Notes |
|
|
I can confirm this, and it looks like it's a case of the old Window compartment not closing down properly:
(1.020374) build/Linux-gtk/duktape/window.c:189 window_alloc_new_callback: Scheduled callback 0 for 5000 ms from now
(2.716481) content/handlers/javascript/duktape/dukky.c:611 js_newcompartment: Old compartment present, closing it down
(2.722730) content/handlers/javascript/duktape/dukky.c:615 js_newcompartment: Window hopefully released
(2.722994) content/handlers/javascript/duktape/dukky.c:619 js_newcompartment: New javascript/duktape compartment, win_priv=0x55bf7a143360, doc_priv=0x55bf7a4ad8e0
Note how the callback isn't cancelled during the closedown of the compartment.
I shall investigate |
|
|
|
This was an interesting problem where the JS compartment didn't close down properly when being replaced. The underlying issue (that the Window object doesn't get finalised) is still there, but from #4646 I have put a workaround in place which resolves this particular issue (an unexpected callback firing because the Window hadn't cleaned up)
|
|
|
|
we believe this issue has been resolved in NetSurf 3.9 |
|