tickets

26 rows where "created" is on date 2008-07-19 and ui_ux = 0 sorted by resolution

✎ View and edit SQL

This data as json, CSV (advanced)

Suggested facets: changetime, stage, status, component, type, owner, reporter, has_patch, needs_better_patch

ui_ux 1 ✖

id created changetime last_pulled_from_trac stage status component type severity version resolution ▼ summary description owner reporter keywords easy has_patch needs_better_patch needs_tests needs_docs ui_ux
7835 2008-07-19 14:41:05 2021-10-01 12:43:47 2022-03-06 03:40:40.896101 Accepted new Testing framework New feature Normal dev   Provide the ability for model definitions that are only availably during testing A current limitation of the unit test framework is that there is no capacity to define 'test models' - that is, models that are only required for the purposes of testing. A regular installation would not know anything about these models - only a test database would have access to them. There are several existing applications that have a need for this capability: For example: * contrib.admin: you can't test the way admin handles models without some models to handle. * contrib.databrowse: you can't test the way the browser works without having models to browse * Django Evolution: you can't evolve models without having some models to evolve. The easiest way to work around this at present is to have a standalone test project which exercises the required functionality. However, these tests aren't integrated into the automated test suite, so they. Another option is to do some app_cache munging during the test - this works, but is very messy. Django should provide a way for model definitions to be defined as part of the test definition, synchronized as part of the test setup, populated and manipulated during test execution, and destroyed along with the test database. nobody russellm feature test models 0 0 0 0 0 0
7822 2008-07-19 00:25:46 2011-09-28 16:12:16 2022-03-06 03:40:38.593343 Unreviewed closed contrib.admin     dev duplicate Creating a OneToOneField with a related model breaks the list_filter interface in the related model This has happened since before the NFA merge. The summary of the ticket pretty much says it all. If you create a model (let's call it "UserProfile") and give that model a OneToOneField to another model (In this case "User"), the list_display interface will disappear on the User model's change list. If you change the o2o to a foreignkey, the list_filter panel comes back. An entire project is included (admin_test.zip). In this case I did User and UserProfile, but it appears to happen with any OneToOnefield. nobody ElliottM admin onetoone o2o list_filter 0 0 0 0 0 0
7827 2008-07-19 01:56:24 2011-09-28 16:12:17 2022-03-06 03:40:39.475504 Unreviewed closed contrib.admin     dev duplicate Admin change_view uses _default_manager in place of self.queryset() The admin change_view uses self.model._default_manager in place of self.queryset(request). nobody Alex   0 1 0 0 0 0
7834 2008-07-19 14:19:42 2008-07-19 14:21:14 2022-03-06 03:40:40.720382 Unreviewed closed Documentation     dev duplicate There's no link from the documentation index to the admin docs Summary says it all. nobody Jan Rademaker <j.rademaker@gmail.com>   0 0 0 0 0 0
7823 2008-07-19 00:50:16 2011-09-28 16:12:17 2022-03-06 03:40:38.786824 Accepted closed Database layer (models, ORM)     dev fixed ForeignKey get_db_prep_lookup doesn't work for custom primary key If a parent model has a user-defined field as a primary key: {{{ class Parent(models.Model): user_id = CutomUserField(primary_key=True) class Child(models.Model): parent = models.ForeignKey(Parent) }}} then this query will fail: {{{ p = Parent(...) Child.objects.filter(parent=p) }}} This happens because get_db_prep_lookup doesn't get called on p.user_id and its value (which is a custom user class) gets passed as an sql parameter into db. Patch follows. P.S. A bug is very practical: we have custom non-model user objects instead of contrib.auth.models.User and can't reference them from models with custom UserField because many queries using those models would break. mtredinnick isagalaev   0 1 0 0 0 0
7824 2008-07-19 00:56:13 2011-09-28 16:12:16 2022-03-06 03:40:38.979260 Accepted closed Uncategorized     dev fixed Missing information about autodiscover and admin.py in the tutorial and backwards incompatible page From what I understand, using admin.py to specify admin related settings instead of entering it in models.py is the desired way in the new trunk version. If this is the case, I suggest that the admin.austodiscover() should be explained in the tutorial. Additionally, if this is the desired setup, maybe the autodiscover should be included but commented out in the urls.py for each project, or even be active if the admin site is imported. jacob tome   0 0 0 0 0 0
7825 2008-07-19 01:06:05 2011-09-28 16:12:16 2022-03-06 03:40:39.134633 Unreviewed closed Metasystem     dev fixed modeltests/delete test fails after NFA merge After the NFA merge, this happens: {{{ ====================================================================== FAIL: Doctest: modeltests.delete.models.__test__.API_TESTS ---------------------------------------------------------------------- Traceback (most recent call last): File "/Users/jacob/Projects/Django/upstream/django/test/_doctest.py", line 2180, in runTest raise self.failureException(self.format_failure(new.getvalue())) AssertionError: Failed doctest test for modeltests.delete.models.__test__.API_TESTS File "/Users/jacob/Projects/Django/upstream/tests/modeltests/delete/models.py", line unknown line number, in API_TESTS ---------------------------------------------------------------------- File "/Users/jacob/Projects/Django/upstream/tests/modeltests/delete/models.py", line ?, in modeltests.delete.models.__test__.API_TESTS Failed example: del D._meta._related_objects_cache Exception raised: Traceback (most recent call last): File "/Users/jacob/Projects/Django/upstream/django/test/_doctest.py", line 1267, in __run compileflags, 1) in test.globs File "<doctest modeltests.delete.models.__test__.API_TESTS[17]>", line 1, in <module> del D._meta._related_objects_cache AttributeError: _related_objects_cache ---------------------------------------------------------------------- Ran 1 test in 0.046s FAILED (failures=1) }}} It appears to be related to the somewhat fishy internals hacking in the test case. The failure may be related to the change in import behavior because of admin registry class. lukeplant jacob   0 0 0 0 0 0
7826 2008-07-19 01:42:13 2011-09-28 16:12:16 2022-03-06 03:40:39.319754 Unreviewed closed Testing framework     dev fixed invalid_models test unhappy after NFA merge Like #7825, this one turns up: {{{ $ ./runtests.py --settings=settings invalid_models ====================================================================== FAIL: runTest (__main__.InvalidModelTestCase) ---------------------------------------------------------------------- Traceback (most recent call last): File "./runtests.py", line 67, in runTest self.fail('Unable to load invalid model module') AssertionError: Unable to load invalid model module ---------------------------------------------------------------------- Ran 11 tests in 0.391s FAILED (failures=1) }}} Note: If you're seeing something else, then try removing all .pyc files in your tree first. nobody uzi   0 0 0 0 0 0
7828 2008-07-19 02:17:54 2008-07-19 02:27:48 2022-03-06 03:40:39.644305 Unreviewed closed Uncategorized     dev fixed Unnecessary import statements in admin/widgets.py   nobody julien   0 0 0 0 0 0
7829 2008-07-19 02:29:42 2011-09-28 16:12:16 2022-03-06 03:40:39.842535 Accepted closed Documentation     dev fixed [djangoproject.com] The admin doc page is not linked to from the main doc page. The admin page has it's own documentation page (http://www.djangoproject.com/documentation/admin/), but a link to that page is nowhere to be found in the main doc page(http://www.djangoproject.com/documentation/). I think the reasons for having it linked are pretty obvious. nobody ElliottM doc admin djangoproject.com website 0 0 0 0 0 0
7830 2008-07-19 05:58:45 2011-09-28 16:12:16 2022-03-06 03:40:40.012862 Accepted closed Uncategorized     dev fixed removal of some deprecated features for 1.0 Attached is a patch that removes: * "simple" cache backend * `ObjectPaginator` * `edit_inline_type` argument for `ForeignKey` fields * QOperator, QNot, QAnd and QOr * maxlength argument Along with mentions of them in the docs. Note, the following should also be removed (if applying the patch doesn't do it): * django/utils/maxlength.py * tests/regressiontests/maxlength/   gwilson   0 0 0 0 0 0
7831 2008-07-19 07:55:10 2011-09-28 16:12:17 2022-03-06 03:40:40.217468 Accepted closed Internationalization     dev fixed get_language_from_request should be case-insensitive while matching settings.LANGUAGES HTTP_ACCEPT_LANGUAGE sent by browser will be different from its vendor or OS, for example: In XP, IE(6 or 7) will send '''zh-tw''', in Vista it will be '''zh-TW''' But firefox2 in some machine will send '''zh-tw''' or '''zh-TW''', In firefox3, it will send '''zh-tw''' always. (In my machine). This will cause this line in trans_real.py failed: {{{ if lang not in supported: continue }}} It should compare it with case-insensitive way to fulfill all situations. Thanks. nobody bear330 locale, language, case, middleware 0 1 0 0 0 0
7832 2008-07-19 10:03:22 2011-09-28 16:12:16 2022-03-06 03:40:40.374058 Unreviewed closed Documentation     dev fixed documentaion syntax error in newforms documentation In the Formsets section, there appears a syntax error message: {{{ System Message: WARNING/2 (<string>, line 2198); backlink Inline literal start-string without end-string. }}} screen capture is attached. nobody omat@gezgin.com   0 0 0 0 0 0
7833 2008-07-19 11:03:20 2012-08-04 13:42:16 2022-03-06 03:40:40.536450 Accepted closed contrib.auth Bug Normal 1.2 fixed UserCreationForm doesn't handle the case where password1 is not set The new UserCreationForm does not check wheter password1 is available in cleaned_data when checking it against the second password. It can be empty when the form is submitted and the user leaves this field blank. The patch has a test and a fix for this issue. Claude Paroz <claude@2xlibre.net> jvloothuis   0 1 1 0 0 0
7836 2008-07-19 15:02:20 2016-04-08 16:08:53 2022-03-06 03:40:41.095429 Accepted closed Testing framework Bug Normal dev fixed django.test.client._session should check for session middleware instead The django test client class allows you to use `self.client.session` only if you have `'django.contrib.sessions'` in your `INSTALLED_APPS` setting. That is, it assumes you are using sessions if and only if you have the sessions app installed. This assumption breaks down when you are using the sessions middleware independently of the sessions app, for instance if you are using a cache backend for sessions. I propose that `django.test.client._session` check for the sessions middleware class in `MIDDLEWARE_CLASSES` instead of the sessions app in `INSTALLED_APPS` to make it easier for people using the session middleware independently of the sessions app to use `self.client.sesssion` in tests. nobody trevor   0 0 0 0 0 0
7837 2008-07-19 15:18:23 2014-06-08 05:29:47 2022-03-06 03:40:41.273035 Accepted closed Forms Bug Normal dev fixed Hierarchy in forms metaclasses can miss declarated fields in forms This was occurs after [7847]. Ok, [7847] is a good commit for make able to metaclass subclassing. This is the scenary: {{{ #!python #### model class FooModel(models.Model): name = models.CharField(max_length=100) #### forms class GenericForm(forms.Form): extra_field = forms.CharField() class FooForm(GenericForm, forms.ModelForm): # to avoid python metaclass conflict error: __metaclass__ = type('FooFormMetaclass', (GenericForm.__metaclass__, forms.ModelForm.__metaclass__), {}) non_existing_field = forms.CharField() class Meta: model = FooModel }}} In previous scenario, {{{FooForm}}} should contain both {{{name}}}, {{{extra_field}}} and {{{nonexisting_field}}}. But only appears {{{name}}} field in {{{base_fields}}}. See this test: {{{ #!python >>> from testapp.forms import FooForm >>> FooForm.base_fields.keys() ['name'] }}} I will attach a patch thats fixes problem. msaelices msaelices   0 1 1 0 0 0
7838 2008-07-19 15:25:58 2008-07-19 15:51:48 2022-03-06 03:40:41.461382 Unreviewed closed Documentation     dev fixed Type inconsistency in part 2 of the tutorial In the "''Customize the admin form''" section, we have the following code snippet: {{{ #!python class PollAdmin(admin.ModelAdmin): fieldsets = [ (None, {'fields': ['question']}), ('Date information', {'fields': ['pub_date'], 'classes': 'pub_date'}), ] }}} But that should be: {{{ #!python class PollAdmin(admin.ModelAdmin): fieldsets = [ (None, {'fields': ['question']}), ('Date information', {'fields': ['pub_date'], 'classes': ['collapse']}), ] }}} There are two issues here with the {{{classes}}} attribute: * the argument should be {{{collapse}}}, not {{{pub_date}}} * and should be an iterable sequence, not a string as in the admin is then rendered as "{{{c o l l a p s e}}}" (iterating over the chars) Now it's not clear what should be the right behaviour: accept both iterable sequences and strings (by "special casing" them, a {{{if type(sequence) == str: sequence = [sequence]}}} should do the work), or raising some error if the wrong type is passed. The same error is in the second example of the "''Adding related objects''" section. nobody kratorius doc documentation newforms-admin nfa 0 1 0 0 0 0
7839 2008-07-19 15:29:10 2008-07-19 15:37:28 2022-03-06 03:40:41.649104 Unreviewed closed Documentation     dev fixed Error in admin documentation and tutorial 2 - 'classes' should be a list The documentation claimes: {{{ 'classes' A string containing extra CSS classes to apply to the fieldset. Apply multiple classes by separating them with spaces. }}} This is false. 'classes' is expected to be a list or tuple. The example in admin.txt is correct, the code in tutorial 2 is wrong. nobody Jan Rademaker <j.rademaker@gmail.com>   0 0 0 0 0 0
7841 2008-07-19 20:21:41 2011-09-28 16:12:16 2022-03-06 03:40:41.987260 Unreviewed closed Documentation     dev fixed A typo in newforms docs, {% form %} is a not needed line I will attached with typo msaelices msaelices   0 1 0 0 0 0
7842 2008-07-19 20:21:54 2008-07-20 18:05:12 2022-03-06 03:40:42.146895 Unreviewed closed Core (Management commands)     dev fixed raw string identifier missing from django-admin.py After creating a new project with django-admin.py the urls.py is missing a 'r' from the `('^admin/(.*)', admin.site.root)` line. Tried it on r7997. nobody yoshi   0 0 0 0 0 0
7843 2008-07-19 21:12:56 2008-07-19 22:26:33 2022-03-06 03:40:42.307135 Unreviewed closed Database layer (models, ORM)     dev fixed r7986 breaks uploaded files stored in date-based directories After updating to a Django revision post- [7986], get_FIELD_url is broken for all FileField and ImageField uploads using an upload_to with date-based directory parts. [7986] updates get_FIELD_url to call field.get_filename on the DB-stored path. field.get_filename only uses the basename of its argument, and calls field.get_directory_name for the directory part. field.get_directory_name uses datetime.now() for interpolating the date parts into upload_to. From django/db/models/fields/__init__.py: {{{ def get_directory_name(self): return os.path.normpath(force_unicode(datetime.datetime.now().strftime( def get_filename(self, filename): from django.utils.text import get_valid_filename f = os.path.join(self.get_directory_name(), get_valid_filename(os.path. return os.path.normpath(f) }}} This means that instead of returning a URL with the correct path that is stored in the DB, get_FIELD_url now returns a URL that always has the current date interpolated into it. For any file uploaded before today, that will be a broken link. This is a showstopper bug; please revert [7986] as quickly as possible. nobody carljm   0 0 0 0 0 0
7844 2008-07-19 22:07:27 2011-11-04 11:52:10 2022-03-06 03:40:42.470347 Accepted closed Core (Other) Bug Normal dev fixed Very small race condition in file field saving Noticed this whilst reading `Model._save_FIELD_file()` for another reason. The directory name used to save the file can be sensitive to the current datetime value, since `upload_to` directories can contain format strings. This directory name is computed a couple of times in this function, from what I can work out: once when creating the directories at the top of the function and once when call `field.get_filename()`. It's possible that the time delay between creating the directory and using the directory lower down could be sufficiently long that the name actually changes. For example, using day-based formats and then clock crosses midnight. The directory name should be the same throughout the code. nobody mtredinnick   0 0 0 0 0 0
7845 2008-07-19 22:36:31 2011-09-28 16:12:17 2022-03-06 03:40:42.646548 Design decision needed closed Core (Management commands)     dev fixed ./manage.py reset fails if app does not define models ./manage.py reset fails with an obscure IndexError if an app does not define models. There seems to be no strong reason why it should fail though, and in fact the problem is a line in ''core/management/sql.py'' that doesn't seem to have any effect at all anyway (see patch). nobody miracle2k   0 1 0 0 0 0
7821 2008-07-19 00:04:54 2008-08-11 10:25:34 2022-03-06 03:40:38.427787 Ready for checkin closed Core (Other)     dev invalid Add exception value to NoReverseMatch exception The NoReverseMatch exception in django.core.urlresolvers.reverse has no string value associated with it. I added a string value to make it display the value of lookup_view, it makes it a little easier to debug URL issues. I've attached a patch, just a quick one-liner. nobody jmohr@bytepulse.net   0 1 0 0 0 0
7846 2008-07-19 22:50:16 2008-07-20 05:18:11 2022-03-06 03:40:42.843357 Unreviewed closed Uncategorized     dev invalid [Cleanup] Unused imported modules in contrib apps I removed all imported statements in contrib apps for modules that weren't used. nobody julien   0 1 0 0 0 0
7840 2008-07-19 20:19:57 2008-07-19 20:33:53 2022-03-06 03:40:41.814671 Unreviewed closed Core (Other)     dev wontfix 'get_list_or_404' does not know the slicing To make a story short, see this code. I have model, Document, it contains 30 objects. >>> get_list_or_404(Document) It returns entire objects of Document, but in this case, >>> get_list_or_404(Document)[:10] Not like a general usage of queryset filtering, get_list_or_404 does not handle the slicing, it just return the all the objects of model. To handle the slicing and return Http404 if there are not objects, wrote new sub-class, 'WrapperQueryset'. See the patched code. nobody spike get_list_or_404 0 1 0 0 0 0

Advanced export

JSON shape: default, array, newline-delimited, object

CSV options: 

CREATE TABLE tickets (
        id int primary key,
        created datetime,
        changetime datetime,
        last_pulled_from_trac datetime,
        stage text,
        status text,
        component text,
        type text,
        severity text,
        version text,
        resolution text,
        summary text,
        description text,
        owner text,
        reporter text,
        keywords text,
        easy boolean,
        has_patch boolean,
        needs_better_patch boolean,
        needs_tests boolean,
        needs_docs boolean,
        ui_ux boolean
    );