We're currently in the process of converting from unittest (pure) to
Nose, and one of the issues I've noticed while doing this was that the
database wasn't flushing correctly when switching from a
TransactionTestCase to a TestCase. I've been digging through the code
this morning, and I honestly
Since not everyone uses Convore, and I wanted to pop in over here and
talk about the direction of the Debug Toolbar.
First off, I want to ask everyone what kind of workflow they prefer.
Previously Rob had been using what is known as the "git-flow", which
basically means master is a stable,
Google seems to have deemed it a good idea to rename this thread
title. This is regarding Sentry (aka django-sentry), a 3rd party
Django app.
On Nov 8, 8:12 pm, David Cramer <dcra...@gmail.com> wrote:
> I'm looking for feedback regarding something I want to implement in
&g
I'm looking for feedback regarding something I want to implement in
Sentry [1].
Currently we allow sorting by a few things, but its mostly based
around the "times_seen" value. Times seen is useful, and always
available, but there are many use cases where log messages would be
better to group by
, is_summary,
**self.extra)
query.aggregate_select[alias] = self
On Mar 22, 4:53 pm, David Cramer <dcra...@gmail.com> wrote:
> One of the recent changes in trunk was a change to how querysets were
> cloned. Due to this, some old code we had is no longer working. This
> was a c
One of the recent changes in trunk was a change to how querysets were
cloned. Due to this, some old code we had is no longer working. This
was a custom aggregate which relied on "aggregate_select" (see below).
I believe the change I'm referring to is what is causing this, and I'm
unsure of what
I've come across the need lately, for my applications to be completely
standalone, and this has presented itself a few problems:
- I need urls.py to be dynamically editable (at least at runtime). I
think I can deal with this one, just haven't gotten to it.
- More importantly, I need urls.py to
3, 2008 at 11:23 PM, David Cramer <[EMAIL PROTECTED]> wrote:
>
> > I'm building a new pluggable ratings app (django-voting doesn't come
> > close to covering what my specs are). It's built around a similar
> > architecture to what we used on Curse. What I'm looking for
In MySQL and PostgreSQL you can go into the database and modify the
auto increment sequence value to start at something else. There's no
way to set that via Django at the moment.
On May 13, 4:49 pm, jwwest <[EMAIL PROTECTED]> wrote:
> Hi all,
>
> Is there a way to create a column that has a
I'm building a new pluggable ratings app (django-voting doesn't come
close to covering what my specs are). It's built around a similar
architecture to what we used on Curse. What I'm looking for right now
is feedback on my proposed API and usage:
http://code.google.com/p/django-ratings/
The
I know some of you out there use my full-text layer relying on Sphinx
in Django. I released a new version yesterday, which on my initial
tests is fully backwards compatible.
So the point, what I'm looking for, is has anyone put the new version
to use yet? If not, can you try it on your working
I'm not too familiar with .NET, but with middleware and decorators it
would be possible, there would just need to be some awareness. One
thing we had done on Curse, was wrap the jQuery base AJAX call, and
our login_required methods. If a JS method was sent in (all JS methods
were to /js/ or had
lanning on using a library
> for this tool and if so which one?
>
> Thanks for the work. I look forward to seeing the tool,
>
> Michael
>
> On Thu, Mar 27, 2008 at 10:13 AM, David Cramer <[EMAIL PROTECTED]> wrote:
>
> > As promised, Im going to release some tools that
As promised, Im going to release some tools that make it easier to
integrate client and server-side javascript validation and utilities
for newforms. To do this though, I first need to come up with a list
of what is needed. Here is my current list:
- Validation per-field via client-side and
I've taken the time to upload both our SphinxSearch layer, and our
MediaWiki markup util to Google code today. For more info see my blog
post: http://www.davidcramer.net/code/54/mediawiki-markup-and-sphinxsearch-...
I haven't used Google code before, so if theres any problems just let
me know.
Because this has everything to do with the web 2.0 expo
On Mar 14, 9:33 am, "Chris Brand" <[EMAIL PROTECTED]> wrote:
> What's the best way to restrict access on a per-app basis ?
> I want to have two apps, with some users allowed access to one, others
> allowed access to the other, and some
Any other Django users going to be in the area? I see there's a Django
talk at the conference by Adrian and there's a lot of other great
stuff going to be there as well.
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google
The problem with this is, when the foreignkey can be null the only way
to reference it properly is with a LEFT JOIN. Now, I'm not sure why it
isnt implemented in ORM, but I strongly encourage you to find a
differently solution, as LEFT JOINs can be VERY costly on system
resources.
On the same
The traceback is only within the console and from broken pipe.. theres
no errors on the page :)
On Mar 5, 10:28 am, Tim Chase <[EMAIL PROTECTED]> wrote:
> > To expand on this, sometimes the pages load as they would on our live
> > environment, but it randomly (and happens most of the time) hangs
To expand on this, sometimes the pages load as they would on our live
environment, but it randomly (and happens most of the time) hangs to
where it takes up to a couple minutes to load a page.
On Mar 5, 9:14 am, "David Cramer" <[EMAIL PROTECTED]> wrote:
> I've had
I've had issues as of lately with runserver hanging upon using it for
the dev server. I've been unable to diagnose this myself so here's my
plea :)
The issue seems to be on the system as a whole, and I've tried going
through the list of middleware and context processors and none of that
seemed
Anyone else around who's interested in meeting up before Monday? We
are staying in the event hotel and I'm very interested in talking with
other Django or even Python developers who manage high load, or even
smaller websites, to see what they have come up with for solutions as
well as standards
We've run into a pretty annoying issue lately with Django. Being that
our site is VERY large, and the codebase has become a framework for the
smaller sites, which all run off the same load balanced servers, we
began seperating content.
For example:
- We have www.curse-gaming.com, which is all
We have two database apps, in two different directories. Django sees
these both as database_. I thought this was ok except we
obviously have issues w/ permissions and content types, which we'll be
changing, except tonight I came across the fun error that it's showing
data from the wrong tables.
What it should be (I believe):
SELECT `files_version`.`id` , `files_version`.`file_id` ,
`files_version`.`md5` , `files_version`.`name` ,
`files_version`.`post_date` , `files_version`.`archive` ,
`files_version`.`change_log` , `files_version`.`author_id` ,
`files_version`.`downloads` ,
AIL PROTECTED]> wrote:
> David Cramer schrieb:
>
> > As we were having issues lately with handling database load, I'm
> > looking at alternatives for MySQL/Innodb in the future. One thing that
> > came up was solidDB. I'm curious as to if this would be possible to use
>
> Why don't you just generate lists and dictionaries in the view and send
> those to your templates instead of data objects?
I like having access to the models attributes, such as
get_absolute_url, we use things like this quite a bit in our templates
> 1. It'd be pretty easy to write a template
As we were having issues lately with handling database load, I'm
looking at alternatives for MySQL/Innodb in the future. One thing that
came up was solidDB. I'm curious as to if this would be possible to use
with Django, or if anyone has used this in the past and has experiences
they could tell
select_related() info moved:
http://groups-beta.google.com/group/django-developers/browse_thread/thread/400bcdee7fedbb15?hl=en
On Jan 10, 8:37 am, "David Cramer" <[EMAIL PROTECTED]> wrote:
> http://dpaste.com/hold/4539/
>
> I think I still have an issue with related fi
http://dpaste.com/hold/4539/
I think I still have an issue with related fields that werent included
in the query
On Jan 10, 8:15 am, "David Cramer" <[EMAIL PROTECTED]> wrote:
> Use this paste instead, I found an issue :)
>
> http://dpaste.com/hold/4537/
>
> On
Use this paste instead, I found an issue :)
http://dpaste.com/hold/4537/
On Jan 10, 6:41 am, "David Cramer" <[EMAIL PROTECTED]> wrote:
> Alright, anyways, if someone wants to test this, it worked 100% for me,
> the updated select_related(). I'm not sure if we're upd
Using full page caching, or view caching, is not an option for us on
most pages, as things vary such as moderation options.
On Jan 10, 5:26 am, "James Bennett" <[EMAIL PROTECTED]> wrote:
> On 1/9/07, David Cramer <[EMAIL PROTECTED]> wrote:
>
> > I don't th
On Jan 10, 5:27 am, Cliff Wells <[EMAIL PROTECTED]> wrote:
> David Cramer wrote:
> > While working on my select_related changes, I noticed that the db
> > backend was doing COUNT(*) for .count().. why?
>
> > I've changed it on my local copy to count the id column, but
While working on my select_related changes, I noticed that the db
backend was doing COUNT(*) for .count().. why?
I've changed it on my local copy to count the id column, but unless
anyone can give me a specific reason, as far as I recall, it's faster
to just count on the id column.
On Dec 26
, "James Bennett" <[EMAIL PROTECTED]> wrote:
> On 1/9/07, David Cramer <[EMAIL PROTECTED]> wrote:
>
> > I'd like to come up with an optional, or simply, a solution for our
> > framework, where designers do not get access to do the SQL through the
> > tem
select_related is horrible slow on 90% of large database queries, and I
haven't managed to implement the functionality I want in it yet.
I'd like to come up with an optional, or simply, a solution for our
framework, where designers do not get access to do the SQL through the
template on related
I'm talking completely about the related objects accessors, and not
sending models to the templates is not a good solution for us :)
I want to setup some kind of blocking mechanism so that while rendering
a template the models will not do any SQL queries that have not already
been executed, or
I'm petitioning this as a core change, if nothing else, I'm seeking
advice
on how to go about it.
I do *not* like templates having the ability to execute SQL. If I don't
have the data once its past my view (as this is MVC, control doesnt
need to get the data, it just needs to control it), I
38 matches
Mail list logo