Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
Music Media

The Gnutella War: Free vs. Commercial 137

Posted by timothy
from the this-too-shall-pass dept.
Anenga writes "Slyck has an interesting interview with Mike of Shareaza regarding Gnutella2 (see older stories), where he expresses his opinions on how Gnutella2 has been recieved within both the user and developer community. The reaction from the top commercial clients, Limewire and BearShare, on Gnutella2 (as seen in the GDF and elsewhere) is that they will not support it because of how it was presented, however, Gnucleus (free, open source) plans to support it and feels the GDF is not seeing the bigger picture. John Marshall of Gnucleus says 'Now it's more like "Free vs Commercial" clients, which [the latter] would rather develop their own next generation protocol (which would probably never happen).' The article in short: Shareaza will keep Gnutella2 open/free, it's already been very successful with a 80-100k growing userbase, Gnutella2 was *not* based on Limewire's GUESS proposal and is in fact very different from it and Shareaza will continue to both support the original Gnutella ('G1') and of course G2."
This discussion has been archived. No new comments can be posted.

The Gnutella War: Free vs. Commercial

Comments Filter:
  • The question is... (Score:4, Interesting)

    by LucidityZero (602202) <sometimesitsalex AT gmail DOT com> on Sunday December 22, 2002 @06:48AM (#4939954) Homepage
    My question would have to be why Limewire/Bearshare/etc have flat out decided to absolutely not support the new protocol, when it seems fairly obvious that both protocols could be implemented within one client. I understand their wish perhaps develop their own proprietary protocol, but this seems like treading water to me. SPECIFICALLY in a P2P architecture, wouldn't more protocols be a direct correlation to access to more files, and in that case, an increase in popularity, quality and then, in turn, profitability?
    • by gazbo (517111) on Sunday December 22, 2002 @06:59AM (#4939977)
      Well maybe the proposed protocol just wasn't very good? When the companies saw this presentation they will have looked at it with a critical eye and may have found flaws that mean it is hardly better than the current protocol and yet would cost much developer time to implement.

      Remember that these are businesses, so just because He Is Your Brother is no reason to blindly follow his lead. They may well have a totally reasonable, above-board justification.

      • by nick-less (307628)
        Well maybe the proposed protocol just wasn't very good?

        I just tried their beta and must say: it might not be perfect, but its alot better than standard gnutella.
      • Or it may well be that the companies rejected this protocol because it was not theirs and/or they could not completely control it. Logic and reason can be suprisingly scarce.
    • LimeWire/BearShare said they would implement the parts of the protocol that they like but not the whole thing. E.g. Shareaza's idea to prevent DDOS attacks using resources of the gnutella network will probably be used in some way.
    • I'm no math-guy, but it seems that if the new protocol uses a different type of search, then adding users from that protocol would enslow your searches.
    • by number11 (129686)
      My question would have to be why Limewire/Bearshare/etc have flat out decided to absolutely not support the new protocol
      Seems to be two reasons. One is, the new protocol is (at this date) proprietary and not open, though it is being heavily marketed. There are promises that it will be open, some day.

      The other reason is that they're seriously annoyed that the new protocol was named "Gnutella2", which implies that it is superior to Gnutella. Maybe it is, but naming it thus was a marketing coup at the expense of personal relations with the other developers. Kind of like calling your linux distribution "ImprovedRedHat" even though it never was RedHat to begin with.
  • In recent years p2p systems have caught the fancy of CS researchers. Gnutella always gave the feeling of being designed more by hackers than people with a sound theoretical base.

    Has Gnutella2 taken cue from the recent research publications?
    • all i know is that i'm using it and it's better (and free-er) than kazaa which was my previous best p2p program.
      • I disagree... Kazaa works far better for me and for most people I know...
        • Kazaa has one huge problem for me: large downloads invariably corrupt. That doesn't really matter for movies since it just means you have a few seconds of distorted sound/picture. However, if you're downloading things that need to be 100% intact like (cough) a linux iso (cough) it is very convenient if the p2p client bothers to verify that what it is downloading is the same as that should be downloaded.

          Shareaza solves this problem, kazaa doesn't. However, the shareaza network is so small currently that there isn't much to download. With 3 million users at any time, the Kazaa network is unbeatable in that respect.
      • Gnutella basically uses flooding (limited by some TTL)... basically the network expends a large effort to execute queries, and you still don't query all of the network (or even necessarily a very large part of it)

        Gnutella might work OKish if you're connected by a fat-enough pipe, but otherwise it does very poorly.
      • Have you tried any of the stripped down version of Kazaa like KazaaLite? I feel that Kazaa has a better engine in practice than the big G but without being stripped down to the bare essentials, Gnutella definately wins. However, look at Kazaalite vs. Gnutella and there's quite a bit to like on the streamlined kazaa. Mostly tho, Gnutella (when I used it) was a bit bigger of a bandwidth hog than Kaz is. (however, i am NOT an ethernet techie, nor do I profess to know if this is due to my ignorance at implementing either program).
        • Re:Kazaa (Score:2, Insightful)

          Kazaa is based on the concept of "supernodes"... computers which have high resources keep track of info about the "little guys" in their neighbourhood. When you do a querry, you're really only querrying the supernodes directly.... makes for great bandwidth savings exactly where they are needed.

          I think that given the fact that bandwidth of internet users vary by a factor of a 1000 or more (compare a 33.6 kbps modem to a 100Mbps ethernet), any network (like Gnutella) which treats all computers the same way isn't going to perform very well.

          Also, gnutella is almost defenseless against DOS attacks... because it uses flooding (thereby allowing an attacker to instantly turn one packet into thousands or millions). I don't know enough about Kazaa implementations to know how well it resists DOS attacks....
          • Re:Kazaa (Score:2, Informative)

            by patchmaster (463431)
            Gnutella has implemented a system similar to the Kazaa supernodes. Someone suffering a severe case of oxymoronism dubbed them "ultrapeers". It doesn't fully cache the inferior node content like Kazaa, but it does keep a giant hash table for each "leaf node". (Oxymorons are best when flavored with mixed metaphor.) The ultrapeers shield the leaf nodes from most of the query traffic, routing to the leaf node only the queries that appear to have a high likelihood of being successful.

            There are still people using older versions of modern clients and some antiquated clients are still in use, but most of the gnutella network has moved away from the "all computers are the same" model.

            The new protocol implemented by Shareaza (and hopefully by several other open-minded developers) provides for an exhaustive search of the entire network using a combination of direct client-to-hub and hub-to-hub communication. The hubs and clients can both still communicate with standard gnutella ultrapeers (and regular peers if one desires), so there is full integration of the new protocol with the old.

            The new protocol is much more resistant to DOS attacks because queries do not flood the network. Also, a client must establish a trust relationship with each hub with which it intends to communicate. This won't prevent a DOS attack, but it will most certainly slow it down.
  • why should I care? (Score:3, Insightful)

    by el_mindwarp (593805) on Sunday December 22, 2002 @06:52AM (#4939963)
    What I don't understand is how desperate you have to be to go commercial with application which is used mainly for illegal file sharing. I mean, cmon people are sharing mp3's, divx-ripps, applications and games. Not like somebody would actually download mpegs of my pets or my kid brother's birthday. Not that I am preaching, but it was kinda gray activity, we all know what it is used for. Going commercial is going to be the death of it, but hey they, are just lazy and are trying to make a living without having a proper job. I wish somebody would pay me money to change desktop backgrounds in my blackbox, and play around with my Eterm, because I think it's fun. I kinda grow out of that idea long time ago, and had a nice job since... Get a life.
    • You make a good point, and it also goes along with one of my biggest gripes.

      Suppose you DO in fact want to find an MPeg of a dog or something (for whatever reason). The file sharing protocals ARE so overwhelmed with copyrighted material it is becoming next to impossible to find public domain material. Pirates are hurting the file sharing services from both directions.
  • by HealYourChurchWebSit (615198) on Sunday December 22, 2002 @06:59AM (#4939978) Homepage
    I think the paragraph that pays for me is:
    "The GDF's first reaction was negative because they claimed it used the same ideas from other proposals. Once the protocol specs were released this was obviously false, but the GDFs reaction was still negative so Mike has not bothered to release the rest of the specs.

    What it really sounds like is that the commercial entities are balking for something. That is, they are negotiating with their veto.What specifically they want out of this, whether it is a voice in the process or perhaps a cut of the action, I'm not entirely clear. I'd like more on what the author of the article called the 'backstory'.

    • I dunno... a *lot* of people I know are still using spyware-infested KaZaa because they haven't heard about KaZaa Lite. Marketing, marketing, marketing!
  • ...Napster?

    I just say: "In former times..."

    ;)
    • I never liked napster very much. It may have been because I didn't have broadband back then, but to this day I still have a number of 98% complete mp3 files, thanks to stupid napster bugs.

      Now Audiogalaxy... THAT was the holy grail.
  • by darkov (261309) on Sunday December 22, 2002 @07:19AM (#4939998)
    The published interview is about as balanced as a Linux press-release issued by Microsoft. A couple of points to consider:

    - Shareeza has implemented a new protocol and released it soely on it's own client. It hasn't publish any hard details and it's dubbed it "Gnutella2" with no support from existing supporters of the Gnutella protocol.

    - The existing clients are a bit upset that Mike has done this, and his actions since, but doesn't really have an opinion on the technology becuase its specifications have not been released.

    This seems to me as a fairly egotistical kid hijacking the Gnutella name for his own purposes, then charactising eveybody else as bitter about his wonderful new tchnology.

    Theworse thing is that the GDF even pay attention to this fellow, they should just igonre him rather than waste their energy on being upset at his lack of manners.
    • It sounds about as scripted as such a press-release, too. This Mike guy is a first class wanker.

      Of course, it is funny to hear people whining about co-opting the name, when the name itself is a pun on GNU and a yummy chocolate spread. If they're going to use a silent G in their name, they ought to honor that convention and conduct development of the protocol in an open manner, and GPL their reference implementation. It wouldn't stop the bickering, but it would give a clear moral high ground to those who release actual code.

      Why don't they nickname this new protocol "mutella". It's catchy, likely to carry greater mindshare than "gnutella2", and incorporates Mike's name while offering him a suggestion.
      • Of course, it is funny to hear people whining about co-opting the name, when the name itself is a pun on GNU and a yummy chocolate spread. If they're going to use a silent G in their name, they ought to honor that convention and conduct development of the protocol in an open manner, and GPL their reference implementation.


        If they did that, would they have to call it GNU/Gnutella? The redundancy might be too much even for RMS! (Sorry, I couldn't resist.)

        --K.
    • But to use the Gnutella name without permission would be a trademark violation... the intellectual property lawyers should be easily able to tear them apart.

      Oh, yeah, that's right... these are all people who do nothing but violate intellectual property laws all day anyway...
    • The published interview is about as balanced as a Linux press-release issued by Microsoft.
      Why would you expect an interview to be balanced? An interview is an expression of one person's opinion.

      The article clearly stated that this was part one of two parts, the second part presenting the opposing views.
      This seems to me as a fairly egotistical kid hijacking the Gnutella name for his own purposes, then charactising eveybody else as bitter about his wonderful new tchnology.
      If you had read the venom pouring forth from some members of the_gdf you'd realize "bitter" is a rather severe understatement. There are definitely some egotistical kids involved in all this, but I think your labelling has been misdirected.
      • If you had read the venom pouring forth from some members of the_gdf you'd realize "bitter" is a rather severe understatement. There are definitely some egotistical kids involved in all this, but I think your labelling has been misdirected.

        I've worked on gtk-gnutella with Raphael Manfretti, and while I haven't been in touch with him for a bit, your comment sort of stopped me cold.

        Raphael and the other main developers have put *huge* amounts of time into developing and *carefully documenting* a protocol that's pretty backwards compatible. They've been in contact with each other to ensure that their clients work well, and are one of the more impressive examples of competitors working together for everyone's good.

        Now, after all this work to avoid protocol fragmentation, one guy makes a new protocol. He uses some work from existing members, and refuses to publish any specifications. He then *takes* the name from this project that has seen *so* much work to be open (because he wants to grab a bit more PR and a few more $$$), and uses it.

        Now, these developers are, more than understandably, frusterated. This leads to lots of end-user confusion. It's bad for *everyone involved*. Had Make said "I'm making a new protocol and calling it Sharella" or something, it would have been okay. But he created incompatibility, *he* refused to publish specs to let other developers remedy that, and *he* is out trying to profit off the users of the network.

        So, I have to disagree. I've seen a lot of Raph's writing, and while sometimes he turns something down, he acts a lot more mature than, say, Linus does.

        Your criticism of him is unfounded.
    • The published interview is about as balanced as a Linux press-release issued by Microsoft.

      I believe you missed this paragraph... the last of the introductory text before the actual interview begins:

      Our goal with this article is not to take sides on this situation. Instead, we're publishing Shareaza's interview today, and in the following days we'll present the opposing viewpoint. Contribution to the opposition will include interviews with BearShare, LimeWire and XoloX.
    • Just in fairness, has anyone asked the guys & gals in the backroom of nullsoft what they think, being as how they invented it and all?
  • by flopsy mopsalon (635863) on Sunday December 22, 2002 @07:38AM (#4940017)
    The whole controversy over Gnutella 2 is being hashed out over on the GDF message boards. Basically, complaints are that Mike is coopting the Gnutella name for his own benefit, and is not bieng forthcoming with the details of the protocol. Some quotes:

    Raphael Manfred: [yahoo.com] "I'm speaking only for myself here, but I will NOT support Mike's protocol unless I'm forced to do it. I was neutral-positive when this saga started, but Mike ruined it all by his stubborn attitude, and I'm now rather hostile-negative.
    If there are good ideas in what he did, you can rest assure that we'll end up using them. He'll even get credit for those ideas, but it will be within the Gnutella network."

    "tonygeek": [yahoo.com] I am sceptical. All signs are there that (Gnutella 2) is one strictly commercially driven undertaking/experiment (possibly by a very large company pulling all the strings from behind) and that somebody wants to experiment with his own network attached to one that is already up and functional.

    "fungusbuttocks": [yahoo.com] I am another who is against Mike's use of the "Gnutella 2" name, because he simply did it as a marketing strategy (...)
    If Mike releases a nicely documented unambiguous protocol doc when he's finished testing the beta, and the protocol appears to have some sort of mathematical merit to it, we should support the protocol.

    Looks like the situation's less one-sided than the writeup makes it seem.

  • Nice to see that people are still out there chipping away at making Gnutella a smooth P2P network. It sure as hell needed some work. I haven't had to use Gnutella since KaZaA was released, but I would be interested in the opinions of anyone who has used Gnutella recently.
  • Some clarification (Score:5, Informative)

    by John W. Lindh (607757) on Sunday December 22, 2002 @07:47AM (#4940034) Homepage
    1) The developers opposing Gnutella2 seem to be the LimeWire developers (their client is open-source under the GPL see www.limewire.org), gtk-gnutella (GPL as well, see gtk-gnutella.sf.net) and BearShare (not open-source). So calling this a war between free and commercial is stupid, especially since Shareaza IS NOT open-source.

    2) LimeWire and the other opposed Gnutella2 for a variety of reasons. They didn't want a new message format where the old would still work, they preferred the GUESS search algorithm over the Gnutella2 search and they said they would not accept the name because if there ever was a Gnutella2 it should be announced by the whole GDF (Gnutella Developer Forum) and not by a single developer.

    3) After Shareaza developer Mike Stokes has shown an attitude towards the GDF that could very well be called hostile, things got a little out of hand. The GDF now demands that Mike hands the Gnutella2.com domain to the people running Gnutella.com. Mike won't do so and Raphael Manfraedi (gtk-gnutella) has even proposed to start blocking gnutella2 enabled clients.

    4) Shareaza fan's like the one who posted this news story helped a great deal to create the current situation by flaming on the GDF, posting rumors and lies (like Shareaza had 80k-100k users - even Mike Stokes denied that) on various news sites and in gnutella-centric forums.

    5) The Gnutella2 protocol is still an undocumented proprietary extension.
    • by imr (106517)
      about 4, he took it from the article:
      In terms of the network size, the highest Gnutella2 node capacity I've seen so far is about 200,000 nodes and growing. Node capacity is a bit different to the number of unique users, as the network has a high level of built-in redundancy. That would equate to somewhere between 60,000 and 100,000 actual unique users online.
      that's Shareaza talking here, so that would be Mike Stokes, isnt it?
      The article isnt very clear as it is, anyway.
    • > The Gnutella2 protocol is still an undocumented proprietary extension.

      how outrageous, that someone uses a word with GNU in its name, appends a 2 to it and make it non open-source. Try doing that to M$ stuff and see :)
    • After Shareaza developer Mike Stokes has shown an attitude towards the GDF that could very well be called hostile, things got a little out of hand.
      I believe this seriously mischaracterizes the events. I've been following the_gdf traffic on this topic since day one. Mike Stokes has never posted anything on the_gdf that wasn't reasoned, professional, and intended to foster cooperation in implementing what he sees as a better protocol. The attitude and hostility in the_gdf came entirely from elsewhere. He clearly took an approach to this that was not entirely in keeping with the_gdf tradition, but there has never been any "attitude" from Mike.

      As you mentioned in point 4), the same, unfortunately, can not be said for some of the Shareaza fans. Most of the problems started when members of the_gdf took postings by the fans as being representative of the opinions of Mike Stokes.
    • 1) The developers opposing Gnutella2 seem to be the LimeWire developers (their client is open-source under the GPL see www.limewire.org), gtk-gnutella (GPL as well, see gtk-gnutella.sf.net) and BearShare (not open-source). So calling this a war between free and commercial is stupid, especially since Shareaza IS NOT open-source.
      GTK cannot really be called a legitimate voice in the "war" because their client is very outdated. It doesn't even support Ultrapeers. Limewire and BearShare are "commercial". I didn't say closed source. Limewire has been thrown through the mud because it included "Stealware" (as seen previously here too) and BearShare for being the first Gnutella client to include spyware/ads. Gnucleus and Shareaza are both non-profit, and seem to care more about the user experience than the highly esteemed ego's of the GDF's "leaders". So that's why it was dubbed "War of Free vs. Commercial" by John Marshall.
      2) LimeWire and the other opposed Gnutella2 for a variety of reasons. They didn't want a new message format where the old would still work, they preferred the GUESS search algorithm over the Gnutella2 search and they said they would not accept the name because if there ever was a Gnutella2 it should be announced by the whole GDF (Gnutella Developer Forum) and not by a single developer.
      None of their opposition stemmed from legitiamte technical reasons. All of those complaints were debunked and addressed by Mike, and he commented on why the community should better choose his methods. And the GDF did not respond with technical reasonings, but rather emotional outburts. Mike didn't expect Gnutella2 to be mass implamented by all developers that quickly, he of course expected the members to debate on all aspects of it, and work towards a common goal off of the Gnutella2 proposal. But immidiately, the GDF saw it as a thret, and labeled it as such.
      3) After Shareaza developer Mike Stokes has shown an attitude towards the GDF that could very well be called hostile, things got a little out of hand. The GDF now demands that Mike hands the Gnutella2.com domain to the people running Gnutella.com. Mike won't do so and Raphael Manfraedi (gtk-gnutella) has even proposed to start blocking gnutella2 enabled clients.
      That's crap. Mike is clearly the only developer in the GDF who acts professional. I'm guessing what your calling hostile is his inability to cooperate with the "supposed" leaders of the GDF. And he has his reasons to, and has made them clear on countless occassions. Blocking Gnutella2 clients is silly. They're doing a disservice to their users if they block Shareaza, which works seamlessly with Gnutella1. It uploads to BearShare, Limewire and other clients. There is no technical/logical reason why Shareaza should be banned - just simply out of spite. What kind of message does it send banning Shareaza? "Never mess with Limewire or BearShare, or your BANNED!". Give me a break.
      4) Shareaza fan's like the one who posted this news story helped a great deal to create the current situation by flaming on the GDF, posting rumors and lies (like Shareaza had 80k-100k users - even Mike Stokes denied that) on various news sites and in gnutella-centric forums.
      Whenever there is a person supporting Shareaza, they're immidiately labeled as a rabid insane fool. The negative-ness put upon Shareaza fans just shows how blind the GDF is with jeliousy. Vinnie of BearShare (who is suppost to be a professional sane individual) complains of people on his forum saying his client sucks. SO WHAT? That's life. Don't you think Pepsi Co. gets letters saying their product sucks in comparison to Coca-cola? This is buisness! Jeez, grow up. My slashdot submission was very valid and non-bias, IMO.
      5) The Gnutella2 protocol is still an undocumented proprietary extension.
      It is not an extension. It's an entirly different protocol meant to replace the current Gnutella. It is documented, and specs are completed, but not yet fully avaliable. Mike tried to release specs to the GDF, but all he got in return were rabid demands like "SURRENDER" and "GO TO HELL" and what not. So he has suspended those submissions. Perhaps if the GDF were more level-headed, and emotionally stable, he could finish releasing the specs. It is my belief that he will release them before v1.7 final of SHareaza is released, but not directly to the GDF. The Gnutella2 protocol is meant to be open/free just as Gnutella is. It was meant to heal the problems with Gnutella, and bring it up to speed with other conventional/current P2P networks.
      • If the proposed protocol is entirely new, why unilaterally declare it the new Gnutella protocol by calling it Gnutella2? What if I propose something that is called Gnutella3 and is crap, from a technical point of view (I don't suggest Gnutella2 is)?

        If the protocol is really that good, why not give it an entirely new name and let it become popular by its technical merits?

        Just as the Limewire and Bearshare developers shouldn't be the ones to decide what becomes the next Gnutella, nobody else should do that on their own.
      • GTKG cannot really be called a legitimate voice in the "war" because their client is very outdated. It doesn't even support Ultrapeers.

        Your idea of legitimate needs revisiting. By the same line of reasonning, all the clients that do not support Mike's Protocol are very outdated and do not have the right to voice their opinion?

        Any developer wishing to start a Gnutella servent today will necessarily have a lot of work to do, and his servent will be outdated right from the start. You would also want this developer to shut up until he has at least as many features as the others?

        What Gnutella client are you developing yourself exactly? You're not even a Gnutella developer, so why are you voicing your opinion about things you obviously do not control, and perhaps don't even fully grasp?

  • Mike makes the point that due to the increase in the intensity of compettion between the different p2p clients and networks that each network will have a limit to how big they can grow.

    I have used Shareaza 1.7 beta and am very impressed with it's performance (it tells you when you are connected to a G1 hub or G2 hub so it's very easy to see the difference). I strongly believe that if Kazaa gets litigated out of existance that the best alternative for Kazaa users is Shareaza. Shareaza could very soon be king of the hill.
    • The problem is that you cannot see the merits of Mike's Protocol (MP for short) since some Shareaza Ultrapeers are talking to both MP and Gnutella.

      It's hard to know the exact level of bridging achieved between MP and Gnutella, but the fact that this bridging exists at all means that whatever wonders you can claim MP has are tainted by the fact MP uses Gnutella to bootstrap.

      That's why I'm advocating you stop this bridging and play with your MP-net for a while. Then we'll be able to compare the relative merits of the two networks.

      And no, simply disabling Gnutella support from your servent is not good enough. Someone else among MP-net may be doing the bridging even if you are not.
  • by wdebruij (239038) on Sunday December 22, 2002 @08:10AM (#4940067) Homepage
    What I don't understand is why everyone actually talks about `gnutella2' here. There are many different P2P protocols
    available. This guy named his gnutella2 and now we should believe it actually is the second version of the gnutella protocol?

    Shouldn't we have a discussion about what makes a good protocol before adopting it as a (pseudo)standard?

    In this context I'm afraid the commercial vendors might have a point.
  • Profit? Dream on. (Score:4, Informative)

    by Mulletproof (513805) on Sunday December 22, 2002 @08:21AM (#4940083) Homepage Journal
    Free vs. Commercial...?

    And when has commercial ever won out to free when it comes to file sharing, music specifically? Doesn't mean they're not welcome to try, but if history is any indication, somebody is on the wrong side of the profit-8ball (spyware and banner adverts not withstanding, of course...)

  • by Anonymous Coward
    Shazaa is so feature-rich and works so well because:

    ***The developer(s) don't wait for committee to implement backwards-compatible extensions to the protocols***

    A lot of features that have been proposed and discussed to death on the gdf and here, the Shazaa developer(s) has already made happen!

    Some such features:

    1. Compression of gnutella peer/ultrapeer/leaf traffic a la zlib. (my little cablemodem that used to be able to support up to 110 connections now supports up to 290 connections as ultrapeer with compressed streams.)

    2. Tigertree hashing - tigertree, as well as e-donkey2k, sha1 and md5 hashes (i believe) are all supported. Not sure if shazaa actually verifies each chunk against the tigertree, but it _should_.

    3. Ultrapeer "crawling" via udp queries.

    4. support for gnutella: , ed2k: , and magnet: urls, all of which are also integrated into the shell.

    The moral:

    **As long as you don't break the existing stuff,**

    If standards groups can't decide on standards, go ahead and add your own improvements. Let the improvements be accepted or rejected by the natural meritocratic movements of the open-source/open-standards worlds.

    Lots of neat-o technology has been wasted effort because standards groups couldn't agree on the standard before the technology was outdated. (e.g. ip/lan.e over atm).

    The makers of those other two *crippled*,*crapwar-burdened* semi-commercial clients are probably upset because their client doesn't have these nifty new features. And the best part: Shareaza hasn't (yet) broken the existing gnetwork.
    (I have no problem with commercial products. they're great. support is great. my problem is with "free"ware that bundles crap-ware for the uneducated masses.).

    Gnucleus and Shareaza have been the best gnutella peer apps on the windows side because they
    1) implement new features and new standards promptly,
    2) don't abuse the gnet, and
    3) don't carry any excess baggage.

    and for the moment, this long-time-gnucleus-faithful-user has gone shareaza untill gnucleus implements these new features. and shareaza is ooohhh-soo-stable.

    Again, the point:

    **As long as you don't break the existing stuff,**
    implement your own improvements to the specs. Let the improvements be accepted or rejected by the natural meritocratic movements of the open-source and open-standards worlds.

    Committees are good when they are productive. In this case, the gnutella standards have been outgrown in some respects by a relative new-comer with great ideas, fast fingers, and healthy respect for existing standards and infrastructure.
    • 1. Compression of gnutella peer/ultrapeer/leaf traffic a la zlib. (my little cablemodem that used to be able to support up to 110 connections now supports up to 290 connections as ultrapeer with compressed streams.) Proposed and implemented first by gtk-gnutella. However LimeWire is also using some form of compression. 2. Tigertree hashing - tigertree, as well as e-donkey2k, sha1 and md5 hashes (i believe) are all supported. Not sure if shazaa actually verifies each chunk against the tigertree, but it _should_. md4/md5 hashes won't be used by others because it creates a huge redundancy. If you have two files, one with a md4 has the other one with a sha-1 hash you can't make sure if they have the same content or not. As far as tigertree hashing is concerned, nobody ever said it wouldn't be implemented after it was proposed by Gordon Mohr. LimeWire has it on their to-do list for example. 3. Ultrapeer "crawling" via udp queries. Even that was decided to be used before Gnutella2 was released. My problem is, that Mike Stokes knew those features would be implemented but he didn't take part in the discussion, he kept his ideas for himself to be the first one implementing them. The GDF was productive (it produces the proposals more quickly then the GDF members are implementing them). Shareaza hasn't (yet) broken the existing gnetwork. The way you say it, it sounds like that is just a matter of time. - By the way, Shareaza is sending corrupt alternate locations, so it is breaking the network.
    • by ram4 (636018)
      ***The developer(s) don't wait for committee to implement backwards-compatible extensions to the protocols***

      That's just great. Yes, by doing things alone, just for yourself, you don't have to wait for others to agree. But then how do you ensure your ideas will be inter-operable with others? This approach can only work in a single-vendor world.

      1. Compression of gnutella peer/ultrapeer/leaf traffic a la zlib.

      You sound fairly ignorant about the current state of Gnutella. Compression is not something new, it has been implemented for almost six months by gtk-gnutella and Swapper (at least, forgive me if I forgot another vendor).

      3. Ultrapeer "crawling" via udp queries.

      This is exactly what LimeWire's GUESS proposal is about yet. But LimeWire, contrary to Shareaza, has discussed the matter openly before implementing it, taking the feedback of most developers.

      Gnucleus and Shareaza have been the best gnutella peer apps on the windows side because they 1) implement new features and new standards promptly,

      I could not say for Shareaza, since I don't use it, but I looked at the source code of Gnucleus and I can tell you there are many things that are not standard within Gnucleus. Yet Gnucleus currently behaves as a decent Gnutella client.

      To summarize, I think your post is more pro-Shareaza (blindly) than well-informed. I'm not sure you fully grasp what is at stake here.

    • 1. Compression of gnutella peer/ultrapeer/leaf traffic a la zlib. (my little cablemodem that used to be able to support up to 110 connections now supports up to 290 connections as ultrapeer with compressed streams.)
      Proposed and implemented first by gtk-gnutella. However LimeWire is also using some form of compression.

      2. Tigertree hashing - tigertree, as well as e-donkey2k, sha1 and md5 hashes (i believe) are all supported. Not sure if shazaa actually verifies each chunk against the tigertree, but it _should_.
      md4/md5 hashes won't be used by others because it creates a huge redundancy. If you have two files, one with a md4 has the other one with a sha-1 hash you can't make sure if they have the same content or not. As far as tigertree hashing is concerned, nobody ever said it wouldn't be implemented after it was proposed by Gordon Mohr. LimeWire has it on their to-do list for example.

      3. Ultrapeer "crawling" via udp queries. Even that was decided to be used before Gnutella2 was released.

      My problem is, that Mike Stokes knew those features would be implemented but he didn't take part in the discussion, he kept his ideas for himself to be the first one implementing them. The GDF was productive (it produces the proposals more quickly then the GDF members are implementing them).

      Shareaza hasn't (yet) broken the existing gnetwork.
      The way you say it, it sounds like that is just a matter of time. - By the way, Shareaza is sending corrupt alternate locations, so it is breaking the network.
  • How many networks? (Score:3, Informative)

    by NineNine (235196) on Sunday December 22, 2002 @08:50AM (#4940128)
    What I'm trying to understand is why does everybody and their brother build a brand new P2P network (or try to)? What's the point? There's exactly $0 to be made off of it. More network just mean that each one is not as good, since files are spread out across multiple networks, not just one. Kazaalite works just fine. Why switch? This is another geek vs. businessman thing where a bunch of geeks are creating things for no apparent reason whatsoever other than the fact they may think it's "cool".
    • 1) Honor. It's good to be know as the guy or the team that created the best technology.

      2) Employment. Imagine coming to a company and when someone asks you what you did you just say "Got MP3? Thanks to me."

      3) Sell out. There ARE commercial uses for a P2P application. Just think about finding a document, with metadata, divided by category and by hash (so it wasn't modified). And finding it in seconds...
    • Napster crashed into bankruptcy and burned in RIAA hell, yet despite that Sean Fanning got to keep his pay and the money he got from selling off his company before the crash.

      Free P2P sharing of RIAA music as a business is never going to be a legal business model. However, if you base yourself in a place where it's hard for the USA laws to get to you, and you're willing to be shameless with what spyware you inflict on your users, there is quite a bit of money available in the short term.
    • by Snaller (147050) on Sunday December 22, 2002 @11:26AM (#4940468) Journal
      Kazaalite works just fine. Why switch?

      Yeah, the authorities love it too, its soo easy to find your IP number and slap a fine on you.
    • by Yi Ding (635572)

      What I'm trying to understand is why does everybody and their brother build a brand new P2P network (or try to)? This is another geek vs. businessman thing where a bunch of geeks are creating things for no apparent reason whatsoever other than the fact they may think it's "cool".

      What Gnutella is trying to do is make an open, operable standard which then can be implemented by many clients, thus simultaneously making all of those clients better, and therefore making so that even though there are many clients, the network they are on is good.

      In a sense P2P networks are similar to encryption methods. New compression methods will constantly be discovered. The reason why people choose to use them even though it will break compatiblity with old clients is that they're better (smaller, faster, etc...). P2P is at the same stage compression methods were 10 years ago. Some people come up with new P2P networks as an academic exercise. Others implement them because they are better than what they have. Your argument may make sense 10 years from now when the P2P process is nearly optimized (similar to compression now), and then, the more people we have sharing one network, the better. However, when there is signifcant room for improvement, it is always better to implement the new standards instead of trying to maintain compatibility with the old ones.

  • by Kunta Kinte (323399) on Sunday December 22, 2002 @09:06AM (#4940158) Journal

    I'm not a Shareaza fan, But I think Mike is within his rights to call his protocol Gnutella2 if he wishes.

    I've been following this thing for a while now and this is my view. Gnutella was made by a group of developers at nullsoft, right? They never trademarked the name and eventually abandoned the technology all together, I believe.

    GDF is an ad hoc group put together to continue the development, but have no special rights concerning gnutella.

    Love him or Hate him, I think Mike is perfectly in his rights to call his protocol Gnutella2. It's not a very nice thing to do, but he is within his rights.

    The GDF should accept this, realize that at any time someone can create a 'Gnutellan' and all the GDF need to do is that when describing their protocols, specify the version that they created and/or endorse. eg 0.6, etc.

    • GDF is an ad hoc group put together to continue the development, but have no special rights concerning gnutella.

      Love him or Hate him, I think Mike is perfectly in his rights to call his protocol Gnutella2. It's not a very nice thing to do, but he is within his rights.

      If by rights you mean legal rights, then you're probably right.

      However, just because something is OK legally does not make it necessarily right morally and ethically. There isn't a law for every situation you will face in your life, nonetheless you will have to take some decisions. And those decisions are based on your ethics.

      By using the Gnutella 2 moniker, Mike probably initially made a mistake (although I think it was a carefully thought out move, but I have no proof), and he should have rectified immediately when the GDF reacted.

      The fact that he did not says a lot about his values.

      • However, just because something is OK legally does not make it necessarily right morally and ethically

        I agree entirely. not a nice thing to do was an intentional understatement. But legal rules are all we can force people to abid by.

        Mike has clearly shown that he has trouble working with others. Why make this more drawn out and debated than it needs to be?

        Here's a simple solution...
        (i) GDF core members get together and choose a name for their protocol. Basically rename the GDF version of gnutella.
        (ii) They trademark that name. It's not that expensive. I'd suggest GDF register as a non-profit whilst they're at it.
        (iii) Legal counsel for GDF commercial members sue the pants of Mike or any other person that tries to bastardise this new protocol.

        This, I believe is the proper way to fix this mess, instead of 'duking it out' on message boards and mailing lists.

        PS. thanks for the work on gtk-gnutella.

  • IANAD (developer) (Score:2, Insightful)

    by br00tus (528477)
    I am developing [sf.net] a Gnutella client currently, and have been following this on the GDF, although not that closely

    First of all, if this is "open" (free) versus "closed" (commercial), WHERE is the Gnutella 2 specification? "It's coming". I mean that's one of the oldest notions in the free software community, it is NOT open source (or protocol) until the source (or protocol) is actually open! When (if) the specs come out, I'll believe it.

    Currently the Gnutella "1" (aka v0.6) specs are published, and functioning in many clients, and the Gnutella 2 protocol is not to be found anywhere. It's true that Shareaza does not (yet) have ads and Limewire and Bearshare do, but Shareaza's source is closed, unlike Limewire's. Calling Shareaza open and free and Limewire closed and commercial is kind of silly, especially since Shareaza source is closed and Shareaza G2 protocol is (currently) closed.

    Second of all, Gnutella is a coalition of the most popular Gnutella developers - Limewire, Bearshare, Gnucleus, Xolox, GTK-Gnutella, Morpheus (sort of) and so forth. Currently, they call the Gnutella version they have version 0.6. Along comes a new client Shareaza, and they try to hijack the Gnutella name and call it "Gnutella 2".

    I hope Mike comes to his senses. Shareaza is a decent p2p client, and has been a positive thing for Gnutella, and he can do what he wants, but I am uninterested in any new protocol until protocol specifications are published, and trying to seize the Gnutella name is kind of silly as well, especially since the protocol he wrote (and has yet to share specifications of) is so radically different than Gnutella. He can switch to his new protocol if he wants, but he should stop calling it an open protocol until he publishes the specifications, and he should consider a name aside from "Gnutella 2".

    • Along comes a new client Shareaza, and they try to hijack the Gnutella name and call it "Gnutella 2".

      I really fail to see what the big deal is. As far as I'm concerned, the only people with any right to complain are those who wrote the 0.4 protocol. The GDF situation is no different than Nutella shutting down gnutella.de and newtella.de.

      'Hijacked', as the opposition likes to put it, can easily be read as 'inspired by'.

      I hope Mike comes to his senses. Shareaza is a decent p2p client, and has been a positive thing for Gnutella

      I agree completely. I will not promote G2 or add support to my client until I have an open specification in my hand, though I am quite eager. I'll be emailing Mike to conditionally pledge my support after I post this message, I advise everyone else to do so (while remembering to be courteous!) as well so that we can perhaps end this debacle on a positive, friendly, and mutually beneficial note.

      Niko

  • by Moritz Moeller - Her (3704) <mmh.gmx@net> on Sunday December 22, 2002 @10:24AM (#4940319)
    GNU Internet File Transfer

    http://gift.sourceforge.net/
    " What is giFT, you ask? giFT is a modular daemon capable of abstracting the communication between the end user and specific filesharing protocols (peer-to-peer or otherwise). The giFT project differs from many other similar projects in that it is a distribution of a standalone (platform-independent) daemon, a library for client/frontend development, and our own homegrown network OpenFT. "

    This is a great network, where you find many oggs, downloads actually work (up to 600kb/s!!) and finding files is really fast. Lots of altruists are using it. Plus: You have to compile it from CVS, which prevents idiots from using the network. On the average each user shares 8 GB!
    • While it may be great in theory to say that having to do compile keeps the idiots off of the network, the strength of p2p lies in numbers. As I'm typing this, KazaaLite is showing me 3.5 million users and over 650 million files on the network. You just can't get numbers like that unless it's easy to install and use.
      • yah, but a network with fewer higher-quality nodes are often more useful than many crappy nodes. For lots of reasons.

        If you break down the 650 million files, mostly these are the same ones - there is huge redundancy in like the top 1000 files. This is your benefit in exchange for the network being loaded with 3.5 million users.

        Example of small high-quality network: back with napster, the smaller openNap networks had a MUCH better signal-to-noise ratio, and as a result were much more useful. The reason was that it was fairly obscure, users were anal-retentive about having "correct" filenames, and tended to have fat connections. P2P heaven :-)
      • giFT just has not yet had a stable release, even though it works very good already. In the long run, I expect it to be able to support the 10s of thousands of users kazaa or gnutella have. In fact just try it, you'd be surprised.

        Also, from the fact that you use kazaa, I presume, that you use windows, which is not an option for me. In fact it makes you look like a fool on /..
    • Plus: You have to compile it from CVS, which prevents idiots from using the network.

      # make install
      ===> giFT-0.9.7 'Does not currently work'.

      Apparently, it prevents more than just idiots.

  • by n1k0 (553546) on Sunday December 22, 2002 @10:26AM (#4940327)
    The anti-G2 lot isn't just lacking reason, they're contradicting it. They would cut off their nose to spite their face (or Mike Stokes). This is the mentality we're dealing with:

    "...as long as gnutella2.com exists in its current form, you are all fifty cent whores that got pimped by Slimy Mike."

    Childish and spiteful. This is not the attitude with which to reach compromise and solve problems. Name calling? Are we freakin' twelve?

    In the end, the G2 opponents are being obstinately selfish, and the heated, illogical emotion they're injecting into this discussion is hurting far more than helping the community by delaying the widespread implementation and deployment of a superior protocol that has already been tested and used effectively in public hands.

    I don't care what anyone personally thinks of Mike, his opinions, or how he's handled the unleashing of Shareaza on the world. The only working reason in this argument is the Gnutella2 mark, and the G2 opponents seem to be desperately clinging to it, as if this is the only way they can oppose G2 without showing their true feelings, which I suspect have to do either with personal conflicts such as Shareaza stepping on the feet of other Gnutella players. Surely its a sign that the Loud Voices complain that G2 will only cause a schism, but then turn around and talk about Gnutella3 as a way of battling G2. Would I be wrong to construe this as an indicator of where certain peoples' intentions really lie?

    The G2 opponents are so busy thinking of creative ways to sanction Shareaza and Mike Stokes that they've not given a single thought to what course of action would most benefit the community of users in this situation. As a Gnutella user and developer, I say this spurious, wasteful behaviour must stop now before it gets any worse. I suggest they regain their emotional composure and grow the hell up.

    For me, this is an awkward situation. I'm a proponent of all things open and Free, and I should support the GDF in this conflict else violate my principles. But my principles also encompass the proper behavior of a rationally thinking human being who wants to successfully communicate with others of his kind. In this regard, I find the behaviour of some GDF key players to be so repulsive that I like myself less when I support a protocol backed by propaganda-spewing, egotistical drama queens than when I support a protocol that's only quasi-open, or less, as is currently the case. (I _really_ hope Mike opens up G2. It would be such a graceful way of pulling the rug out from under these fools, and beyond that I don't know how long I can endure taking sides here without having an ethical melt-down.)

    Niko

    PS

    Replace the phrase 'G2 opponents' with the name 'Vinnie', who seems to think calling people 'slimey' and 'imbicile' demonstrate one's argumentative superiority. His contradictions of logic, hypocrisy, personal attacks on the character of those making neutral observations or expressing neutral opinions, all serve as wonderful examples of how not to effectively influence people or raise support for a cause. He's like a politician who's election campaign consists of insulting the mothers of his constituents.

    > All of this is a moot point. Mike wont change the name, no one here
    > can make him. Get over it. This is so rediculously unimportant in
    > the scheme of things I hate to see so much list traffic dedicated to
    > it.

    Amen.
    • ...and it makes me sick beyond words that the GDF would actually advocate blocking or lower the service priority of Shareaza clients. Are you taking lessons from the Evil Empire or what? How many of you hypocrites would be crying louder than a fog horn if Microsoft decided to, say, rig IIS to refuse to service non-I.E. browsers?

      ...simply disgusting...

      Niko

      • .and it makes me sick beyond words that the GDF would actually advocate blocking or lower the service priority of Shareaza clients.

        You're wrong here. Noone is suggesting to block Shareaza in particular. That would be against my own ethics.

        Instead, I'm advocating for the ombilical between the Gnutella network and Mike's network to be cut. That is, I don't want to discriminate against a particular servent but against a feature that is not necessarily welcome, i.e. ones that support Mike's Protocol.

        It's only a proposal for now, and the fact that Shareaza is the only servent implementing Mike's Protocol is due to the closed nature of that protocol.

        The reason I'm in favor of drawing the line between the two is because of its closed nature, there is no telling how Mike's Protocol will evolve. It could end-up messing up with the Gnutella network carelessly.

        • Noone is suggesting to block Shareaza in particular. That would be against my own ethics.

          I saw two different people advocating this as I looked over the GDF archives for the last few days.

          It's only a proposal for now...

          That it is even considered is more than enough.

          Instead, I'm advocating for the ombilical between the Gnutella network and Mike's network to be cut. That is, I don't want to discriminate against a particular servent but against a feature that is not necessarily welcome, i.e. ones that support Mike's Protocol.

          You want to disallow support for the 'feature' that is the G2 protocol in your client? Am I stupid, or is that achieved simply by not adding G2 support to your client? Or do you mean the feature of relaying hits from the G2 network to the G1 network? Aside from G2 providing that as an option to G1 clients, I don't see how its possible without ignoring all G2 clients, which you ethically oppose. :-) Am I just missing something?

          Unconditionally blocking Shareaza will also limit (perhaps severely, depending on the G2 nets) the amount of data your client can see. You're making a decision for your users that I'm not sure is yours to make. Perhaps your users don't believe G2 is evil or that Mike is our pimp? I know I wouldn't use any client that involuntarily imposed such restrictions, or would hack them away first thing.

          Niko

          • Am I stupid, or is that achieved simply by not adding G2 support to your client? No, what he meant is to disallow bridging from G2 clients to G1 clients. In other words: force G2 to use only G2 clients, not both G1 and G2. You're making a decision for your users that I'm not sure is yours to make. RAM develops his client as he sees fit. If you don't like the changes he makes, use another one. If you like the changes, great!
            • RAM develops his client as he sees fit. If you don't like the changes he makes, use another one. If you like the changes, great!

              If I used his software regularly and such an unfortunate decision were made, I'd remove or rewrite the offending code.

              Niko

    • I'm a proponent of all things open and Free, and I should support the GDF in this conflict else violate my principles.

      Until Mike truely releases complete documentation on the protocol, in a manner that permits anyone to implement it, I can't see how supporting "gnutella2" is aligned with your principles.

      • I can't see how...

        Perhaps you didn't read the entire paragraph. If you need to be spoon-fed the point, principles of human decency are more important than principles of licensing. For all you or I know, we'd have a Free specification in our hands if certain people were able to be civil.

        Niko

  • "John Marshall has made his decision. Now let's see him enforce it."
  • by schlach (228441)

    Gnutella Developers Front?

    =p
  • by havaloc (50551) on Sunday December 22, 2002 @05:02PM (#4941695) Homepage
    Mike has said on numerous occasions that he will release the G2 protocol when it is finished. Since the protocol is still in beta and being changed, it's still not finshed. When he works out the bugs and makes it as good as it can be, it will be released.
  • The more pretentious a corporate name, the smaller the organization. (For
    instance, The Murphy Center for Codification of Human and Organizational Law,
    contrasted to IBM, GM, AT&T ...)

    - this post brought to you by the Automated Last Post Generator...

Power corrupts. And atomic power corrupts atomically.

Working...