[Greasemonkey] "Invalid document,
could not get global object" error?
Aaron Boodman
zboogs at gmail.com
Sun Aug 28 17:16:27 EDT 2005
No, 0.3.x was pretty braindead simple as far as implementation. The
basic problem is that the only available "right" way for greasemonkey
to run in in 1.0.x is really really slow. Too slow to really be
useful.
So instead, Greasemonkey tries to do this really complex thing where
it gets a reference to these methods it needs before they can have
been tampered with. But it turns out that it's pretty brittle.
:: sigh ::
--
Aaron
On 8/28/05, Michael Kraft <mjk4984 at yahoo.com> wrote:
> --- Aaron Boodman <zboogs at gmail.com> wrote:
>
> > In my testing, the first progress change event after document_start
> > had the correct document in place already, but it seems like sometimes
> > document unload is slower than that :-(.
> >
> I don't know why it occurs mostly on this page. The server is sometimes lagged
> down, but even when it isn't it seems to have problems. I've seen that error
> in the log for other pages on occasions.
>
> > You seem proficient with extension dev. If you can come up with a
> > patch that does this (Grabs whatever is in new
> > XPCNativeWrapper(unsafeWindow, "document").document and wait until the
> > first progresschange that it is different, and fixes your problem, and
> > doesn't seem to break anything else, I'll add it.
> >
>
> I supposed I can try though it will take me a while to understand what GM is
> actually doing. As far as I can tell it intercept statechanges on page, but
> trying to follow this in the log is very difficult, especially for pages with
> frames on them since it doesn't log what page initiated what window/frame or
> URL/location caused the event.
>
> Did it used to do all this in 0.3.1 because I never had problems with that version?
> _______________________________________________
> Greasemonkey mailing list
> Greasemonkey at mozdev.org
> http://mozdev.org/mailman/listinfo/greasemonkey
>
More information about the Greasemonkey
mailing list