-----Original Message----- From: Jeff Chan [mailto:jeffc@surbl.org] Sent: Sunday, July 18, 2004 5:07 PM To: SURBL Discuss; SpamAssassin Users Subject: 6dos data into ds.surbl.org
Chris Santerre asked on the spamassassin-users list whether ds.surbl.org was being updated with the 6dos data. The answer is no, it's static now based on that snapshot Chris gave me a couple weeks ago. To be honest, the 6dos data seems too over-inclusive to me for production use, but perhaps Chris or Rich could help me figure out something closer to a live feed as we beta test it.
Jeff C.
LOL, I don't remember half the stuff I ask :) Actually the initial data that was used wasn't even the freshest from the source. This data is used in a much different way and is quite large! Well beyond a simple list of domains. I'm actually thinking it should not be used. It does contain a vast amount of info.
I'm hoping instead, that this data source could be used for SURBL research in the future. When and if we get a nicer interface to it.
So my official recomendation is for ds.surbl.org to go away.
--Chris
Hi!
LOL, I don't remember half the stuff I ask :) Actually the initial data that was used wasn't even the freshest from the source. This data is used in a much different way and is quite large! Well beyond a simple list of domains. I'm actually thinking it should not be used. It does contain a vast amount of info.
I'm hoping instead, that this data source could be used for SURBL research in the future. When and if we get a nicer interface to it.
So my official recomendation is for ds.surbl.org to go away.
We are still in testing phase, but i feel the same. Anyone else can speak up? I have been running the list in production some time now, and i am aware its a older list with data, but the ammount of hits is pretty low compared to the size...
We should not forget that data is static, and to have a real test it should be dynamic, but ... :)
Bye, Raymond.
On Monday, July 19, 2004, 3:03:50 PM, Raymond Dijkxhoorn wrote:
Hi!
LOL, I don't remember half the stuff I ask :) Actually the initial data that was used wasn't even the freshest from the source. This data is used in a much different way and is quite large! Well beyond a simple list of domains. I'm actually thinking it should not be used. It does contain a vast amount of info.
I'm hoping instead, that this data source could be used for SURBL research in the future. When and if we get a nicer interface to it.
So my official recomendation is for ds.surbl.org to go away.
We are still in testing phase, but i feel the same. Anyone else can speak up? I have been running the list in production some time now, and i am aware its a older list with data, but the ammount of hits is pretty low compared to the size...
We should not forget that data is static, and to have a real test it should be dynamic, but ... :)
Generally speaking, I feel the same way. There's some good data in there, but it seems "polluted" with too many aggressively included domains.
The best ones Chris can pull into ws too.....
I'd like to hear what anyone else has to say before we pull the plug on ds though.
Comments?
Jeff C.
-
Chris Santerre -
Jeff Chan -
Raymond Dijkxhoorn