搜尋 Mozilla 技術支援網站

防止技術支援詐騙。我們絕對不會要求您撥打電話或發送簡訊,或是提供個人資訊。請用「回報濫用」功能回報可疑的行為。

了解更多

Current performance in blocking phishing sites?

more options

In a 2017 test, Firefox performed much worse than Chrome and especially Edge when it came to blocking phishing sites. See https://www.techradar.com/news/microsoft-edge-is-the-best-browser-for-blocking-phishing-websites. How is the current performance? Have more/other third-party tests been done?

In a 2017 test, Firefox performed much worse than Chrome and especially Edge when it came to blocking phishing sites. See https://www.techradar.com/news/microsoft-edge-is-the-best-browser-for-blocking-phishing-websites. How is the current performance? Have more/other third-party tests been done?

由 FrankC 於 修改

所有回覆 (2)

more options

There's a 2018 test, which costs 4500 USD on the NSS Labs site, but you can also download it for free from Microsoft (It states "This report is Confidential and is expressly limited to NSS Labs’ licensed users." I'll leave it up to the reader to draw conclusions about NSS's relation to Microsoft!)

In it, you'll see Firefox and Chrome give very similar results, which makes sense as we both use the same database - provided by Google.

I'm not sure why the 2017 one might have had a large difference, but you can see the development of the feature in this bug: https://bugzilla.mozilla.org/show_bug.cgi?id=662819 and there was quite a bit of activity around that time, so they might have tested Firefox just as we were re-aligning with Google's changes. That's just one hypothesis, though. The 2017 report actually alternately shows Firefox beating Chrome (so quite the opposite of the headline!) but also the reverse, by large margins, which doesn't really make a lot of sense.

In the 2018 tests the differences aren't really meaningful between all browsers, i.e. 99.9% vs 99.9% in one test and 96% vs 98% in the other. I didn't want to fill in a bunch of forms and give them my email just to get their methodology, but last time I looked at the test in detail, it favored giving false positives over accuracy, so that's another thing to factor in when comparing "detection rates".

Unfortunately I don't know of any real third party testing of these features, so despite the known issues, the NSS report gets very widely parroted.

The good news is that, according to NSS at least, all browsers are now very good at this.

more options