On 29/09/11 23:39, Alex Gaynor wrote:
> When I was writing it, I suspect my objective was to move the minimal
> amount as possible to make the caching valid. I've got no problem with
> having them be attributes, rather than closures. The patch looks fine
> to me, feel free to commit.
To keep op
On Thu, Sep 29, 2011 at 6:36 PM, Luke Plant wrote:
> On 29/09/11 18:23, Alex Gaynor wrote:
>
> > Your right, there should be an additional assert. FWIW it's not even a
> > time/memory tradeoff, it's just savings. Type objects on every Python
> > VM are some of the most expensive objects, and ha
On 29/09/11 18:23, Alex Gaynor wrote:
> Your right, there should be an additional assert. FWIW it's not even a
> time/memory tradeoff, it's just savings. Type objects on every Python
> VM are some of the most expensive objects, and have various caches with
> them, allowing those caches to stay f
On Thu, Sep 29, 2011 at 1:20 PM, Luke Plant wrote:
> On 29/09/11 17:13, Alex Gaynor wrote:
> > Hi Luke (and the rest of the list ;)),
> >
> > Just saw r16912 and the subsequent commits, I wonder if you
> > saw https://code.djangoproject.com/ticket/14270? I think most of the
> > changes in that p
On 29/09/11 17:13, Alex Gaynor wrote:
> Hi Luke (and the rest of the list ;)),
>
> Just saw r16912 and the subsequent commits, I wonder if you
> saw https://code.djangoproject.com/ticket/14270? I think most of the
> changes in that patch and that you committed are the same, (which really
> just pa
Hi Luke (and the rest of the list ;)),
Just saw r16912 and the subsequent commits, I wonder if you saw
https://code.djangoproject.com/ticket/14270? I think most of the changes in
that patch and that you committed are the same, (which really just paves the
way for the caching I was going at).
Tha
