Gnutella has a problem

[The Boogerman]

i dont know about you, but i get a woody every time a bearshare brownoser is attacked, oooooooohhhh do it again!

[spooning leads to forking]

Exactly, give it a name: Bearshare fans started it, tolerated from Vinnie. BS fans trashing other clients were often made a VIP (very important person) or a moderator on bearshare.net. Don't forget Vinnies position: badmouthing anything that isn't Bearshare. Doesn't sound like a improvement for me.

BS/LW do not strongen the network, they try to control the network and make it a BS/LW network. Meanwhile they disadvantage other clients, create obstacles, puting fuel on the fire of clients badmouthing, discourage and prevent other developers and spread serious scumware to pay their own bills. That's the problem.

[CobraKitNu]

Moak your ideals for gnutella are just your ideals.. You dont own the network as much as the next guy. Posts like this, blaming all the popular programs, and other tyrates dont help no matter how much you claim to have done for guntella. The fact that you do nothing to improve the network but instead just bashings against a few programs where you want to split the network apart and encrouage developers to not work with eachother in improving gnutella which is really in everyone's best interest.

Your ideals is really just the same vague description over and over which you use in threads like this where you bash programs and the current network. Not once have you actually proved your ideals.

So instead of the bashing against programs and the network as it is today please post how wonderful your ideal will be and what it would mean for the guntella and most importantly why do you think it will work? Whats the incentive for developers? and what direction does it take gnutella? Please provide some actualy data and examples to support your ideals.

Also without bashing other programs, developers or the gdf explain how gnutella has gotten worst from a year ago? Please explain that to me because IMHO gnutella has gotten alot better..

Do i think you will actually finally post what your "ideals" will mean for gnutella no. I expect you and your groupies to post some insults, call me a troll and etc.. Which make me wonder if your "great" ideals really exist at all.

[Gini.]

Dear developers,

there is a problem. Let me describe my personal dissapointment with Gnutella: I collected
voices from different developers and have spoken it out (I prefer to speak out problems
to solve them). Now I wonder, many have been complaining about Vinnie, Bearshare and
Limewire... but many stay silent. Well, I don't have a own client and it's not my problem
anymore. But it's very inefficient and insincerely IMHO to complain behind the lines and
officialy speak 180

[~Duhh Random~]

Detail: BearShare connected to Gnutella. 'nuff said.
I don't think I have to go into how the BearShare BrowNosers are screwing things up too.

[IUG]

There isn't any, that's the point. Your favorite BearShare author caused all this so you may want to think about how your support for that client has taken Gnutella to new lows.

[Paxton W]

Sorry, it seems that moak left the GF, if you want to know why (if you don't already know), please come to the irc channels and ask him.



You just asked for what's wrong with the network, and i listed some facts that are discussed in many threads, they are not only my opinion...



If someone really doesn't care, he can install as much as he want, but i c it on this pc (a multi user pc) where i have to remove accidently installed spyware (e.g. by my sister) every 2 days.

The problem are not people who don't care, the problem are people wo don't KNOW it!

If you'd install a programm, would you uncheck "CommonName Toolbar" for example (or would you try to remove it if it's automatically installed)?



The first question is easy- if you want to implement it, you can do it, if you want to add your ideas, you have to be member of the small community of developers, and if the protocol doesn't fit your needs, you don't have to use it.

bad features...hm... i haven't thought much about that yet, maybe it could work like this: If a member of this community implements a bad feature, he'll no longer be a member; if someone else implements a bad feature, his client will simply be incompatible.


I know it's not possible, but it should be a main goal to get near the 100%.
Developers don't need to implement all features, but if they are to far away from the other clients, they are incompatible because some features are *NEEDED*
They then have to search a solution, e.g. finding another protocol, or impementing these features.


That's really a problem, maybe we could switch some clients into a both-net mode, that they use the normal net, and have some ports free for building a test net, which is not directly connected to the user net (it would be about 1% more CPU usage to the users) - that's the price for new updated software.


No i meant GDC, the GnutellaDev Community, founded in #gnutelladev. It was a try to collect some developers and develop the clients with equal feature decisions etc.



There is a reason to leave the GDF, i repeat it: Vinnie!! Look in my posting you quoted to get the answer why he's the reason.



i forgot to say, pollings should be made in the member area, where the developers are, and here it is how it could work:

Someone has an idea, say SuperduperGIGApeers. He writes a description about it. Then, it will be discussed and changed in CHAT or FORUM, with a time limit, let's say 2 weeks. Then there's a poll where every member can chose what's best for him. At the end, say another week later, it's ready for implementing into the Developers.net, or if noone wants it, it will not be used.

Any questions?

[Muzik]

Morgwen, please remove the unecessary flooding of Sephiroth quoting the entire post of Phantom81. Sephiroth, this type of deliberate obfuscation of the forum is what gets people the name of 'troll'. If you behave like one, people will call you one. Why don't you edit your post and remove the 30 lines of quote before Morgwen has to? The original post is immediately above your post, you should not quote the whole thing AGAIN immediately below your post.

I hate 'the' GDF and I've only read two messages there. I hate it because it is 'behind closed doors' - they do not promote it's existence, it is not listed in Yahoo - I found two or three other P2P discussion groups there, but not 'the' gdf.

Accept firstly that there is no 'the' gdf. There will never be a 'the' gdf. There are already too many other gdfs .. these need to be consolidated if other developers want to get anywhere. Make it someplace open and friendly. Don't go with this idea of having restricted members, expelling people etc. That is just childish.

The one criteria for whether something is considered is whether it makes sense. Post references and background information so people can understand your suggestions, and they will be considered if they make sense. If people post nonsense, contentless propoganda, just ignore them.

I make the above observations because I think that an open protocol should be developed in the open - it seems the only way. And so a culture needs to be adopted which is defensive of important information and ideas, and which reduces the flooding we see so much of in this forum.

By all means people should be able to give their opinions, but when it becomes repetition of things previously stated, please at most give a link to where it was previously stated.

In response to the question of how do you enforce well behaved clients in an open network, for the answer lets look at the internet in general. It is an open network - the only requirement for participation is to speak TCP/IP or your packets won't get routed onto the network.

So why is it that people aren't able to just bombard servers, send wrong packets etc? Well, they can, but the servers and all good network software which interfaces with the network are built defensively. They sanity check packets before passing or accepting them. A lot of time goes into this, and gnutella it seems is too young to have developed a lot of this, so 'exploits' such as passing encrypted packets are tolerated by the clients, when in fact they should not be.

Then there is the 'organised defensive response' technique - how does the internet deal with spammers? Well, in fact people who adminsiter mail gateways put a lot of time and effort into protecting you from spam, believe it or not. There is a forum news.admin.net-abuse on usenet with many subbranches for particular protocols. One of these deals with email spam. People who ALLOW the use of their email gateways for a significant amount of spam, find that mail coming from their gateway is blocked. This makes it annoying if they host people other than spammerss as well, eg if ibm allowed spammers to relay through their gateway, all the ibm staff would be up in arms because their business emails would not be answered or would bounce.

I hope you see the analogy I am trying to make here, Sephiroth - gnutella is growing up and needs to learn to start to behave itself, or it will die. Sad fact of life. There are plenty of competing protocols now with the same idea. The only thing gnutella has going for it (believe me!) is the large number of open source clients. If these guys die, then you have two or three big commercial clients, and that is no different to say fast-track - except fast-track was designed properly from the beginning, and works. There are lots of fast-tracks. There is only one gnutella.

So when the open source clients die, gnutella will be eaten by fast-track or something similar, and then you will be back to square one - lots of small clients trying to build a protocol from scraps of opensource software.

Make yourself a shortcut - go with the opensource software now and build on it sensibly, not higgledy piggledy two-weeks-thinking-before-we-implement-something-fashion.

For people eager to get 'things underway' there are plenty of things that can still be worked on while other things are being thrashed out.

I think LimeWire were unconsionable in bringing out SuperDuperPeers the way they did - effectively blocking older clients from 98% of their clients overnight. They should have phased it in gradually until other clients had accepted that the implementation they had chosen of ultrapeers was the right one and had started to build compatible clients, then reduce the number of connections to non-ultrapeer clients.

Yes, this results in less 'benefit' to the limewire clients who have done "ooh so much work" to 'improve the gnutella network'. But it is polite. Sorry, you may respond 'but this is business' - I respond 'b*ll*&%'. The most successful business people are also the most polite people you will meet. Having an eye for an opportunity is not the same thing as trying to bend the world to your will and pulling the wool over the eyes of your customers. That is very common, but by no means necassary for success.

Spamware IS evil. Spyware IS evil. In current implementations, the two are the same. Advertisers want to 'target' their ads, so the customer is spied on. Sorry, but it is a fact. So far this has been happening for too short a time to see the consequences. And like most computer crime, the consequences will be under-reported even in the rare cases that they are detected, because they are embarrassing.

A network free of commercial clients seems the only response at the moment, to all these problems. We are fighting a threats. It is not that the network is completely screwed right now, it is the fact that it looks like it is getting screwed, and if we don't do anything it will be screwed. It is not that people are completely spied on but it is the threat that they will be endlessly watched and marketed to in return for the priveledge of communication.

As you said, most people don't care. Most people don't care when they buy petrol that the environment is being screwed for profit. But luckily some do, and I'm thankful for it. The less that ordinary people care, the more extreme those who do care have to become when the threat is real and dire.

Nos