The problem that I, and a lot of people have, is that Amazon has many different sites and you need to sign up to the affiliate program for each site separately.

For example, if you are registered with the amazon.co.uk affiliate program, you cannot use that affiliate tag on a link to amazon.com.

Make money all over the world

Perhaps you think “why not just sign up to each affiliate program separately? This is not a problem. The problem comes with making sure your website visitors see a link to the correct Amazon store for them.

For example, visitors from the US should see a link to amazon.com, with your amazon.com affiliate tag. Visitors from France should see a link to amazon.fr with your amazon.fr affiliate tag.

Identifiying The User’s Country With Maxmind

Maxmind is a service that takes any IP address supplied by you and tells you the country, state (or region), city and latitude and longitude. The service is not free, but the prices are very reasonable. You can get 200,000 IP address lookups for just $20. There is nothing to download as you simply make your lookup request to the Maxmind servers. Also, the credits never expire so you don’t have to use them up within a certain time limit.

You have to supply the service with the IP address to look up first. There are many ways to detect a user’s IP address in PHP, but this is how I go about doing it:

function getRealIpAddr()
{
    if (!empty($_SERVER['HTTP_CLIENT_IP']))   //check ip from share internet
    {
      $ip=$_SERVER['HTTP_CLIENT_IP'];
    }
    elseif (!empty($_SERVER['HTTP_X_FORWARDED_FOR']))   //to check ip is pass from proxy
    {
      $ip=$_SERVER['HTTP_X_FORWARDED_FOR'];
    }
    else
    {
      $ip=$_SERVER['REMOTE_ADDR'];
    }
    return $ip;
}

Once you have detected the user’s IP address you need to access the Maxmind web service. There is some sample code on the Maxmind website, but I will guide you through it now using PHP.

Maxmind offer two services, one that returns the location and the other which returns the location and ISP details. We will just use the location one as all we are looking for is the country.

The service is accessed with the following URL:

http://geoip3.maxmind.com/b

It takes 2 parameters:

• i (lowercase I) – which is the IP address you are checking
• l (lowercase L) – your own licence key

Here is an example with a dummy IP address and licence key

http://geoip3.maxmind.com/b?i=123.54.566.4&l=123abc456def

This is the PHP code that we are going to use to return our information:

$query = "http://geoip3.maxmind.com/b?l=" . $license_key . "&i=" . $ipaddress;
$url = parse_url($query);
$host = $url["host"];
$path = $url["path"] . "?" . $url["query"];
$timeout = 1;
$fp = fsockopen ($host, 80, $errno, $errstr, $timeout)
	or die('Can not open connection to server.');
if ($fp) {
  fputs ($fp, "GET $path HTTP/1.0\nHost: " . $host . "\n\n");
  while (!feof($fp)) {
    $buf .= fgets($fp, 128);
  }
  $lines = split("\n", $buf);
  $data = $lines[count($lines)-1];
  fclose($fp);
} else {
  # enter error handing code here
}
echo $data;
$geo = explode(",",$data);
$country = $geo[0];
$state = $geo[1];
$city = $geo[2];
$lat = $geo[3];
$lon = $geo[4];

After all this effort we get what we were trying to find.

Contained inside the variable $country is a two letter country code that indicates the visitor’s location.

Those that are relevant to us are:

• US – USA
• GB – Great Britain (UK)
• CA – Canada
• FR – France
• DE – Germany
• JP – Japan
• CN – China

These are all the countries that Amazon operates websites in. The ones that you choose to use depends on the target audience of your website. Most English langage sites will just focus on the US, UK and Canadian markets.

How to Display The Correct Product Links

Now we finally get to the meat of the article. We have joined our appropriate affiliate programs, detected where our visitor is, so now we need to display the correct code.

if($country == "UK")
{
     // Display our UK affiliate code here
}
elseif($country == "CA")
{
     // Display our Canadian affiliate code here
}
else
{
     // Display our US affiliate code here
}

What this code does is displays the Canadian or UK affiliate code if the user is in either one of those countries. If not, then display the US code. The US code will be displayed as a fall back if the script cannot detect the location, or the visitor is in one of those countries we are not catering for.

Geographic targetting is not 100% reliable and you should always have a fallback incase the scripts encounter any problems (like displaying affiliate code for the US site if the script is not sure where the visitor is from).

I hope this code helps you to make a lot more money from Amazon. Feel free to post any comments or questions below.

Currently, 2.5% of this website’s search traffic comes form people searching for “plagirism.”

According to Google’s Adwords tool, plagirism has 4 400 searches per month. This may be just a tenth of what plagiarism gets, but plagirism (incorrect spelling) has 10 500 results in Google, while plagiarism (correct spelling) has 6 360 000. There are a lot of numbers to look at, but can you see which would be easier to penetrate the first page of search results for?

Here is the data summarised in a table:

I then decided to look at a list of commonly mispelled words to see how I could gain traffic by taking advantage of these. There are a lot of words that would be tricky to drop into your website, the likes of “consensus,” “embarass” and “foreward” would be difficult to find a place for.

Here are some good words you might want to look at, together with their monthly search volume and number of results in Google. You can also take a look at a list of commonly mispelled words in Wikipedia.

Some Caveats

As you probably know, whenever Google encounters a spelling error it will offer to redirect you to the correct spelling. Obviously this will take away some of your traffic and you need to bear this is mind, particularly when looking at the number of montly Google searches.

A Classic Captcha

It is probably not unreasonable to say that everybody who has ever used the Internet has come across a Captcha before, even if they did not know that is what it is called. Just to get everybody up to speed, a Captcha is an image of a word or random text that you have to type into a box to verify that you are a human.

Captchas on the web are a pain for visitors. As hackers and crackers become more sophisticated, Captchas need to become more difficult to break. If you regularly see Captchas that you cannot solve then you are not alone. It is not just people with physical or visual impairments that struggle with Captchas, what about people with javascript or images disabled? Or what about people using mobile phones?

So we know what the problems are with Captchas, but what are the alternatives?

Image Recognition

Kitten Auth – Asking you to Recognise Images

Instead of having an image of some text that you type into a box, you will see a selection of pictures or photographs. You will be instructed to “click on the pig” or “choose the red flower.” So long as you follow this instruction correctly, the system will verify you as a human.

For

• Harder for a computer bot to recognise a picture than a string of text

Against

• Still requires the use of images, so a problem for visually impaired users
• Usually requires javascript to be enabled
• Only a certain number of images can be used, so eventually they will start to be repeated

Sound Recognition

This is often provided as an alternative to regular text Captchas. A user listens to a word or a string of letters that are spoken and types them into the box for verification.

For

• Using it as an alternative for text Captchas doesn’t impede visually impaired users

Against

• Very difficult to implement technically
• Still open to attack by bots that can recognise sound
• If used by itself will discriminate against deaf users

Multiple Choice Question

You add a short sentence with a missing word to your form and give your users a few options to choose from to fill in the gap.

For example you could use “I have a ______ car.” The options could be “red”, “house”, “monkey” and “spoon.” Only a human would be guaranteed of getting the answer right.

For

• No requirements for javascript, sound or images

Against

• If a bot randomly guessed an answer, ther is still a 25% chance it will guess correctly.
• There is a limit to how many questions you will be able to use. A bot could be programmed with all the correct answers.
• It may prove to be a problem for users who do not speak you native language

Hidden Form Elements

This solution sounds technical but it is actually quite simple to implement. Many bots scan the source code of your page for form elements. They give all these elements a value depending on what the field name is. For example if you have a field name called “email,” it is not difficult for the bot to guess what type of data is required in that field.

If you introduce a dummy form field into your form and hide it with a style sheet, the bot will not know it is an invisible field and will insert a value into it. You will find it easier to trick the bot into filling in this field by calling it something like “Name” or “URL.” You can hide a form field by using the style attribute display:none;

On processing the form, any submissions that have a value for the hidden field can be regarded as submissions by a bot and investigated or discarded.

For

• Technically, it is very easy to implement
• No reliance of javascript or images, but does require stylesheets to be enabled

Against

• As soon as a spammer becomes aware of your system, the bot can be modified to avoid filling in the field

Filtering Content

Akismet Can Cleanup Your User Generated Content

In my opinion, the future of preventing spam is about analysing what is being posted. Spammers are using the form on your website in order to post their content and links on your website. The most common areas are blog comments, wikis and forums. Therefore it is useful to think about filtering these posts themselves rather than validating the form that was submitted.

Akismet is a system designed to do just this. It is produced by the people who created WordPress and it analyses the content of what is posted to determine if it is likely to be spam. The system has analysed billions of messages and is highly effective. You do not need to store anything on your own server because Akismet handles eveything remotely.

Akismet is very popular on WordPress blogs but can be integrated to analyse any user submitted content. It won’t stop bots from submitting registration forms, but you can still block any content that they might submit.

For

• No need for installation or storage of messages
• Very, very accurate

Against

• Relies on an external service, so it is beyond your control
• Not free for commercial purposes

In Summary

Google’s Difficult Captchas

Despite the many problems associated with their use, Captchas are very prevalent on the web. This article is intended to make you think twice before you blindly drop in a Captcha. How many potential visitors (or even customers) are you going to lose by putting such a barrier in their way? You need to decide if the trade off is worth alienating a large group of potential visitors, or is a more intelligent solution available?

After all that work it can be very down-heartening to see your users requesting to leave your mailing list. After all, it is potentially a good way to appeal to repeat customers and inform them of any special offers and new products you have available. When they want to leave there is a temptation to make it very, very difficult for them to do so. My advice for you today is one word: don’t.

It is entirely unacceptable to put obstacles in the way of people unsubscribing. In certain countries there are also laws dictating how the process must work and time limits for removing email address from your database. Unfortunately, only the laws of the country where the sender lives can be enforced. So if you live in a different country, your options may be limited.

I took a look through my Google Mail archive today to unsubscribe from all the mailing lists I am on that I don’t really use any more. We will take a look at the unsubscribe process for each and give it a rating out of 10.

First up is Catalink, a company that sends you catalogues about various topics you are interested in. This is what their unsubscribe link looks like in the email.

Catalink's unsubscribe link in the email

The unsubscribe link seemed pretty standard, a small link at the bottom of the mail telling me to log into my account and unsubscribe from there. No direct link to unsubscribe, but that isn’t unusual. Now let’s take a look at removing ourselves from the lists. After you log into your account you will find a list like this.

The Catalink unsubscribe process

I was absolutely amazed by the inconvenience of this page. In order to unsubscribe, you have to check the box, select a reason and click “Unsubscribe” for each newsletter. So for the eight newsletters in this example, I had to repeat the process eight times. This is entirely unacceptable and a clear attempt at trying to make us stay by making the removal process too much work. They obviously think that we won’t bother repeating the process so we will end up staying subscribed.

To add insult to injury they also state at the bottom that it can take a month to be “fully unsubsribed from our system.” What that means I really don’t know. Will you receive emails for a month after unsubscribing?

Convenience rating (out of 10; 1 being bad and 10 being good) 1/10.

Making us log into our account is one thing, but making us unsubscribe from our newsletters one at a time is entirely unacceptable, as is making us wait a month to be “fully unsubscribed.”.

EDIT: Catalink have taken the time to respond to my words and have posted in the comments below. They explain that they do have other unsubscribing mechanisms and an “unsubscribe from all” feature in the pipeline. Since the rating I posted is based on my experience, the score will stand as it is. However, I am always willing to take a look at it again when the changes have been made.

The next service we will look at is Prospects, a website for recent graduatess looking for jobs and training course. As with Catalink, they tell you to log into your account a find the unsubscribe option there. It is quite easy to find and the unsubscribe page is simple and straightforward.

They have multiple newsletters but there is an option to remove yourself from them all at once, which is an essential feature for multiple newsletters.

Convenience rating 5/10.

The next list we are unsubscribing from is an ab builder affiliate newsletter. It uses a commercial email network call Get Response. Since emailing people is what their business is all about I was expecting some very good unsubscribe options from this company. I was not disappointed. There is a very straightforward and direct message at the bottom of the email, “To unsubscribe or to change your contact details, visit…” Visiting this link took us to this page.

Unsubscribing from the abs affiliates newsletter couldn’t be easier

By clicking a single link we are unsubscribed. No having to log into an account. We didn’t even have to click a confirmation button. There is even an option to resubscribe if we have left the list by mistake.

While the praise here is probably more due to the mailing company than the people who send the newsletter, I was still very impressed.

Convenience rating 10 out of 10.

Our next company is Tool Station, a UK based equipment and machine supplier. The message at the bottom of their email looks like this.

Unsubscribing from Tool Station’s newsletter

This is another website that requires us to log into our account. Unfortunately there is no direct link to our account page so we have to follow the link to the homepage. All these steps just to remove us from a newsletter.

Convenience rating 4

The final company we will be looking at is ING Direct, a Dutch bank that is one of the world’s biggest. They send out marketing emails to their customers but they are not very frequent. This is part of the text displayed at the bottom of each email.

The ING Direct unsubscribe process

Not exactly convenient. We all know that people are much less likely to unsubscribe from something if you have to speak to a person to do it. That’s why all those free month trial offers that make you ring the company to cancel your trial are so successful. It seems entirely unnecessary to require customers to do this and it is a blatant attempt to stop people unsubscribing.

Convenience rating 2 out of 10

So What Have We Learned?

The first thing we have learned is that if some of these companies were located in the USA, they would be breaking the law. The US CAN-SPAM Act gives several key requirements:

• Subject lines cannot mislead about the contents of the email,
• Recipients must be given a web-based method of removing themselves from the mailing list, and
• Unsubscribe requests must be honoured within 10 days.

While most of the companies looked at today are outside the USA, most of them failed on at least one of the above points. While they may not be laws in certain countries, they are not exactly unreasonable requests.

If a customer or visitor has requested to leave your newsletter, they are already lost. At least make the process as easy as possible and your company can escape with some reputation in tact.

It is also important to realise who is responsible for what is posted on your website. Firstly, the contributor of the comments is responsible. Secondly, the owner of the site is responsible and thirdly, the ISP or webhost can potentially be responsible.

First of all we should explore the more common abuses. The first is copyright theft.

Copyright Theft and Plagiarism

While these two have been grouped together, they are not always the same thing. Plagiarism refers to using the words or ideas of somebody else and passing them off as your own. Direct quoting without acknowledging the source is plagiarism. It is not limited, however, to direct copying. Using the ideas or original thoughts of somebody else without acknowledgement is also plagiarism. You can read a lot more about plagiarism in our plagiarism article.

Copyright theft usually refers to taking a creative idea of somebody else and reusing it for your own purposes. Examples would include using an image or video and posting them on your website without permission, either in the original or altered state. Note, however that there is a principle of “fair use.” As with everything this differs throughout the world but as a general example if you posted a corporate logo on a website along with a story about them, it is unlikely that the company would be able to sue you. There is an excellent article about Copyright Myths by Brad Templeton if you want to read more.

I don’t want this article to get stuck in legal jargon and defining legal terms, but copyright theft and plagiarism are without a doubt the two biggest offences committed on the internet. It is very easy to cut and paste the work of somebody else without giving credit and there are even special search engines that exist to detect plagiarism.

Posting Software

While this comes under copyright theft, posting downloads of other people’s software is so rife on the Internet it deserves a special mention. Beware of the definition of copyright theft. It will differ based on which country you are talking about. Hosting a copy of Microsoft Windows for download in the USA is most certainly illegal. If the website’s server was in Afghanistan, this would not be a crime as copyright laws are basically non-existant in Afghanistan.

Websites such as The Pirate Bay have exploited such copyright law loopholes and have previously looked to purchase the Principality of Sealand. They could host their servers there and be immune to copyright restrictions.

You should also be aware of free downloads that are taken from a non-free source. With very few exceptions this would contravene the original licence agreement. For example, if you purchased a WordPress template and distributed it to others for free, you would probably violate the agreement.

For such reasons, you should always be sure of ownership of any downloads hosted on your website. Any user contributed submissions should also be screened for any copyright violations.

Violating Terms and Conditions

It is the text that everybody agrees to but very few actually read. When you use practically any service or software, you must agree to the terms and conditions. When you sign up for a webhost, when you purchase Photoshop, when you use PHP or Apache and when you purchase forum software you must agree to the terms and conditions set out by the publishers.

For instance, say you download the latest PHP release and make some modifications to it. You then release it as PHPfoo. This would be a violation of the PHP licence as it makes it clear that any derivative of PHP cannot use “PHP” in its name.

In another example, you purchase a copy of vBulletin. You install it on two of your websites. This is a breach of the licence you agreed to as it only granted you permission to use one instance of the software on one domain or server.

Copying Designs

Two copies of a design. From pirated-sites.com

Again, technically covered under copyright theft, imitating a design deserves a special mention too. Legally speaking this is a very tricky subject even in countries with clear copyright laws. While stealing a logo or photograph is very easy to prove, copying a design is not. It is not uncommon for a designer to look at a website and feel that it is an imitation of their own. It is often difficult to prove it in court. For this reason it is best to contact the site owner directly. If this fails, contact their web host. They are often very keen to avoid any legal issues and most will give you the benefit of the doubt and remove the offending content. There are full details about how to deal with content theives on our plagiarism article.

Violating Accessibility Laws

The UK has the Disability Discrimination Act, USA has section 508, Ireland has the Disability Act 2005 and Canada has the Human Rights Act 1977. All of these laws include sections on accessibility of services for people with disabilities. They all include the Internet, although may not mention it specifically. The W3 maintains a page of links to the relevant discrimination laws for various countries.

The problem with most of these laws is that they were not written with the Web in mind, which means they are not very specific with what needs to be done to make a website accessible. They tend to include ambiguous clauses such as to ensure a website is not “unreasonably difficult for disabled people to make use of its services.” What is defined as a service is not clear either. It is unlikely that a website such as this one would be liable for prosecution. But a website for a cinema chain that requires the use of Flash to book tickets certainly would.

Prosecutions against organisations have been successful under the various laws. In 1999 a blind man called Bruce Maguire lodged a complaint against the website of the organisers of the Sydney Olympic Games. The complaint was upheld and the website was ordered to make changes. They failed to comply and were fined AU$20 000.

The W3 maintains a checklist of points to follow to ensure accessibility.

Security Breaches

Losing customer data or having it stolen from your servers is not only bad for business and customer confidence, it can lead to prosecution. TJK Companies Inc (the owner of TJ Maxx and TK Maxx in the UK), is currently the subject of several class action lawsuits after the details of 45.7 million debit and credit cards were stolen. It is unlikely the full extent of the breach will ever be known but it is regarded as the biggest data breach in history.

This is the main reason why very few online businesses process credit card information themselves. They leave it to a clearing service to do it.

Failure to Maintain a Privacy Policy

The 2003 California Online Privacy Protection Act is a law which requires any website that collects “personally identifiable information” to “conspicuously” post a privacy policy. The policy must describe the types of information that is collected, any methods that exist to alter that information and describe how users are notified of a change in the policy.

Although this law only applies in California similar laws are being prepared in other US states.

Sites that don’t comply will have 30 days to rectify the situation. You can view the full text of the law at this URL.

This article was designed to give you an overview of what actions can cause a website to break the law. It is not designed to scare you and always remember that what is illegal in one country may not be illegal in another. Now you know you have the chance to go back to your website and put things right that you have discovered may be wrong.

Avoid offending people on your website

In the days of simple access to the Internet, defamation of people and companies can prove a significant problem. This problem can often be exasparated by the apparent anonymous nature of the Internet. This assumption is very rarely true, in fact very little of the Internet is truly anonymous. This can lead to people making false accusations, exagerting claims, and just plain lying in the hope that they will not be discovered.

Definitions

Defamation is the making of a statement that makes a false claim which damages the reputation of an individual or a company. When it is in a “transitionary” form such as speech it is referred to as slander. When it is in a “fixed” form such as written or in a photograph is is libel. Due to the nature of how information is published, libel is the more common form on the Internet.

Where Does Libel Take Place on the Internet?

Anywhere where statements or images are “published” is under scrutiny in terms of libel. This includes email, newsgroups, mailing lists, and the World Wide Web itself.

The speed and simplicity of the Internet allows messages to be published instantaneously:

• Email can be sent in a few keystrokes and forwarded to many other with the click of a mouse.
• Messages in newsgroups that are replied to will often go to everybody in the group rather than just the author of the original message.
• Message boards and Internet newsgroups are very rarely anonymous, despite how they may look.

All of these features of the Internet make it remarkable easy to publish statements about people and organisations without really thinking through what you are saying or checking any facts you have stated. Unfortunately for some, this cannot be used as a defence against libel in itself.

Who Is Responsible for What Is Published?

As with most of the libel law discussed, the specific aspects vary from country to country. With most countries, however, the same people are usually responsible for what is published. Firstly, the author of the statement is responsible for what they have written. Secondly, the owner of the website is responsible for what they are publishing. The ISP or webhost is also responsible for what is published on their servers. Often, the first you will hear about any accusations of libel is an email or letter from your ISP saying that the offending page or message has been removed. They very often play on the safe side and remove the material when asked rather than waiting for the decision of a court.

What Can I Do On My Website?

Anything that is published on your website is your responsibility, even if you did not write it yourself.

Taking a proactive approach is the best way to avoid running into any potential situation where you are accused of defamation. As you will see from the possible defences listed below, simply saying “I didn’t know what was going on as anybody can contribute” is rarely a suitable defence. You should pay particular attention to:

• Chat rooms – treat any claims of defamation and harassment seriously. Check the logs and take steps to ban anybody who is caught doing this. You should also make it easy for people to report harassment or abuse. Chat rooms can be particularly problematic as people often (falsely) believe they are anonymous.
• Message boards – generally, message boards (or forums) are moderated, but posts are rarely screened before they are published. If a potentially defamatory post is found it should be deleted and steps taken to warn or ban the poster. Most decent message board sofware comes with built in reporting features. Users can “report this post” if it causes offence. Again, make sure you take these reports seriously. If you have been notified of an abusive post and you do not act then you cannot use innocence as a defence.
• Product reviews and comments – users who post factual comments are OK. This comes under the fair comment rule. Generally, you can say what you like so long as it is true. As usual, there are exceptions. Take a look at the defences section below. Care should be taken when users start to post their own personal opinions, especially if they are negative. “I bought this product and it broke” is a fair, factual comment. “I bought this product and it broke. The company are lying thieves” is probably defamatory as the user’s opinion is involved. Product reviews should always be monitored and contentious comments should be removed.

The line between free speech and defamation on the Internet is often very narrow so most people tend to go on the side of caution. If you are unsure if what is published is libellous, you need to decide if you are going to remove it or not while you investigate.

Who or What Cannot Be Libelled?

Albert Einstein cannot be libelled

In most countries the dead cannot be libelled. The clearest exception to this is the Philippines. Under their law, the dead can be libelled.

For example, were he still alive and I claimed that Albert Einstein was a thief, I could be prosecuted for libel. If a historian wrote a book claiming the same thing after he had died, there would be no risk of prosecution for libel. There are some exceptions to this however.

After the death of former British Prime Minister William Gladstone, author Peter Wright published accusations in 1927 that he used to sleep with prostitutes. The family of Gladstone publicly called him a liar and so he sued for defamation. When Wright couldn’t prove his accusations about Gladstone in court, his case was thrown out.

Additionally, if false accusations are made about somebody who is dead that affect his estate or surviving family, this can be considered libel.

Another contentious point is accusations made about people who have not yet been proven to be dead. Care must be taken when assuming that somebody is dead just because they have been missing for so long.

Defences Against Libel

There are various different defences that people commonly employ to defend themselves against accusations of libel. Since we are just looking at how libel laws affect Internet publishers, the most useful ones in this situation are listed below.

• Truth – in many countries proving a statement to be true will be sufficient defence against an accusation. Usually, the information being revealed must be shown to be in the public interest. This means that publishing something simply because it is true cannot always be guaranteed to be a suitable defence. Again in the Phillipines, truth itself is not a suitable defence.
• Good faith – if a statement was made in good faith it may be regarded as suitable defence. However, the belief must be reasonable.
• Innocence – similar to good faith, if a defendant acted without knowledge that the statement was made they cannot be held responsible. For example an ISP cannot be held responsible for a message being posted on an unmonitored newsgroup. However, if they are notified and do not act, they cannot use innocence as a defence.

Examples of Libel Action

Keith-Smith v Williams

In 2006 the case of political candidate Michael Keith-Smith established that English libel laws apply to two individuals on the Internet rather than an individual and an ISP or publisher.

After an argument with Tracy Williams in an Internet chat room hosted by Yahoo, comments were posted by Williams calling him “Nazi”, “racist bigot” and “nonce.” His solicitors petitioned the court to obtain the poster’s details from Yahoo.

Despite court proceedings, she continued the abuse. She was ordered to pay £10,000 GB compensation and £17,200 GB costs.

Godfrey v. Demon Internet Service

Laurence Godfrey, A British physics lecturer, alleged that somebody had been posting messages to the newsgroup soc.culture.thai. The message were forged to make them look like they were sent by Godfrey.

He alerted the ISP, Demon, and asked them to remove the messages. They did not remove the offending material and it remained there for 10 days, before it was automatically deleted like any other messages.

He sued for libel and cited Demon’s failure to remove the messages as the reason. As pointed out above, innocence is only a defence if the defendant is unaware of the comments. Under English law, ISPs are not regarded as publishers unless they have been notified of the material.

Demon settled out of court and paid £15,000 GB damages and about £250,000 GB in expenses.

Godfrey went on to bring about a number of successful libel actions including against Toronto Star, University of Minnesota and Cornell University.