Propagate active property to tail

[TimvdLippe]

This is an update of #125 to incorporate the feedback of @e111077 and to rebase onto current master. As the repository was moved after I made the original PR, I was unable to push to the original PR.

Supersedes #125
Fixes #110
Closes #77

[scbergeron]

Any update on this pull request ?

[RackhamLeNoir]

For the time being, I use a workaround which is fine in my situation. I used to bind the tail data to an iron-pages and the default was set with fallback-selection="home". So I changed all my links /foo to /foo/home. In this case I always overwrite the previous value. Hope it can help others.

[vinerz]

Almost two years later, any news about this?

[e111077]

Hello, we considered this when we created app-route, but initially went against doing it largely because of thrashing the databinding system. Thus, the element is working as intended.

We have since put our elements on feature freeze in an effort to lessen our support surface for our understaffed team, and since this element is working as intended, this PR is considered a feature. I have left this PR open as an easy way for users of this element to find this fork.

[mercmobily]

When you say "working as intended"... can you please expand on it? Does the new binding system that comes with P2 make it possible to have this implemented correctly?

[e111077]

Well it isn't necessarily thrashing, but it would cause the databinding system to fire for an entire subtree of active / inactive elements. App route is already computationally inefficient so we took this definicency into account when creating the element.

[mercmobily]

@e111077 ok but... What is the point of having the tail/active if they are 100% unreliable? Or, is there a recommended set of workarounds/best practices to still use it effectively?

[e111077]

It should be reliable at the level of the app-route tree that was made inactive which should always be up the tree.

With this interaction model, the view handling that newly-deactivated app-route should be in charge of deactivating the view below it.

[e111077]

Perhaps "active" is a bad name for this as it is a declarative representation of a "deactivate" callback

[mercmobily]

Well I guess none of this matters anymore since the two-way binding elements are effectively legacy now... I guess the routing elements will never make it out of beta now!

…

On Wed, 18 Apr 2018 at 23:32, Elliott Marquez ***@***.***> wrote: Perhaps "active" is a bad name for this as it is a declarative representation of a "deactivate" callback — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#176 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACB7XpdC1wrdTKS8L6o1yK0_HMIXhj6-ks5tp1xkgaJpZM4LqE_O> .

[TimvdLippe]

This PR is very old and I am no longer pursuing this change. Therefore I am closing this PR.