> [...]

> Jumping in a little late here, but the things I remember thinking about
> (and discussing briefly with Adrian when he described his plan to me) is
> that app_label collisions are often only the tip of the iceberg in some
> respect. They also tend to be an indicator of upcoming template
> directory name collisions -- for example if you use the
> templates/<app_name>/* convention -- and custom templatetag name
> collisions.

> None of those problems are in any way insurmountable, but we need to
> think through what will work practically for those situations. For
> example, the template app_loader, if it remains, should not use the
> "label", in Adrian's scheme, to load the template because that would
> mean an application cannot load its own templates (it won't know what
> its own label will be because that is set by the user at installation
> time and unknown to the app writer). I think we're probably fine here:
> app_loader continues to just merge all template directories at the top
> level and application writers still strive for unique names for their
> template directories (or stop using app_loader, which is something I've
> been tending towards more and more lately).