I don’t think Acid3 has been officially changed. It just has been abandoned, with no replacement.
Both Acid2 and Acid3 tests have aimed high to force browsers to clean up some “inelegant” behaviors, and unfortunately in both cases this turned out to be a stretch (IIRC Acid2 wanted SGML-compliant parsing of <p><table>, and in retrospect that was a needless breakage). In case of the things that Acid3 tested, the decision was to change the specs to align with browsers and web content, instead of forcing browsers again to match the test.
Note that the Acid tests are not conformance tests. They don’t test that you correctly implement a large set of the standards, they test whether you get certain painful corner cases right. It’s a huge achievement to pass them, but it doesn’t necessarily mean that the browser is usable on a wide variety of sites. I suspect a clean-slate browser from an experienced WebKit developer will find it easier to pass these tests by avoiding the implementation choices that led to the failure modes that these tests were intended to check. That sounds very negative, so I want to reiterate: getting to this point is a massive amount of incredibly impressive work.
Is this browser going to be usable outside of SerenityOS? If it can be separated out as a library, in a few years this could finally be the new browser engine that breaks the WebKit/Gecko duopoly.
I honestly doubt they’ll reach production quality, because that would require a lot more like sandboxing, speed & performance optimization etc. Maybe they’ll reach that, but I doubt they will reach the feature completeness of a chromium or gecko engine.
The whole philosophy of SerenityOS is that everything is developed in tandem and as one cohesive system. To give the Browser a distinct name and brand would be in some sense to depart from this ideal.
Keep in mind that the first site when you google “acid test browser” (and the link on the Wikipedia page) is no longer the right site: http://acid3.acidtests.org/.
The real site is http://wpt.live/acid/acid3/test.html.
Before knowing this I was confused why Chrome on Mac only gets a 97/100. But that’s on the old test.
Does anyone know why they didn’t create a new Acid4 test instead of changing the existing test that was already in widespread use?
I don’t think Acid3 has been officially changed. It just has been abandoned, with no replacement.
Both Acid2 and Acid3 tests have aimed high to force browsers to clean up some “inelegant” behaviors, and unfortunately in both cases this turned out to be a stretch (IIRC Acid2 wanted SGML-compliant parsing of
<p><table>, and in retrospect that was a needless breakage). In case of the things that Acid3 tested, the decision was to change the specs to align with browsers and web content, instead of forcing browsers again to match the test.Note that the Acid tests are not conformance tests. They don’t test that you correctly implement a large set of the standards, they test whether you get certain painful corner cases right. It’s a huge achievement to pass them, but it doesn’t necessarily mean that the browser is usable on a wide variety of sites. I suspect a clean-slate browser from an experienced WebKit developer will find it easier to pass these tests by avoiding the implementation choices that led to the failure modes that these tests were intended to check. That sounds very negative, so I want to reiterate: getting to this point is a massive amount of incredibly impressive work.
Thanks for clarifying that! I have been wondering about how expansive the tests were.
Is this browser going to be usable outside of SerenityOS? If it can be separated out as a library, in a few years this could finally be the new browser engine that breaks the WebKit/Gecko duopoly.
I honestly doubt they’ll reach production quality, because that would require a lot more like sandboxing, speed & performance optimization etc. Maybe they’ll reach that, but I doubt they will reach the feature completeness of a chromium or gecko engine.
I am pretty sure it already includes some sandboxing, it seems to do multiprocess tabs already.
interesting, though what I meant with sandboxing is a little more sophisticated than spawning one worker per process
They also use pledge and unveil to disable system calls and filesystem access in the sandboxes
*one process per tab
You forget Blink? The fork was in 2013, so it’s not exactly news.
Blink/Gecko duopoly then :)
I wouldn’t expect so, that seems to cut against the whole idea of the project.
Maybe it’s time for a SerenityOS tag?
Impressive! I hadn’t thought any new browser engine would ever get this far.
Note that there is a reason the tweet is qualified to new open source browser, not new browser. Ekioh’s Flow Browser already passed Acid3 in 2020.
Honestly, I don’t care about anything that’s not open source.
It appears it’s currently named “SerenityOS Browser”. I think an actual name could go a long way.
Why? “Browser” is the logical next step after “Chrome” :D
I’m positive this is why everyone still says “Epiphany” instead of “(GNOME) Web”
The whole philosophy of SerenityOS is that everything is developed in tandem and as one cohesive system. To give the Browser a distinct name and brand would be in some sense to depart from this ideal.