User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.0; sv-SE; rv:1.9.1) Gecko/20090624 Firefox/3.5 Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.2a1pre) Gecko/20090708 Minefield/3.6a1pre When HTML5 parser is enabled, bug 457809 still shows Reproducible: Always Steps to Reproduce: 1. Go to http://stevesouders.com/efws/stylesheet-before-iframe.php with HTML5 parser disabled, load time will be a little over 4000 ms 2. Enable HTML5 parser and go to url again Actual Results: Load time is over 8000 ms Expected Results: Load time should be approximately the same

Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.2a1pre) Gecko/20090709 Minefield/3.6a1pre For me it's always between 4410 ms and 4620 ms and the value of html5.enable doesn't seem to influence this on my computer.

Seems that the 'reload page', and reload buttons don't work just right, I have to close the tab and then click the link again to start a new test run. HTML5 - On page load is just over 8000ms - 9500ms HTML5 - Off page load is just over 4600ms Loading the elements in the box: HTML5 - On page is just over 4500ms HTML5 - Off page is around 350ms My system specs: DSL 3 meg AMD Phenom x4 quad 2.3ghz 8 gig RAM HD3200 integrated video Win7 x-64 latest hourly: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2a1pre) Gecko/20090709 Minefield/3.6a1pre Firefox/3.0.11 ID:20090709010542

I created new profiles for all tests. Maybe only a problem with some hardware.

Ok, I see, for me the patch hasn't improved anything. Builds from before the fix of bug 457809 give exactly the same results. But wasn't the bug Linux-only?

Testing with a new profile, no addons, and clearing Recent History between test runs is giving inconclusive data. For the most part, it does appear that turning on/off HTML5 parser makes, has no difference in the test page load results. With it off, I've seen much lower numbers on 2 occasions. HTML5 TEST RUNS PAGE LOAD FRAME LOAD OFF 4535 246 8949 4529 8696 4410 4634 334 8621 4334 8712 4409 8666 4341 8751 4456 8753 4453 8682 4323 8640 4333 ON 8981 4448 8613 4332 8600 4318 8689 4332 Looks like it didn't paste the table from Excel, but you get the idea.

(In reply to comment #4) > Ok, I see, for me the patch hasn't improved anything. Builds from before the > fix of bug 457809 give exactly the same results. But wasn't the bug Linux-only? No, it's OS all

I consistently get results in the range 4000-5000 ms with html5 disabled, and 8000-9000 with html5 enabled.

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2a1pre) Gecko/20090708 Minefield/3.6a1pre ID:20090708042452 same as comment #7

Speculative loading of any type has not yet been implemented for the HTML5 parser. It's one of the reasons it's not yet on by default.

Note that the plan for getting this working with the HTML5 parser is bug 482918.

This doesn't affect HTML5 compliance, AFAICT.

Is this a duplicate of bug 482919?