How do I compare incomplete tests?

[Anonymous]

I have a bunch of tests I'd like to compare, but not all have the same things. For example, I don't always run the CD test because I could care less about CD performance and don't always have a CD handy. I just bought a machine for a server which failed the 3D test. I could care less because there won't even be a monitor hooked up to it most of the time, but it means that I can't easily compare the overall speed of my various machines. And, the only time I'm ever going to stick a CD in it is when I'm installing new software (and I think I'm done with that already). Is there a way to have PerformanceTest give me the total rating while ignoring those tests which were incomplete for one or more of the machines I'm comparing at the moment?

[David (PassMark)]

There are individual ratings for each test suite. So you should ignore the total result and just concentrate on the CPU Mark, Memory Mark and Disk Mark results.

You could even make up your own rating system by adding up the individual results.

---------
David
Passmark

[Anonymous]

<<There are individual ratings for each test suite. So you should ignore the total result and just concentrate on the CPU Mark, Memory Mark and Disk Mark results.

You could even make up your own rating system by adding up the individual results. >>

Yeah, that's pretty obvious, but I have results from a whole bunch of machines. Copying it all to Excel and making my own spreadsheet seems pretty clunky and I bought PerformanceTest so that it would make it easier for me to comapre machines. It seems to me that what you currently do with missing tests (giving them a score of 0) is never meaningful. You compare apples and oranges as if they were both apples. At least providing an option to ignore the missing tests would make it easier (and, if you do this, don't call it a PassMark score -- express it as a % comparison or something to avoid confusion).

Can you put it on the list for version 6?

[Anonymous]

Another example: As I stated earlier, I never test the CD. Well, I did on one machine, a clunker, and every time I compare machines, that machine always looks better than it really is because it gets a bonus. In essence, giving unrun tests a score of 0 and then calling the resultant score a "PassMark score" is inaccurate.