The block above this line fetches the latest matching version, so it's
pointless to then use that to get the latest version's URL.

Instead, we get the URL of the object we already have.

This will allow us to add custom queryset methods in an upcoming commit.

Django models shouldn't know anything about URLs. This is a step in the
direction of removing `get_absolute_url` from the models.

Before this change, we used a model helper method to fetch the URL of
the latest version of a matching class.

Now we fetch the latest version, and then get its URL.

Part of the process of removing the models is ensuring that we don't
rely on any methods that live on them.

This change removes most references to `get_absolute_url` on `Klass`,
leaving behind those in the templates.

We're in the process of moving logic off the models.
This model method was only used by one view.

This moves the logic from the model to the view.

We were previously manually doing what defaultdict does for us.

Before this change, we didn't have a test which covered rendering
overridden attributes.

This change adds UpdateView to the snapshot tests, to catch regressions.

We previously tested against a couple of simple artificial examples,
but I didn't trust it to cover all the possible cases.

This change ensures that we're testing against a complicated real-world
example, UpdateView, which I believe to represent sufficient complexity.