Wednesday, November 28

When you are Googling... (What Google doesn't want you to know about)

When you start a query in Google.com, the website creates a serial number to record this query, including your IP address and your searching keyword. If you are logged in with your Google account, the keyword will be saved with your account. So Google knows all your searching history, and it knows your history and your recent activity in real life.

This is not the end of story yet. After the searching, you browse through the results and short description, decide to click number 4th link of the result. Google creates a hidden query to send the serial number kEI="aE5ER_7xA6T2gQKJ0szoDA" and the item number cd=4 to the server so that Google knows what you are really interested in, from the link you click.

When you move mouse over an item of search result, from the status bar you can see the link of that item. But when you click it, the webpage activates a Javascript program to report the item you clicked before it navigates to the link. If you ever use Yahoo, you can see something different: Moving mouse over an item of Yahoo search result will shows a link like "http://rds.yahoo.com/_ylt=A0oGkxx7Nk1Hc_gAIwdXNy...", so you have a very clear idea that Yahoo will process the click (save your click history) before forwarding to the searched item. They both save your search history and click history to know what pages you are interested to improve their search result and advertisement. Yahoo is like a "real bad guy" because it tells you that, but Google is like a "fake good guy" because it tries to hide the truth.


PS: Today (Feb 18) when I was reading Google News, I saw a new "Recommended for you" section:

That is what I am talking about.

Labels:

Tuesday, November 27

HTML Page Break

In Word/Open Office Writer, you can easily insert a Page Break to start a new page. How to make that in webpage?

If you are reading a webpage from monitor, you actually don't care about page break, because you can easily turn page using the keyboard or scroll the page using your mouse. But when you print it out, you get pages like the one on the left. The second table is separated into 2 pages, which makes the printout look weired.

By adding < p style="page-break-before: always" / > before the second form, we can avoid this weired printout because a page break is inserted for printing, and the result is like the one on the right. When you are reading the web page, you can only feel an empty line between the 2 tables; but printout will be different.

page-break-before and page-break-after are standard CSS2 elements. They work fine both in IE and Firefox. A simple way to add page break into your webpage (printout) is to modify H2 element's CSS:
H2 {page-break-before: always; }
H2.nobreak {page-break-before: avoid; }

because it is very likely that you want to start the header in a new page, and H2.nobreak is prepared for the very first header...

W3C site introduces some other CSS elements that can control printout size, margin, and orientation. It's good to take a look.

Labels: ,

Sunday, November 25

亩产万斤的讨论:回方舟子,Yush等网友

上文贴出之后,方舟子回复
因为1.2%这个数字碰巧约等于800斤/5.85万斤,你就理解成我是根据钱的数字推导,然后就大喊我荒谬?
我那个数字引用的是袁隆平的说法,和钱学森有什么关系?他说:“从理论上讲,水稻的光合作用对地表太阳能的利用率可以达到5%。目前全国的水稻平均亩产只有800斤,只相当于利用光能的1.2%,通过科技进步,只要把光能利用率提高到理论水平的一半即2.5%左右,即意味着亩产的翻番。”
袁隆平搞了一辈子水稻,还会不知道水稻的这些基本数据?你有什么资格来质疑它?实际的光合作用效率是可以测量出来的(测量叶片对二氧化碳的吸收量、水分、光照强度等即可算出),其理论上限也是可以估算出来的,你连基本的植物生理学常识都没有,这么牛气干什么?


我的文章中每个数字都是有出处的,所以我以为方舟子的也是如此。他的《误区》一文,关于1.2%的上下文如此:

...如果考虑到水稻高产品种的光合作用效率大约为1%,那么上面的计算结果就变成了亩产量只有五、六百斤,接近实际亩产量。

光合作用效率恰恰是制约水稻产量的瓶颈。目前全国的水稻平均亩产为800斤,只相当于利用太阳光能的1.2%。...根据袁隆平的说法,从理论上讲,水稻的光合作用对地表太阳能的利用率可以达到5%,相当于亩产3000多斤,这应该是水稻亩产的极限。

正常逻辑的人当然会认为这个1.2%是从上面的1%延伸过来的,全文至此从来没有提起袁隆平,谁能想到这个1.2%是从“原湖南杂交水稻研究中心党委副书记谢长江”所写的短篇报道《袁隆平故事》中摘抄出来的呢?我不需要有“基本的植物生理学常识”,因为你的文章是给拥有正常逻辑的人看的。而且我很谨慎地说“我的理解是...”,就是给你提供一个解释/改错的机会。教主,您经常教导我们“吾爱吾师,吾更爱真理”,这次到来您头上,您就从了吧。“教主宝训,时刻在心,建功克敌,无事不成!”至于我说“荒谬”,这是一个口误;15分钟之后我就修改过来了,请不要介意。

存疑:袁隆平是如何算出“相当于利用太阳光能的1.2%”呢?也许也是继承于钱学森的算法?不知有无论文出现,希望内行能查一查。

由此看来,方的《误区》一文是建立在党委副书记的袁隆平报道基础上的。这个基础很不牢固,因为它不权威。希望方能续写《误区》,介绍科学家是从何时,如何认识到光和作用效率,当前是如何计算出(而不是“相当于”出)一亩田对地表太阳能的利用率。

另外,《袁隆平故事》中说,水稻的光合作用对地表太阳能的利用率可以达到5%;xj引用的文献说,The net result being an overall photosynthetic efficiency of between 3 and 6% of total solar radiation.鹏归引用的文献说:The overall efficiency is then .286x.43x.8x.67 = .066 or 6.6%。
从3%到6.6%相差120%,影响到我们到底最多可以有1000公斤还是2200公斤粮食的问题,不知谁能给一个权威的结论?

Yush: 吴礼说“松鼠“应用”的不是我的原文。 ”我的blog上面有他帖子的链接,他抵赖不了。他只能管好自己的帖子,不能控制我如何“应用”,只要我没有歪曲他的原意。
况且,吴礼只是一个网友,也不是什么著名科学家,如果你要反对他的观点就直接反对好了。即使他是著名科学家,根据“吾爱吾师,吾更爱真理”的宝训,只要你有能驳倒他的理由,我们也会接受,对不对?

ping_max,师太:就我所知,方舟子这篇《误区》是第一篇面向大众普及“光和作用效率”驳倒亩产万斤理论的文章。如果有其他人在此之前提出这个观点,请您指出来。

经xj和鹏归科普,得知6%左右的光合作用效率可经由公式推导出来。

xj:我说过“网上没有《农业中的力学问题》的全文,只有一半拷贝”,你还不信?

勘误:上文说“钱学森的文章发表27年之后才由方舟子提出”,应该是47年。

松鼠,你这是怎么了?理解错误啥的先不说,你那段教主云云实在不应该。你还没等老方说话呢,就说他教主。现在他反驳了,你也拿不出更权威的数据,这话都收不回来。

我相信你没有什么恶意。但是那话实在过重。
 
哦,我是bluesea
 
哈,德国著名哲学家卡尔。马克思说,缤纷色彩闪出的美丽,是因他没有分开每种色彩。就是因为每个人的想法都不一样,这个世界才如此精彩。


谢谢你的关心。现在找代理上来不容易吧?
 
去xys和洗脚城逛了一圈,刚看了你的这篇回应. 怎么了,很失水准啊. 摆事实讲道理,存疑就存疑,错了就错了,中间插那么一段教主啊什么的,很没风度. 每个人想法不一样很正常,但要打嘴仗写成文字,还是严谨一点好. 还是冷静几天,多读读资料再写回应吧.
 
“缤纷色彩闪出的美丽,是因他没有分开每种色彩。”您是“童年稚氣夢未污染”?如果方先生理性地理解了您的上贴,我猜他可能会动容。

谢谢您的好意,是因为您没有恶意。^-^
 

Thursday, November 22

亩产万斤的讨论:质疑方舟子文

1958年6月,钱学森发表了著名的《粮食亩产量会有多少?》,预告亩产量达5、6万斤,客观上刺激了大跃进的疯狂。



其中论证产量的过程是这样:
把每年射到一亩地上的太阳光能的30%作为植物以利用的部分,而植物利用这些太阳光能把空气里的二氧化碳和水分制造成自己的养料,供给自己发育、生长结实,再把其中的1/5算是可吃的粮食,那么稻麦每年的亩产量就不仅仅是现在的2000多斤或3000多斤,而是2000斤的20多倍!

第二年他又发表了《农业中的力学问题》,给了更详细的论证,结论是5.85万斤(全年)或3.9万斤(考虑冬季作物不生长)。

方舟子的《“亩产万斤”的误区》质疑这个论证过程,因为其中没有考虑植物通过光合作用,把吸收来的太阳光能转化成有机物所含的化学能的光合作用效率。文中说
光合作用效率恰恰是制约水稻产量的瓶颈。目前全国的水稻平均亩产为800斤,只相当于利用太阳光能的1.2%。

1.2%这个数字约等于800斤/5.85万斤,我的理解是:方舟子承认钱学森的“30%光照到作物叶子上“和”作物的1/5是粮食“,采用钱学森的结论5.85万斤来计算光合作用效率,因此得到1.2%这个数字。

在这里,你能看到:方舟子错用了5.85万斤这个全年生产作物的数字!如果800斤是海南的产量(袁隆平的试验田就是在海南)还可以理解,但是这是全国的平均亩产,很显然不应该用全年操作的数字。改用3.9万斤来计算,得到的转化率是2%.

方舟子的下一句:
根据袁隆平的说法,从理论上讲,水稻的光合作用对地表太阳能的利用率可以达到5%,相当于亩产3000多斤,这应该是水稻亩产的极限。
也很有疑问。
1,袁隆平是杂交育种专家,不是细胞研究专家,所以他在植物光合作用理论上的言论不太权威。
2,这个理论有没有论文发表?同行论证?
3,这个5%是否也是在“30%光照到作物叶子上”的基础上?
4,同上,如果改成用3.9万斤来计算,亩产2000斤是极限(如果上面的答案都是肯定)

JFF的《钱学森欠的血债会有多少?》一文的最主要观点是
钱应该知道《粮食亩产量会有多少?》一文中的论证,明显的最重要
的环节——射到植物叶上的光能和储藏在食物里的能量之间的转化率已被毫无根
据地最大限度地夸大了——100%,而实际上,根据专家说水稻的这个转化率只有
大约1%-5%。作为一个有20多年科研经验的科学大师应该知道,他那样得到的结
论毫无可靠度可言。

可是你想想,钱学森的文章发表2747年之后才由方舟子提出这个转化率问题,可见这个转化率并非人人皆知;JFF用这个来抨击钱,可不厚道。
杨周的《一个没有良心的科学家及其他》很不严谨。“模模糊糊还记得当年我们在中学里学习热机转换效率时,知道单机热转换效率的最大值大约是百分之四十到五十。”我不记得有这个最大值(我很信任我的中学物理学习),而且现在在网上也找不到这句话的任何佐证。而且他还抨击钱文不除掉秸秆,明显属于“回帖不看帖”的那种。钱文的“1/5算是可吃的粮食”,不就是把桔梗放到4/5中了么?

吴礼在洗脚池介绍了Modem的上限不断提高的故事
拿电话线来说,香农极限是大约46千比特/秒,经过研究,人们认识到通常电话线路中的噪声其实是设备设计中造成的。它不是完全随机,而是可以在一定程度上预测的。通过改变Modem的接入方法,可以把噪声进一步降低。于是1996年56千比特/秒的“超香农”Modem诞生了。

可见,理论上的极限还是可以突破。


钱学森的亩产千斤是他的一个污点,但是在当时形式下所说的“今天条件不具备,明天就会创造出来,今天还没有,明天一定会有!”中的明天,不等于现实中的第二天。钱的系统工程在人体科学上的应用也是一大笑话。可是我希望看到更有实质内容的评论(希望比方舟子的《“亩产万斤”的误区》更丰富),而不是JFF这样的大字报。

另外,网上没有《农业中的力学问题》的全文,只有一半拷贝,所以没法看到论证的全过程。希望多事者到国家图书馆去翻《知识就是力量》1959年8-9月合刊上传。

你的质疑很stupid, 中小学,化学,物理,生物没学好。
 
anonymous: 不如你介绍一下你所学的中小学,化学,物理,生物是怎么教你不用证据就下结论的?
 

Wednesday, November 21

Basic of Ontology

This is opening of Chapter 5, Gene Ontology, of my thesis Empirical Study of Gene Ontology based Microarray Clustering. You can check abstract from here.

5.1 Basic of Ontology

Ontology” has been a philosophy jargon since Aristotle times, and it means the nature of existence. Computer scientists adopted this word to express a formally structured vocabulary in a discipline. In this vocabulary, items and relations between two items are well defined to present the knowledge in this discipline.

Ontologies offer a mechanism by which knowledge can be represented in a form capable of machine processing[Lord 2003]. Ontologies can be provided in Rational Database format or XML format.

Now ontology becomes the core of Semantic Web, because the geographically distributed Web forms information islands in the Internet, and the use of ontology can interpret meanings of information in different islands, reduce the confusion, and integrate data automatically. The decentralized infrastructure makes the communication and collaboration over Internet easy. Every one can focus on her own part of the project independently and integration of their work will be streamlined since every part of the collaboration follows the same ontology and plays her own role. Every one can also build new ontologies, and cooperate with the third part without the permission of her collaborators. The collaboration will be stronger and stronger as more and more collaborators join in and share their knowledge. This decentralized infrastructure breaks down the barrier between languages, geographical distance, automates the integration of knowledge, and leads to the evolution of knowledge.

Labels: ,

Tuesday, November 20

Configuration of Virtual Host in Apache 2.2

I have been using Apache for 5 years, and I can literally tell each directive in httpd.conf. But yesterday I failed configuring Vitual Host in Apache 2.2.

Thanks for Google. David reminds me that Apache 2.2 imposes stricter permissions, so I moved all virtual hosts into one folder "\working\remoteserver" and added directive:
< Directory \working\remoteserver>
Order Deny,Allow
Allow from all
< /Directory>

Then modify the httpd-vhosts.conf as
< VirtualHost *:80>

#AddDefaultCharset utf-8

ServerName private.fadshop.net

ServerAdmin webmaster@fadshop.net
DocumentRoot \Working\RemoteServer\private.fadshop.net

ErrorLog "|rotatelogs logs/private.fadshop.net-error.%Y-%m-%d.log 604800"

CustomLog "rotatelogs logs/private.fadshop.net-access.%Y-%m-%d.log 604800" combined

# ForceType text/html

< /VirtualHost>


and other virtual hosts. One more thing to remember: After you deploy virtual hosts, the "original" host will be ignore. So you should create first virtual host as the original host.

Labels: ,

Sunday, November 4

Modifying Orage: Setup Alarm for your event.

Orage is a lightweight calendar in Xfce desktop. Yes, it is the default calendar in Xubuntu.

I don't like adding timer after adding an event. I think whenever I add an event, the system should alarm me before it happen. The alarm is the reason why I would possibly use calendar: Remind me to eat at lunch time, so that I won't be starve.

By default Orage won't add a timer, unless you select a number from pull-down menu. So I downloaded the source code using svn. BTW, the Xubuntu 7.10 has Orage 4.4.1, when svn gives you the source code of orage 4.5.10.
In appointment.c, modify line 2029:
apptw->Alarm_spin_mm = gtk_spin_button_new_with_range(0, 59, 5);

as
apptw->Alarm_spin_mm = gtk_spin_button_new_with_range(1, 59, 5);

The added event will have a pop-up reminder 5 minutes before the event, by default.


For Evolution, you only need to enable "Show a reminder":

Sharing files between Ubuntu and Windows

At first I was using Xubuntu, Ubuntu with Xfec. Following this post I was able to visit shared folder of one computer in the same network, after I modified the WORKGROUP of Xubuntu. But I couldn't get access to my Win2003 Server, because it is password protected.. I had to switch to Ubuntu with Gnome. Nautilus, the file explorer, is more friendly than Thunar. Select Go->Network, I am able to visit my Win2003 Server using username and password.

Now it's time to share files in Ubuntu. After modifying the /etc/samba/smb.conf, I could see the shared folder from Nautilus, but not from Windows box. When I tried to access this folder from Windows box, I was prompted to input username and password, and my username/password is always rejected. The reason is that the system is asking for username/password of SAMBA, not u/p of Ubuntu. The solutions is to create a samba password:
sudo smbpasswd -a ben

After doing so, I can visit Ubuntu shared folder from Windows box.

Labels: ,