ike709
ike709
`#pragma UnsafeClientAccess disabled // NOTE: Only checks for unsafe accesses like "client.foobar" and doesn't consider if the client was already null-checked earlier in the proc`
Why would we close an issue because of a dead PR?
Unit tests plz
please god merge this
This was fixed for TG but Paradise must handle it differently because it's still an issue there.
Paradise seems to handle it via cookies: ```js $('#toggleDarkChat').click(function(e) { internalOutput('Dark Chat toggled. Reconnecting to chat.', 'internal'); var backlog = $messages.html() if(getCookie('darkChat') == "on"){ setCookie('darkChat', "off", 365) } else {...
This seems to be an issue with us canceling the reload if the paths are identical, but _not_ canceling identical paths causes goonchat to have a seizure. I'm not sure...
This _may_ (I haven't checked) be related to the cache relying on IP address instead of hashing, since local dev servers all have the same IP (`127.0.0.1`) regardless of codebase....
Actually, clearing the cache doesn't seem to fix this, so it must be something else.
Isn't this not fixed until we actually bump RT again?