许式伟
2006-12-28
概要
提起Google Code，我需要很惭愧地承认一点，我一开始的时候“把它看扁了”。初接触Google Code的时候，我简单地认为，Google Code只是一个半成品，从各方面的功能来讲，似乎远不能够与Sourceforge的完备相比。但是，随着对Google Code 了解的深入，我不能不承认我错了，Google Code再一次证明“简单是美”的法则：Google的服务似乎永远是看似简单（甚至让你觉得不够用），而实则非常完备。
如果你有开源的计划，或者已经开源，或者你有私人的项目希望有个完备的平台去管理它，那么本文值得你一读，因为这是我在这方面的一些心得。到目前为止，我开源了WINX和DocX已经4个月多了，和Google Code、Sourceforge均有了较长时间的接触，希望借机会和大家交流下。
选择Google Code，还是Sourceforge？总体来讲，两者各有各的优势。从进入的门槛来讲，使用Google Code会比Sourceforge容易很多。
Google Code的优势与劣势
Google Code的优势
* 速度第一。Google的服务以快速著称，Google Code亦不例外。当你上传新的Release时，你会特别感受到这一点。
* 操作简洁。如果Sourceforge和Google Code提供了同样的功能，那么我可以毫不思索的讲，Google Code比Sourceforge快捷易用。我们后文将详细比较各个功能。
Google Code的劣势
* 容易被盾。以我的经验，Google Code被盾比较频繁。这导致我有时无法浏览Bug列表，发布新的Release等等。所幸的是，Google Code的SVN连接从来没有出现访问障碍。因此这不算太大的问题。另外，按Google Code的说法，有些功能可以在命令行进行，而不需要到Web上，但是我本人尚未进行详细的验证。
* 100M的容量限制。 Google Code限制每个项目的SVN空间大小不能超过100M，发布的所有Release亦不能超过100M。由于目前我的项目刚刚开始，这一点还不是一个障碍。但是这是一个比较容易到达的数字。可以预见，你的项目要面临搬家的风险。我试图寻找办法来清除一部分SVN历史版本数据、删除一些历史 Release版本，未果。
* 无数据统计功能。你无法得知你的Release被下载的情况（下载总数以及趋势）。
Sourceforge的优势与劣势
Sourceforge的优势
* 更多的受众，已经形成第一开源社区。简单统计结果表明，Sourceforge上大约每5分钟就有一个新的Release。也就是说，一天大约有300个新的发布（这不是精确的统计，只是简单估算）。
* 功能更加完备。目前来看，Google Code在功能上并不比Sourceforge成熟。如果Google Code是一个完整、快捷、功能足够的平台，那么Sourceforge则是一个方方面面均已经相当完备的开发平台。
Sourceforge的劣势
* 操作繁复，速度较慢。
详细功能对比
下面我们展开来对比一下两者的功能：
* 建立帐号。Google Code只要你有Gmail帐号即可。Sourceforge要求你进行注册。而且你发现，Sourceforge的注册流程十分罗嗦。
* 建立项目。Google Code只要你简单填写一下项目的相关信息（http://code.google.com/hosting/createProject），仅一个页面即可完成申请。而Sourceforge的新Project申请流程是繁复的，有很多的下一步，并需要你阅读一些英文细节。另外，你的项目并不马上生效，而是需要进过审批。不出意外的话，在申请之后的第二天再上Sourceforge，你的项目应该通过审批了。
* 代码管理。Google Code仅支持SVN库，Sourceforge支持SVN和CVS。Google SVN限制在100M内，Sourceforge则尚未见到明确的容量限制。
* 代码库的权限管理。Google Code的代码库权限管理非常简单，在你登陆后，http://code.google.com/hosting/settings 将显示Google Code为你自动生成的随机密码。通过你的用户名和这个随机密码，你就可以访问SVN代码库，对其进行修改。sourceforge推荐使用putty套件，详细参考sourceforge上的说明，总的来说，相当繁复。
* 发布版本。Google Code限制所有Release包不能超过100M，单个Release不能超过10M。Sourceforge尚未见到明确的限制。
Google Code发布一个新的Release流程，仍然只需要一个页面即可完成。你只需要选择上传的文件，并为该文件加一些预定义的标签（用于告诉别人你的Release支持的平台或其他信息，这些标签Google Code已经预定义部分，项目管理人员亦可以定义），确认即开始上传。特别地，如果一个Release被打上 Featured 标签，那么它将出现在项目首页。
Sourceforge的流程则比较复杂：
o 通过ftp上传你要发布的文件。上传地址：ftp://upload.sourceforge.net/incoming，这个过程匿名进行。上传的文件在你后续的步骤中进行“认领”，如果某个包长时间没人认领，Sourceforge将清理之。
o 进入Sourceforge。用你的Sourceforge帐号登陆。
o 进入项目主页，进入Admin -> FileReleases。
o 选择已有的Package或者新建Package，向该Package中添加一个Release。在你的项目发布的文件比较多（不是发布的历史版本多，而是一次发布的文件多，你可能希望对他们进行归类）时，Package可以更加有条理地管理这些文件。
o 为该Release添加ReleaseNote和ChangeLog。确认。
o 认领你上传到ftp上的文件。打上勾，确认即可。
o 为你发布的文件添加说明（支持的平台、包的格式及其他规格）。
o 如果有人关注这个项目的Release，最后一步，你可以向这些人发送信息，告诉他们发布了新的Release。这一步很有意思，是Google Code不能做到的。
* Bug报告。Google Code中的issues，Sourceforge的tracker。由于用的还不算多，暂时不做详细比较。
* 论坛。Google Code中并不直接提供论坛功能，当然，这是因为它推荐你使用强大的Google Groups。Sourceforge则除了提供给用户的论坛（并把用户论坛分为两个，一个讨论产品功能的，一个属于头脑风暴型的开放式讨论），也提供开发者专用的论坛（对非项目成员不可见）。
* Project Web。 Google Code中没有提供项目主页，而是颇具创意的提供了一个Wiki Pages。Sourceforge提供了项目主页，项目主页的地址是："项目名.sourceforge.net"。你可以利用这个网址来宣传项目、提供项目信息。这个空间是是php+mysql的，并且你可以通过远程 Shell控制，极其灵活。
* 团队成员管理。Google Code你可以通过添加其他Gmail帐号为项目成员（Members）或管理员（Administers），但似乎还不能进行更加详细的权限管理。Sourceforge提供了非常精细的项目成员的权限管理。
* 下载与访问统计。Google Code无此功能。Sourceforge提供了最近7日、最近2个月、年、全部访问/下载统计。非常详细。
* 软件截图（ScreenShots）。可以让用户迅速获得项目的第一印象，吸引眼球。Google Code无此功能。
2007年12月13日星期四
2007年11月16日星期五
Google Finance不错的股票网站
中国股市如火如荼，偶然发现Google的Finance也是创意十足！
不仅可以方便的查找股票，而且利用最先进的Flash技术将新闻
与股票走势图联系在一起，真是太清晰了！
还有页面布置，基本上把相关的数据都显示出来了，一目了然！
目前Finance还处于Beta阶段，不知道正式版出来后是什么样子啊，
期待ing..
不仅可以方便的查找股票，而且利用最先进的Flash技术将新闻
与股票走势图联系在一起，真是太清晰了！
还有页面布置，基本上把相关的数据都显示出来了，一目了然！
目前Finance还处于Beta阶段，不知道正式版出来后是什么样子啊，
期待ing..
2007年8月12日星期日
Israeli researcher discovers theoretical model for time travel
By ISRAEL21c staff August 07, 2007
Fantasized about traveling through time ever since you saw the movie Back to the Future or read H.G. Wells? Well, one Israeli scientist has brought this science fiction fantasy just a tiny bit closer to reality.
Professor Amos Ori of the Technion - Israel Institute of Technology has developed a theoretical model of a time machine that could enable future generations to travel in the past. In his article published in the latest issue of Physical Review, Ori provides solutions to a number of difficulties that has long vexed researchers.
But don't make plans just yet to go back in time. According to Ori, very specific conditions are needed to allow for time travel. In a 2004 paper, Ori described a set of conditions needed to create a time loop. A donut shaped vacuum, the loop would curve back onto itself, so that a person traveling could go further back in time with each loop. Ori's theory is a set of mathematical equations that describe hypothetical conditions. If the hypothetical conditions are met, the formation of a closed time loop would create the possibility of time travel.
"If we were to create an area with a warp like this in space that would enable time lines to close on themselves, it might enable future generations to return to visit our time," he conjured. "We, however, could not return to previous ages because our predecessors did not create this infrastructure for us."
Stephen Hawking has described time travel as "an important subject for research" but has also challenged the concept. While the idea of time travel has never been eliminated, scientists have identified the physical challenges required, including the need for exotic matter with negative density.
Ori's latest work has eliminated the need for the exotic matter, as well as non-exotic but unidentified matter that he proposed in the 2004 paper. His new calculations show that the envelope can in fact be filled with dust, a simple modeling of which is used regularly in theoretical physics, while still allowing for the evolution of a time machine. His research, conducted with Technion Ph.D. student Dana Levanony, suggest that the creation of a time machine can only be achieved under a very narrow range of conditions.
"If the proper initial conditions were achieved, the time machine would evolve on its own without any further intervention," explained Ori. "It can be likened to shooting a ship with a cannon. Once the cannon is aimed properly and fired, the cannonball hits the ship on its own, driven solely by the laws of physics."
Research on time travel has existed for over a centuries. Described in fantasy novels as early as the eighteenth century, previous theories are grounded in Albert Einstein's general theory of relativity. Einstein's theory states that the gravitational pull of large objects can bend time and space. Most research on time travel is based on the concept of bending space-time so far that the time lines actually form a loop. "We know that bending does happen all the time, but we want the bending to be strong enough and to take a special form where the lines of time make closed loops," said Ori.
"We are trying to find out if it is possible to manipulate space-time to develop in such a way."
Fantasized about traveling through time ever since you saw the movie Back to the Future or read H.G. Wells? Well, one Israeli scientist has brought this science fiction fantasy just a tiny bit closer to reality.
Professor Amos Ori of the Technion - Israel Institute of Technology has developed a theoretical model of a time machine that could enable future generations to travel in the past. In his article published in the latest issue of Physical Review, Ori provides solutions to a number of difficulties that has long vexed researchers.
But don't make plans just yet to go back in time. According to Ori, very specific conditions are needed to allow for time travel. In a 2004 paper, Ori described a set of conditions needed to create a time loop. A donut shaped vacuum, the loop would curve back onto itself, so that a person traveling could go further back in time with each loop. Ori's theory is a set of mathematical equations that describe hypothetical conditions. If the hypothetical conditions are met, the formation of a closed time loop would create the possibility of time travel.
"If we were to create an area with a warp like this in space that would enable time lines to close on themselves, it might enable future generations to return to visit our time," he conjured. "We, however, could not return to previous ages because our predecessors did not create this infrastructure for us."
Stephen Hawking has described time travel as "an important subject for research" but has also challenged the concept. While the idea of time travel has never been eliminated, scientists have identified the physical challenges required, including the need for exotic matter with negative density.
Ori's latest work has eliminated the need for the exotic matter, as well as non-exotic but unidentified matter that he proposed in the 2004 paper. His new calculations show that the envelope can in fact be filled with dust, a simple modeling of which is used regularly in theoretical physics, while still allowing for the evolution of a time machine. His research, conducted with Technion Ph.D. student Dana Levanony, suggest that the creation of a time machine can only be achieved under a very narrow range of conditions.
"If the proper initial conditions were achieved, the time machine would evolve on its own without any further intervention," explained Ori. "It can be likened to shooting a ship with a cannon. Once the cannon is aimed properly and fired, the cannonball hits the ship on its own, driven solely by the laws of physics."
Research on time travel has existed for over a centuries. Described in fantasy novels as early as the eighteenth century, previous theories are grounded in Albert Einstein's general theory of relativity. Einstein's theory states that the gravitational pull of large objects can bend time and space. Most research on time travel is based on the concept of bending space-time so far that the time lines actually form a loop. "We know that bending does happen all the time, but we want the bending to be strong enough and to take a special form where the lines of time make closed loops," said Ori.
"We are trying to find out if it is possible to manipulate space-time to develop in such a way."
2007年8月11日星期六
望文生义的典型：research到底是什么意思？
作者：eng
金拓在讲知识产权的时候说“研究的英文为：Re-Search，即search, and
search again。”
这是典型的望文生义。Research这个词源自法语，前面的re-可不是again的
意思。它在这里表示intensive force的意思。从字面上来看，可以理解为to
investigate thoroughly。这个词本身并没有包括“原创性”的意义，只是在做
科学研究时，才有原创性的要求。
顺便说一下，金拓把北美的博士系统和国内的做比较，来说明博士多于三年
毕业很正常。但是，每个地方的系统都不同，这样的简单比较是没有意义的。比
方说，我也可以说英国的博士大都三年毕业，是不是就说明国内的系统是有道理
的呢？很明显，我们应该考虑到国内的硕士还有两年到三年，而美国本地的博士
生通常是没有经过硕士这一步，或者中途转成博士。从本科毕业到拿到博士学位
同样是五到六年。
金拓在讲知识产权的时候说“研究的英文为：Re-Search，即search, and
search again。”
这是典型的望文生义。Research这个词源自法语，前面的re-可不是again的
意思。它在这里表示intensive force的意思。从字面上来看，可以理解为to
investigate thoroughly。这个词本身并没有包括“原创性”的意义，只是在做
科学研究时，才有原创性的要求。
顺便说一下，金拓把北美的博士系统和国内的做比较，来说明博士多于三年
毕业很正常。但是，每个地方的系统都不同，这样的简单比较是没有意义的。比
方说，我也可以说英国的博士大都三年毕业，是不是就说明国内的系统是有道理
的呢？很明显，我们应该考虑到国内的硕士还有两年到三年，而美国本地的博士
生通常是没有经过硕士这一步，或者中途转成博士。从本科毕业到拿到博士学位
同样是五到六年。
2007年8月10日星期五
nasa航天飞机奋进号的网上在线直播!
http://www.nasa.gov/multimedia/nasatv/index.html
直接用Watch with RealPlayer可以观看,但是声音有点不清晰
Watch with RealAudio弥补了这个缺陷
直接用Watch with RealPlayer可以观看,但是声音有点不清晰
Watch with RealAudio弥补了这个缺陷
发现Google Reader还蛮好用的！
作者:落人村 2007年8月11日
最近总是上各种各样的新闻网站，有cnn,bbc,nasa,etc.还包括google,gmail等的博客。
每次总要同时开好多窗口，好不麻烦！而且无法保存新闻。最后决定还是用个在线RSS阅读器吧。还不用装任何软件。
尽管Google Reader还在试用阶段。但经过2天的使用，简洁的风格，超强的搜索功能还是那么出色。nod,敬仰 !
最近总是上各种各样的新闻网站，有cnn,bbc,nasa,etc.还包括google,gmail等的博客。
每次总要同时开好多窗口，好不麻烦！而且无法保存新闻。最后决定还是用个在线RSS阅读器吧。还不用装任何软件。
尽管Google Reader还在试用阶段。但经过2天的使用，简洁的风格，超强的搜索功能还是那么出色。nod,敬仰 !
订阅：
博文 (Atom)
