[edonkeyAdministrator]

Dear Community, as long I am providing my eDonkeyServer I was not minded to do any request on emule forum.

But actually the whole eDonkey network has very tough times, not only the legacy issues which force a lot
of eDonkey server administrator to stop their server.

Also we have "internal" problems now.

The biggest problem is the financial aid, actually it is only possible to earn money about control the welcome message on visiting the eDonkey server. But only a few clients read this message and we have a very very small amount of clicks on our website.

So it is financially very tight calculated, we have server to pay, traffic and especially a hugh amount of lawyers' fees to say nothing about the mentally stress to provide such a server, it is not easy to be part of the enemys of RIAA, BREIN and all the other guys which would preferably send me in prison for some years.

But this time it is about the lugdunum server software, with due respect to lugdunum, I don´t want to use his server software anymore.
The reasons are powerful and I don´t think it is a good idea to talk about it, but this is a situation I can not welcome.
In my honest opinion and researches I found fatally self interests which are being practised for as long his server software exists through his backdoor system, while he was also of course did positiv things to keep the edonkey network working.

So I can just complain WE NEED OPEN SOURCE SERVER SOFTWARE, as fast as possible to be able to keep the edonkey network clean and without self interests.
Also I am ready to spend a fair amount of money for good server software, which source code is available.

By the way my server is freezing and crashing all the day, I hope it has nothing to do with my disputation to lugdunum some hours ago

This post has been edited by edonkeyAdministrator: 23 November 2008 - 08:10 PM

[NanQuan]

While I cannot offer any insight on the server software issue, I'd like to thank you for what you're doing! It's greatly appreciated!!!!

[Andu]

Well this may just be my opinion but since the server software seems to be on the brink of ending it's lifespan it might not be bad to plan ahead.

We are approaching eMule 0.50 and it might be both symbolically and from the timing POV the ideal time to make a cut here and overhaul the entire network. Taking up the server software development would make it possible to reinvent the network and get rid of all the legacy problems that the ed2k network means. That would mean supporting 3 networks (ed2k, Kad and the new "emule" server network) but it would allow a slow transition. You could even put out new documentation about the new software as the project develops. That way other ed2k projects would be able to be ready to join the new network right away and they could make suggestions to improve it.

All it would really need is a bunch of dedicated developers that would be willing to take up such a project.

But I'd understand if the devs are not too thrilled about the idea either.

[PAKnorm]

edonkeyAdministrator, on Nov 23 2008, 04:09 PM, said:

Dear Community, as long I am providing my eDonkeyServer I was not minded to do any request on emule forum.

But actually the whole eDonkey network has very tough times, not only the legacy issues which force a lot
of eDonkey server administrator to stop their server.

Also we have "internal" problems now.

The biggest problem is the financial aid, actually it is only possible to earn money about control the welcome message on visiting the eDonkey server. But only a few clients read this message and we have a very very small amount of clicks on our website.

So it is financially very tight calculated, we have server to pay, traffic and especially a hugh amount of lawyers' fees to say nothing about the mentally stress to provide such a server, it is not easy to be part of the enemys of RIAA, BREIN and all the other guys which would preferably send me in prison for some years.

But this time it is about the lugdunum server software, with due respect to lugdunum, I don´t want to use his server software anymore.
The reasons are powerful and I don´t think it is a good idea to talk about it, but this is a situation I can not welcome.
In my honest opinion and researches I found fatally self interests which are being practised for as long his server software exists through his backdoor system, while he was also of course did positiv things to keep the edonkey network working.

So I can just complain WE NEED OPEN SOURCE SERVER SOFTWARE, as fast as possible to be able to keep the edonkey network clean and without self interests.
Also I am ready to spend a fair amount of money for good server software, which source code is available.

By the way my server is freezing and crashing all the day, I hope it has nothing to do with my disputation to lugdunum some hours ago

Thank You for what you do.

[AUG]

http://forum.emule-p...s...35636&st=20

This guy said he got code but he can not release it.

[fox88]

AUG, on Nov 25 2008, 01:14 AM, said:

This guy

As far as I know 'this guy' wrote this code

PS. And his nick is already mentioned in this topic.

This post has been edited by fox88: 24 November 2008 - 10:49 PM

[AUG]

server version 17.15 (lugdunum)
We relocate eDonkeyServer No1 IP: 92.241.164.30 Port: 4242
Please use our new eDonkeyServer No1 ed2k://|server|92.241.164.30|4242|/
. __ ___ ______ ______ ________ _________ __ ________ __ __
. / / / | /__ __/__ __/ ____/ | / /__ __/ /__/ / __ / | / / / /
. / / / / | | / / / / / /__ / | / / / / __ / / / / | / / / /
. /_ / / ___ | / / / / / /___ / /| / / / / / / /__/ / /| / /_ /
.( _ ) / _ / |__\/ _ / / _ / / _____/ __/ |_ / / _ / / _ / /_____ / _/ |_ / ( _ )
We relocate eDonkeyServer No1 IP: 92.241.164.30 Port: 4242
Please use our new eDonkeyServer No1 ed2k://|server|92.241.164.30|4242|/

This post has been edited by AUG: 27 November 2008 - 03:55 AM

[eremini]

As a server administrator (TVU donkey server) I would like to join this complaint/request. There is no support for software, which has random crashes for unexplained reasons and some other problems... Maintaining an ed2k server is really a problem... Am ready to pitch in money as well

[Some Support]

Quote

Taking up the server software development would make it possible to reinvent the network and get rid of all the legacy problems that the ed2k network means.

There really aren't any major legacy related server problems which would justify another network.

About the initial request: Not wanting to comment the complaints about the current software, i can say that it would be of course nice to have a maintained and open source server software. However i can also say that we aren't going to work on this, as we don't have the time ressources for a second software project at this point.

[Andu]

The point is not to solve server related issues but to bring the ed2k protocol up to speed. It would allow to do progress on a level which is simply impossible with the need to keep backward compatibility.

[Some Support]

Quote

It would allow to do progress on a level which is simply impossible with the need to keep backward compatibility.

Like which kind of progress? There isn't any problem with the protocol itself, in the opposite it was fairly well made in the first place.

[Andu]

I don't know. Let me see if I can come up with something.

1) New chunk sizes. No need to bother with 9.28MB chunks. Instead you can have chunk sizes of your own choice. Maybe even use variable chunk sizes depending on the size of the shared file like they do on BT by default.
2) Free choice of the hashing scheme without having to bother with 2 hashes for ICH and AICH.

I bet there is more but I'm just too lazy to make a proper list.

[Some Support]

Andu, on Nov 27 2008, 11:23 PM, said:

I don't know. Let me see if I can come up with something.

1) New chunk sizes. No need to bother with 9.28MB chunks. Instead you can have chunk sizes of your own choice. Maybe even use variable chunk sizes depending on the size of the shared file like they do on BT by default.
2) Free choice of the hashing scheme without having to bother with 2 hashes for ICH and AICH.

I bet there is more but I'm just too lazy to make a proper list.

I see. The major backwards compability problem for such changes aren't the server however, but the existing users and the data they contribute to the network - we didn't deceided to keep focusing on backwards compability because the servers hindered us (this isn't even the case with Kad). So the question if to develop a new server or not isn't really related to the question if we want to create a new network.
Also in my (personal) oppinion the answer to the second question is quite clear: No we don't. The protocol is flexible enough to do all changes we need and want. For example the hashing system is going to change sooner or later in the direction of AICH (which is a sha1 tree and therefore also allows flexible blocksizes) - it is just more work (and of cause some overhead) to do this in a way which keep compability with the old standards, but this work is well worth it considering that the ed2k network is one of the largest arround while a complete new network starts with 0 users.

[Andu]

Well the new network wouldn't exactly be new. It would essentially be ed2k v2. An evolution of the current network. Incremental development has it's upsides but it also has it's downsides. At some point development usually needs to make a clean cut and put everything on a new clean footing. Not to make a change for change's sake but just to get rid of all the bloat that accumulates over the years and clean the mess out.

Unless you are saying there is no bloat/inefficiency in the current protocol and you'd write it exactly as it is if you had to write it from ground up right now. But if everything would be so peachy I don't see why so many mods think they need to make "improvements" that need protocol extensions. Like netfinity's block exchange or all the different techniques that are supposed to improve the release process.

A new server software wouldn't necessarily be required but if eserver doesn't develop much it would make sense to start a new project.

[Some Support]

Quote

Unless you are saying there is no bloat/inefficiency in the current protocol and you'd write it exactly as it is if you had to write it from ground up right now.

I say that the "bloat" which is mostly code wise to keep compatible to older clients, rarely overhead wise is for sure not worth starting a new network. The protocol is exendable fair enough and i'm pretty sure it looks very well compared to most other application protocols. Even if it would be rewritten to drop all backwards compability related stuff, you wouldn't see any faster speeds or anything - it's not that protocol overhead is eating any significant amount of bandwidth. There seems to be the misjudgement that the protocol itself is less effecitve than others for example BitTorrent. This is not the case (i would say the opposite, but my oppinion might be biased ) and mostly results from people downloading some file on Bittorrent faster than on Ed2k and contributing it to the protocol itself (to the technical aspect instead the network philosophy aspect).

Quote

But if everything would be so peachy I don't see why so many mods think they need to make "improvements" that need protocol extensions

Me neither But honestly, I didn't said the protocol nor eMule is perefect or finished and can't be improved. No software will ever be and so devs try out new things, code new stuff here and there. That's what mods do too, that's what we do. The fact that they can easily implement their changes into the protocol rather shows that it is extendable without having to rewrite everything.

[R3Qu13M]

Hi all!
Just for curiosity, for which reason the page offered by eDonkeyServer [www.eDonkey.to] is redirected towards the Web of usexnet [http://www.usenext.com/]?
is also usexnet server administered by edonkey?

[PacoBell]

R3Qu13M, on Dec 5 2008, 11:56 AM, said:

Just for curiosity, for which reason the page offered by eDonkeyServer [www.eDonkey.to] is redirected towards the Web of usexnet [http://www.usenext.com/]?
is also usexnet server administered by edonkey?

edonkeyAdministrator, on Nov 23 2008, 12:09 PM, said:

The biggest problem is the financial aid, actually it is only possible to earn money about control the welcome message [read: usenext advertisement] on visiting the eDonkey server. But only a few clients read this message and we have a very very small amount of clicks on our website.

[kobolds]

I though with kads we don't need server anymore . what for at this time to start on a project that doom to extinct ?

[Dick_Manitoba]

It isn't doomed. The centralized server will always prove superior if that server is legitimately functioning as an eD2K server. It is not composed of thousands of clients all interacting, and therefore, does not suffer any loss of data when clients disconnect or become unavailable to each other for any of the dozens of reasons that can occur in KAD. Think of how many times you've fired off a simple one or two common word search in KAD and came up empty, or nearly so. That sort of thing does not happen with a server. You'll get it's responses, and all of them if it's willing. KAD will never provide that capability because it was never intended to allow you to fire off a dozen "More" requests after it's sent you hundreds of hashes.

So while KAD provides many useful functions, and all of them self-reliant, it isn't perfect either, and a working server offers up features that are valuable to anyone willing to use them.

[lordpake]

Very nice post from D. Manitoba . Personally, I have always seen them complementing each other. While Kad may provide a fallback solution for a while, I do believe there's merit in what Dick said about properly functioning servers.