Windows7之家（www.orchn.com）：质疑InfoWorld的Windows 7评测

 我们知道自从Windows 7有版本泄露开始，已经有不少人对其进行了测试，而其中比较著名的就是InfoWorld的测试了

*第一点我要质疑的就是测试中所用的工具-DMS Clarity Studio
www.orchn.com

附部分原文：

    * My first issue is with Kennedy’s choice of benchmark tool - DMS Clarity Studio. Bottom line, I have trouble accepting this tool as a valid benchmark. OK, as part of a comprehensive roundup, this tool does provide a useful data point, but when used alone I view it with suspicion. There are far more established benchmark suites that can be used.
    * I don’t understand Kennedy’s obsession with thread count. There’s very little that you can take away from totting up the OS thread count, and to jump to the conclusion that more threads is bad is naive. You might as well throw a cup of tea into the PC and read the leaves.
    * Equally, looking at process lists to try to determine how much the OS has changed is pointless.
    * RAM consumption is also a weak metric. It would be better to benchmark the same system with different amounts of memory installed. That tells you far more about how the OS uses the available RAM.
    * The compatibility issues that are raised in the article are interesting, but it’s important to remember that Windows 7 is in a pre-beta state and that apps like Skype and Daemon Tools will probably be updated many times before Windows 7 goes RTM. However, application compatibility is something that Microsoft needs to keep a very close eye on.
    * From the comments it seems that Kennedy is testing build 6801 of Windows 7. This is a pretty old build. Once a new build is released any data you collected from old builds is instantly obsolete.
....
Windows7之家（www.Win7china.com），爱上网，爱上www.Win7china.com

Windows 7现在还只是pre-beta阶段，而在Windows 7 RTM之前，像Skype和Daemon tools这样的软件还不知道要更新多少次

*公平地说，仅仅从进程列表中来决定系统究竟做了哪些改变显然是不可取的

*内存占用测试也并不可靠，以不同大小的内存来测试同一款系统的内存占用显然要更好

*对于测试中线程数的计算方式我很是不理解，本来从操作系统中统计线程数就并非易事，而随后得出的结论就更加幼稚了

*最后，测试中用的是Windows 7 Build 6801，这版本已经很老了

*至于测试中提及的兼容性问题那就更滑稽了
一旦某个新版本出现了，之前你从旧版上收集的数据也就失效了
不过，兼容性确实是微软必须重点关注的问题之一
事实上，我并不认为这是一款比较实用的测试工具，虽然它也能提供不少详细的数据，但作为单独的测试软件，我就对其测试结果有些怀疑了
而国外某IT写手显然对该评测不太赞同并提出了几点异议
除此之外，有太多可用的测试软件了