Hi!
In the course of implementing `natural_key` for many different models, I've
noticed that the implementation is fairly predictable; it tends to use the
fields already marked as unique. To avoid writing a separate implementation
for each model, I've written a patch that implements the relevant l
On Tue, May 06, 2014 at 05:57:28AM -0700, Piotr Gosławski wrote:
> W dniu wtorek, 6 maja 2014 11:47:17 UTC+2 użytkownik Florian Apolloner
> napisał:
> >
> > [...]
> > Memcached doesn't provide access to the remaining TTL, and I don't see how
> > we can reasonably fake this without writing an extr
On Tue, May 6, 2014 at 8:57 AM, Piotr Gosławski
wrote:
> W dniu wtorek, 6 maja 2014 11:47:17 UTC+2 użytkownik Florian Apolloner
> napisał:
>>
>> [...]
>>
>> Memcached doesn't provide access to the remaining TTL, and I don't see
>> how we can reasonably fake this without writing an extra key contai
W dniu wtorek, 6 maja 2014 11:47:17 UTC+2 użytkownik Florian Apolloner
napisał:
>
> [...]
> Memcached doesn't provide access to the remaining TTL, and I don't see how
> we can reasonably fake this without writing an extra key containing the
> expiration date.
>
> Would that be unacceptable to a
On Tue, May 6, 2014 at 5:47 PM, Florian Apolloner wrote:
> Hi Russ,
>
>
> On Tuesday, May 6, 2014 2:54:52 AM UTC+2, Russell Keith-Magee wrote:
>>
>> As far as the NotImplemented bit goes - I'd rather see this implemented
>> for all officially supported backends, rather than only one backend. The
>
Hi Russ,
On Tuesday, May 6, 2014 2:54:52 AM UTC+2, Russell Keith-Magee wrote:
>
> As far as the NotImplemented bit goes - I'd rather see this implemented
> for all officially supported backends, rather than only one backend. The
> only exception to this would be if there is a fundamental technic
