Re: [Declude.JunkMail] Experience with 4.x

[Matt]

Sanford Whiteman wrote: I am not denying _you_ shared credit for the concept with the Postfix people, but the idea that your "friend" the vendor can claim it as intellectual property, when you spec'd it out and have documentation of same -- and that that's why you're playing it close to the vest -- is ridiculous. If it's yours or Brian's, it's yours; in reality, it's nobody's. I came up with the idea before I ever heard mention of it anywhere else, and long before 8/2005.  I was searching for someone to develop a gateway from scratch with various improvements/adaptations of existing techniques as well as new ones since the first half of 2004.  Although ORF saved my arse from dictionary attacks, I realized quickly that it wasn't ever going to become a reliable way of pre-scanning E-mail and I wanted more.  With something like 6 billion people in this world, and the concept being as obvious as it seems to me, I definitely wouldn't suggest that I was the only one to have thought of it, and I'm sure that I wasn't the first to ever think of it either.  This is not something to waste bytes on. What does matter in the context though is that Brian made many of these things work...and then some.  His code is his company's property.  He is not the type of guy though to claim rights to such functionality being the good netizen that he is, and I'm sure that we would all agree that much of the software technique patent stuff is damaging to advancements.  I don't want to make too much of this one little thing though because it isn't all that earth shattering to think that it could be improved.  It would something like suggesting that SpamCop could clean up their false positives by qualifying IP's for listing, or dequalifying them, sort of like CBL does.  It's just my opinion that there are a lot of good ideas that don't get developed into mature ideas in this field, and thus far greylisting has been one of them. Besides greylisting, groups of us have talked about or even implemented things before they existed to the public.  Keith Anderson gets credit for identifying what a "payload" is and how to extract that information and use it for blocking E-mail.  Group discussions started in 9/2003, six months before SURBL.org was even registered as a domain, and clearly the concept pre-existed that day by even longer.  We as a group also came up with enhancements such as resolving payload links to IP's, or even MX records, NS records to check against URIBL lists as well as RHSBL lists, and then resolve those again to IP's and look them up in IP4R lists.  Keith even created a product that did this called Eradispam before anything else that I know of did (besides the straight URIBL stuff).  Not everyone got these ideas from us though, some clearly came up with it on their own as well. There are mounds of other inventive if not unique ideas that power users around here came up with, and some of them we all benefited from.  Scott used to design functionality around community development and I'm sure that many of us would love to see a return to that atmosphere.  Unfortunately there are more good ideas than there are great programmers that can make the maximum use of this stuff. Brian is not only a great programmer, he also brought something very unique to the table in this framework that made it immensely better than anything else that does this could be.  MXRate when combined with greylisting is very powerful and allows for greater accuracy as well as flexibility.  MXRate in it's native form isn't just a three zoned IP4R list, it's probability based and the IP4R version is just backwards compatibility.  This allows users to choose the level of greylisting and blocking to their own tastes so that they can block more spam and avoid more false positives/delays.  Here's the net result from all 4 of my MX records now being bound to Alligate on Tuesday:       Incoming message attempts: 708,777 Blocked by gateway pre-scanning: 676,397  *likely to be 3/4 dictionary attacks.     Blocked by Declude JunkMail:   9,228        Blocked by Declude Virus:      90          Delivered to customers:  23,062          Landed in Review range:     980 This change in my infrastructure is the single best thing that I have done since starting to use Declude.  Even though I stood up for Brian's character when he posted to the list about his product, I didn't even give the idea of using his product a second thought until we started exchanging E-mails and I grew to understand that he was an even better person that I had thought, that he was a good programmer, and he gets "it". I think that we should embrace development like this because it is for the good of everyone, including Declude.  Now non-experts can place a address validating and pre-scanning gateway in front of any Declude setup whereas before you nearly had to be an expert at this stuff and be aware of the obscure.  The fact that it kills over 99% of zombie spam with near perfect accuracy also flies in the face of frequent postings to the Declude, IMail and even Sniffer lists about stock image spam (not to belittle Sniffer which hits 97% of all spam and is incredibly accurate for it's volume overall, but the volume of zombie spam is just tremendous and images without links can be hard to block). No offset cost, no payment in kind? Okay, that's good. Apology accepted :)  I reserve the right though to discuss business arrangements with anyone at any time though.  I'm certainly not driving a Lamborghini you know. Flatly, no. The reason you have this misconception is that no other vendor spammed the list. Others earn an audience the old-fashioned way: spending money, not just time, to make money. The other vendors rely on search-sensitive advertising, banners, print ads, PR, and in the cases of cash-strapped authors and free products, simply hoping to be indexed and found by dogged sysadmins -- and then further promoted by unaffiliated users and reviewers. There are surely some unethical apples in that other bunch, of course, as everywhere, but they don't spam. You can't judge a person on one mistake, and it's not like he committed murder.  Brian knows that he made a mistake in how he approached the topic on this list, but you can't fry him in Hell for the rest of his life because of this.  Brian is an old-timer that has been fighting the good fight against the same crooks that are spamming us for more than a decade.  Don't think for a second that some specialized gateway app is ever going to move his company's fortune from CyberSitter and the ubiquity of the desktop market.  This is more of a hobby project for him, and I'll bet that the only financial target he has is to justify continuing to work on it so people like us can benefit.  Gushy as it may sound, there are in fact people out there that care still. And just in case you were wondering, he didn't pay me to say that either :)  In fact, it's a shame that neither one of us will even bring up this thread for fear of violating some sort of imaginary rule about what is proper and what isn't. I do wish that you would learn to forgive and forget.  There is so much more good outside of dwelling on this. As far as how appropriate the continued discussion is of Alligate, I will, with no misgivings, never talk of it again if Declude even suggests that it is not in their best interest to have it talked about here. Matt