diff options
| author | Stephen Blott | 2016-04-27 10:26:24 +0100 |
|---|---|---|
| committer | Stephen Blott | 2016-04-27 10:26:24 +0100 |
| commit | ac340e2346247b5cd1878e1a814d4c151df3e892 (patch) | |
| tree | 829bd5b8c2ee27dca752e39ae79b2d4a904993f7 /content_scripts/vimium_frontend.coffee | |
| parent | 637a03c74b56bfb8879049c4e7da90c654276ec0 (diff) | |
| download | vimium-ac340e2346247b5cd1878e1a814d4c151df3e892.tar.bz2 | |
Unregister frame port on onDisconnect.
I seems that we cannot rely upon either the "unload" event in the
content script or the port's onDisconnect handler there to unregister
frames.
To see this, go to a frame-busy page like this one:
- https://www.theguardian.com/technology/2016/apr/26/apple-iphone-first-revenue-decline-13-years
and then navigate to any other simple page (in the same tab). Navigate
back and forward with `H` and `L`.
If you watch frames registering anf unregistering, almost all of the
frames from the frame-busy page do not unregister.
Here, we unregister frames on onDisconnect in the background page too.
It is possible that this is the source of the problem mentioned as point
1 in this comment:
- https://github.com/philc/vimium/issues/2081#issuecomment-210980903
And for which #2116 is a workaround.
Diffstat (limited to 'content_scripts/vimium_frontend.coffee')
0 files changed, 0 insertions, 0 deletions