rugk wrote:It is closed, because they say it is a bug in NoScript, which I can understand…
Hmm, yeah, but it is more surprising that this works at all. Because should not WebExtensions have no access to about:newtab as it is a Firefox-internal site? At least it looks as if they do not have, as all add-ons do not display any information about the site. Same with NoScript. It is labelled "NoScript" there and clicking on it takes you to the settings…
So why does the XSS detection would work there?
I very much think it is entirely ok for no script to be able to block XSS calls that Mozilla in their god given freedom decided to push on us.
If something talks to the web without explicetly being asked to, I would like to be able to be notified about that and block it.
Why would Mizillas new tab page be different?
Again, I still don't understand why you are talking "addon" and not "webpage". It's prefetching a thumbnail WITHOUT ASKING YOU from a webpage that you have visited (in you case COINCIDENTALLY the webpage that you got an addon from, but for others just SOME webpage they visited, among them some news site, or a gamer site they visited).
If you have visited a questionable site, that may cause someone to abuse this functionality to trigger something you might not want, and this a warning is entirely warranted.
What is your argument that these CSS shenanigans should NOT be monitored by a tool that promises to do exactly that?
Where is the bug here? And they closed it because you phrased it entirely wrong. The correct bug would be "FF initiates XSS calls from new "newtab" functionality" and initiates these in the background even when not explicitly calling the newtab page.
And how to you call it "FF internal" when it explicitly calls out to webpages. That is the core issue. It isn't INTERNAL if it just calls out to wherever.
Basically if you have "snippets" or "highlight" activated on your newtab page (which is the DEFAULT BEHAVIOUR of FF no less), Mozilla might at an point initiate contact to webpages and load data via XSS.
How is that acceptable to the point of you insisting on calling monitoring and blocking them "a bug".