed5e9f by Benjamin Peterson in branch '3.3':
weaken callback count inequality (closes #20901)
http://hg.python.org/cpython/rev/dbc9e3ed5e9f

New changeset 4d626a9df062 by Benjamin Peterson in branch 'default':
merge 3.3 (#20901)
http://hg.python.org/cpython/rev/4d626a9df062

Updating versions to match branches fix was committed in