| Age | Commit message (Collapse) | Author |
|
|
|
perf testing shows that in chrome this change adds 5-15% overhead
when destroying 10k nested scopes where each scope has a $destroy listener
|
|
|
|
Closes #793
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
#feature
- ngForm directive can now be used with element, class, and attributes
|
|
- compiler would rewrite boolean attributes on all elements. This is too aggressive and interferes with some third-party frameworks
|
|
- increase ng-non-bindable prioirity to 1000 so that it prevents attribute interpolation on same level.
|
|
- The validationErrorKeys are now published as CSS for easy styling. The errorKeys should be in
camelCase and the CSS will be in snake-case
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
the forms/controls code refactored not to depend on events which forced
us to create new scope for each form element.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This reverts commit 8fb34f008eca4d1e15e6fa63902aaa36d199c4af.
|
|
|
|
|
|
so that we can render better usage examples in docs
|
|
the original code magically worked for ng:foo but for nothing else
|
|
we can't provide this functionality because the directives are lazy
loaded when the module loads, which is too late for the shiv to do
anything useful.
|
|
|
|
|
|
It registers a provider class, so this makes more sense.
Breaks Rename $provide.service -> $provide.provider
|
|
Each widget (ng-model directive) emits $newFormControl event instead of getting hold of parent form
and calling form.registerWidget(this);
|
|
|