except bad guys will always opt of having their content inspected.

so it just doesn't work in this case.



On May 18, 2013, at 10:46 AM, Jeffrey Walton <[email protected]> wrote:

> On Sat, May 18, 2013 at 1:24 PM, mark seiden <[email protected]> wrote:
>> ...
>> there are numerous other IM systems that are server centric and do a lot of 
>> work
>> to look for and filter "bad" urls sent in the message stream.
>> 
>> this is intended to be for the benefit of the users in filtering spam, 
>> phishing, malware links,
>> particularly those that spread virally through buddy lists of taken over 
>> accounts.
>> sometimes these links (when believed to be malicious) are simply (and 
>> silently) not
>> forwarded to the receiving user.
>> 
>> this involves databases of link and site reputation, testing of new links, 
>> velocity and
>> acceleration measurements, etc.    the usual spam filtering technology.
>> 
>> my impression is that almost all users thank us for doing that job of 
>> keeping them safe.
>> they understand that IM is yet another channel for transmitting spam.
>> 
>> the url filtering is aggressive enough (and unreliable enough) in some cases 
>> that
>> you have to check with your counterparty in conversation if they got that 
>> link you
>> just sent.  so users are aware of it, if only as an annoyance.  (once again, 
>> spam filtering
>> gets in the way of productive communication)
>> 
>> i am merely telling you how it is.  obviously user expectations differ on 
>> AIM, Yahoo Messenger,
>> etc. from those of users on Skype, some of whom believe there is magic fairy 
>> dust sprinkled on it, and that
>> it is easier to use than something else with OTR as a plugin.
> Perhaps the user should be given a choice.
> 
> The security dialog could have three mutually exclusive choices:
> 
>  * Scan IM messages for dangerous content from everyone. This means
> <company> will read (and possibly retain) all of your messages to
> determine if some (or all) of the message is dangerous.
> 
>  * Scan IM messages for dangerous content from people you don't know.
> This means <company> will read (and possibly retain) some of your
> messages to determine if some (or all) of the message is dangerous.
> 
>  * Don't scan IM messages for dangerous content . This means only you
> and the sender will read your messages.
> 
> Give an choice, it seems like selection two is a good balance.
> 
> Jeff

_______________________________________________
cryptography mailing list
[email protected]
http://lists.randombit.net/mailman/listinfo/cryptography

Reply via email to