http://blog.topix.com/archives/000016.html
Google has 100,000 servers. [nytimes] If a server/disk dies, they leave it dead in the rack, to be reclaimed/replaced later. Hardware failures need to be instantly routed around by software.
文中提到，google用大量的比较“差”到机器而不是用少量的“好”机器。这样做降低了成本。于是在成本相同的情况下他们获得了更高的计算能力–用来实现更好质量的搜索，这需要很高的计算能力。 聪明的google工程师们用软件和程序的方法来进行分布式的计算，即使没有稳定可靠的服务器，也能提供稳定到计算结果–而且更便宜。分布式系统将是未来计算技术发展的方向啊。
http://www.cnblogs.com/wj/archive/2006/04/22/381851.html
楼主的文章确实有问题，prototype也是可以访问私有变量的。不过对于用OO语言的传统思想对这些作用域的分类却是不错的。
just recently I read this article about Scrum adoption in china,which I think is a valuable reference/evaluation for projects being done by China development teams.
http://www.infoq.com/articles/Agile-adoption-study-china
Compared to the benefit/success story of Srcum in China, I am more interested in the failure examples. I think from David and Fiona’s description their failures are both due to Communication problem. Every experienced leader would advocate that “Communication is very important….”,but we need to define communication. it’s not a regular meeting or formal report in some format/template.There is no BEST template for communication.
Communication is all about convey/receive information. If you think someone is not clear about something and he needs to know that, you keep telling him,if you think you don’t understand something and you feel you should know it,then you ask.But many cases are that we don’t have a concrete thing to talk or we don’t feel we really talked about the key issue.In Chinese,”没说到点子上”.
From the aspect of communication , I think the benefit that Scrum can bring about is it greatly increases the volume of information communicated during a project process.
Because with circular sprint,we deliver many versions of product to customers, then we can have much more feedback from them for a long term. Just like when you’re talking to someone about a car,there is always more to talk/hear when you really have a car by your side.