What Really Slows Windows Down

Glad to see that Norton straightened themselves out a little bit; however, that boot time is still way too high for me to worry about purchasing a copy of their software. I'll stick to "being careful."

In reference to your forthcoming article, i've found the best way to speed windows up, is to un-install xp, install 98, and play the random delete file game from back when I only had a 800mb had drive and had to squeeze windows and half-life both on it! Delete files until something breaks, when it breaks, you know for the next install you cant delete that file.. ive had it around 20ish MB before :)...

Sorry, thats completely irrelevant, but I miss the good old days! Excellent article as always sir, your results proved the rough idea I had in my head of how the security products performed. I have to speed up a lot of Norton PC's for private customers when im freelancing. Ive recently gotten rid of Mcafee Enterprise 8 and replaced it with Avast much to my happyness! For speeding up windows, as well as the service tweaks, ive found TweakNow Powerpack2006 to be relatively useful, makes enough of a difference to notice, including boot times.

Cheers dude!

Really great job on the "what slows windows" down. I've been recently trying to find on my own system what is causing large slow bootup times. I'm thinking that Google Desktop Sidebar / Search may be something interesting to check. ...and if you decide to check out antivirus suites, be sure to take a look at F-secure/Kasperky, recently rated as the "best" detecting antivirus and antispyware suite in its Internet Security offerings.

I would like to see yoru review the free Avira AntiVir PersonalEdition Classic. http://www.free-av.com/

Glad my computer doesn't need any of that crappy antivirus, spyware garbage...

its a Mac, and sweet as a nut.

Why do you guys put up with all that sh*t?

My opinion with watching Norton Internet Security over the last couple of years is that it is getting more and more bloated every release. Every release what they seem to be doing, is adding new features (which is fine by itself), but they add one or two more (new) background services in doing so every time, I am not sure how many services Norton Internet security runs on now, but it must be at least over 6 or 8 or even more, which is ridiculous really.

What I am experiencing in the real world is that a lot of people only have 256Mb of RAM and Norton Internet Security 2006 requires so much RAM that these PC's have no more physical RAM left over (even when no programs are running) and start using the swap file constantly and the PC comes to a crawl, taking Nortons off and using the XP2 firewall + AVG free edition uses much less RAM, runs less processes and the machines become much faster. Norton Internet Security 2006 appears to require at least 512Mb of RAM to prevent the swap file from being used constantly, which is far too much RAM used.

In my mind, what needs to happen with Norton Internet Security, is they need to combine a lot of the tasks into only a few services (i.e. optimize), instead of heaps of separate services/background tasks like it is now, I am sure a lot of RAM can be saved by optimizing sections of the code too. I have looked at PC's with only 256Mb of RAM running Norton Internet Security 2006, and Norton's is actually using MORE ram that Windows XP itself, which is getting ridiculous really, in my mind, Windows XP SP2 firewall + AVG free edition + Firefox to prevent spyware that comes through ActiveX holes is more than enough and doesn't require half as much RAM, but then again I primarily run Ubuntu myself.

Great Article, Was wondering if you could also run some tests on P2P apps, Emule/utorrent and some other media players like PowerDvd,WinDvd,Media Player Classic.

Also some other suggestions on software would be imgburn/dvddecrypter/photoshop/Kaspersky AV/Kerio Firewall/sygate/outpost

also to test it would be cool if a test was conducted which was ran before and after a defrag? That would be aswome

You should not do performance tests in a virtual machine. Main reason is the condition of the host machine at the given moment. You know that that VM is just another application on the system that shares cpu and IO with the main system. True performance tests are done on a real machine where all tests are run on the same restored each time image.
There is one more thing that you should have mentioned in your tests. Do all security suites have a boot time driver that starts scanning when no user mode application has started? I doubt - namely, McAfee and Microsoft both scan in user mode meaning their boot time performance degradation in theory should be 0% but it is not. Norton's driver starts at the system time which means it scans during the boot file IO intensive process. This protects the machine for worms such as FunLove that infect just when the network drivers are up and functioning among other things. I think you should set all software to have the same options and then only run your tests.
Cheers,
Mike

why NAV 2002 and the rest of the progs are new?