http://blog.csdn.net/snow_fox_yaya/category/321962.aspxzh-CNTue, 25 Mar 2008 23:08:39 GMT60snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706284.aspxTue, 24 Jul 2007 21:23:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706284.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706284.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706284.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706284.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706284前段时间一直在忙于工作流的改版，经过将3个多月的努力，公司基于我们的核心框架framework4.0（spring、hibernate、webwork）的javabean版工作流终于推出了，在此发文庆祝一下。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706284.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706262.aspxTue, 24 Jul 2007 21:20:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706262.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706262.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706262.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706262.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706262在mule中提供了tcp、vm、http、soap、jms、udp、file等多种协议，那么这些协议在运行期是怎样查找的呢？<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706262.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706254.aspxTue, 24 Jul 2007 21:19:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706254.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706254.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706254.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706254.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706254在spring中使用mule时，有两种主要的方式，一种是将mule中的bean直接配置在spring的bean.xml文件中，统一由spring来管理；SpringConfigurationBuilder类在加载mule-config.xml文件时，由MuleApplicationContext类调用mule-to-spring.xsl文件，对mule-config.xml文件进行transform(转换)，实际上就是在mule-config.xml文件的头上加入三个bean的定义，以实现对mule的启动、初始化、服务的注册（mule-config.xml文件中的服务），bean的定义如下（可以查看mule-to-spring.xsl）：<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706254.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706245.aspxTue, 24 Jul 2007 21:18:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706245.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706245.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706245.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706245.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706245类中有几种变量，分别是：局部变量（英文可以为：local\automatic\temporary\stack variable）是定义在方法里的变量；任何变量在使用前都必须初始化,但是需要指出的是局部变量必须显式初始化，而实例变量不必，原始类型的实例变量在该类的构造方法被调用时为它分配的缺省的值，整型是0，布尔型是false，而浮点型是0.0f，引用类型（类类型）的实例变量的缺省值是null（没有进行实际的初始化，对它的使用将引起NullPointException），类变量的规则和实例变量一样，不同的是类变量的初始化是在类被加载时。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706245.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706226.aspxTue, 24 Jul 2007 21:16:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706226.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706226.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706226.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706226.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706226J2EE的模块化部署<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706226.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706196.aspxTue, 24 Jul 2007 21:12:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706196.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706196.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706196.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706196.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706196applet的运行<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706196.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706169.aspxTue, 24 Jul 2007 21:09:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706169.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706169.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706169.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706169.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706169static块的执行<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706169.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706151.aspxTue, 24 Jul 2007 21:06:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706151.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706151.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706151.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706151.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706151Artix是一系列高性能Web 服务集成产品，它使企业能够使用Web 服务解决复杂的、要求苛刻的集成问题。ESB是一种新型中间件，可以把以前不同类型的中间件整合在一起，它支持Web 服务和基础通信中间件，上提供高附加值服务。2004年领导型企业在其SOA项目中使用ESB,主流企业应当在其IT战略中考虑ESB，并在2005年开始使用ESB。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706151.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706142.aspxTue, 24 Jul 2007 21:05:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706142.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706142.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706142.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706142.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706142可自管理的分布式工作流引擎的设计与实现<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706142.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706133.aspxTue, 24 Jul 2007 21:03:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706133.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706133.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706133.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706133.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706133这两个月一直忙于写工作流论文的事，所以也没有记什么东西，上一周，自己研究了一下spring和hibernate，发现spring尤其是在aop方面的强大，在javaeye上看了N多N多的关于transaction和session的讨论，其实这些spring已经全部给封装好了，又看了看它的源码，也算是搞明白了它的实现原理，最后花了几天的时间又自己封装了一个简单的framework,这一周打算把公司的workflow engine集成进来，，然后再把公司的platform全部移植到这个框架上来。这样我们就有了一个比较不错的集成了workflow engine的开发框架。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706133.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706114.aspxTue, 24 Jul 2007 21:01:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706114.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706114.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706114.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706114.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706114到底分布式的工作流机要解决的问题是什么？我觉得分布式工作流机要解决的根本问题不是性能问题，性能问题可以通过cluster解决，分布式工作流机要解决的还是一个分布式的问题，也就是解决分布式应用的协作问题，举个列子：对于一个大型的企业（或者是跨国公司）它有销售部、研发部、客户支持部，三个部门都分布在不同的城市（甚至是不同的国家），而有一个业务流程需要这三个部门协作完成，那么此时怎么办？<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706114.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706109.aspxTue, 24 Jul 2007 21:00:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706109.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706109.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706109.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706109.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706109工作流机的执行分为集中式和分布式两种，对于分布式的工作流机，其分布式的调度算法是关键所在。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706109.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706107.aspxTue, 24 Jul 2007 20:59:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706107.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706107.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706107.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706107.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706107最近准备写一篇关于分布式工作流引擎的文章，于是查阅了一些已有产品和相关论文的实现方式，主要有以下几种实现：一、基于COBRA的分布式引擎；工作流管理系统的分布式分为三层：工作流建模的分布式、工作流引擎的分布式、工作流定义的分布式（即流程定义采用分布式存储）。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706107.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706105.aspxTue, 24 Jul 2007 20:58:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706105.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706105.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706105.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706105.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706105影子clone和深度clone的问题。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706105.aspx" width = "1" height = "1" />snow_fox_yayahttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706101.aspxTue, 24 Jul 2007 20:54:00 GMThttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706101.aspxhttp://blog.csdn.net/snow_fox_yaya/comments/1706101.aspxhttp://blog.csdn.net/snow_fox_yaya/archive/2007/07/24/1706101.aspx#Feedback0http://blog.csdn.net/snow_fox_yaya/comments/commentRss/1706101.aspxhttp://tb.blog.csdn.net/TrackBack.aspx?PostId=1706101公司的platform产品中用到了基于EJB的工作流，而platform每次调用工作流时，都是通过JNDI对home接口进行lookup，这个动作是很耗性能的，而以前竟然没有进行缓存，汗ing........,刚刚看到了Brett McLaughlin写的一篇。<img src ="http://blog.csdn.net/snow_fox_yaya/aggbug/1706101.aspx" width = "1" height = "1" />