[loverboy]

hooligan3000, on 06 February 2012 - 04:17 PM, said:

my ed2k server is no fake
also from peerates and emule-security ;-)

Maybe the problem is not that "the server" is not fake (tested by peerates)but that in any moment its data can be taken away by FBI or similar to sue people.

A couple of weeks ago megaupload was raided, and similar sites have closed to prevent similar charges.
Today btjunkie has closed (piratebay closed long time ago).

So hitting emule servers, after filesharing sites and bittorrent sites, is not so unbelievable nowadays.
The only (partial) solution seems Kad network.

[DavidXanatos]

hooligan3000, on 06 February 2012 - 03:17 PM, said:

my ed2k server is no fake
also from peerates and emule-security ;-)

So you want to say that ed2k server support is still useful?

Is there any developement of teh server software? cause from what i know the only dev that head the sources quit after a law change in france.

[tHeWiZaRdOfDoS]

Yeah and - unfortunately - the src wasn't released... what a pity...
But you don't know... maybe someone picks up development... RIAA etc. seem to be really interested in working servers *g*

[hooligan3000]

loverboy, on 06 February 2012 - 04:18 PM, said:

...data can be taken away by FBI or similar to sue people.

A couple of weeks ago megaupload was raided, and similar sites have closed to prevent similar charges.
Today btjunkie has closed (piratebay closed long time ago).

So hitting emule servers, after filesharing sites and bittorrent sites, is not so unbelievable nowadays.
The only (partial) solution seems Kad network.

ed2kserver is only for source exchange no data log or other on the server
megaupload is closed 4 money laundering and a project called megasound(they wanted to employ artist by there self)
btj as no eggs:-D
and tpb is still online 4 me o.0

@david for what u need server development?
okey 1reson ipv6 but not realy for other things

the most will die out
emule, torrent, och, dc and later usenet
also i think little communitys will survive

so long...

This post has been edited by hooligan3000: 06 February 2012 - 08:06 PM

[DavidXanatos]

hooligan3000, on 06 February 2012 - 08:05 PM, said:

@david for what u need server development?
okey 1reson ipv6 but not realy for other things

for example,
or AICH hash exchange,
or some modern relevance based search method

hooligan3000, on 06 February 2012 - 08:05 PM, said:

the most will die out
emule, torrent, och, dc and later usenet
also i think little communitys will survive

I dont think so, since the dawn of time, the only reason that caused a fielsharing technike to go down was a better one, as long as you cant think about something better, more efectice than OCH, torrent, ed2k, it means it wont die for now.

The content MAFIAA form the United Police States of America is loosing thair battle as more and more pirate parties rise to power.

David X.

[]Axl[]

DavidXanatos, on 06 February 2012 - 05:30 PM, said:

Is there any developement of teh server software?

Is there any development of the eMule software???
I agree that a software without updates is dead, emule is not update by almoust 2 years, and in this forum you speck that it is not a problem.

This is the moment to relase a new version if we want to get back same users: megaupload and similars are closed, even btjunkie, the legal history of tpb etc etc.
With wich changes?
If in 2 years the developers have no news, could be enough to change default settings: only clean server (peerates list) or no server by default, uPnP default enable (a lot of beginner users are low ids only because dont use this setting), Upload Speed Sense and unlimitend upload speed enabled in default and some cosmetic changes...

Sincer I hope that in 2 years someone has done something more, but even this could be sufficent.

We bet that if a new version will be relased quickly emule would have a lot of new users??? ("A new version of emule is out, so it is not dead!I want to tried it after a lot of time!")
...and this is really possible, Kad is stronger that torrent sites about legal attacks.

just 2 cents of a user that have emule always open from 8 years

[Tuxman]

From 08 February 2012 - 11:26 AM:

This is the moment to relase a new version

Why? What should it do? What is missing? Name it!

The vi editor has not been updated since 1979. So what?

This post has been edited by Tuxman: 08 February 2012 - 11:06 AM

[DavidXanatos]

Tuxman, on 08 February 2012 - 11:05 AM, said:

From 08 February 2012 - 11:26 AM:

This is the moment to relase a new version

Why? What should it do? What is missing? Name it!

IPv6 and/or Nat-T

There is also a lot to improve on keyword searches to better mitigate the impact of fakes.

May by some vote managed distribution "channel" system like BitTorrent Tribler client.

Improving KAD to be resistant to Sybil attacks as well as eclipse attacks.

Etc....

David X.

[Some Support]

Quote

IPv6

IPv6 won't get into eMule anytime soon (if at all). Not because it is not needed, but because of the unsolved problems explained in the feature request thread. Possibly the only realy solution would be reworking Kad completely, but I doubt that we are going to do this.

Quote

Improving KAD to be resistant to Sybil attacks as well as eclipse attacks.

Due to the design of Kad there can be no real protection against attacks, but of course stuff can be improved. It is quite a bit a cat and mouse game. The problem however is that if the attacker has enough ressources (IP addresses) then it gets pretty tricky.

What should be in the next version is however a fix for faster uploading together with asynchron / threaded reading of partfiles. This is on CVS already for quite a while, waiting to be tested in a beta. But still I cant say when the next version will be released. Pretty sure within this year, maybe before eMule's anniversary.

[DavidXanatos]

Some Support, on 09 February 2012 - 09:49 AM, said:

Quote

IPv6

IPv6 won't get into eMule anytime soon (if at all). Not because it is not needed, but because of the unsolved problems explained in the feature request thread. Possibly the only realy solution would be reworking Kad completely, but I doubt that we are going to do this.

I guess the new client would have to get a new KAD for IPv6, as you wrote earlier an own KAD would be anyways a good idea for adding new own features.

Some Support, on 09 February 2012 - 09:49 AM, said:

Quote

Improving KAD to be resistant to Sybil attacks as well as eclipse attacks.

Due to the design of Kad there can be no real protection against attacks, but of course stuff can be improved. It is quite a bit a cat and mouse game. The problem however is that if the attacker has enough ressources (IP addresses) then it gets pretty tricky.

While the issue may not have a final solution, there might be a lot improvements that can make the live of an attacker difficult enough to not be able to cause major problems with a reasonable afford.

David X.

[Some Support]

DavidXanatos, on 09 February 2012 - 10:33 AM, said:

I guess the new client would have to get a new KAD for IPv6, as you wrote earlier an own KAD would be anyways a good idea for adding new own features.

Well first you have to solve the (security) problems of course. The easiest solution would be adding a centralized aspect, which would actually solve most attack scenarios, but of course goes against the basic idea of kad.

Quote

While the issue may not have a final solution, there might be a lot improvements that can make the live of an attacker difficult enough to not be able to cause major problems with a reasonable afford.

eMule has already that kind of protection, but when an attacker starts to have thousend - millions of IPs at hand, there is no easy response. Sure you could block the whole IP range, but that isn't really a solution to the problem.

There are some other ideas, described in some reasearch papers linked in kads forum, but they are harder to implement and in the end also only a step to make it a bit harder.

[DavidXanatos]

The way how to handle the problem depands on how the IP v6 ranges are being distributed.

As long as the adversary has somewhat adjacent IP ranges i guess this can be detected with some statistical analysis.

The only major problem would be an adversary that uses bot nets and has milions of equaly distributed Ip adresses.

[Tuxman]

DavidXanatos, on 09 February 2012 - 08:28 AM, said:

IPv6 and/or Nat-T

You have IPv6 down in Austria?

[DavidXanatos]

Tuxman, on 09 February 2012 - 08:05 PM, said:

DavidXanatos, on 09 February 2012 - 08:28 AM, said:

IPv6 and/or Nat-T

You have IPv6 down in Austria?

My international anonymous VPN provider has it.

[Tuxman]

As long as your ISP gives you an IPv4 address, there is no reason to have IPv6.

[fox88]

Tuxman, on 10 February 2012 - 12:51 AM, said:

As long as your ISP gives you an IPv4 address, there is no reason to have IPv6.

First, it might be an interesting problem to solve (for someone). Second, IPv6 probably can help in overcoming Low ID troubles.

This post has been edited by fox88: 10 February 2012 - 11:41 AM

[DavidXanatos]

Tuxman, on 09 February 2012 - 09:51 PM, said:

As long as your ISP gives you an IPv4 address, there is no reason to have IPv6.

You, a person leaving under German law is telling me to use my ISP's IP for filesharing, that completely insane!
Why would a person in his or her right mind do that in this day and age?

This post has been edited by DavidXanatos: 10 February 2012 - 07:46 AM

[Stulle]

Because cops follow the money trail and if I am asked if I would rather pay twice to be caught I would prefer to only pay once. Do that or just share stuff that is legit and free or simply of no interest to anyone that would bother suing you in your country.

Either way, this is paranoid talk. Two wrongs don't make a right and hiding your ass is admitting wrongdoing to some degree.

Tuxman has a very good point. A majority of the (European) ISPs do not support IPv6 for their regular customers and as long as they don't it is doubtful there even is a chance to get good natural test cases for an IPv6 implementation of KAD or eMule. Also, Some Supports reasoning on the scarcity of IPs being one of the major security mechanisms in KAD makes a good point. Unless you solve that issue you will either be stuck with bad hacks (i.e. blocking whole ranges) or a network that will be attacked with incredible ease. Now combining these two with what you said about the distribution of IPv6 ranges should show you that it might actually be bad to just blindly hasten an implementation of an IPv6 KAD. The future topology of the network is not quite clear at this point which opens any code written at this point to shortcomings unless it uses entirely generic algorithms to banish attackers while allowing legit clients in.

This post has been edited by Stulle: 10 February 2012 - 08:37 AM

[Tuxman]

DavidXanatos, on 10 February 2012 - 08:44 AM, said:

You, a person leaving under German law is telling me to use my ISP's IP for filesharing, that completely insane!

I use a VPN located in Luxembourg. Still, I connect to it through IPv4. So where is your point?

[Omnithec]

I am of the opinion that moving to Ipv6 now would be a really, really bad idea. However, I believe things shouldn't be left stagnated. There is a saying in my country, “you don't tinker with a winning team”, that is to say, if something is working and doing it well, you shouldn't mess with it. The reason, obviously, is that you might actually make it worse instead of improving it.

So, following this logic, it would make sense to let eMule as it is and use it to the best of its capabilities. It seems logic, doesn't it?

The problem is, eMule has stopped on its tracks but things around it are moving on. The team isn't winning any more, at least not as it did before, because it is disappointing the fans and is not devising new tactics to counter the other teams' agressive moves. Why the need for so serious mods and why no requested features? Why does no one tweaks the network and the client to help root out fake files, fake clients and fake servers? There are other issues.

This is why I think things should start moving again. If compatibility is kept with prior versions or if there is support for two versions until the new version of the client/network is deemed completely safe, there would be a smooth transition, wouldn't it?

If there was a problem, we could always turn back to the earlier version and continue as we are, planning another attempt after finding out what went wrong. This version of eMule should be sacred, in a sense, and kept around in the sites for a long time, and that would be a fail safe for problems with new versions.



By the way, is it proper for me to explain here how things work in my country so you know how simple it would be to protect the servers here? I think you'd be amazed.