On Oct 23, 2009, at 4:38 AM, Fritz Borgstedt wrote: > GrayHat <[email protected]> schreibt: >> that those servers send me ham but also spam > > Purely IP based - checks do not make sense for IPs which have mixed > (ham/spam) output per definition. > This is als valid for the ISPIP list. > So - many people run the nodelay list also as ispip list.
Hi Fritz, and everyone else in this thread. I want to make sure I understand this. If you add an ip or set up ip's to nodelay, the following will essentially pass? - nodamping - SMTPdomainIP - DoFrequencyIP - no MessageID check - no RBL check - no PBExtreme - no add to PB (black or white) My experience with other servers, and how I would prefer this to behave, as well as what I think is logical, is that greylist is a good front end defense, it will catch a lot of spammers since they use machines that do not know how to try again. This happens pretty early on and there is little overhead to the system as a result. But well before they hit the greylist, I should be able to RBL block them first shouldn't I? Let's take some large free email providers as an example, gmail, hotmail, yahoo, aol, etc. I would add their IP ranges to nodelay. Google has a very large pool, if they were in delay, it would take a long time for all IP's to be hit and get past. But according to the list above, by adding google to the nodelay, I also give up a message id check. With gmail allowing direct SMTP, I believe I would want the message id check. It also will skip the RBL, which would normally seem ok, since if I am going to allow a host to not be delayed, you could assume that I want their email. However, there are many different types of RBL now, which are not a block or no block cases. I would want to apply weighting to the incoming email based on a set of RBL scores, which seems it would stop if I add to nodelay. I am not entirely sure what all the reset of the above settings do, but I believe that nodelay should only prevent the host from hitting the delay in the greylist, all other tests should be performed. Does this mean that if a host naturally goes from delayed to an OK state, that it is also going to then be skipping the above tests? If not, I suppose a workaround would be to simply not enter the ranges in the nodelay list, and instead, somehow, if possible, add them in as OK state hosts in the delay database. Thanks for any clarification on this. -- Scott * If you contact me off list replace talklists@ with scott@ * ------------------------------------------------------------------------------ Come build with us! The BlackBerry(R) Developer Conference in SF, CA is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9 - 12, 2009. Register now! http://p.sf.net/sfu/devconference _______________________________________________ Assp-test mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-test
