> > But... If they wanted to use SA they would use SA. Since they are > > You assume that people using SA or Razor do so by choice and/or are > technically savvy enough to set them up.
By choice, yes. Your point about technically savvy is interesting since I think both SA and Razor have similar technical thresholds of ability required to install and set up. I would not say either is easier or harder than the other. I used to use Razor before finding SA and find that SA is easier to set up. YMMV. > > not using SA when it is available to them but are using Razor it leads > > me to believe that they do not want to use it. They are exercising > > Or they don't know about it, or their ISP hasn't set it up, or ... Certainly their ISP would be technically savvy enough to install and configure either. Certainly their ISP would be able to make an informed choice! I say that tougue in cheek since we all know ISP's that are as clueless as any newbie. But I still disagree that just because we think SA is the best tool available for fighting spam that we should put that view upon everyone else too. Other people have other ideas. It is okay to have other and different ideas. > > Why should SA users force Razor users to use SA and take that choice > > away from them? > > What are you talking about? The statement was "just because SA caught a > spam doesn't mean you shouldn't report it to Razor." How is that taking > away a choice? I let SA determine if something is spam instead of me > having to look at the message. Putting on a Razor centric user hat and personae: It is taking away choice since I expect those messages to be spam verified by a user and not be a false positive generated by a tool. If you are automatically reporting to razor any message that your SA tool is tagging as spam then when I use Razor I am really using an extension of your SA tool. Anything your SA tool tags as spam will be listed in the razor db. I as a razor user no longer have a choice and must live with your SA tool even if I don't want it. Here is another tactic. As a Razor user how do I avoid checking my mail against your SA rules? If your SA rules are bad, I as a razor user suffer since your SA will pollute the Razor database with legitimate mail message signatures. If I can't avoid your SA rules then you are taking away my choice. The default SA threshold to tag as spam is five. That is good enough for me and my mailbox and for most people and their mailbox. But that is a configurable threshold. Many on this list have modified it for their own needs. If it were perfect why would anyone need to modify it? And as a razor user how do I modify the SA threshold? As a Razor user if I think it should be 7 or 10 instead how do I choose that? What? I don't have a choice? I also think it pollutes the web of trust being developed for Razor. With manual reporting there will always be the bad setup where people forward all mail to razor. They will be ignored because their trust factor will be low. But if you are using SA then you will generally have a high trust factor since SA is good at this. In that case in particular if SA missidentifies a messages it will go into the database with a high trust factor. I believe that to be particularly insidious and will be much harder to weed out. > In the end, the message is still spam and should be (if you're so > inclined) reported to Razor. You are assuming that SA is perfect. If it were then I would agree. But it is not and I cannot agree with you. In the end the message may not be spam. SA has been known to fail to properly categorize messages. False positives happen. I know, I know, we don't like to admit that our favorite tool is not perfect. But it is true and it is not perfect. The default report to razor threshold was 30 and very conservative. It is extremely unlikely that threshold would be exceeded and the message not be spam. That seems safe enough and can't cause too much trouble. The rate of false positives there is probably less than the error rate of humans scanning the messages. But to say that the default spam threshold is foolproof is incorrect. At that level sending all tagged messages to razor is bound to cause trouble. A long standing problem with razor has been people that forward mail to the razor db incorrectly that is legitimate mail. Then users that rely upon the razor db to contain only spam messages will get those messages filed incorrectly. Automatically reporting messages tagged by spamassassin will only generate bad press for spamassassin and will hurt it in the long run. Bob
msg07059/pgp00000.pgp
Description: PGP signature
