ikeyan

> It could be implemented for `strlen()` and `mb_strlen()` for strings too. No. `'0'` is also evaluated as false.

https://github.com/FiloSottile/mkcert/blob/0a3190b1659e514d6e9b03eedfa25049d046000b/cert.go#L59-L62 expiration period is hardcoded here, and the comment tells the reason of the value.

if this is a feature request, maybe duplicate of https://github.com/FiloSottile/mkcert/issues/339.

@trusktr The problem is, > callbacks (and records) for that target and > future callbacks and records in place for other targets can have nodes in common. For example, if...

1. disconnect() semantics: the method should discard all mutations in the queue (@ArkadiuszMichalski mentioned [here](https://github.com/whatwg/dom/issues/126#issuecomment-167374814)) * I think this is error prone and surprising behavior

2. unobserve() with queue filtering semantics: the method should behave as if there was no observe() call(s) on the node. (@trusktr mentioned [here](https://github.com/whatwg/dom/issues/126#issuecomment-318555824)) * mutations in the queue should be...

3. unobserve() semantics without queue filtering: same as [3.](https://github.com/whatwg/dom/issues/126#issuecomment-494597985) but no queue filtering ( @justinfagnani mentioned [here](https://github.com/whatwg/dom/issues/126#issuecomment-228455689), @smaug---- mentioned [here](https://github.com/whatwg/dom/issues/126#issuecomment-167597753)) * mutations already in the queue are left as they...

So I think [unobserve() semantics without queue filtering](https://github.com/whatwg/dom/issues/126#issuecomment-494605671) is the most promising behavior to implement.

@trusktr So, the records should know which nodes they are for? Then, I also think it's a sound and simpler behavior. ```js // node A is a descendant of node...

The page says "You have agreed to the CLA for uber/nebula.gl", but even after clicking `recheck` CLAassistant won't recognize it