Skip to page content or skip to Accesskey List.
Search evolt.org
evolt.org login: or register

Work

Main Page Content

Spam-Proofing Your Website

Rated 3.79 (Ratings: 17) (Add your rating)

Log in to add a comment
(36 comments so far)

Want more?

 
Picture of optimization

Dan Thies

Member info | Full bio

User since: September 28, 2002

Last login: September 28, 2002

Articles written: 1

Spam-Proofing Your Website

Anyone who operates their own website knows that you need to provide a way for visitors to contact you by email. The big challenge is providing easy email access to your visitors, without letting junk mail (SPAM) flood your email inbox. The techniques described in this article have enabled me to dramatically reduce the amount of junk mail I receive on all of my websites.

Preparing and Preempting

You need a couple things before you can really take effective action against SPAM. Your email software must be capable of filtering incoming email. All of the major email applications (such as Eudora, Outlook, and Pegasus) support filtering. We will use multiple email addresses to allow us to filter out SPAM and identify the source - you can't combat SPAM effectively without them.

You need to use a website hosting provide that allows unlimited email aliases or addresses, and/or a catch-all email address. An "alias" is an email address that forwards to some other address (for example, webmaster@domain.com forwarding to your real email address). A "catch-all" email address will forward any emails sent to unknown addresses in your domain.

For my own websites, I just use the catch-all, so that every message goes to my real email address. If you have more than a one-person operation, however, multiple email accounts and aliases are pretty much a necessity. Any email address you use online could become a target of spam. If your hosting provider is especially good, you may even be able to create email aliases that automatically delete all incoming messages.

Fighting Back

The first step in fighting back against the spammers is understanding where they get your email address. You must diligently protect your email address, if you ever hope to stop them. Once your email address gets into the wrong hands, it will be sold on CD-ROM (via junk mail, of course) to thousands of spammers. Once that happens, you've lost the fight.

Spam Source #1: Domain Name Registrations

When you register a domain name, you must provide a contact email address. If you give them your real email address, you've just given it to everyone, including the spammers. Instead, use a portable email address (like Hotmail) to set up your domain.

If you have multiple domains, you can also use an alias (domains@yourdomain.com) on your primary domain for all registrations. With an alias, you can use your email software to filter out and save any emails that come to that address from your registrar's domain.

Spam Source #2: Web Forms & Email Newsletters

If you give your real email address on any web form, or use it to subscribe to an email newsletter, you are asking for trouble. Instead, create a unique email address for each website or newsletter. I just use the website's domain name for this.

For example, if you subscribe to the "evolt.org" mailing list as "evolt.org@yourdomain.com" and let your catch-all address route it to you, you will always know where the email came from. If that address ever starts receiving junk mail, you can filter it out using your email software.

If you submit to search engines or free-for-all links pages (FFA's), use a unique email address every time. FFAs, in particular, are famous for flooding the world with junk mail. Once you've given an email address to an FFA, you may as well forget about ever using it again.

Spam Source #3: Your Website

The biggest source of email addresses used by spammers is your website. Most websites list multiple contact addresses, etc. Any time an email address appears on your website in plain text, even if it's hidden in a form field, you're opening yourself up to having that email address captured.

To combat this menace, I've developed a set of JavaScript snippets that will meet almost every need you have to display your email address to the public, without allowing spambots to see it.

The Big Battle: Securing Your Website From Spambots

Almost every website operator wants search engine spiders to visit. After all, search engines are the best source of free traffic on the web. In the event that you don't want them to visit, they are easily kept at bay with a properly formatted "robots.txt" file.

Unfortunately, there's another group of spiders out there crawling the web, with an entirely different purpose. These are the spiders that visit site after site, collecting email addresses. You may know them as spambots, email harvesters, or any number of unpublishable names.

When it comes to controlling these rogue spiders, a robots.txt file simply won't get the job done. In fact, most spam robots ignore robots.txt. That doesn't mean you have to give up, and just let them have their way. The following techniques will stop these spiders in their tracks.

Technique #1: Use JavaScript To Mask Email Addresses

One of the weaknesses that spiders of all kinds suffer from is an inability to process scripts. Adding a small snippet of JavaScript in place of an email address effectively renders the address invisible to spiders, while leaving it accessible to your visitors with all but the most primitive web browsers.

In the examples below, simply substitute your username (the first half of your email address, everything before the @ symbol) and your hostname (everything after the @ symbol). To use the scripts, just insert them into your page's HTML wherever you need them to be displayed.

Example 1: Creating A Spam-Proof Mailto Link

This snippet of JavaScript code creates a clickable link that launches the visitor's email application, assuming that their system is configured to work with "mailto:" hyperlinks. You can replace the link text with your own message, but see example 2 if you want to display your email address as the link text.

<script language=javascript>
<!--
var username = "username";
var hostname = "yourdomain.com";
var linktext = "Click Here To Send Me Email";
document.write("<a href=" + "mail" + "to:" + username +
"@" + hostname + ">" + linktext + "</a>")
//-->
</script>

Example 2: A Spam-Proof Mailto Link With Your Email Address Showing

Some visitors won't be able to use a mailto link. This snippet shows your email address in the link so they can copy and paste, or type it by hand:

<script language=javascript>
<!--
var username = "username";
var hostname = "yourdomain.com";
var linktext = username + "@" + hostname;
document.write("<a href=" + "mail" + "to:" + username +
"@" + hostname + ">" + linktext + "</a>")
//-->
</script>

Example 3: Display Your Email Address Without A Mailto Link

Here's a snippet that displays your email address a clickable link:

<script language=javascript>
<!--
var username = "username";
var hostname = "yourdomain.com";
var linktext = username + "@" + hostname;
document.write(username + "@" + hostname)
//-->
</script>

Technique #2: Use A Contact Form

Sometimes, the sheer volume of legitimate email from real visitors can become a burden. In this case, a simple solution is to remove your email address from your site entirely, and use a contact form. There are dozens of free ASP, Perl, and PHP scripts available online that will allow your users to fill in a form, and send you an email. Most hosting providers now offer this service for free to their customers.

A contact form can enable you to deal with a higher volume of mail, by allowing you to pre-sort different types of message. This is easily accomplished by creating a drop-down menu with different options (e.g. customer service, billing, tech support, etc.) that will populate the subject line of the email message, and/or change the email address to which the form is sent.

Since many spambots simply read the entire HTML source of the page, looking for anything that looks like an email address, your contact form will not protect you, if you include your email address in the HTML for your contact form (for example, as a hidden field). You can use JavaScript, as in the example below, to mask the address, or if you have the skill, you can embed the email address in your form processing script, where nobody can find it.

Example 4: Masking The Email Address In A Form Field

Instead of simply listing your email address in a form field, use the snippet below to replace the form field that contains your email address.

<script language=javascript>
<!--
var username = "username";
var hostname = "yourdomain.com";
var linktext = username + "@" + hostname;
document.write("<input type=hidden name=email value=" +username + "@" + hostname" + ">";
document.write(username + "@" + hostname);
//-->
</script>

All contact forms, regardless of the language used, will work more or less the same way. Users will fill out a form, which is processed by a script on your server that emails the submissions to you. Because the script that runs on your server, your visitors never see the contents of that script. Hiding your email address in the script provides the greatest security, but this does require some programming knowledge. Form-to-email scripts that are offered by hosting companies almost always require the email address to be included as a hidden form field in your web page.

Disadvantages of an all-JavaScript approach

The main drawback to using JavaScript is, of course, browser compatibility. While the most popular web browsers all support JavaScript, a small percentage of users will be unable to see them. Only you can make the ultimate decision on whether the needs of these users are greater than the need to stop spammers. Offering a contact form to those users can reduce the problem for these users.

If you take this approach, it is much better to hide your email address within the script itself, rather than in the HTML code of your page. If you can't hide the address in the script, use an email alias, so that you can change the email address in the form from time to time, whenever the spam gets out of hand.

Advanced Techniques: URL Rewriting

Both the Apache and IIS web servers have plug-in URL-rewriting modules that can be used to provide additional protection to your website, by redirecting queries from known spambots to a blank page, or to another website. These techniques are beyond the scope of this article, and using them will slow your server down, if only a little. For a good discussion on using this technique, including its use to combat spambots, see http://www.webmasterworld.com/forum13/687.htm.

URL-rewriting is a powerful technique, however, and should not be overlooked. In addition to its potential value in deterring spam, it can also be used to prevent users from downloading your website with offline browsers, MS FrontPage, etc. If your content must be protected from unauthorized copying or other misuse, judicious use of URL-rewriting may be exactly what the doctor ordered. Be aware, though, that not everyone attempting to download your website is doing so with bad intentions.

For a good example of what you might want to show those who try to download your site, see this page: http://www.purplemath.com/terms.htm.

Thanks for reading...

I hope that this tutorial has given you a clear understanding of how to protect your website, and your email address, from spammers and spambots. Your feedback is welcome. If you have any questions about this article, feel free to contact me through my website. The (spam-proof) email link can be found at the bottom of my home page.

I wish you success...

Dan Thies has been helping his clients (and friends) promote their websites since 1996. His latest book, "Search Engine Optimization Fast Start," offers a simple, step by step plan to increase your website's search engine traffic.

Other interesting reads

Submitted by garrett on October 23, 2002 - 03:43.

Good article with a lot of useful tips.

It's also worth having a look at Daniel Cody's two articles on stopping spambots:

Lot's of information on how to configure Apache to stop spambots in their tracks from a server administrator point of view.

login or register to post comments

For IIS admins, yes you can

Submitted by optimization on October 23, 2002 - 07:05.

Although I don't use anything but Apache, you can also do URL-rewriting on the Microsoft web server with IISRewrite:
http://www.qwerksoft.com/products/iisrewrite/
It's compatible with mod_rewrite, which should make life easier for you puppets of the Microsoft empire, since there's plenty of Apache documentation all over the web.

login or register to post comments

accessibility issues

Submitted by branko on October 23, 2002 - 08:03.

Some of the methods you mention make one less accessible. It is all fine and dandy to use a Hotmail account for registering a domain, but if you never read that mail, you've just made yourself that bit less accessible. (Not to mention that you may have broken the rules of the domain controller by doing so.)

I use the 'unique' address myself a lot. So now I know that most of the spam I get on my business account I get through my listing on the yellow pages. Then again, most of my prospects come through the yellow pages, so you won't hear me complain.

What makes spammers different from almost everyone else, is that they use automated means to harvest your address. Perhaps there are legitimite reasons for harvesting e-mail addresses, but I don't know of any.

Cameron Gregory of bloke.com (see the bottom of the page) makes handy use of this fact, by offering e-mail addresses on his web site that he will not read. A human visitor will read this and refrain from using the fake addresses. The spammer will send his warez to both the real and the fake addresses, and will identify himself as a spammer this way.

In the end, there are only two effective methods for fighting spam, and those are to A) fight the spammers, and B) make spamming cost prohibitive. All other methods are at most damage control (spam filters) and in the worst case burying your head in the sand like an ostrich (spam blocks using JavaScript).

(A) requires that, for instance, you will only vote for politicians who take an extremely hard stance on spammers. Even if they do not get voted into office, this will at least give a strong signal that people won't stand for spam. A lot of politicians still see spam as a legitimate way to advertise. It should be made clear to them that it is not.

(Ironic anecdote: spammers recently won a court case in the Netherlands where the judges found that changing an e-mail address is so easy that they saw no reason to prohibit spam. Following this opinion, a spammer started spamming politicians, and suddenly a different judge found that spam should be prohibited. I guess it matters who you harass. That is not a nice signal for any democracy to emit.)

(B) requires a different e-mail protocol. There is some talk about a protocol in which messages remain on the senders' e-mail server until they get collected by the user.

Both methods take time and require us all to take our responsibility.

You have to wonder whether it is fair to punish your users for the behaviour of spammers though.

login or register to post comments

no argument here, but..

Submitted by optimization on October 23, 2002 - 09:10.

I don't really see this as "punishing users." It's important to be accessible, but I just don't see it in the extremes either way. If you have an email address on your site that uses Javascript and a contact form for 99% of the other folks, it's hard to argue that you can't be reached.

These scripts aren't really pushing the envelope of browser compatibility. Most users who would have a problem with this approach are also unable to use the majority of commercial websites, period.

My goal here was to offer some help to folks who don't necessarily have your level of skill. It's easy enough to implement, that the do-it-yourselfer can usually handle it. These folks can't "make spam cost prohibitive," (can you?) so damage control (what you would call the 'ostrich' method) is the only option.

WRT domain registrations, you obviously can't use an email address you never read - renewal notices, etc. are going to come to that address, but a unique address is essential. I have an email account that receives over 300 junk mails a day, and it has never been published anywhere except WHOIS.

That's enough for now, gotta go bury my head in the sand again, but I'm taking the straw man with me.

login or register to post comments

A server-side approach

Submitted by simonc on October 24, 2002 - 10:50.

I've come up with a server-side approach that uses a couple of tricks to generate a live mailto link which doesn't depend on javascript. I'm not completely happy with it yet, but it does offer a reasonable solution.

Feedback and suggestions welcome.

login or register to post comments

Attributes

Submitted by g1smd on October 24, 2002 - 13:50.

Don't forget to add the type="text/javascript" attribute to the script tags if you want the code to validate.

It seems like nearly every JavaScript tutorial on the web forgets to mention this point.

login or register to post comments

An almost bullet proof server side technique

Submitted by skunk on October 24, 2002 - 14:33.

I came up with an idea a few months ago to stop spam bots which (in theory) should be pretty muich bullet proof: Hide your address behind a web form using the POST method:

http://simon.incutio.com/archive/2002/09/11/newFormOfSpamProtection

In order to grab my address spam harvester bots would have to evolve to the point where they identify ALL forms on a page and submit every single on of them - this seems like way too much hassle to be worth bothering with, especially as the kind of people who take that much effort to hide their email addresses are almost certinaly not the kind of people to fall for Nigerican con artists or bizzare get-rich-quick schemes.

login or register to post comments

Free contact form scripts

Submitted by Martin Tsachev on October 25, 2002 - 06:25.

I wouldn't use a free contact form scrpt. Defining the recepient email as a hidden form field is unsecure enough for me.

Otherwise I wouldn't call Opera (or any other feature rich browser) with JavaScript turned off one of the most primitive web browsers.

login or register to post comments

Very useful!

Submitted by k1 on October 27, 2002 - 05:15.

I had a low profile site, and do not get a lot of exposure. Of course, (maybe) unfortunately, I submit my site and got listed. (at the time i was happy)...Then some wierd mail started coming in to which I have no idea where or when I ever give out my mail. Luckily, I've got only a few such mail, and your technique my help me in the longer run... Good article!

login or register to post comments

Another tool

Submitted by cre8d on October 29, 2002 - 12:27.

Check out Mean Dean's Anti-spam, eMail Obfuscation Tool to Ward-Off Nasty Spambots -- very handy.

login or register to post comments

Only compromises

Submitted by jma on October 30, 2002 - 13:28.

While non-JavaScript solutions might sound promising, they can not really provide spam-proofed web pages. Only the most basic spambots will fail on them. Stripping tags (after looking for possible email addresses), lowercasing for easy text search and converting HTML entities like &amp;#64; or other escapes is not taxing enough to warrant their exclusion in address detection algorithms. In fact, they're the easiest to code. It's much harder to extract obfuscated email addresses like "account at domain dot com." And even that won't be hard to break -- after all, the email addresses must be usable by real people. Even replacing the at-sign and dots may be too much.

Only JavaScript can be considered proof against email address gatherers. But email addresses protected that way won't stay protected forever. There are already a number of easily integrateable JavaScript translators which could be integrated into email gatherers. The opportunities for spammers are wide; had they not an abundant supply of plain text email addresses they would propably invest more money on email address detection algorithms. With the current "techniques," we can consider ourselves only lucky, not clever.

login or register to post comments

Re: Only compromises

Submitted by Martin Tsachev on October 30, 2002 - 13:46.

Well that leaves out only choice to not display any email addresses at all. The problem is that you sometimes have to do it and if you provide any information you can expect that one day or another someone may use it for things you didn't want to.

Does anybody have stats of how many search spiders support extracting emails encoded by some of the methods mentioned and what are the trends in technology advancement?

login or register to post comments

Compromise

Submitted by bobince on November 5, 2002 - 16:53.

It's perfectly possible to use a JavaScript deobfuscator that still gives non-JavaScript users a fighting chance. For example: http://and.doxdesk.com/personal/posts/wd/20020611-spam.html

login or register to post comments

Will this work?

Submitted by k1 on November 12, 2002 - 09:15.

Eversince I believed that my site was visited by a spam bot, (coz i received (luckily) 1 spam mail, for every e-mail link on my site), I started thinking about how to counter it and came up with this

First I make a file, email.php which contain the e-mail information. For example:

code:
--------------------------------------------------------------------------------

<?php
$domain1 = "example.com";
$k1 = "k1@".$domain1;
?>

--------------------------------------------------------------------------------

This file resides in a folder called email which is CHMODed 111.

Then in the file that needs to show the email, I include the file above, and use


code:
--------------------------------------------------------------------------------

<?php
print $k1;
?>

--------------------------------------------------------------------------------

to where ever I need to print out the e-mail link...

But I really wonder if it will really work?

login or register to post comments

Just realised the above won't work!

Submitted by k1 on November 12, 2002 - 10:15.

Sorry.../me making a fool of myself...(at least I've been told it won't work)

login or register to post comments

cfabort

Submitted by leftCorner on January 14, 2003 - 12:56.

I wrote a Cold Fusion script, which one can include in application.cfm for Cold Fusion web sites, that simply doesn't load the site for Email Harvesting spiders.

Kill Email Bots

It requires you to resreach and add to the spider list for time to time, and of course you need Cold Fusion, but it certainingly can improve your odds (and anyone who posts their email address to your site) for a spam free existence.

login or register to post comments

How do you identify the spider?

Submitted by optimization on January 14, 2003 - 19:13.

There are a lot of cloaking scripts, and you can use URL-rewriting with mod_rewrite etc, but that assumes that the spider gives you its identity. Most of them are capable of posing as whatever user agent the spammer chooses. How do you tell which "Internet Explorer 5" user agent is a real user, and which one is a spider?

login or register to post comments

Posers

Submitted by leftCorner on January 15, 2003 - 12:42.

You can't catch them all. If I can stop one bot, and maybe 100s of spam emails down the road, it's worth it...

login or register to post comments

one more layer of defense

Submitted by optimization on January 15, 2003 - 13:40.

Agreed - the server is one layer of defense. Everything you can do to slow them down helps. Once your email address makes it onto enough of the CD-ROMs, you can expect dozens or hundreds of junk mails a day. Apache, ColdFusion, and IIS all have the ability to rewrite or redirect based on the user-agent, although it may cost a few bucks with IIS. Anything else and you can still run CGI scripts that do the same thing.

login or register to post comments

blah [ at ] blah [ dot ] blah?

Submitted by trfc791 on January 19, 2003 - 05:46.

For me, I see only one really spiderproof method right now: posting your email in an address such as:

trfc791 [ at ] f2o [ dot ] org
trfc791 ( at ) f2o ( dot ) org
trfc791 { at } f2o { dot } org
or something along those lines. You could perhaps argue that in future spambots may be able to filter out characters that should not occur in email addresses when they appear in the format blah {at} blah {dot} blah and harvest the address anyway, but I don't see it coming in the next few months or so. Even if that is the case, you could still post your address as trfc791.at.f2o.dot.org which would be quite readable to the user, but will get messed up by spambots who will probably intepret it as trfc791.@.f2o...org or something like that. The day spambots become that smart will be the day I use dots in catchall addresses (my.name.is.grace.at.trfc791.f2o.org) which, hopefully, will not be soon. Another currently plausible way is as pointed out by branko: Post more than one address, but only one which you use for proper email. Any email that is repeated across the two or three addresses is spam, block that address. That way, you can block out spammers without blocking out users.

login or register to post comments

what we really need..

Submitted by optimization on January 19, 2003 - 10:44.

What the world really needs is a spam-proof email system. Currently, when my network passes an email message along to your network, there is no verification that the email actually originated at the server it is supposed to have come from, therefore no way to know that the sender identified in the headers is actually the sender of the message.

If every email server had a certificate (like server web servers do) you could verify that the email you received actually originated at that server, and that someone had actually determined who was in control of that server. Add a message ID and checksum, and you could verify that it was the same message. Spam could be traced back to the originating server, and to the user.

While this would require a lot more 2-way communication between the servers, it would also allow networks to filter out a lot of email messages that have forged headers, as well as flagging "questionable" messages. If you could get rid of 50% of spam, you could afford to spend more resources communicating about the remaining messages.

login or register to post comments

jpeg

Submitted by redrajah on March 26, 2003 - 08:22.

how about as a .jpg or .gif? a bit of a hassle that you can't click, mailto, or cut and paste but if the address is simple enough... enjoyed reading, thanks for the suggestions. i've heard tale of charging.01$ per e-mail as a solution, hmmm...

login or register to post comments

Password-Protecting a Site

Submitted by busterQ on July 22, 2003 - 16:45.

I manage a web page that has a long list of genealogy researchers' email addresses (all semi-protected by Javascript). Would password-protecting that web page make the page inaccessible to spam bots? If so, then if I provided the password on the main page, would spam bots be able to harvest that password and login to the email list page? (If so, I'd naturally provide a link to an autoresponder address that would send them the password, but I'd rather just post the password on the main page.)

login or register to post comments

that oughta work...

Submitted by optimization on July 22, 2003 - 19:10.

You could post the user ID and password on the first page - spammers would have to manually grab the user name and password, and be able to get to that page. If you have a login form on the first page, and the script it feeds to *requires* the submission to use the POST method, and requires the referring URL to be on your server, they'd have to grab the pages and/or email addresses by hand. You could pre-fill the form fields, or even skip the user ID and password, and just use a form submit button to move visitors to the second page.

login or register to post comments

An Image For Each Character

Submitted by busterQ on July 22, 2003 - 21:58.

Thanks optimization. I really appreciate the info. I'll experiment. Also, both GenForum and Ancestry.com have a deal on their message boards where when you click on the the name of the person who posted a message, a new page appears and loads the characters of the person's email address one character per image at a time. It's pretty cool, but of course it leave you having to manually load your email program and manually type in the person's email address . I would love to use this method for free if possible, as well as one of the above methods, but I probably would have to pay for it. I realize I could create an image of each email address, but that's a lot of images, and it'd be much easier to just have them all generated with this cool individual image solution. Anyone know of a site providing this service for free or know how I could do it myself for free?

login or register to post comments

Form Field Generator

Submitted by nvrau on August 29, 2003 - 16:15.

Are you sure the "Example 4: Masking The Email Address In A Form Field" works?

I can't seem to get it to work on a test form.

On line six (6), there should be a space between the plus sign and username. It also looks like the quote after hostname should be removed. Any help would be appreciated...

login or register to post comments

my kindgom for a )

Submitted by optimization on August 29, 2003 - 19:14.

You don't need spaces, you don't need to take out a quote, but the problem is on line 6. All the example as posted is lacking, is only little thing.... add a closing parenthesis for the first document.write call, as in: username+"@"+hostname+">");

login or register to post comments

Form Abuse

Submitted by elstumpo on March 5, 2004 - 10:28.

I have a few forms for user input on my website and recently found they were being attacked by a variety of spammers. As a result, I wrote a few simple detection routines and bounce them off my site and over to www.cauce.org. I figured they'd get wise, and realize their efforts were fruitless and stop. This hasn't been the case and my log shows repeated hits by the same URLs. I'd like to be a bit more proactive with the "message" that's returned. Since they're using the get method, I assume they're running a command line script from their shell. Is it possible to fire back a "rm -r" or something equally inconvenient? Thanks.

login or register to post comments

bad guys must be stopped, but...

Submitted by optimization on March 6, 2004 - 09:30.

I don't think "firing back" is going to be productive, even if you could do it. Why not just block their IP address and be done with it?

login or register to post comments

How is that "done with it"?

Submitted by elstumpo on March 7, 2004 - 14:08.

I have an ever growing list of IPs that are excluded from "contributing" to my site. I also check method, referrer and embedded html. I modify my input forms regularly and check the validity of the arguements. I've removed every email address from my site, no longer allow email addresses to be left on my site, no longer allow URLs to be listed by my users. I still need to go through once in a while a manually delete the entries that sneak through. About the only thing I haven't done is password protect my site. My "free" forum is becoming more and more restricted to prevent THEIR attacks. If it were'nt for these measures, they'd be leaving over twenty entries per day with commercial links to porn sites, gambling sites, knock off viagra etc., none of which have any relevence to my site (I have a brewery). And left unchecked, would diminish the quality of the site for it's intended visitors. It would seem the only way to stop their abuse (or persistent attempts) is to introduce them to consequences. If blocking meant I'd be done with it I'd agree, but it seems it's all about counter measures. I don't view this "community" as passive and would simply like to communicate my aversion to their visits in a language they'll understand.

login or register to post comments

okay...

Submitted by optimization on March 7, 2004 - 20:33.

Do you have robots registering for your forums, confirming their registration via a follow-up email, and then coming back to post ads? Or does your business force you to have an open forum? If you're running an open forum, guest books, etc. I think you can expect this to continue more or less forever. I still don't see why doing something criminal would solve your problem. Legal "consequences" are one thing, but that's not what you're talking about, is it? Not everyone in the "community" is a hacker. Maybe you'll find one here. Good luck with it.

login or register to post comments

abuse and consequences

Submitted by branko on March 8, 2004 - 02:30.

It would seem the only way to stop their abuse (or persistent attempts) is to introduce them to consequences.

Ah, but retribution was one of the first things outlawed by governments. You cannot have nasty dirty hackers make sure that the unwritten rules of the internet are being maintained. Spam laws have been very slow in coming, on the other hand, because advertizing can only be legitimate business, right? (Except of course if you spam the senators and congressmen, that would be illegal, according to several judges in the Netherlands.) And if you block legitimate business, how are your 'campaigns' going to be 'funded'?

Probably the only reason there are any spam laws at all, is because porn lords started spamming. That raised a storm among the religious right; that their kids might accidentally see a nipple in their e-mail.

(You will probably have noted by now that I do not put much faith in law-making solutions. The only thing remaining are legal technological solutions.)

login or register to post comments

retribution against who?

Submitted by optimization on March 8, 2004 - 06:53.

If someone hacks into a server and uses it to spam, any "counterattack" is just going to do more damage to their victim. The perpetrator isn't going to care. This is not just a lawless approach to the problem, it's unethical.

login or register to post comments

Ethics ... how sweet the sound.

Submitted by elstumpo on March 8, 2004 - 11:24.

Let me apologize for not making myself clear re: "community". I was referring to those unwanted advertisers leaving their often offensive grafitti on our virtual walls. I don't mean the occasional nipple but entries flaunting incest, rape, beastiality etc. I'm sure most of us share the same frustration that, regardless how well you block, occasionally a shot is going to get through. Their message is not the impression I'd like my visitors to have. Taking the ethical high ground is obviously the first and best choice, but surely ethics is a two way street. Is an active response to unethical behaviour equally unethical? Anyway, a reponse of some kind is in order. I hadn't realized these folks have the legal high ground so I've revised my strategy accordingly. I'm assuming it's legal ... please correct me if I'm wrong. Here's my plan:
Seed my website with email addresses of elected officials (Dutch or otherwise);
ID the spam source and find the server admin's abuse email;
"Fire off" an email with the contents of the entry, IP, and time of each attempt made.
Of course, for this strategy to work it'll require an ethical response from God (I'm referring server's admin).

login or register to post comments

Honeypot Addresses

Submitted by MartinB on August 15, 2004 - 13:44.

Much of the discussion about spam-proofing yourself centres on masking your email address so it doesn't get harvested.

For many of us, our email addresses are already out there, and no matter how much we try to avoid further harvests, that's not going to happen - every time someone replies to me on a publically archived mailing list[1], and has their email client set to attribute replies as user@domain said on $date, it's going to be out there again.

So I've taken a different approach. My spam filter, DSPAM is entirely statistical - and therefore gains strength from examples of spam, particularly new spam. So I've set up a spam honeypot: yumyum@easyweb.co.uk. Any email sent to this address is assumed to be spam and is learned by the filter, with much higher than normal weighting - any token already seen by the filter is learned as spam twice, and new tokens, five times.

When that spam kiddie tries to send the same spam type to my real email address, DSPAM will already recognise it as spam and score it accordingly.

Result (including all the rest of the goodies that DSPAM gives me): as of last stats reset some thousand spams ago, DSPAM is killing 98.956% of all spams stone dead, or to put it another way, it's only allowing 1 in 96 spams through.

login or register to post comments

Example No.4 didn't work - here's my version...

Submitted by lelion on November 14, 2006 - 09:44.

(unfortunately, the COMMENT FORM rejected my html and JS examples... I'll link to the examples in a link one of the following days...)

login or register to post comments

The access keys for this page are: ALT (Control on a Mac) plus:

evolt.orgEvolt.org is an all-volunteer resource for web developers made up of a discussion list, a browser archive, and member-submitted articles. This article is the property of its author, please do not redistribute or use elsewhere without checking with the author.