Re: Ticket 2273: django.contrib.auth.models.User: username is case-sensitive

2015-08-30 Thread Florian Apolloner
On Saturday, August 29, 2015 at 9:59:30 PM UTC+2, Tim Graham wrote: > > The suggested change is to not allow a new user if there's a > username__iexact match. > Yes, but as change to what? To the admin interface? this will only cover admin usage. To a full_clean of the user model [This is ofte

Re: Adding more __repr__() methods

2015-08-30 Thread Ned Batchelder
The eval criterion rarely seems like a useful rule of thumb to me. Does anyone actually use this to make objects? I think the useful rule is, a repr should be an unambiguous representation of the object that is useful in debugging. Certainly the "eval" rule is sufficient for the "unambiguous"

Re: Ticket 2273: django.contrib.auth.models.User: username is case-sensitive

2015-08-30 Thread Collin Anderson
I propose adding a check on UserCreationForm (used by the admin). The current implementation relies on the model fields unique=True check which is case-sensitive (except on mysql). Whenever I've made a login form, I've always used or made a copy of or at least studied UserCreationForm. I think if

Re: Making the test suite run faster

2015-08-30 Thread Aymeric Augustin
Hello, I polished my test parallelization patch a bit. I'd like to merge it before 1.9 alpha. I rewrote history heavily to make the changes easier to review: https://github.com/django/django/pull/4761 Eventually I settled for the approach I dismissed in a previous email: run tests in workers, pas

Re: Making the test suite run faster

2015-08-30 Thread Andrew Godwin
I'm on board merging this without Oracle support - if I understand right, it means we just have to run those non-parallel for now, so we're not losing anything, right? Andrew On Sun, Aug 30, 2015 at 1:19 PM, Aymeric Augustin < aymeric.augus...@polytechnique.org> wrote: > Hello, > > I polished my

Re: Making the test suite run faster

2015-08-30 Thread Aymeric Augustin
On 31 août 2015, at 06:12, Andrew Godwin wrote: > I'm on board merging this without Oracle support - if I understand right, it > means we just have to run those non-parallel for now, so we're not losing > anything, right? Right. -- Aymeric. -- You received this message because you are subs

Re: Making the test suite run faster

2015-08-30 Thread Tino de Bruijn
Hi Aymeric, Really cool that you pushed this further. Out of curiosity I have two questions: - What happens when two SerializeMixin tests try to lock the same file? Does one wait for the other (probably not), or is a lockfile exception raised? - How does this work in combination with the --keepdb