起点

Marc: This idea that companies had to build their own data centers and buy or build their own applications -- even though this was what everyone was doing then -- this would go away. And it would create the end of software.
And the idea of the end of software is not that there is no software -- because of course there's some software on our side. We're building that nuclear power plant. But for a customer of ours, they don't have to do that heavy lifting. They don't have to buy the hardware; they don't have to buy the databases and the application servers. They don't have to hire the people to put all that together.

说实话，偶喜欢最后这一段：用户不需要去买硬件，不需要去买数据库和那些应用程序服务器，更不需要雇佣别人来把这些东西装在一起……sigh, 绝对的诱人！只要他们购买了网上服务，这一切都有可能实现！
这让我马上想起了去年用过的一个online版的进度管理工具。很像MS的Project 2000。可以自定义项目起始时间，结束时间，各个任务以及每个成员的工作进展状态。而这一切都是通过B/S结构实现得。对于这个软件用户来说，他再也不用为了维护server的良好运行状态而坐卧不安了……
这就是用户最大的收益点，所以，就成了风险投资商们理所当然的偏爱。

当然，他的理论在很多业界人士看来，是过于美丽的一个神话，一个理想国。大家都乐见其成，但其中有多少的困难，只能边走边看了。

其实，微软还是一个肯于接受新理念并愿意付诸实现得公司。看看live.com,以及汇集于Windows Live旗下的各个产品。你就可以清晰地看出MS想如何去做Software as Service了。只是，Marc不是很认同MS的这种模式。不过，谁知道，以后会是谁说了算呢。