[Some Support]

DavidXanatos, on 05 February 2016 - 07:55 AM, said:

But than I would guess that will not be a "viable fork / mod /new client which has a responsible development", right?

Indeed and posts like this which show a good portion of paranoia and insults only underline the fact. But aside from this and all "visions" you have for your client, it's not open source which would be a requirement.

[DavidXanatos]

Some Support, on 05 February 2016 - 08:49 AM, said:

DavidXanatos, on 05 February 2016 - 07:55 AM, said:

But than I would guess that will not be a "viable fork / mod /new client which has a responsible development", right?

Indeed and posts like this which show a good portion of paranoia and insults only underline the fact. But aside from this and all "visions" you have for your client, it's not open source which would be a requirement.

Why paranoia? I'm not even the first to mention in this thread the offi point on adding or rather not adding NAT-T.

About NeoLoader source: a NeoLoader version without hosters support is available as source at https://github.com/D...natos/NeoLoader

As explained in other threads hosters needs link protection and this is only possible as closed source, see jDownloader, while they put 99% of their code under the GPL the binarys they distribute contains 1% closed source code with the DLC keys. And every contributor that wants to participate official has to give them written permission to use his otherwise GPL'ed code in their closed source public release.

This post has been edited by DavidXanatos: 05 February 2016 - 08:57 AM

[eremini]

Nah yet another torrent client is stupid too, there's 100500 of those. There just needs to be an actively maintained ed2k client. If you at least githubed the source and accepted pull requests and had a bug tracker would be a nice start

[xSTHNSx]

@eremini
Thats not what David meant by his post. The message he was trying to convey was that these days in age no one would waste time developing another standalone client for ED2K/KAD network. Since our current userbase is near half million active at any given time with rotational userbase of near million+. This is why you see hybrid multi-network bridge clients like mldonkey, Shareaza and David's own NeoLoader. Which uses ED2K/KAD to harness its power in conjunction with the de facto P2P protocol Torrent as part of additional complementary extension. As you can see there is no point of reinventing the wheel as no one is even flying any more let alone driving, as everyone is currently beamed up via teleporter.

@Some Support
Bit ironic you of all people talking about "paranoia" as you were the only leading voice obstructing future development decade ago in absence of innovative developmental realistic goals which were backed up by foolish idiotic idealism as it held no merit then as it doesn't now decade after the proven fact.

You Sir are the main reason why it will be "Last eMule" as caption addressed under your name. Just because you babble about network overhead does not make it true. How do you think other network operates with DHT/PEX? magic pixie dust (scarcasm)? So you think people would stop using eMule for Windows due to higher network overhead? RIGHT (sarcastic) yet they dropped eMule (ED2K/KAD) completely and moved on with torrent. Ironically whose traffic footprint is much larger huh? So what’s your answer to hybrid network that supports Torrent+GT1/GT2/DC/ED2K/KAD/I2P? As you claim people would leave due to higher overhead BUT contrary to the BS you preach YET they bootstap into 2-4+ additional networks?

1) What are the names of these subjects you speak of who takes part in official stance?
2) What are your future goals regarding client development for network resurrection?

That's right you don't have any. With all due respect man I question your credentials for alternative motives. What better way to destroy something than from within as your our very own ANTI-Merkur, I don't want to be to harsh now.

[eremini]

Well I definitely wouldn't' want torrent crapolla in my ed2k client

[Mobandi]

I stand with Some Support, let eMule fade into obscurity, fork the client or restart from scrap.
My opinion is that the first thing to do is to develop a cross-platform/multi-architecture shared library for ed2k, somebody started doing something like that on github but I think he abandoned the project ( hxxps://github.com/qmule/libed2k last serious commit is from 1 year ago ).
After that, develop clients that rely on that library, this way the users can have all the options in the world: headless client for linux servers, pretty GUI client for win10, shitty GUI client for win95, etc.etc.
I always wanted a linux headless client different than aMule, and nobody does that
Another thing, about overhead: some people may say that we're living in times where internet connections don't suffer too much from overhead, well I might say that they're wrong, not everybody has a 100 mbit... Hell, I don't even reach 1 mbit in upstream, wanna talk about it?
My post may sound naive, but I'm open to critics, if there is something wrong with my vision, please let's discuss about it. I don't have any problems to change my mind if I'm wrong

This post has been edited by Mobandi: 06 February 2016 - 10:40 AM

[xSTHNSx]

@Mobandi
Now lets try to understand first part of your post where you claimed you stand with Some Support in favour of eMule to fade into obscurity. As I see you have been member in this forum for decade+ but mostly inactive so I do assume you read this forum and don't bother to just participate in threads. Now the idea to abandon the main client of the network fully not just its support in favour of another fork is naive and dangerous, as this kind of thinking is very puzzling to me as I see no rational approach to it. I mean in absence of active overall developers interested in our network to support its client eMule (Win) and aMule (*nix) is scarce to begin with since mostly all work is done on de facto P2P protocol Torrent. At the sometime a new fork wouldn't fracture the current userbase or reduce it but complement it. As it will increase it's userbase marginal just enough not to have any overall impact as you can currently witness with 3rd-party mldonkey, Shareaza, lphant for past decade, does it have any serious impact or was it game changer? No. I mean you don't even have to look at 3rd-party clients just look at 2rd-party clients like aMule and eMule-Mods, do you think it has any overall impact since eMule is cause of the whole problem in ED2K/KAD network for past decade? Only way to fix it is by having the mainline client slowly bring its core standard upto date. Now in your case if you don't want to deploy monolithic aMuleGUI or it's headless aMuleD, aMuleCLI, aMuleWEB. why don't you use 3rd-party client mldonkey for headless as its superior in terms of control. But I highly suggest you stick to aMule's for mainline support as its the de facto ED2K/KAD client for all UNIX/LINUX platform.

For the second part of your post regarding bandwidth lets tackle that as well. It is very true not everyone has 1mbps upstream tier let alone broadband for that matter due to regional restrictions and service availabilities from provider. At the same time you need to accept the opposite fact that some of us has much higher tier as in some region its very common for 1gbps tier. I will give you example back in 2004 when I had ADSL1 (Rev2) from BellSouth(AT&T) with their highest tier Xtreme (6mbps/512kbps). I was still using 95% of my upstream line capacities from home to serve in ED2K. The 512kbps translates to 64kB/s with actual 56kB/s usage where I had 52kB/s allocated at any given time. Due to physical restriction of higher tier availabilities from Teleco I took matters in my own hand and paired 2xADSL1 circuits via multi-PPP for active shared full path load balance. This was well before they deployed UVerse which was HTTC/N to ATM and then twisted dual copper to home. If you go to DSLReports forum you will see all my posts in detail on breakdown howto setup. Point is a decade after as my tier is now Extreme 150mbps/25mbps from Comcast Xfinity residential fully unmetered no cap via DOCSIS3. I have local0 (STHNSv4x0) allocated 3MB/s (24mbps) with remote1-remote5 (STHNSv4x1-STHNSv4x5), which are hosted on 1gbps shared optimized links all allocated to ED2K/KAD. So far I have never seen 1MB/s upload as mostly they all are in 600-700kB/s range to single user since ED2K/KAD userbase are higher in EU which has slower connection. I share about 350-400+ GB per day now. As I have NEVER used eMule in my life and never will but if I were operating on Windows and even if were to run it then it will NEVER fully utilize my high-speed link as eMule is hindered and can't handle it. You may want to refer forum search regarding this matter if your unaware of the issue as it was addressed half decade ago and still not fixed in official version as most MODers have released mods which supports it.


app.box.com/s/ofody4ti7wew9ty7242oq24zc5u0qclb


app.box.com/s/mwabxbljw01vepbzjab391km59yasaqp


Your entitled to your own opinion in general vision but its not shared by us here as we trying to push for change for past decade+ for better and not worse. There is very old saying that those who does not understand changes fears to accept it as they fear what it will bring. With that comes the burden of being out of touch with reality and falling back from common core standard. Which has serious implication in the long run in survival.

[Mobandi]

@xSTHNSHx
Thank you for your very detailed response.
I'll do my best to answer all of it, even though you know a lot of technical things that I don't understand

1) Let's talk about your conclusion: «Only way to fix it is by having the mainline client slowly bring its core standard upto date».
How would you do that? Some Support won't let anybody do that.
So, assuming you can't/won't steal his brand, why isn't a fork the solution? History lists many forks that had more success than the mainstream code.
Another point of view is that once the users notice that there's a new eMule fork being developed, that works better than the former, and that implements all (or most of) the functions implemented by eMule mods, then: some users will migrate to the fork; mods developers will migrate to the fork too (that requires that all devs get along, can they do that?).

2) I don't understand this part, you are saying that you can upload to 1 MB/s, but you actually (from your screenshot) upload at 860 kB/s. Why is that? eMule problems or other users' bandwidth isn't enough?

P.S. Nice workstation!

This post has been edited by Mobandi: 06 February 2016 - 06:41 PM

[RejZoR]

eremini, on 05 February 2016 - 11:35 AM, said:

Well I definitely wouldn't' want torrent crapolla in my ed2k client

I don't want bittorent in my eMule either, but I want eMule itself to evolve.

[xSTHNSx]

@Mobandi
#1
Which is why I made this thread with active poll so others along with Some Support is aware of it once again. As this is not the first time this issue has been discussed in past decade and it won't be the last time. As he himself knows people here does not support nor share his backward way of thinking which has openly plagued this network for over decade with eMule being the cause of the problem. Which was the original debate started by eDonkey2000 accusing eMule client for network slowdown, ironically over decade and half later it propelled us into brink of extinction with our network's traffic footprint virtually wiped out.

As his name states he provides "Some Support" but not fully. Which is obsolete now as its obstructing and hindering developmental growth. He keeps using "our" and "we" as he claims other took part in decision regarding official stance but can't name them. Nor can he can't give detail explanation regarding anything for that matter or why it was then how it reflects now. Even if he did it doesn't hold any merit specially today. I think at this point he is in denial and maybe borderline delusional, as its his own personal problem that he can't cope with it decade after all the points were virtually proven wrong. So he should step down and let other take the torch since he is not fit nor qualified any more since he lacks overall vision to move forward into the future with current standards. Which is why this thread was created to raise awareness among MODers/developers and users to push for change in mainline client and not promoting another fork.

You don't need to tell me about importance of forking as I derived from UNIX world as I have strong background in networking and programming. Far as file trading goes before 1G-P2P (Gnutella/Napster) was even developed I wrote XDCC bot in perl/tcl for IRC to host on IRN. As that’s how we did it back then aftermath of UseNET being dropped by local ISPs natively came our scene warez era. Moving forward into mainstream 2G-P2P file sharing era I was part of eDonkey-BlackNET releaser project fork from xMule later aMule tree under Shareaza's mod revision.

PS: Damn can't post fully in single post.

This post has been edited by xSTHNSx: 06 February 2016 - 10:25 PM

[xSTHNSx]

Odd I think I came across a bug or something I can't make post.

@Mobandi
#2
http://pastebin.com/FGPW1n2D

This post has been edited by xSTHNSx: 06 February 2016 - 10:48 PM

[pier4r]

xSTHNSx, on 31 January 2016 - 12:33 PM, said:

pier4r if it was so easy why wasn't it done before?

Because - imo - (a) people have ideas but no time ( it is difficult to make people cooperate © someone has to provide a domain to let the fork have a bit of authority, like amule is not released here but on its own channel/site.

If making, already just the developers, cooperate would be easy, you will find one mod accepting push from others in a subversion/git/whatever repository . Instead most of the time developers works on personal basis, integrating ideas from others, and sometimes they works together for little time.

Now add to this the management of ideas that should flow from the community to the code.
It would be great, but for me it is difficult. And this won't change even if you change Some support, because the organization infrastructure between people has to be designed, maintained and used.

[nemoW]

I have an idea.
1. Prepare crowdfunding campaign to restart eMule development and motivate developers (where people can vote with their money for particular features and overall eMule development direction).
2. Release minor eMule update full of advertising about crowdfunding campaign (during install and in client).
3. Start crowdfunding campaign to check whether people really need new eMule and what they demand from it.

But wait, this is senseless without point 0: Find developers interested in eMule renaissance and further development.
Sadly I think these people are no longer there. Because if they existed and wanted to develop a modern & popular eMule successor, they had enough time to do it yourself.
Since we do not see the eMule successor except BitTorrent, it's withering look natural and inevitable.

This post has been edited by nemoW: 08 February 2016 - 09:05 PM

[DavidXanatos]

Mobandi, on 06 February 2016 - 10:40 AM, said:

My opinion is that the first thing to do is to develop a cross-platform/multi-architecture shared library for ed2k, somebody started doing something like that on github but I think he abandoned the project ( hxxps://github.com/qmule/libed2k last serious commit is from 1 year ago ).

Well if some one wants to develop a new modern eMule only client he can grab the Ed2k/Mule code from my NeoLoader source, https://github.com/D...natos/NeoLoader its implemented very modular such that separating that from torrent and other parts is virtually no work at all.
The only limitation is that the license is not compatible with the GPL so in his choice of additional libraries the person would have to stick to LGPL, BSD, MIT, etc licenses, you know really liberal licenses.

nemoW, on 08 February 2016 - 09:02 PM, said:

But wait, this is senseless without point 0: Find developers interested in eMule renaissance and further development.
Sadly I think these people are no longer there. Because if they existed and wanted to develop a modern & popular eMule successor, they had enough time to do it yourself.
Since we do not see the eMule successor except BitTorrent, it's withering look natural and inevitable.

"modern & popular eMule successor"

I think the problem is the "popular" part, there are modern clients that support the Ed2k/Mule protocol they just are not popular due to bad/no marketing.

[xSTHNSx]

You guys are missing the bigger picture. As long as we all remember the history of where and how eMule came to be and share its core value. Which will be adopted as inherited from old to new project for underline principle.

Current core value:
A) New clients backward compatible with all clients
Uploaders must not able to ban Downloaders
C) No crapware embedded within the client
D) Open transparent source code for viewing
I think we all agree with those as no one has any objections to those other than leechers.

Now for it to work all the scarce resource now that's dedicated to variant of mods has to be allocated to single project. Where all the MODer/DEVer play key role in some abstract of the project. For example David can work on NAT-T, IPv6, LowID2LowID as he done under his NeoLoader. Stulle can work on CS and DLP incorporating from his other mods. Fox can submit bugfix as he is doing now. Along with Tux, Wizard, ect all who each will play over seeing something. This will enforce raw burst project with unified progress under single brench thus overtaking current project even with objection from Some Support. Since his crippling stance backed up by foolish idealism to not tackle these problem for past decade propelled us into brink of extinction. with right leadership for full fledged development from DEVer/MODer pushed by active community and media to go with the network backed up by scene releaser with dedicated bandwidth and rich contents we can start a full pledge campaign to resurrect eD2K for this era.

[hooligan3000]

and there it starts

dlp

but the other things and facts sounds nice

[DavidXanatos]

Indeed!.

In my opinion eMule should get rid of the credit sytsem and a waiting queue in favoure of a more random upload scheme like torrents use.
The upload should be distributed more egalitatian, ensuring that when someone starts a new download it wil start downloading right away with next to no waiting.

[tHeWiZaRdOfDoS]

... and in my opinion the credit system should be enforced
Also, a random queue system will not ensure that the download will start ASAP, either.
Well, but that's just the usual "your opinion - my opinion" stuff

About the topic:
It would be great if development would be more active.
There was a point where you @Some Support could have literally picked between multiple talented developers to join the dev team. With nice ideas, with nice mods... but you missed that point - be it on purpose or accidentally.
Nowadays there aren't many devs left. They moved on after being stuck with a stubborn Vanilla client... other projects, real life... whatever happened.
It's true that eMule can be extended but as @RejZoR already stated: Extensions are useless unless supported by a big part of the network.

It'd be cool if an extension system would have been introduced, e.g. a dev posts and extension, an official dev reviews it and it's added to the next (bi-monthly? ) maintenance release of the Vanilla client. Thus, stuff like ICS or even NAT-T could have made their way into the client without too much effort.
I also understand that the project needs some guidelines but some stuff (upload slots, per-slot speed) could really have been adapted years ago - IMHO you simply waited too long.

Still, that doesn't mean that I think you should "pass the torch" but instead you should rally some "junior devs", sit down with them, talk about the direction the project should take and let them do the work




PS: All this discussion really makes me wanna compile another kMule soon

[RejZoR]

Also, I find it funny how official eMule team isn't (and never was) looking for simplicity and ease of use which would suggest overly automated and simplified client that anyone can pick up and use. And yet they still resisted picking up all the cool more advanced features developed by the modders and mostly looked after by advanced users. Which just hits me straight in the face. And not in a good way. NeoMule, kMule, Beba, MorphXT, XtremeMod, these are some fine, really well coded mods. And to me it seems like ZERO features were borrowed from it to enrich the official client. Something mods do all the time between each other. Like wizard said, there was a time when official eMule could borrow cool things from mods. I think those times have passed already. In the past, active modders would regularly indirectly contribute to the eMule project because official client could be picking up stuff from mods. But now that mods are basically dead and official dev team non existent, it just all stalled.

I mean, if someone actually put just 10% of overall needed effort, interface could be arranged better, less unnecessary over-complicated bloat thrown into faces of users, simplified first time wizard, improved UPnP detection and application (because it sometimes keeps on failing for no logical reason) and bloody hell, update the god damn server list and nodes file. eMule could be improved 100x better just with that alone. And yet no one seems to be bothered doing even that. Don't tell me sticking a fresh server list and nodes.dat file into the client is something that needs 5 years of testing and R&D. It's just a total BS if anyone will say that. Most people give up because right from the start default servers are all dead, nodes are non existent because they are so old. I had problems when I re-started usage of eMule a year or two ago. And I'm a bloody eMule veteran so to speak. And then people maintaining eMule are expecting casual people to go full haxor on it. Sorry, just no. I may not be a programmer, but if working with software and customers thought me anything is that you have to treat users as the dumbest people you can imagine. Only that way you can establish a functional base and then build advanced features on top of that. Not the other way around.

[Tuxman]

Bullshit polls are bullshit. "Should Some Support step down or do you want to see everything here burn in flames?", suggesting that eMule just needs a different "main developer" (or whatever he prefers to be called) and everything will be better.

If you think eMule is dead, go away and stop bothering us. We're zombies to you anyway. Goodbye.
(I could bring you some very vital evidence of eMule being alive and kicking, but I am not willing to violate this board's link rules for that.)

xSTHNSx, on 21 Januar 2016 - 09:02 , said:

I'm sure DavidXanatos, Tuxman, Stulle, tHeWiZaRdOfDoS, or someone trusted in the community can bring eMule for Windows upto date.

It totally depends on what you would call "up to date". eMule can run on Windows 10, after all.

xSTHNSx, on 21 Januar 2016 - 09:02 , said:

Some Support contributed much as he could now if he can't keep up why doesn't he pass the torch?

Why do you need anyone to "keep up"? What is missing in eMule and why don't you contribute it?

xSTHNSx, on 21 Januar 2016 - 09:02 , said:

Don't tell me to code.

Ah, precisely. I think you get the point even without me explicitly naming it.

Mannnein, on 22 Januar 2016 - 07:29 , said:

Even David, Tuxman and Wizard have so different opinions about the future of emule. So which of them has the right opinion, which is wrong.

I do. (That was easy.)

Mannnein, on 22 Januar 2016 - 07:29 , said:

Be p2p, be more like torrent, focus to an healthy network, focus an speed.

If you think P2P is about speed, you did not understand anything of P2P.

This post has been edited by Tuxman: 09 February 2016 - 11:36 PM