You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Eric Gillespie <ep...@google.com> on 2007/10/11 04:43:31 UTC
Minimum Python for swig-py bindings
glasser@tigris.org writes:
> Author: glasser
> Date: Mon Oct 8 14:24:21 2007
> New Revision: 27029
>
> Log:
> For SWIG-Python tests, use unique names for temp repositories, so that
> multiple executions of the test suite don't stomp on each other (or on
> future executions: client.py doesn't really clean up after itself).
>
> (tempfile.mkdtemp would be better than tempfile.mktemp, but
> subversion/bindings/swig/INSTALL claims we need to support way back to
> Python 2.0.)
mkstemp is new in 2.3, which is over four years old. How many
people are upgrading to svn 1.5 while keeping a four-year-old
Python? Not that many. If we could at least move up to 2.2,
then we could use new-style classes for classes like
svn.delta.Editor and svn.ra.Callbacks .
I see dlr talking in another thread about bumping the
requirements for the Java binding.
I see no reason for the Python binding to be stuck in 2000.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Minimum Python for swig-py bindings
Posted by Max Bowsher <ma...@ukf.net>.
Eric Gillespie wrote:
> glasser@tigris.org writes:
>
>> Author: glasser
>> Date: Mon Oct 8 14:24:21 2007
>> New Revision: 27029
>>
>> Log:
>> For SWIG-Python tests, use unique names for temp repositories, so that
>> multiple executions of the test suite don't stomp on each other (or on
>> future executions: client.py doesn't really clean up after itself).
>>
>> (tempfile.mkdtemp would be better than tempfile.mktemp, but
>> subversion/bindings/swig/INSTALL claims we need to support way back to
>> Python 2.0.)
>
> mkstemp is new in 2.3, which is over four years old. How many
> people are upgrading to svn 1.5 while keeping a four-year-old
> Python? Not that many. If we could at least move up to 2.2,
> then we could use new-style classes for classes like
> svn.delta.Editor and svn.ra.Callbacks .
>
> I see dlr talking in another thread about bumping the
> requirements for the Java binding.
>
> I see no reason for the Python binding to be stuck in 2000.
I think we should *definitely* move up to Python 2.2 at least, and would
be totally ok with us moving up to Python 2.3.
Max.
Re: Minimum Python for swig-py bindings
Posted by David Glasser <gl...@davidglasser.net>.
On 10/10/07, Eric Gillespie <ep...@google.com> wrote:
> glasser@tigris.org writes:
>
> > Author: glasser
> > Date: Mon Oct 8 14:24:21 2007
> > New Revision: 27029
> >
> > Log:
> > For SWIG-Python tests, use unique names for temp repositories, so that
> > multiple executions of the test suite don't stomp on each other (or on
> > future executions: client.py doesn't really clean up after itself).
> >
> > (tempfile.mkdtemp would be better than tempfile.mktemp, but
> > subversion/bindings/swig/INSTALL claims we need to support way back to
> > Python 2.0.)
>
> mkstemp is new in 2.3, which is over four years old. How many
> people are upgrading to svn 1.5 while keeping a four-year-old
> Python? Not that many. If we could at least move up to 2.2,
> then we could use new-style classes for classes like
> svn.delta.Editor and svn.ra.Callbacks .
>
> I see dlr talking in another thread about bumping the
> requirements for the Java binding.
>
> I see no reason for the Python binding to be stuck in 2000.
Indeed, I just wanted to make the tests pass now without invoking a
mailing list thread first. This sounds reasonable to me.
--dave
--
David Glasser | glasser@davidglasser.net | http://www.davidglasser.net/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org