Remote Desktop for Linux - Reloaded : lxpages.com blog

Remote Desktop for Linux - Reloaded : lxpages.com blog:

My first article on remote desktop became popluar among Linux users
and some Windows people as well. I received many comments on what I had missed from the short list I had compiled. Therefore, I’ve written this updated article that contains additional remote desktop apps that can be used for Linux to Linux / Windows to Linux and also Windows to Mac and Mac to Windows.

Linux to Linux

1) VNC (Virtual Network Computing) is a remote display system which
allows the user to view the desktop of a remote machine anywhere on
the internet. It can also be directed through SSH for security. There
are 3 types of VNC, they are:

1. http://www.realvnc.com/download.html
2. http://www.tightvnc.com/download.html
3. http://www.uvnc.com/

While you’re at it, get VNC Scan from (http://www.vncscan.com/) - a popular Network Administration tool that allows you to remotely manage all of your desktops and servers from a single location.

2) FreeNX is a system that allows you to access your desktop from
another machine over the internet. Some helpful links on how to
install:

1. https://help.ubuntu.com/community/FreeNX
2. http://ubuntuforums.org/showthread.php?t=97277&highlight=freenx
3. http://freenx.berlios.de/ (FreeNX homepage)

3) The third free application is 2X Terminal Server for Linux.

1. Download at http://www.2x.com/terminalserver/download.htm
2. Forum is at http://www.2x.com/forums

4) XDMCP. The original way of doing this on Linux. This link will
help get you started:

1. http://www.tldp.org/HOWTO/XDMCP-HOWTO/intro.html

5) CygwinX. A complete Linux emulation on Windows. You’ll find every
tool and app that you have on Linux on Cygwin.

Get it from http://www.cygwin.com/

6) XRDP. RDP server that runs on Linux, thus allowing you to use
Windows Remote Desktop Client or rdesktop to connect.

Get it from http://sourceforge.net/projects/xrdp/

7) x2vnc - great little utility that allows you to tie a linux and windows (or anything that can run the vncserver) together with a ingle keyboard/mouse, avoiding the need for a switcher box. Mousing cross screens transparently switches between machines, and cut and aste works.

Get it from http://fredrik.hubbe.net/x2vnc.html

8 ) Xming - t’s a great and lightweight implementation of X11 for
Windows that allows you to connect to a Linux box. You can read about
it at http://freedesktop.org/wiki/Xming

9) KDE Desktop Sharing (formerly krfb) - part of KDE since version 3.1. It is located in the kdenetwork package. If your distribution splits the KDE applications into separate packets, you may find the client as ‘krdc’ and the server as ‘krfb’. Also uses VNC technology.

10) X-Win32 - Top rated PC X server solutions for Windows PCs connecting to remote Unix and Linux host systems. Works well over SSH.

Get it from http://www.starnet.com/products/xwin32/

11) Single Click UltraVNC - In case you would like to remote control without any software installed on the target computer you need UltraVNC SC. The user on the to be controlled computer needs to simply click on a web page and remote controlling begins.

Get it from here http://ultravnc.sourceforge.net/addons/singleclick.html

12) CrossLoop - CrossLoop is a FREE secure screen sharing utility designed for people of all technical skill levels. CrossLoop extends the boundaries of VNC’s traditional screen sharing by enabling non-technical users to get connected from anywhere on the Internet in seconds without changing any firewall or router settings.

Get it from http://www.crossloop.com/

13) Thinstation - Although not a remote desktop app but worth mentioning here. Thin client linux distro for terminals using std. x86 hw. It can boot from network, pxe, syslinux,loadlin, CD, floppy or flash-disk and connect to servers using VNC, RDP, XDM, SSH and etc.

Get it from http://sourceforge.net/projects/thinstation/

14) rdesktop - an open source client for Windows NT Terminal Server and Windows 2000/2003 Terminal Services, capable of natively speaking Remote Desktop Protocol (RDP) in order to present the user’s NT desktop. rdesktop currently runs on most UNIX based platforms with the X Window System, and other ports should be fairly straightforward.

Download from: http://www.rdesktop.org/

While you’re at it, get grdesktop from (http://www.nongnu.org/grdesktop/). It is a GNOME frontend, for rdesktop. It can save several connections (including their options), and browse the network for available terminal servers.

15) ssh -X - You can check out this great article written by a slashdot user sometime ago.

Read it here http://slashdot.org/~Trolling4Dollars/journal/69740

Windows to Mac / Mac to Windows

1) RDP Client for Mac allows you to connect to a Windows-based computer and work with programs and files on that computer from your Macintosh computer.

Download it here

2) OSXVnc - Vine Server is a full featured VNC server for Mac OS X providing remote access to the GUI, keyboard and mouse using Vine Viewer or any other VNC client.

Download it from http://www.redstonesoftware.com/products/vine/server/vineosx/index.html

3) Chicken of the VNC - A VNC client allows one to display and interact with a remote computer screen. In other words, you can use Chicken of the VNC to interact with a remote computer as though it’s right next to you.

Download it from http://sourceforge.net/projects/cotvnc/

Unfortunately I was not able to find too many available to connect to Mac from Windows other than VNC. I think Windows need to support RDP into Mac. Many people would benefit from this.

If I am missing anything from the list, please let me know.

Visual Studio 2008 Is Pretty Damn Slow… « Williamo’s Blog.

Visual Studio 2008 Is Pretty Damn Slow… « Williamo’s Blog.

Update: The speed issues I complained about in the original version of this post have been remedied (as far as I can tell) by “Hotfix for Microsoft Visual Studio 2008 - KB946581″ which you can find here: Downloadable Hotfix: Performance and Editor fixes for Microsoft Visual Studio 2008 and Visual Web Developer Express 2008. Thanks to David and Cameron from Microsoft for helping me out with this!

If you’d like to read the original post, you can find it below, but keep in mind that most of this rant is now moot. Enjoy!

---

Disclaimer: I will *happily* post a retraction to this rant if *anyone* can tell me how to make Visual Studio 2008 perform like Visual Studio 2005 did, or better. My intention here is not to offend (despite the title of this post) but to pique some interest in the community as to how to speed up this beast of a program.I’m a software engineer for a web development company, developing code in C# and ASP.NET. Thus, the tool I use most in my daily work, of course, is Visual Studio. I’ve been happily using Visual Studio 2005, well, since 2005, and have enjoyed it thoroughly. Everything it does (for me), it does well, from IntelliSense to the integration with source control. So, when my company decided to upgrade all its developers’ copies of Visual Studio 2005 to Visual Studio 2008, I had no reservations. I’d heard about things like better JavaScript handling (with IntelliSense as well), split Design/Code view in Web Developer, and, well, it just looked slicker than 2005 in general.

Now, less than a couple months after our upgrade, I use a text editor (Notepad2, for what it’s worth) almost exclusively for writing and modifying code. Why? Because Visual Studio 2008 is slow, unwieldy, slow, memory hogging and *really damn slow*.

(I see this the majority of the time Visual Studio is open.)

I have tried everything I can possibly think of to wrestle with this beast. Cleaned my hard drive, reinstalled Windows XP, installed VS2008, removed all components but Web Developer and C#/ASP.net (no C++, VB, mobile development tools, etc.). Before you ask, my development machine has the following specs:

• Intel Centrino (Core Duo T2300) @ 1.66GHz
• 2GB of DDR2-667 RAM
• 80GB SATA2 Hard Drive

Not the fastest machine on the block, but it should be more than sufficient to do web development. Typically, when I’m doing web development, I have four programs open: Outlook, Visual Studio, Firefox and Internet Explorer. Even with only these four programs open and only two tabs open in each browser, I see the (Not Responding) message in Visual Studio’s title bar more often than not.

Loading the environment takes anywhere from 30 seconds (fast) to two minutes (most of the time). Closing the program takes about the same amount of time. Opening a C# source code file takes 15-20 seconds, with the hourglass spinning and most of the time the window won’t repaint at all during this time. If I make the rare mistake of attempting to click into “Design” mode when viewing an ASPX page, my whole system locks up. No response from any windows, and even Ctrl+Alt+Delete takes 5-10 seconds to slowly paint the dialog with its options.

Possibly the most disappointing part of all this is the fact that Visual Studio 2008 was touted for its speed increases.

So, to wrap this up, if you have some good tips or tricks for (drastically) speeding up the performance of Visual Studio 2008, please leave a comment on this post. I’d love to hear from you, and I’m sure a lot of other Visual Studio 2008 users would love to hear from you as well.

---

Possibly related posts: (automatically generated)

• Performance and Editor fixes for VS2008…

Posted by willwm Filed in Uncategorized

http://www.freeantennas.com/projects/template2/index.html

Ez-12 Parabolic Reflector Template

1. Download Template
2. Open in MS Paint and resize as desired. Ensure that you keep the square on the diagram square.
   
3. Print image on business card stock or acetate (document protector).
4. Cut it out and open the six slots with sissors or an X-Acto knife.
   
5. Glue tin foil to the back side of the reflector surface
6. Assemble by placing the six tabs in the six slots.
   

Tell a friend about this idea. Note: If printed at the download size you will see about 9 dBi of gain. If you double the size of the image before you print it you will see about 12 dBi of gain. If you place two reflectors on an AP with two antennas you will see an additional increase in performance.


Six inch reflector pattern at 2.4 GHz

Grass trimmer maintain and repair

To start a cold engine:
1. Place the on/off switch in the ON position.
2. Slowly press the primer bulb seven times.
3. Rotate the choke dial clockwise to the FULL CHOKE position.
4. While squeezing both the trigger lock and throttle trigger, push in the throttle lock button.
Release the throttle trigger. This latches the trigger for starting.
5. Pull the starter grip sharply until engine attempts to run (no more than 4 pulls).
6. Rotate the choke dial to the counterclockwise HALF CHOKE position.
7. Pull the starter grip sharply until the engine runs.
8. Allow the engine to run for 10 seconds.
9. Rotate the choke dial counterclockwise to the RUN position.
10. Squeeze the trigger to disengage the throttle lock and return the engine to idle.

PREMIUM EXACT MIX (50:1) 1 gallon gas (US) / 2.6 oz oil, or 1 Liter gas /20 mL oil

Usage
You should change the gas/oil mixture and replace it with new stuff if you did not already do so. Insure the choke is set to full choke when you start a cold machine. Move the choke to full opened after the machine has warmed up. If you machine has a primer bulb, usually a small plastic type button, push it at least 8 times before you attempt to start it. Try a few pulls with choke on - then try a few with choke open, keep the throttle fully open and pull the starter cord. If it doesn't fire, spray a little starter fluid into the air inlet and repeat the above sequence. You can get the starter fluid spray at a 'box' store or auto parts store. If it starts, let it run for about 30 seconds and then gently release the throttle and and then adjust the choke to the off position, as they are very temperamental, and just need a bit of coaxing.

Repair
All these small 2 stroke engines are difficult to start if the fuel-air mix is not right and they are prone to over choking.

If the engine is not running, the cause can be poor spark, lack of fuel or air delivered to cylinder, or some sort of mechanical failure such as low compression. Check the easy stuff first, i.e. fresh fuel mix in tank; on/off switch on; air filter reasonably clean.

1. Check the fuel system to ensure it is ok and no blocked filters etc. Clean out the fuel tank and put some fresh fuel in it.
2. Air filters. Ensure they are clean and allow unrestricted airflow into the carb.
3. Spark Plugs. Take out the spark plug, clean and dry it, if you know the plug gap check that it's right, if you don't know the gap and it hasn't been altered leave it as it is, If the plug looks anything less than perfect replace it.
If the plug is wet with fuel, dry the plug and do a few pulls before reinstalling the plug (keep spark plug wire away from the open hole). Reinstall the spark plug and try the above again.
Check for spark by taking it out, putting it back in the boot and grounding the threads to the head and watch for a blue snap.
4. I will agree that clean plug and air filter are necessary for superior running condition. However, the two most common problems that I encounter in 2 cycle products is carboned up anti spark screens/exhaust ports and varnished up in tank fuel filters. If, after checking the above four items, try adjusting the high and low speed screws counter clowckwise half a turn if the carb has them. If adjustment doesn't help I rebuild the carb. If no adjustment screws are present, the diagphram pump in the carb is likely hard and not capable of pumping proper fuel.
5. If at any point in this process it fires but won't run - the carburetor is probably dirty/gummed up etc. This can be caused by using old fuel. If this is the case, spray some carb cleaner into the air intake after you removed the air filter.
6. If the above does not fix the problem adjust the jets in the carburettor. It is possible that with all this pollution stuff happening that the formulation of the fuel you are using has changed causing the unit to change tune.

Repeat above sequence.

change the plug and air cleaner don't use starting fluid in a two stroke engine their is no oil in it and you could cause piston/ring/cylinder damage.

http://forum.doityourself.com/showthread.php?t=270292
Seatech:
Hi, all. I'm looking for some help with my trimmer. I bought a Homelite String Trimmer (UT20699) at a garage sale ($5 woohoo!). The kicker is that it won't start. When I got it home, I noticed that the 2 fuel lines were out of the tank, so I got new lines and a new filter (I think I have them going to the right lines on the carb as there is fuel showing up in the primer bulb). But the trimmer won't start. It doesn't really sound like it's trying to catch. First I tried spraying a little starter fluid in the carb and then a little in the spark plug hole but it won't go. I checked the plug and it's all black on the tip (not sure if that counts as fouled or not). Also, I noticed that when I pull the throttle trigger the throttle on the carb isn't moving....seems to be stuck but I was able to movie with a screw driver. Anyway, I'm pretty much a complete newb when it comes to anything mechanical so any help or advice you can offer is definitely appreciated.

v8driver:
You need to buy a carb kit for that particular carb, tear the carb apart and soak it in carb cleaner and blow it out through any and all holes real good. Sounds like the throttle cable might be broken if you could move the butterfly on the carb with ease. And of course you could have a fouled spark plug but check for spark by taking it out, putting it back in the boot and grounding the threads to the head and watch for a blue snap. Nothing? remove the kill wire..... still nothing then a new coil...
Also you should check the cylinder and piston through either the spark plug hole or exhaust port for scoring.... any scoring, probabally will have too low of compression to even start, and a new cylinder, piston and rings will need to be bought... not really worth it either. A 2 cycle like a trimmer, under 90 psi of compression is bad, even at 90 it'll be hard to start, over is good.

Seatech:
Thanks for the response, V8.
To check for spark, when you say to ground the threads of the spark plug to the head, do you just run a wire from the threads to the engine block or to the tip of the spark plug? Also, after it's grounded properly I assume I just pull the cord and see if there is spark, right? Sorry, I'm really a complete novice when it comes to all of this, so sorry for my ignorance.
Also, I can move the butterfly but it doesn't move easily at all. When I pull the trigger, the cable does seem to be moving but the butterfly seems to be stuck, so I'm guessing it needs to be cleaned?

cheese:
It can be hard to ground a plug on some trimmers with all the plastic they have on them. Sometimes I wedge a screwdriver next to the cylinder, and lay the plug on it in such a way so that the metal on the plug is touching the metal on the screwdriver, and the screwdriver is touching metal on the engine. You can use your wire method too, or whatever it takes. Sounds like the carb does need cleaning.

Seatech:
I got it started!!!!! woohoo!
Ok, so after reading your tips on checking for spark, I went home after work and pulled the plug out and for the ground I cut the cord off an old charger for a cell phone I no longer have and taped the wire to the threads on the spark plug and then to the engine. I got spark, and it looked good and blue, so I put the plug back in and took the muffler off and cleaned the screen.

I put it all back together, closed the choke, completely and gave it 6 or 7 pulls on the cord. Switched the choke to the next step open and pulled again...and it started! (kinda scared me actually. lol)
So, thanks for all of your help. Gotta love the internet!

I do still have a problem with the throttle not moving by itself. It's basically stuck wide open. I can move it with a screw driver, but it's not easy at all. Any tips on getting that working besides liberal use of wd-40 or some other lubricant?

v8driver:
You still need to clean the carb..... sounds right now that the butterfly is stuck in place maybe by gummy gas. Or the shaft is broken on it and its rubbing (unlikely but possible).

Install VirtualBox Additions for an Ubuntu 8.04 Guest | Tombuntu

Install VirtualBox Additions for an Ubuntu 8.04 Guest | Tombuntu

First you must install VirtualBox guest addition then you simply need to edit this file :
sudo gedit /etc/X11/xorg.conf
Then you make something equivalent to this :
Section "InputDevice"
Identifier "Configured Mouse"
Driver "vboxmouse" <-- this important for Ubuntu 8.04
Option "CorePointer"
EndSection

Section "Device"
Identifier "Configured Video Device"
Driver "vboxvideo" <-- this is important for Ubuntu 8.04
EndSection

Section "Monitor"
Identifier "Configured Monitor"
EndSection

Section "Screen"
Identifier "Default Screen"
Monitor "Configured Monitor"
Device "Configured Video Device"
DefaultDepth 24
SubSection "Display"
Modes "1280x1024" "1024x768" "800x600" "640x480" <-- add the resolution you want
EndSubSection
EndSection

Then save and exit.
Then go in system/preference/screen resolution and select the good one.
That's all.

Have a nice day

罗伯特·马修斯（Robert Matthews）的基础概率问题

罗伯特·马修斯（Robert Matthews）的基础概率问题
http://web.wenxuecity.com/BBSView.php?SubID=kxtw&MsgID=7768

假定一项疾病检验的准确率为95%，即患病者的检验结果呈阳性的概率，和未患病者的检验结果呈阴性的概率都是95%。已知被测人群中这种疾病的患病率为千分之一.那么，已知一个人的检验结果呈阳性，则他确实患病的可能性有多大？半数受试人完全没有意识到基础概率的影响，他们的答案是95%. 只有不到五分之一的人给出了正确答案：此人患病的概率不足2%.。’

这个例子猛一看似乎很有道理，但总觉得广泛使用的辅助检查，大部分的都有假阳性和假阴性的问题，如果阳性结果对诊断疾病如此没有帮助（可能性不足2%），全世界的医院岂不是都在开玩笑？这样的玩笑可真是国际玩笑了。

问题究竟出在哪里？罗伯特的概率计算没有问题。按照他的计算方法，应该这么理解：随机选取一个人，他的患病概率是 千分之一，即：0.1%. 而呈现阳性的人群中患病的概率是2%，换句话说，呈现阳性的被检查者，患病的可能性是 普通人群的20倍。从这个意义上来看，并不像作者所说的那么悲观。

作者还混淆了一个概念，阳性率95%指的是：本来应该阳性的5%没有呈现阳性，即有5%的假阴性。假如检查阳性就可以诊断某种疾病，将有5%的漏诊。而正常人群中，假阳性率为5%，也就是说被检验者中将有5%的可能性被误诊。因此，对病人来讲，如果检查是阳性的一定是患病。检查阳性的病例中有多大可能患病哪？那就是，100%-减去假阳性的可能性5%，即100%-5%=95%。（这里和漏诊没有关系）我认为，马氏错误地使用的基础概率的整体。临床检验不是人群的普查。临床检验时有高度针对性的，也就是说：医院检查的这个人群已经不是普通人群，这个人群的患病率一定大大高于普通人群！
考虑的对否，请概率知识丰富的同学指正。

附录：

你的预测有多准？

作者：罗伯特·马修斯（Robert Matthews）
翻译：yimin

原文：
http://www.newscientist.com/article/mg15320724.000-how-right-can-you-be.html
　　译文首发在译言网上：http://www.yeeyan.com/articles/view/yimin/9204

　　你准备去集市上买点东西，大概需要一个小时左右就可以回来。不过天气预
报说有雨，你该怎么办？假如你知道，预报的准确率在80%左右。那么，你需要
一把雨伞的可能性是80%，不是吗？非也，下雨的可能性其实只有30%.

　　这个结论看起来有些奇怪，不过这与预报员的夸大其词或天意弄人可没什么
关系。人们在理解概率问题时，有一种奇特的数学效应总会使我们出错，刚才提
到的只是其中一个例子。这种效应影响极为广泛，不仅阴晴雨雪这类日常预测受
其影响，从地震的预测到谋杀案的目击证词，没一个逃得出它的掌心。

　　隐藏在幕后的那股力量就是"基础概率效应"。简单地说，我们经常需要预测
某件事在将来是否发生，而这种预测的效果会受到基础概率的影响。所谓基础概
率，就是指我们从经验数据中得到的某件事的实际发生概率。当你预测任何罕发
事件时，这种影响就会非常明显。即使你认为自己的预测是非常准确的，你的准
确预测也会淹没在大量错误预测之中。

　　降雨的预报是个经典的例子。英国气象局对于降雨的预报准确率为80%左右，
这很容易让人产生这样的预期：如果气象局预报有雨，那么八成就真的要下雨。
这种推测之所以是错的，是因为它忽视了降雨的基础概率。

　　仔细一想就会发现这一点有多重要。即使是最蠢的预报员也能以惊人的准确
率预报智利阿塔卡马沙漠的降雨：在那里，数十年才下一场雨。因此，如果想以
近乎100%的准确率预报那里的天气，你只需说："各位观众，明天不会下雨。"

　　对英国变化无常的天气来说，情况就要复杂一些，但基础概率对天气预报的
可靠性仍然有巨大的影响。英国的小时降雨基础概率是0.1，就是说，在任何一
个小时内，降雨的可能性均为十分之一。这个基础概率决定了我们应该对80%准
确率的降雨预报抱有几分信任。

　　为什么这么说呢？假设你一年内会进行100次这样的"一小时购物"。10%的小
时降雨基础概率意味着你的90次购物之旅不会碰上下雨，另外10次则没那么幸运。
在这10次下雨天气中，天气预报将会准确预报其中的8次，因为它的准确率为80%.

　　不过80%的准确率同时也意味着20%的不准确率——因此气象局将会把18次晴
好天气预报成有雨（译者注：18=90*20%）。加起来，共有26次预报有雨，其中8
次是准确的。所以，尽管预报的准确率高达80%，但在预报有雨的日子里，出门
的一个小时内真碰上雨的可能性只有30%.

　　被放大的误差

　　实际上，大量晴好天气将降雨预报中的微小误差放大了，以致准确的预报被
淹没在错误预报之中。（参看本文"计算罕发事件的发生概率"）这清楚地提醒我
们是否该认真对待天气预报。在最近的一期《自然》（Nature）杂志上，我证明
了：如果你只需出门一两个小时，而且可以忍受偶尔被淋湿，那么，即使气象局
预报说有暴雨，最佳的做法仍是：不带伞。

　　除了帮助我们决定是否带伞，基础概率效应也解释了为何当今准确率颇高的
天气预报仍然饱受质疑。在最新一期的《当代数学》（Mathematics Today）中，
我认为这是因为人们出门的次数和呆在户外的时间不够长，等不到下雨的那一刻。
如果你打算在户外待上一整天，那么你可得认真看天气预报：日降雨的基础概率
是0.4，这就使得在预报有雨的日子里，下雨的概率比不下雨的概率高出近两倍
（译者注：给定预报的准确率为80%，此时当天下雨的概率是73%，是不下雨的概
率27%的近三倍）。可以这么说，一件事越常见，就越容易准确地预测。

　　这么一说，基础概率效应似乎是显而易见的。可让心理学家们不解的是：既
然如此，为何我们在碰到类似问题时总是出错？斯坦福大学已故的阿莫斯·特沃
斯基（Amos Tversky）和普林斯顿大学的丹尼尔·卡纳曼（Daniel Kahneman）
开创性地研究了人类对于不确定性信息的认知能力。他们的研究成果长期以来被
广泛引用，证明人类在处理概率问题时会变得无可救药，尤其在处理与基础概率
有关的问题时。经常被引用的一个例子就是所谓的"出租车问题"（Cab
Problem）。

　　蓝车还是绿车？

　　一个城镇中发生一起出租车夜间肇事逃逸案，这个城镇只有两家出租车公司
营运：一家的车子是绿色的，数量占所有出租车总数的85%，另一家的车是蓝色
的，占15%。一个目击者声称肇事车是蓝车。警察在出事当夜相同的环境下测试
得到目击者的判断准确率为80%. 那么，肇事车确实是蓝车的可能性有多大？

　　如果对基础概率一无所知，很可能回答：80%——这是人们直觉倾向的答案。
但正确的答案却是41%（用列联表可以很容易得到这个结果）（译者注：列联表
的使用见本文最后一部分）。占多数的绿色出租车意味着：被目击者错认为蓝车
的绿车数量会大于他认对的蓝车数。结论是：警察抛硬币来决定肇事者更靠谱。

　　有人会把这样的问题仅仅当作智力游戏，但医生在做出生死攸关的决定时，
也会受到基础概率的影响。哈佛大学医学院发表于1978年的一项研究中，60名受
试的教师和学生被问到以下问题：假定一项疾病检验的准确率为95%，即患病者
的检验结果呈阳性的概率，和未患病者的检验结果呈阴性的概率都是95%。已知
被测人群中这种疾病的患病率为千分之一. 那么，已知一个人的检验结果呈阳性，
则他确实患病的可能性有多大？

　　半数受试人完全没有意识到基础概率的影响，他们的答案是95%. 只有不到
五分之一的人给出了正确答案：此人患病的概率不足2%.

　　令人担心的是，对于医务工作者的其它类似研究都得出同一结论：受试者普
遍头脑混乱。在诊断阶段如果忽略基础概率的影响，将会造成大量的过度医疗。
不过，一些心理学家开始问："如果这种研究以更加浅显的形式进行呢？"最近发
表在《行为科学和脑科学》（Behavioral and Brain Sciences）上的一篇基础
概率综述文章中，德克萨斯大学奥斯汀分校的乔纳森·科勒（Jonathan Koehler）
指出，许多研究都预先设定受试者无法正确回答问题，这些研究与其说揭示了我
们认知概率信息的能力，还不如说暴露出他们提问的方式有问题。

　　他认为类似"出租车问题"的那些难题总是语焉不详。比如，你可以说：重要
的基础概率不是蓝车在车辆总数中所占的比例，而应该是蓝车在有夜间事故记录
的车辆中所占的比例。毕竟，蓝车虽少，但也许他们的安全记录普遍更差呢？这
个基础概率并没有给出，受试者只能瞎猜。科勒说，如果你没有给出受试者认为
重要的信息，那么他们犯错误就没什么好奇怪的。

　　以正确的方式提问

　　科勒补充道，最近研究显示，如果以频率的形式而非概率的术语提问，受试
者对基础概率的认知会好许多。比如，不要求他们估计检验呈阳性的人确实患病
的概率，而是要求他们估计100个检验呈阳性的人当中，有几个确实患了病。

　　这种措辞的微妙变化带来了全新的结果，至少看起来是这样。在加州大学圣
巴巴拉分校的勒达·考斯米德和约翰·托比（Leda Cosmides and John Tooby）
去年发表的研究中，受试学生回答了与哈佛大学研究中类似的疾病诊断问题。当
问题以概率的术语提出来时，他们表现同样糟糕；可当问题以频率的形式提出来
时，许多学生都意识到了基础概率的影响。

　　这个发现对于如何训练医生解读检验结果有着清晰且重要的意义。俗话说"
罕见的病不容易诊断"，话是没错，可这对诊断没什么帮助，何况陪审团也不会
认可对医疗事故的这种解释。

　　陪审团和法官经常碰到包含概率信息的证据。更好地理解基础概率效应会对
他们的工作颇有助益。有关DNA鉴定的证据特别容易受到基础概率的影响。许多
法学专家对这种证据在法庭上的出示表示担忧，因为陪审团、法官和法医专家很
可能误解DNA匹配概率的真实含义。（参见《增加公正的机率？》，《新科学家》
杂志，1994年6月15日，12－13页）"Improving the odds on justice?", New
Scientist）

　　即使证据显得无可置疑，忽略基础概率也会导致误判。如果对被告不利的证
据非常少——即其犯罪的基础概率很低——那么就算DNA匹配程度极高，我们仍
有权利对"被告就是罪犯"的论断表示合理的怀疑。

　　回到现实中来

　　忍辱负重的气象局如果更加重视基础概率，他们也能从中获益。考虑到天气
预报的难度，气象局已经做得相当不错了。如果他们能在预报时考虑基础概率，
比如在预报时说："如果您只出门一小会儿，那么很可能不用带伞。"这样他们可
能会得到更多的认可。

　　理解基础概率效应能帮我们更好地预测天气，处理法庭上的证据，以及诊断
疾病，不仅如此，它给我们的核心启示——罕发事件很难预测——还能帮我们节
省一大笔很可能有去无回的科研经费。

　　以地震预测为例。过去100年里，地震学家将大笔的经费投入到地震预测研
究中，收获廖廖。尽管如此，研究者们仍固执地寻找那些可能帮我们预测大地震
的种种"前兆"。

　　基础概率效应清楚地告诉我们，这种执着误入了歧途。预测像神户地震这种
强度的大地震当然很了不起，可这样的地震极其罕见，大约50到100年发生一次。
这么低的基础概率意味着，如果要让决策者下定决心，命令大规模人口转移的话，
任何地震前兆都必须极其可靠。

　　粗略的计算显示，任何有价值的地震预测手段，其观测的地震前兆的准确率
必须达到98%以上。到目前为止，任何所谓的地震前兆都远未达到这个准确率。
而且，不断有证据表明，地震就像雪崩一样，本质上是极不稳定的"临界"现象
（译者注：关于临界现象，可以参看
http://www.-reader.org/blogs/fangzhouzi/2008/06/05/%E5%83%8F%E6%B2%
99%E5%A0%86%E4%B8%80%E6%A0%B7%E5%B4%A9%E5%A1%8C/）。所以，找到高度准确
的地震前兆的可能性微乎其微。

　　虽然长期以来基础概率效应主要出现在心理学研究当中，但它远非仅供学者
娱乐消遣的简单逻辑游戏。理解它可以帮助我们更好地做出决策，并避免无用功。

　　* * *

　　计算罕发事件的发生概率

　　如果有预报说某件事会发生，则可以用概率论来计算其发生概率，可这办法
不仅枯燥而且不易理解。"列联表"是个更便捷、易懂的工具。最简单的列联表中，
两列代表两种可能的实际状态，比如下雨和没下雨；两行代表相应的预测，比如
预报有雨和预报无雨，只要几次简单的算术运算，你可以填满这个表格，并从中
得到任何你感兴趣的事件概率。

　　以降雨预报为例。有两项关键数据：每小时的降雨基础概率10%，以及预报
准确率80%. 这意味着在100次为期一小时的外出中，有10次会碰上下雨，90次无
雨。把这写到每列的标题中去。

　　先来填第一列，标题为"下雨"，我们知道，在确实下雨的情况下，给定预报
的准确率为80%，则预报有雨的次数为8次（10*0.8），预报无雨的次数为2次
（10*0.2），依次填入第一列。

　　同样地，在90次无雨的外出中，预报有80%是准确的，即预报无雨72次
（90*0.8），预报有雨18次（90*0.2），填入相应单元格内。这样表格就完整了，
所有信息一目了然。

　　比如，从第一行中，我们马上可以知道在这100次外出中，共有26次预报有
雨，可实际上只有8次是准确预报的，准确率仅为30%. 不过，在74次无雨预报中，
有72次是准确的预报，准确率高达97%. 所有的预测系统，从地震预测到癌症诊
断，只要给定事件基础概率和预测准确率，都可以用这种方法分析相关的概率。

　　 下雨(10) 没下雨(90)
　　预报 有雨 8=10*80% 18=90*20%
　　预报 无雨 2=10*20% 72=90*80%

预报有雨次数 8+18=26
准确预报次数 8
准确率 8/26=30.8%

回复 1：我觉得罗伯特·马修斯错误应用了条件概率概念. 0.1%的患病率,与95%的准确率,不是相关事件.
有病的人如果检查是95%阳性，5%阴性。假定是“有病”。
没病的人如果检查是95%阴性，5%阳性。假定是“无病”。
都是条件概率。
另外，你也可以看出来，95%的准确率是很好的检验手段了。但，如果只有2%地把握判断有没有病，几乎等于没说。要是这样的话，上医院还不如烧香。 ^_^

回复 2：
假设1000万人，1万人患病。
检验准确率95%。一万病人中有9500阳性，500阴性。999万没患病的人中，9490500人阴性，499500人阳性。
检验阳性人数：509000人，其中患病人数：9500人，站检验阳性人数的1.8664%。

患病 (10000) 没病 (9990000)
检验 阳性 (10000*95%=9500) (9990000*5%=499500)
检验 阴性 (10000*5%=500) (9990000*95%=9490500)

检验阳性人数 499500+9500=509000
检验准确人数 95000
准确率 95000/509000=18.664%

问题的关键是：临床检验不同于普查，在诸多检验中，进行千万人普查的机会是不多的。举个例子：甲胎球蛋白阳性来诊断肝癌，在10000万人中，永远不可能所有5%的假阳性全被查出来。换个说法：临床检查一共查出100例阳性样品，问其中的假阳性占多少？我觉得答案应该是5%。临床检查并不是人群的随机抽样，因此整体人群的患病率在这里不起作用，换句话：临床检查的人群总体不等于正常人群总体。医生只会给肝区疼痛，有慢性肝炎和肝功不正常等怀疑可能患肝癌的人开甲胎球蛋白的专门检查。

可以取代Nero的最佳免费DVD/CD刻录软件!

http://web.wenxuecity.com/BBSView.php?SubID=computer&MsgID=150348

Nero is one of the best piece of softwares available in the market for CD/DVD Burning Purposes, but it’s quite expensive. It’s bulky and comes with a lot of programs that you’ll never really use.

Why not try some freeware CD DVD Burning Apps applications that can serve as an alternative to Nero Burning ROM for your Windows PC?

1. CDBurnerXP

This is by far the most popular free alternative to the expensive Nero Software. Works on Vista as well, and here’s a features-highlight from CDBurnerXP:

• Multi Language Interface
• Support for Blu Ray/HD DVD
• BIN > ISO converter included

Download CDBurnerXP

2. InfraRecorder

This is the software that I currently use on my computer as a Nero alternative, and I must tell, it’s damn simple and straightforward. Has all that you might expect in a CD DVD Burning software like Nero. Some highlights from this Nero alternative include:

• Support for Multi-Session
• Very Light on Resources
• Support for not just ISO, but also BIN and CUE images
• Burning on Dual Layer DVDs is supported

Download InfraRecorder

3. DeepBurner (Free Edition)

Another good alternative. Available both as a regular and portable app. So if you want something you can carry around on your USB Flash Drive then check this one out.

• Supports both internal and external CD/DVD writers.
• Burn any data, copy any disc.
• Create multiple disk copies.
• Make ISO CDs.
• Print CD labels.

Download DeepBurner

4. Ultimate CD/DVD Burner

This simple solution can serve all your basic CD burning needs. Similar to above Deepburner, there is a portable version. The only thing I don’t like about this one is interface, it reminds of Windows 95.

• CD Photo Albums creation
• Easy Backup CD creation
• Disk copying
• ISO Creation and Burning

Download Ultimate CD/DVD Burner

5. Burn At Once

Burn At Once is another good CD/DVD Burning Application. It can copy discs on the fly and has the following unique features:

• Tagging of media by importing data from FreeDB
• Multi Language Support
• Drag and Drop Interface

However, it looks like support for this application is stagnant at the moment.

Download Burn At Once

6. Burn Aware Free Edition

This relatively new software is creating waves, and it’s intuitive interface is attracting users. It’s very light on resources, and supports every other format that you can imagine (including Blu Ray). Burning is fast because this software burns data directly on to the medium, instead of waiting for ‘hard disk staging’

Download BurnAware

7. ImgBurn

Another software that’s worth mentioning here is ImgBurn. It’s a lightweight CD / DVD / HD DVD / Blu-ray burning application that can be installed both on Windows and Linux PCs. It’s probably the most extensive out of all. If you’re looking for something advanced and with lots of features ImgBurn seems like a good candidate.

Download ImgBurn