Indeed RAV for linux is a good piece of software.
BUT was close source. There are some Open Source
solutions but not as good as RAV. We must admit that.
What keeps us from developing an good opensource solution ? They are a few we are many.
A long time ago i was developing an antivirus myself .
Since i was developing it by myself and for myself it wasn't too much. But it worked.
Main problems of an opensource anvtivirus will be:
1) we need viruses in order to code the antiviral part.
2) I don't see why to keep in a database viruses from
5 or 6 years ago that can be found ( i guess ) in collections only. So we don't need to code 70000 of
routines of cleaning but instead let's say 1000 of the actual viruses.
3) Speed of the development is essential. if a virus is
found on the wild it will be essential to have an update
as soon as possible.
4) IMHO an antivirus combined with antispam techniques will cover at least 80% of the ISP providers needs so main focus has to be on this issue.
5) the antivirus have to have some method to detect / eliminate polimorphic viruses. (a virtual machine to emulate the virus until it can be sure that is a virus) ,
or per virus specific routine ( but this method requires more daily work from the coders )
6) the antivirus have to be written in C or C++ due to
speed / memory considerations.
7) the antivirus should integrate easy with the popular
mailing systems ( qmai , sendmail , postfix )
Slashdot Mirror
Indeed RAV for linux is a good piece of software . : . .
BUT was close source. There are some Open Source solutions but not as good as RAV. We must admit that.
What keeps us from developing an good opensource solution ? They are a few we are many.
A long time ago i was developing an antivirus myself . Since i was developing it by myself and for myself it wasn't too much. But it worked.
Main problems of an opensource anvtivirus will be
1) we need viruses in order to code the antiviral part.
2) I don't see why to keep in a database viruses from 5 or 6 years ago that can be found ( i guess ) in collections only. So we don't need to code 70000 of routines of cleaning but instead let's say 1000 of the actual viruses
3) Speed of the development is essential. if a virus is found on the wild it will be essential to have an update as soon as possible.
4) IMHO an antivirus combined with antispam techniques will cover at least 80% of the ISP providers needs so main focus has to be on this issue
5) the antivirus have to have some method to detect / eliminate polimorphic viruses. (a virtual machine to emulate the virus until it can be sure that is a virus) , or per virus specific routine ( but this method requires more daily work from the coders )
6) the antivirus have to be written in C or C++ due to speed / memory considerations.
7) the antivirus should integrate easy with the popular mailing systems ( qmai , sendmail , postfix )
Just my oppinion .