Shareaza

Any chance of KAD network being implemented?
With more and more legal atacks against e2k network, users are moving forward to Kamdelia, being it a major source of files.

Thanks.

There has already been done work on this, but as we lack developers ATM, we don't have the needed people to work on the issue.
This means, it will be done sometime in the future, but it will take its time, if we don't find more devs to help us on this task.

Well would help if you could ask skilled programers to become devs, that could help.

I am currently using an eMule mod, largely precisely in order to see how bad we need kad, obfuscation, secure id, and other such improvements. At the moment the most useful of these seems the secure id, which in theory can sometimes lead to lower speed with some mods. (The official client doesn't care AFAIK.) Kad is at the moment still most useful as a backup for the servers to deal better with server changes. Only 10% or so of clients is Kad only. This is only slightly larger than when I last checked a year ago. And these clients are still accessible thru source exchange. So there is still time for Kad.

True, but may I ask what good eD2k servers are there left?

My server list has 25 that respond thru PeerGuardian... Of these about half report significant user counts. Three report good user counts and 7500 file soft limit. I'd say that right now the situation is better than it has been in a while.

By the way since the servers report a soft file limit Shareaza should use that instead of the default 1000 file hard limit if possible. I haven't installed 2.5 yet for obvious reasons (testing something else) so I don't know if this has already been done, but it would be quite valuable for people with larger libraries. And it would stop people from accidentally setting the limit too high as well.

Should the code been implemented from scratch?
I mean, maybe some dev can take the implementation on official eMule client source, or any other open source client.
Would that help shorten implementation time?

The problem is that it would be better if the same kademlia code supported BitTorrent and Gnutella DHTs in addition to Kad. Also the way Shareaza currently works is quite different from eMule. I am sure that the eMule code can be reused, but whoever does it should think about lots of issues that eMule devs didn't need to. So it needs a skilled developer with the time to really think about what he is doing whether using eMule or something else as a base or doing an entirely new implementation.