Looks like the other thing they could do for this particular vector is prevent https requests from being intercepted and replaced - why should the user be permitted to override certificate errors in this case? Doesn't make sense.
Looks like the other thing they could do for this particular vector is prevent https requests from being intercepted and replaced - why should the user be permitted to override certificate errors in this case? Doesn't make sense.
Funnily enough, I think the CVE they're saying it is is wrong? The link to the project zero itw tracking page is for CVE-2021-30858 which is an IndexDB UAF, which is a completely seperate bug; the MITRE tracking page for CVE-2021-30858 just has the boilerplate "a malicious page can cause an exploit" from Apple, and doesn't say if it's for the IndexDB or FontFaceCache bug, so I think the Jailbreak people are just confused on this one.
That seems like the wrong main focus for a security project. I'd have preferred if the "most important" thing was to provide a secure browsing experience, but, to each their own...
They are specifically calling out that it's not ready for prime time, and that it will have an impact on users. In practice, many very effective security controls for browsers (and other software) get tuned down or disabled by default because they break too much of how the web works (SameSite cookies are a good example), or because it creates friction in customer acquisition or growth.
I don’t think I’ve ever opened a web browser on a console except on dreamcast for the novelty of it.