Gary Robinson's Spam Rants
A Weblog containing links and occasional rants on spam.


Gary Robinson is CEO of Transpose, LLC, and a Research Director at ActiveState. Main blog here.

Subscribe to "Gary Robinson's Spam Rants" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.



  Thursday, October 2, 2003

Ray Ozzie thinks email is going away:

Think about it.  Think about the rate of increase of "noise" in email over the past two years, which is a very short time.  Think about where we'll be in as short as five years.  Can you imagine?

Right now, every major enterprise has a "content-scanning gateway" that processes every incoming email, looking for Dangerous Stuff.  Many individuals do the same thing on their own computers.  Some enterprises are beginning to quarantine incoming email for extended periods - sometimes an hour or more.  Maybe you'll get too much junk, or maybe you won't get what you're supposed to get.  Maybe you'll get it, but it'll be too late.  It depends upon where they turn the knob on the software ... and it's insane.

If you're hoping for some super-duper neo-PIM to or super-filter or super-law to come along to make your life easier, spare yourself the agony and just think ahead: it is NOT a sustainable solution if it is still called "eMail".  eMail is thirty years old, and we owe it a great debt of honor, but it has been pushed well beyond its design center and it's time to move on.  Incrementally, progressively, but most definitely.

And his solution?

If you have work to do with others, online, try workspaces.  There are many different types - from Groove if you like client-based mobility, to SharePoint if you like using Websites.  No noise, no spam, tuned to save your time.  Of course, you can't give up on eMail, and likely never will.  As time goes on, though, you'll only visit eMail as a low-priority background task, much as you do when sorting through your physical mail at home.  You'd never do important work through your home mailbox, would you?

The problem with working with others in Ozzie's vision of a workspace is that we all have to buy into a particular product, because they aren't compatible. So if I want to use workspace to work with my lawyer with have to bothhave Groove or SharePoint or whatever. The end of that path is either that everyone has a number of different workspace applications that they've paid for, learned how to use, and are willing to keep updating over time, or one company dominates the space ala Microsoft Office. I think it's clear that the latter is the more likely solution, for pretty much the same reason that most people in business use Office rather than a motley collection of premium word-processing tools and spreadsheets. Unfortunately for Ozzie, who runs Groove, I think such a solution is more likely to come from Microsoft than from Ozzie.

But in reality email has more legs than Ozzie thinks. In the quotes above, he is quite passionate about how rapidly email is degenerating. But you know what? I run SpamSieve and I don't have a spam problem any more. Yes, I get about 200 spams a day. But I don't see any but 1 or 2 of them. It's not a problem. And SpamSieve, like a number of other filters, is accurate enough that loss of legitimate mail ("false positives") is not a significant problem.

Now, as long as most people don't know about or don't make the effort to install such filters, there will still be enough profit in the spam business to motivate the sending of billions of spams per day. But one by one, email software manufacturers are adding spam filtering to their software -- most recently Eudora and Microsoft Outlook. (I'm happy to mention that Eudora uses math that is built on the mathematical suggestions of Paul Graham and myself, as does SpamSieve.)

Before long, enough people will have spam filtering -- whether they are aware of it or not -- that spam will become significantly less profitable than it is now. (And not only because of filtering, but also because even the most gullible and insecure among us can only buy so many penis enlargement products before they decide there are better things to do with their money.) Spamming will be less enticing to prospective spammers.

Added to that are increased legal risks due to laws such as recently passed in California, and perhaps most importantly, the likelihood of ISP's and and Internet-using corporations working together to divide the Internet into "two camps, those who can be held accountable for spam, and those whose status is unknown."

When everything is added together, it is unlikely that spamming will continue to make email harder to use. It is much more likely that the noise level of email will be diminishing steadily over the next few years.

Another factor will be the fact that sophisticated users -- those who are most likely to be on a lot of legitimate news mail lists from corporations they do business with -- will increasingly be savvy enough to choose RSS rather than email for most of that sort of news, further increasing email's signal/noise ratio.

So in the end, I think that Ozzie is indulging in a bit of wishful thinking to believe that email will soon be so useless that it will have to be replaced by the likes of Groove. Of course, things like Groove do have benefits; my company in fact has a Groove account, which we received as a benefit of our SBIR/NSF grant. However, we still use an open-source wiki for group document management. The fact that the wiki is free doesn't even play a factor in our choice; our Groove account is free (to us) too. The wiki just works; it's easy and convenient and there is really not much reason for us to use anything else.

While I think that there will advances over email and wikis in the coming years, my guess is that most people will be using products that are open-source and/or conform to open standards. Even though we have had little reason to use Groove, there is also little reason reason not to use a product like Groove if it were opened up so that it didn't lock up the user's data.

I forecast that Groove, and products like it, will be marginalized if they don't create open standards for freely sharing data with other products, but if they do so, compatible open-source products will emerge that make it hard to charge a lot for them. Microsoft would seem to be the company with (by far) the best shot of creating a proprietary, closed solution, but the world is a lot wiser now than when it fell for Office, and will be wary about choosing to get tied into another Microsoft monopoly. And open-source mechanisms for software development have gained power and acceptance in recent years. That's great for consumers of information technology, and less good for people that want to make a lot of money in the software industry.
10:25:35 AM    comment

  Thursday, August 28, 2003

"Dennis K. Berman offers a ray of hope to people frustrated and demoralized by the spread of spam: Spamming could eventually burn itself out by becoming a victim of its own proliferation, he muses." [From the's summary of a Wall Street Journal article. The summary is worth checking out.]
4:07:31 PM    comment

  Saturday, August 23, 2003

Well I'm back in Maine. I actually got a fair amount of work done on my laptop on the drives to and from New York City.

Partly on the drive south, I developed and tested some changes to the chi-square approach to spam detectionand testing so far the changes seem to result in roughly a 5% improvement in accuracy. On the drive back I developed another possible improvement that I hope to find time to test over the weekend.

9:06:35 AM    comment

  Monday, August 11, 2003

SpamBayes, which is based on math suggested by me as an improvement to the approach in Paul Graham's seminal "A Plan For Spam" article, and supplemented with a great observation from Tim Peters that the "Unsure" range of spam/ham classification is very useful, won this comparative testing against PopFile.

The test was slashdotted today.

John Graham-Cumming, who is responsible for PopFile, Tim Peters and I are all members of ActiveState's Anti-Spam Task Force.

I must admit I'm competitive enough to enjoy that sort of thing. ;)

(Note, John is doing a fantastic job at documenting specific tricks used by spammers.)
6:02:19 PM    comment

  Monday, July 28, 2003

"'We’re going to see a lot more spam, instead of less,' said David Sorkin, who teaches a spam law course at the John Marshall Law School in Chicago. 'What most of the state statutes do is legitimize spam as long as you don’t also commit fraud.' []
9:36:23 PM    comment

"Measuring the cost of spam, which takes into account such factors as lost productivity and wasted time, is an imprecise science, and estimates on its total toll vary: Ferris Research reckons that spam cost the United States $10 billion in 2003, while Nucleus Research pegs the U.S. cost around $87 billion." [New York Times (Free registration req'.)]
8:08:00 PM    comment

  Tuesday, July 1, 2003

Lawmakers call for global action on spam. []
10:53:28 AM    comment

  Friday, June 27, 2003

"Junk e-mailers are spreading viruses that let them send spam anonymously through home computers, according to an e-mail security firm.

The company, MessageLabs, operates servers that block spam and viruses for its clients. Its analysis of data shows that mass distributions of junk e-mail are increasingly coming from the Internet addresses of computers that have in the past sent out viruses as e-mail attachments. 'There is a high correlation,' said Matt Sergeant, senior antispam technologist for the New York-based company. 'About 30,000 machines have both open-proxy software and are responsible for sending viruses.'" []
3:12:16 PM    comment

Click here to visit the Radio UserLand website. © Copyright 2006 Gary Robinson.
Last update: 1/30/06; 2:46:08 PM.

January 2006
Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31        
Oct   Feb