Re: Drop Dynamic Querying - It only benefits Spammers these days

Isn't that kind of problem that the current experimental DHT support in LimeWire (Mojito) is about to address?

This would mean that dynamic querying would no longer be needed to find rare files using larger bandwidth, and still allow finding non-rare files using the DHT too (but then the spam problem will try to go also into the DHT tree, so there will be the same problem too, requiring dynamic querying plus adaptation to spam results trying to fill the DHT tree, unless we feed the DHT tree with the content URN only, where it is easy to identify spammed content nodes).

Fighting spam by dropping dynamic querying is clearly not a solution, as it won't stop it. I am convinced of the opposite: spam is most damaging for popular files because it targets the most wanted keywords in searches.

Well, I personnally don't find spam as a real problem, given the antispam filters we can already set in LimeWire and which are just effective for me to eliminate most of them ; But may be others are experiencing it differently. I can still find rare files, as well as popular ones; may be you are not used to make specific searches...

From: "Sam Berlin" <[email protected]>
> We are actively looking into the spam problem, Arne, and intend to
> address it effectively.
> 
> On 1/23/07, Arne Babenhauserheide <[email protected]> wrote:
>> I want to propose dropping dynamic querying, as it only benefits spammers
>> today.
>>
>> It is a really great idea, but it has a disadvantage, which went as far as
>> making Gnutella mostly unuseable for rare file searches today.
>>
>> When it was introduced, it greatly reduced the overhead of popular searches,
>> because it stops queries when they have gathered enough results.
>>
>> Today it mostly serves to keep the bandwidth requirements of effective
>> spamming down, because a spammer just has to be among the first who send you
>> query replies, and you won't find the content you search for.
>>
>> In the case of rare files, searches simply don't get far enough anymore, to
>> find even one version of a file.
>>
>> This means, while it was intended to let rare searches travel far further, it
>> now makes rare files virtually impossible to find.
>>
>> For Creative Commons Content (as for any new content) that's especially bad.
>>
>> So I want to propose dropping Dynamic Querying, until a solution to the
>> spam-problem has been found.