Strange error after qs-rf

2008-04-30 Thread simonbun
Hi, Before I begin: I'm using newforms-admin r7508, which is a couple of commits after the qs-rf update. Here's my stack trace: Traceback (most recent call last): File "/var/www/mydomain/project/django/core/handlers/base.py", line 82, in get_response response = callback(request,

Re: Great article about web2py

2008-04-29 Thread simonbun
It's beyond me why they would classify PHP as a web framework, and on top of that, include CakePHP in the comparison as well... Other than that, it's hardly an independant and unbiased review, coming from the authors of web2py themselves. A thing that does seem handy is their integrated ticketing

Re: Queryset-refactor branch has been merged into trunk

2008-04-27 Thread simonbun
So qs-rf hit trunk and was merged into nfa on the same day! That's excellent news, both for me and for django. Thanks so much for your efforts Malcolm (and Brosner for the quick merge)! regards, Simon Malcolm Tredinnick wrote: > I merged queryset-refactor into trunk just now. This was changeset

Re: Scaling the server

2007-10-26 Thread simonbun
@ Malcolm: I know, I didn't mean you'd need 31 separate Apache processes at once. Most browsers only download a few files at once per subdomain, but that usually still involves spawning new Apache processes. How many is dependent on many factors of course. In the end I still believe that

Re: Scaling the server

2007-10-25 Thread simonbun
> Well, I have another server available now ;) And you're using it for serving static content I hope? Your previous setup with Apache serving everything with KeepAlive off can bring many a server to its knees. If you serve a html page with for example 30 css, js and image files, you make a

Re: apache2 + mod_python vhosts conflict

2007-10-20 Thread simonbun
Yes, apache needs that line for virtualhosting to work at all. If it is not provided, apache just picks the first virtualhost it finds and sends requests its way, regardless of Host header. regards, Simon On Oct 20, 11:36 am, Wiley <[EMAIL PROTECTED]> wrote: > Graham, > > I am not to my

#4165: upload progress in safari

2007-08-21 Thread simonbun
Hello, Has anyone gotten the file upload progress bar from ticket #4165 to work in safari? It works great in IE and FF, but so far no luck on safari. For some reason the xmlhttp requests get stuck on readyState 1 when a file upload is in progress. When I run them without a file upload, the

permissions and many-to-many relationships

2007-08-17 Thread simonbun
I'm working on a sort of integrated email system where an OutboundEmail can be sent to any number of Users. So naturally my OutboundEmail model has an m2m relationship to User. Today I've been fiddling with permissions on the User model. To my amazement I need "change" permissions on

Re: Settings.py variables in CSS file

2007-08-15 Thread simonbun
Why not use relative paths in your css files? They're relative to the css file, not the html file, so you can do url(../images/diag.gif). I'm assuming that all media files are on the same host and that only the base media directory name is adapted to the svn version though. regards, Simon On

Re: Accessing the content-attribute in request.FILES without loading it into memory

2007-08-13 Thread simonbun
Hi, You'll want to take a look at ticket #2070 [1] for streaming uploads. It has a working patch that will make it to trunk pretty soon I think. I'm not sure how you would handle streaming uploads directly to a S3 bucket, but it shouldn't be too hard to hack the patch from the ticket. [1]

Re: How can I run syncdb task without loading any initial_data fixtures?

2007-07-17 Thread simonbun
I'm a bit plagued by slow running tests as well, but i'm using an acceptable workaround (for my uses). This workaround assumes that you run your tests per app and that you only need to test a newly created test case. The idea is to use your keyboard interrupt to stop the test suite after your

Re: reduce execution time of object.count()

2007-06-08 Thread simonbun
If your situation allows it, you could de-normalize and save the count result in a separate table. Then you need to adjust the count whenever you add / delete a row ofcourse. If you're a database purist I might have made your stomach turn, however :) regards, Simon On Jun 7, 8:19 pm, Paul

newforms custom field clean method

2007-06-05 Thread simonbun
I'm having some trouble solving a newforms problem that other people must have run into as well. Feel free to share solutions or workarounds if any. Let's say I have a RegistrationForm that is used for registering new users. It has the usual 'name', 'email', ... etc fields. When a new user

Re: Forcing unit test runner to abort after failed test

2007-05-22 Thread simonbun
, Simon On May 18, 12:55 pm, simonbun <[EMAIL PROTECTED]> wrote: > Ah, should have refreshed before replying...I had placed a test_a, > test_z, test_zz in a particulartestclass and they ran > alphabetically, but that was probably a coincidence apparently. > > Too bad this can't

Re: select_related() and child rows

2007-05-21 Thread simonbun
As far as i know, select_related doesn't work across reverse relationships. Sometimes it's possible to just query the base model to achieve the same results, but not always. Maybe its a good feature for the query.py rewrite, but i'm thinking it could get ugly pretty fast. It would require a new

Re: newforms unique field(s)

2007-05-21 Thread simonbun
Great stuff again Malcolm. Perhaps it should be rehashed a bit and put in the newforms docs under 'custom validation' ? Just a thought :) regards, Simon On May 22, 12:11 am, Malcolm Tredinnick <[EMAIL PROTECTED]> wrote: > On Mon, 2007-05-21 at 14:00 -0500, Tim Chase wrote: > > > What is the

Re: Forcing unit test runner to abort after failed test

2007-05-18 Thread simonbun
Apparently I can't see very far... Unit tests do not run in reversed order. It appears to be alphabetically ordered. I'm still wondering how to abort the test runner without losing my stack traces though. Regards, Simon On May 18, 11:26 am, simonbun <[EMAIL PROTECTED]> wrote: > Hi a

Re: Forcing unit test runner to abort after failed test

2007-05-18 Thread simonbun
ssell Keith-Magee" <[EMAIL PROTECTED]> wrote: > On 5/18/07, simonbun <[EMAIL PROTECTED]> wrote: > > > > > Hi all, > > > I have unit tests for a particular app that take about 2 minutes to > > complete. Needless to say, adding new unit tests to this

Forcing unit test runner to abort after failed test

2007-05-18 Thread simonbun
Hi all, I have unit tests for a particular app that take about 2 minutes to complete. Needless to say, adding new unit tests to this app and running them is a time consuming process. I know that the old unit tests will succeed and am basically only interested in the results for the new unit

Re: QueryDict variables with multiple levels

2007-05-03 Thread simonbun
Ah brilliant. That would do the trick indeed. Every now and then hidden gems like this one pop up that have been in trunk all along, but somehow escaped my attention. Thanks all Simon On May 4, 1:58 am, SmileyChris <[EMAIL PROTECTED]> wrote: > > You could write a function that takes Django's

Re: QueryDict variables with multiple levels

2007-05-03 Thread simonbun
tures of Django that > there is little processing done to the actual request. > > Cheers, > Mike Axiak > > 1:http://www.php-security.org/MOPB/MOPB-03-2007.html > > On May 3, 11:58 am, Malcolm Tredinnick <[EMAIL PROTECTED]> > wrote: > > > On Thu, 2007-05-03 at

Re: Intermittent error since last svn up

2007-04-20 Thread simonbun
, 12:22 pm, Malcolm Tredinnick <[EMAIL PROTECTED]> wrote: > On Fri, 2007-04-20 at 10:03 +, simonbun wrote: > > I've "svn up -r #"d through all the revisions and it seems 4905 > > causes it. Wich is the first import of the code. > > > Normally more people

Re: Intermittent error since last svn up

2007-04-20 Thread simonbun
to what my cause it? I've tried deactivation psyco (grasping at straws here), but no change. On Apr 20, 11:58 am, Malcolm Tredinnick <[EMAIL PROTECTED]> wrote: > On Fri, 2007-04-20 at 09:31 +, simonbun wrote: > > I think this is going to weird you out even more. I wrapped it

Re: Intermittent error since last svn up

2007-04-20 Thread simonbun
Ofcourse it could always be a problem with this dev server, but it's been in use for a couple of years without any problems. Its a pretty standard debian woody setup with python2.4.4 and django over apache2.2 and mod_python3.2.10. On Apr 20, 11:31 am, simonbun <[EMAIL PROTECTED]> wrot

Re: Intermittent error since last svn up

2007-04-20 Thread simonbun
a problem with trunk, somewhere somehow. Very weird error indeed, especially because after 2-3 requests the problem goes away. On Apr 20, 10:16 am, Malcolm Tredinnick <[EMAIL PROTECTED]> wrote: > On Fri, 2007-04-20 at 07:04 +, simonbun wrote: > > Hi all, > > > I'm gett

Intermittent error since last svn up

2007-04-20 Thread simonbun
Hi all, I'm getting a weird intermittent exception since i updated django from trunk yesterday. When I reload apache and request any page I get an exception for about 3 times, and then after that all goes as intended. It seems to be a problem with the delayed_loader that replaces each real_*

Re: Which ajax framework django will support in the upcoming 1.0, prototype/dojo/jquery?

2007-04-10 Thread simonbun
I think that most people that ask for "ajax integration" want to see javascript helpers, as seen in many other frameworks. They're looking for things like for example {% auto_complete_tag .. params ..%}; wich would render an input box with the necessary javascript (inline..) that makes it

Re: Reversed edit_inline?

2007-03-11 Thread simonbun
Indeed, this feature is important to me too. I'm planning to implement it myself, but obviously only after the newforms admin is finished. I'll post it as a patch when it's done. regards, Simon On Mar 12, 1:11 am, "jfagnani" <[EMAIL PROTECTED]> wrote: > > I could add two foreignkey fields to

Re: Python HTML lib_filter

2007-03-07 Thread simonbun
Good stuff! I am going to be needing this at some point and now i have a good base to start from. I have one enhancement that is very important to me (and should be to everyone). Seeing as this is the type of filter that would be used in blog comments for example, i would add a rel="nofollow"

Re: Multilingual content - yet another idea

2007-01-23 Thread simonbun
This kind of translation support is something django really needs in my opinion. If this gets implemented together with the unicode conversion by the time 1.0 gets rolled out, i'd be very, very happy. I've read your blog post and i do have some remarks. I think it's better not to put the

Re: How to find out the id of the last record in a table?

2006-11-15 Thread simonbun
The problem with getting the last record's id and using it, is that someone might have inserted yet another record while you're still working on the previous one. For single user scenario's it's ok, or if you're using table level write locking. Yet afaik, its generally a bad idea.

Re: check for yourself (((;

2006-08-18 Thread simonbun
Exactly... and on that note: http://www.whitehouse.gov/robots.txt --~--~-~--~~~---~--~~ 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