We'll buy it if matches what we have watched our own system monitors show us about what the browsers are doing with CPU utilization and RAM use over years of using Firefox and Opera.
I might add that, as with any experiment, if the "unknown dude" presents the steps he or she took to do the experiment, and they are reproducible, one can then check to see if the same results are obtained. I'll admit I'm not going to do the experiment, but from a quick skim of the webpage, it looks like enough information is given to copy the experiment. What's your problem? Do you only accept a finding if it's done by a "celebrity", rather than by looking at what was done?
You must be using your browser very gently. Have many tabs/windows open, don't control when your JavaScript and/or Flash/Shockwave are running, don't pay attention when a CPU maxes out, and the browsers crash (and race) a lot.
(I also agree with your other responder, as I don't like what the FF developers are or have been doing with location bar, bookmarks, etc. - I don't use them much now, either.)
I'm not disagreeing with the main point of your comment, but
Crash to Desktops are pretty rare these days, and performance isn't actually all that terrible.
What browser are you using? Crashes are common - say once a week at a minimum. Firefox (along with Opera, and sometimes multimedia players and Java) is an application that in my experience actually makes Linux less stable than modern Windows, because it and the other programs I mention somehow get some hook into X, and when they go, X is badly affected as well. (At its worst, I've had it bring the Linux kernel to a grinding halt - a reproducible and infuriating bug two or three Ubuntus back.)
Performance has improved tremendously from FF 3 to 3.5, and over the last few releases of Opera - but it's still nowhere as good as it should be - open a bunch of tabs, and after a bit you'll be maxing out at least one core.
Seriously.
Hmm.... I wonder how many watt/h of energy Flash is wasting in total over the world, CPU's going up to max when could be in low power mode, forcing us to buy a faster CPU with multi cores just to browse the web. I think Adobe should pay some sort of energy tax
And you can take the
An excellent point to consider. I routinely have 2 to 3 times as many tabs and windows open as that, and JavaScript and Shockwave often bring one processor to 100% if not watched. I now have to as a regular thing have a system monitor open to watch the CPU usage for Firefox (and Opera and any other browser I have open). There's even something else in Firefox that runs on and on, as I've recently have a few webpages max out one core using Firefox 3.5 on 64-bit Linux, where Shockwave ads supposedly won't run, with the JavaScript off and yet it races. The couple times I've tried Chrome I've watched it use more and more of the machine over time.
I'm all for Firefox going to multi-process, but they really need to work on something that limits the browser processing - if I'm not looking at a page that page should basically be "off", not doing anything until I come back to that tab, as is the case with a text processor or even an audio or video player when it's paused. How can a browser be the OS when it cannot control its CPU usage or match such to the tasks the user is doing actively? Pages not being looked at must go into the background as far as CPU utilization is concerned unless a job is specifically being run by the user.
It is easier to change the specification to fit the program than vice versa.