Oops, editing error. > Enterprises have revenue, too. And cost structures that are substantially > different (with respect to email) from similarly sized ISPs. This should read "And cost structures that are _NOT_ substantially different..." --Dean On Tue, 27 May 2003, Dean Anderson wrote: > On Tue, 27 May 2003, Vernon Schryver wrote: > > > > From: Dean Anderson <dean@av8.com> > > > > > ... > > > No one has demonstrated any cost to spam, other than annoyance and > > > infrastructure costs which are passed on to users[1], and it seems there > > > is very little to add. > > > > That's a ridiculous overstatement. Spam may be too cheap to meter > > for an individual user, but if you have 30,000,000 users or only > > 30,000, you'll find that the total costs are substantial, particularly > > when you need to double the size of your systems to deal with a doubling > > of spam. > > Whether you have 30 million or 30 thousand, the costs are passed on to the > user. At 30 million, economies of scale make the cost of spam (to the ISP) > even less. It doesn't usually change the price to the user, so big ISPs > tend to make more money at the same price level. > > Enterprises have revenue, too. And cost structures that are substantially > different (with respect to email) from similarly sized ISPs. > > > > Vixie and other radicals also continues to ignore Shannon's theorems. ... > > > > While it's true that proving the non-existence of covert channels > > is hard, you've not related that theorem to anything related to spam. > > Even if covert channels or the relation among power, noise, bandwidth, > > and information have something to do with spam, you've not shown > > any connection to dealing with the spam problem. > > Its not hard, its impossible. > > I've pointed out the spam is a covert channel, with regard to Shannons > theorem. > > > A good rule of thumb is that only people who don't know much about > > major theorems like Shannon's and Godel's quote them as proofs in > > discussions like this. I think you should have chosen some other > > beautiful bit of archana such one of the fixed point theorems or > > Fred J. Cohen's Forcing and Generic instead of something that more > > than a few people around here know about such as covert channels. > > This is a fine exposition, about what I'm not sure.. It does not seem to > demonstrate that I am wrong, or that spam isn't a covert channel, or that > Shannon's theorem (somehow) doesn't apply. > > Apparently, you don't understand anything about covert channels, so > perhaps you should let the ones who do try to dispute my assertions. > > --Dean > >