lists at atrus.org
Thu Aug 18 23:31:57 EDT 2005
Ted Mielczarek wrote:
>On 8/18/05, Jeremy Dunck <jdunck at gmail.com> wrote:
>>So, um, this is the arms race I was talking about. Aaron says there's
>>no arms race because the web is loosely coupled and we can do whatever
>>we want before rendering, and that's true in a theorhetical sense, but
>>it's outside GM's current capability.
>>Nik says override innerHTML, yeah, that works in this instance but
>>there's always n+1 to worry about.
Indeed. Once you enter unsafeWin, it's the content's ball game until we
run GM before content gets a chance.
>I doubt either of these methods are practical for non-static webpages.
> Anything that's dynamic in any sense is not going to have a single
>static DOM, so it'd be pretty hard to determine whether the page DOM
>is exactly what they're expecting.
>And if static webpages start using them, well, isn't un-screwing the
>web what GM is all about?
I've been thinking about how to do this. With the "protected copy"
approach, you just manipulate the copy ;) For the modification
detection, you can remove and re-add the event listener. Both solutions
use namespaces to keep this out of reach of GM.
More information about the Greasemonkey