BrutalDrew said:
I am not thread crapping. I am giving the guy good advice by telling him not to even bother.
Indeed.
Over the past several days you have been quite active debunking the value of disabling services and have provided some documentation to back yourself up.
I applaud the effort.
However, two thoughts have popped up I'd like to share.
First, how certain are you that the tests you are running would even find a performance increase (assuming there is one to be found) attributable to tweaking services? I am not a Windows expert so I have no alternatives to suggest, but you seem extremely confident that the (relatively) small battery of tests (and again, I applaud the effort...I know how time consuming that must have been) you ran completely proved your point.
Second, there is the niggling problem of the "It
feels faster" phenomonon. Certainly, much of this can be dismissed as "I just spent a crapload of money/time...this had damn well better
feel faster!", but at least some of this response has to be considered valid.
You can quantify the performance of the machine till doomsday, but the science ends where the PC-human interface begins.
An example:
I can prove beyond doubt that my Raptor RAID array is almost twice as fast as a single drive, but it doesn't "feel" like it.
The Storage section of the forum is littered with "I just installed Raptors and they aren't any faster...WTF!" threads.
Ignoring the price/performance analysis (irrelevant since we are only concerned with quantifiable speed numbers), the only possible response to the question "What is the best storage solution?" would have to be "Raptors in RAID0", cause that's what the numbers say.
The point being...is it fair, or even scientific, to dismiss the "feeling" factor just because the tests at hand don't/can't/won't put hard numbers to it?
By the way-
Your link to Anandtech leads to FreshDaemon's original post which was cut/pasted on this forum also.
Does posting the same assertion on two forums make it twice as true?
Being goodhearted, I will assume that there is not a specific "anti-BV" agenda in place (although GreNM's post in your second link begs the question a bit) so I wonder if there are plans afoot to continue the testing on a modern machine with an expanded battery of tests?
That would be very interesting.