On Mon, 2008-03-24 at 11:38 -0700, shabda wrote: > Ah, but practicality beats purity :) > > But of course I have not fully profiled, the times so might be not > select_related as default is a good idea.
I think you're also equating your particular use-cases with being the most common and that's not necessarily true (it's also pretty much impossible to measure). I know that on different projects I've worked on, sometimes the code is regularly traversing relations on models, sometimes hardly ever. Uncontrolled select_related() calls are a performance hit because you end up doing database joins against tables you don't need, so we don't want to hit people with the performance penalty when it's not needed. Not everybody needs to access every attribute on their model that has a relation to another table every time, which would be the only case when using select_related() all the time made sense. In fact, in order to provide more control over this, in the "coming soon" features, you can control which of the related tables are pulled in via select_related(), because we realise that pulling in every table related to a model is not a very good idea, particularly on models with high branching degree in their relations. The current "all or nothing" approach is a reasonable first step, but we can do better (particularly with self-referential models). Regards, Malcolm -- Two wrongs are only the beginning. http://www.pointy-stick.com/blog/ --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to django-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/django-users?hl=en -~----------~----~----~----~------~----~------~--~---
