First of all, my "new" computer ... it used to be my brother's computer, he got it about 3 years ago. He finally decided to get internet service, but as his antivirus was 3 years out of date and he had no firewall, his computer lasted about a week before the various worms and adware caused it to grind to a halt. Being who he is, he doesn't try to fix stuff - he looked at it as a good excuse to by a new computer. Since I was still using a 6-year-old computer, he gave me his old one for free.
It's a 1.5 GHz P3 system (my old one was a 500 MHz K6-2) with 512 MB RAM (448 on my old one) and nVidia graphics card. It originally had Windows XP, but as I had to format the drive anyway I installed Linux (Mandriva 2006.0 Free). Obviously the results will be at least 3 times as fast, maybe even faster if they involve graphics.
I'll also note that the website I used in my previous test no longer logs previous results, the results listed for "Last results" are dated back in early July.
Anyway ... I also observed that settings for popup and tab handling could change the results dramatically (for test 2), so I now have several sets of results (besides different versions of various browsers).
So here's my first set of results - Opera 9.01, with various settings for tab/popup handling:
popups as windows popups as tabs foreground background maximized 1.702 1.673 1.724 1.696 4.11 5.927 4.271 1.757 1.612 1.793 1.833 1.641 0.704 0.726 0.725 0.673 0.2 0.189 0.205 0.213 4.944 4.879 4.96 4.721 0.844 0.846 0.899 0.855
There is considerable variability in the tests, even if you rerun the same test; I think we can treat all the results except test 2 as the same. Note that using "Open windows instead of tabs" (referred to as SDI mode by us experienced users) is actually somewhat faster here than opening them as tabs - I was surprised by that. Though setting New tabs as "Maximize all, including popups" was the real surprise. (Note that in this test the content of the popups is injected by the script, there is no loading time for any of them.)
Okay, next set of tests ... Firefox has just released 2.0 Beta 1. I had 2.0 Alpha 3 and 18.104.22.168 installed originally, I added the new beta for completeness:
popups as windows popups as tabs 22.214.171.124 2.0 a3 2.0 b1 126.96.36.199 2.0 a3 2.0 b1 7.11 7.021 6.582 7.281 7.192 6.264 19.629 7.893 7.88 4.652 2.543 3.044 3.415 3.019 3.083 3.464 3.097 2.95 5.481 1.384 2.153 5.668 1.36 2.122 0.582 0.566 0.53 0.619 0.561 0.518 13.74 9.45 9.282 13.83 9.435 8.915 1.883 3.123 2.848 1.874 3.118 2.801
Okay, Firefox 2 is better overall than 1.5, and markedly better at opening new windows and at the layers test (where it assembles a sentence out of floating letters). Oh, and the image swapping test. For some reason, the new version is quite a bit worse at date calculations (test 7 computes day of the week for the next 10,000 Christmases). There actually seems to be (at least, as far as I can tell given the variability of the test results) some improvement from Alpha 3 to Beta 1 in the loop test (Test 1, which shows that progress bar as it counts to 10,000) and also in the date calculations, but they actually lost ground on the image swapping test.
Note if you're trying to compare this with the earlier results that Firefox tabs are always maximized, there is no way to display a popup as a floating child window like there is in Opera. So the first 3 columns compare to Opera's first column, the other 3 columns should be compared to Opera's fourth column.
So how's Firefox 2.0 look? If you're a Firefox user, it looks like it is going to be good. Of course none of your extensions are available for the new beta yet, best wait a little while for the extension authors to update them. I don't use Firefox regularly (in case you couldn't guess), I really can't say how stable the new beta is. Still nowhere near as fast as Opera at these tests, seems like Opera is 2-3 times better on pretty much all of them, but we knew that.