Hi Michael,
I just installed 0.5.4p2 on my Mac 10.5.7 Intel with the Python bundled
pysqlite. Test suite ran without a single error.
>From your post to psqlite, there is a way for memory and gc to be managed.
Hope they implement the change on the upcoming updates. Meanwhile, it's
probably safe t
I did a static build of pysqlite 2.5.5 against sqlite 3.6.14. All
tests pass except the single memory leak test. I'll see if i can
isolate this for them.
On May 27, 2009, at 11:21 AM, Melton Low wrote:
> The first time I installed pysqlite 2.5.5, my Mac 10.4 build fine
> but not on the
DOH, oh right, I made the same mistake and just built pysqlite against
an arbitrary SQLite version. WHEW my day just got shorter again :) .
On May 27, 2009, at 11:21 AM, Melton Low wrote:
> The first time I installed pysqlite 2.5.5, my Mac 10.4 build fine
> but not on the 10.5. The solution
The first time I installed pysqlite 2.5.5, my Mac 10.4 build fine but not on
the 10.5. The solution from the pysqlite people was to build_static on
10.5. It seem to work the first time around. Unfortunately I was not able
to get a clean install last night on either systems. So for the time begi
with the latest pysqlite (2.5.5) I also get a huge (192) amount of
failures with many versions of SQLalchemy, including p1 and p2. I
had noticed that pysqlite's development seemed to be going further
with changes that make the library harder to develop against, but
particularly disturbi
Hi Michael,
I re-build my systems last night - zapped Python, Mercurial, Sqlite,
pysqlite, psycopg, and SQLAlchemy. You 0.5.4p1 test suite ran perfectly as
long as I only have the version of pysqlite included with Python. As I am
going out of town, I decided to leave 0.5.4p2 alone for the time b
the changes between 0.5.4p1 and 0.5.4p2 are extremely minimal and I've
pasted them at http://paste.pocoo.org/show/119367/ .Additionally
here they are passing for py2.4 and py2.5 at http://groovie.org:8012/
which is an intel linux platform. I run the tests with 2.6 here on
an intel
