Is window.onbeforeunload impelementation acceptable?

[andrei0x309]

I made a branch ( https://github.com/andrei0x309/worker-dom/tree/window-on-before-unload ) with window.onbeforeunload implementation and demo, I think it's useful, from what I see onbeforeunload has 3 useful applications:

1. show a loader when navigating to simulate a SPA( works well on modern browsers because they only render elements when they are loaded which means no white pages like in the old days)
2. alerting the user of unsaved data
3. analytics

The only problem I see is that some developers might abuse this feature( since it transfers a function on the main thread), so that's why I am asking if such a feature is acceptable to know if I should write some tests and submit a PR.

Thanks.

[samouri]

Hi @andrei0x309,

This is a very interesting proposal! Also your implementation looks pretty good.

Unfortunately, we do not want to allow for a fully dynamic Function due to the security concerns it comes along with.

[andrei0x309]

Hi @samouri.

Thanks for your feedback, I'll see if I can do it another way.

Maybe you can leave this issue open for a few weeks to see if I can do something.

[andrei0x309]

@samouri

I changed the code now the onbeforeunload user-provided function is executed on the worker thread instead of the main thread, and there's no Function constructor invoked either, the caveat is the worker thread can't directly prevent the navigation so I made a custom API window.onbeforeunloadPreventNavigation( default false, to set if the function should prevent navigation) to support also that user case(personally I don't need to prevent navigation but there may be devs that need that.)

[samouri]

@andrei0x309, I'm out of office for the next 3 weeks. I'll be able to review your updates when I'm back