Dat minuten om te renderen bedoelde ik de browser mee die het ACID3 team zelf gebouwd heeft om te testen of de test klopt. Deze is natuurlijk niet zo gemaakt om snel te zijn, maar alleen om correct te zijn, maar het gaf wel de complexiteit van de test aan (vond ik). Misschien had ik dit iets duidelijker moeten zeggen.
Waarom de ACID3 test niet zo nuttig is als misschien lijkt:
The Acid3 test is a bit of a false positive; as has been noted elsewhere, the WebKit team fixed certain parts of the SMIL standard which were just enough to pass Acid3 but not enough to be useful (that said, they are working on the rest).
While I love the fact we now have Web Fonts, the fixes aren’t of much use to me day-to-day until ALL browsers implement them; that’s the point of a standard, isn’t it?
I think that it’s a good thing that aiming to pass the test causes competition between browsers, but the actual passing of the test is less important.
En een songbird gebruiker (dev?) aan het woord:
quote: Andrew Luecke
Firstly, Songbird is based on the Firefox gecko engine which does not achieve 100/100 yet. When firefox achieves it, the code can be migrated to songbird and we will achieve it..
However, there is a few important things you should know about the ACID testing:
1) A lot of the ACID testing is about handling dodgy code. They are designed with MANY intentional and serious errors which are put in to simply ensure that browsers can handle errors well. That's ok, but developers shouldn't be intentionally putting such errors in their code anyway. So some major aspects of testing are nice, but doesn't say much about the browser
2) ACID1 and ACID2 were based on finalised web standards and so were genuinely useful, ACID3 is actually testing many DRAFT standards. These are standards which developers should not be using at all yet because they aren't finalised. In the past, everyone was rushing to support draft standards, and the end result was a dozen browsers which interpreted things in a different way. It caused massive compatibility problems in some cases. The worst part is that future versions of the browsers then often needed to keep the code to handle the dodgy standard to maintain compatibility
3) Furthermore, by promoting draft standards, it makes it more difficult for the standards to be revised and improved, because users will complain that the standard has changed and their precious browser doesn't pass.
4) One of the tests CANNOT be reliably passed. Its a performance test. Put any well multithreaded browser on a computer with a high clock rate, it will pass it. But put even safari on a 486, it fails. To me, it doesn't make sense to be proudly parading around a test that is dependant on hardware.
5) Finally, it barely tests anything. Believe me, it really is an incredibly small subset of tests. Its more important to chase bugs causing issues in finalised standards, then chase draft standards that can change overnight, and shouldn't be used yet anyway.
And yes, there are plenty of bugs in browsers handling finalised standards still! So if someone works on ACID3 support for Firefox, be aware it will mainly be to quieten Apple. But the tests at this time are useless, and there are so many more important things that should be fixed.
Whilst completion of projects like Mozilla Jetpack (which allows extensions to be added and removed without restarting) could benefit everyone (especially songbird), the reality is that any web developer who is using CSS3 and such already (which is probably tested in ACID3) on production sites, should be ashamed of themselves
Just my opinion.. I'm not an employee of POTI either, so not sure its shared by them. But just be aware that ACID3 is a gimmicky test that has been blown out of proportion by the web crowd somewhat. ACID2 was useful, but ACID3's usefulness is highly questionable.
Regards,
Andrew
[Reactie gewijzigd door roy-t op 22 juli 2024 21:37]