Slashdot Mirror
Sun Launches JXTA
Daniel Rall writes: "Project JXTA, originally a research project spearheaded by Bill Joy and Mike Clary, is now an open source effort, with a vision to
enable developers to create innovative services and applications.
It addresses the need for an open, generalized protocol that
interoperates with any peer on the network including PCs, servers
and other connected devices." Seems to be a combination of Gnutella and Beowulf clustering - lots of potential.
I find your post deeply ironic. You seem to be implying that Bill Joy dislikes open source. Perhaps you don't know much about Bill Joy, so I'll inform. Bill Joy was the principle designer of BSD Unix while a grad student at Berkeley. Berkeley Unix was pretty much *the original* open source project. And it was very influential as well...quite a bit of what we know as "unix" and even "the internet" was defined by BSD. And yes: today's FreeBSD, NetBSD, and OpenBSD projects all descend from this code base and tradition. You could say that Bill Joy is the granddaddy of the BSD world.
Later, Bill Joy left Berkeley to help start Sun. And yes, he took BSD with him and Sun commercialized it into SunOS. Years later it would morph into the SysV type hybrid called Solaris that we still have, but in the beginning Sun basically sold BSD boxes.
No doubt now that he is an executive in a closed source company, he views open source OS's as a bit of a threat. But historically, this man has contributed a great deal to the open source cause. Infact, he's probably up there with RMS in terms of code contribution. This should not be ignored.
--Lenny
Maybe you can still redeem one of the biggest sins ever done in Comp Sci...
One of the people involved with JXTA is Gene Kan, who was involved with Gnutella from very early on. Sun acquired Infrasearch about a month or so ago, which was Kan's project to develop a search engine based on Gnutella concepts. Naturally, Sun saw something of interest.
You, too, can write Silicon Valley press releases!
"Beware by whom you are called sane."
Potato chips are a by-yourself food.
Any reason you needed to steal that article, rather than just point a link to the original, and perhaps a soundbite from it so that we'd be interested enough to click the link?
--
--
Don't like it? Respond with words, not karma.
How the hell do you pronounce JXTA?
...but I'm going to be using Microsoft JXTA instead, because management tells me it's going to be better supported and more reliable, and it will contain all sorts of neat extensions which will link it up with MS Office and Exchange.
-jon
Remember Amalek.
Not tying JXTA to Java is smart, as is Open Sourcing it. It could become a nice end-run around .NET-specific services (client-server is just a degenerate case of P2P, and on the server side, JXTA could be used to load-balance and store data).
Fun stuff is happening.
-jon
Remember Amalek.
There's really no point in buying a sun box if your job doesn't saturate a freenix running on a quad Xeon.
One very real possibility is if you believe that your job could grow past the ability of a quad Xeon processor. If that's the case, then from the Enterprise computing perspective you're better off starting with (and sticking to) the Solaris/Sparc architecture.
But I'm not selling any Sun hardware here -- my original post was probably too harsh. Eventually (probably over the next few years) Linux will make significant progress in the Enterprise environment... then Sun will have a unique decision to make....
--Mid
There is this crazy concept in computer programming called Major releases and Minor releases. A Major release is when you change a Major thing (like the # of processors you can support), whereas a Minor release is when you change a minor thing.
I can't explain it better than that without doing much research. That said, can you concretely tell me that you've seen a 2.4 kernel running on more than 4 processors? No fair lying when you're posting as an AC....
--Mid
Not to pick a fight, but how is this relevent to the JXTA announcement? It sounds like you're moaning just to hear your own voice.
But since you changed the topic, make sure you understand what that Sun employee is talking about -- an "Enterprise-Class" machine is an extremely scaleable SMP box, that can support a huge amount of memory, and has I/O bandwidth out the wazzoo (technical term, of course).
The theoretical limit in the 2.0 Linux kernel (from what I remember) is 16 processors, although I've never seen anyone get more than 4 running at once. Contrast that with a Solaris/Sparc box which, currently, you can buy a 64-way box, and by year's end you'll be able to buy a 128-way box. Oh yeah, and the theoretical limit is over 1024 processors.
As far as I/O bandwidth goes, you're not really going to get all that much out of an x86 box with a single PCI bus. Since IBM is now on the Linux bandwagon, we might start seeing them build high I/O boxen sometime soon (hopefully running on PowerPC, a chip which can scale in MP terms almost as well as the Sparc III). But, that's still in the future.
Finally, you should get some kind of an award for your insinuation that Linux was responsible for Sun's sales to drop 73 percent.... I mean, I love Linux as well, but that's stretching things quite a bit!
My $0.02,
--Mid
Webster defines standard as:
Main Entry: 1standard
Pronunciation: 'stan-d&rd
Function: noun
Etymology: Middle English, from Old French estandard rallying point, standard, of Germanic origin; akin to Old English standan to stand and to Old English ord point -- more at ODD
Date: 12th century
1 : a conspicuous object (as a banner) formerly carried at the top of a pole and used to mark a rallying point especially in battle or to serve as an emblem
2 a : a long narrow tapering flag that is personal to an individual or corporation and bears heraldic devices b : the personal flag of the head of a state or of a member of a royal family c : an organization flag carried by a mounted or motorized military unit d : BANNER
3 : something established by authority, custom, or general consent as a model or example : CRITERION
4 : something set up and established by authority as a rule for the measure of quantity, weight, extent, value, or quality
5 a : the fineness and legally fixed weight of the metal used in coins b : the basis of value in a monetary system
6 : a structure built for or serving as a base or support
7 a : a shrub or herb grown with an erect main stem so that it forms or resembles a tree b : a fruit tree grafted on a stock that does not induce dwarfing
8 a : the large odd upper petal of a papilionaceous flower (as the pea) b : one of the three inner usually erect and incurved petals of an iris
9 : a musical composition (as a song) that has become a part of the standard repertoire
SUN's Jxta set of standards is 'open (speech and beer), assuming they know what they are doing they have some authority and most importantly it is explicitly intended to be picked up by the community. Looks like a standard to me. Besides C++ was only standardized very recently and nobody has sofar bothered to implement that standard in full. The value of a standard by your definition is dubious to say the least. I think you are just pissed off because SUN gets to set the standards.
Jilles
It's obvious you have never used vi.
My point is that he doesn't HAVE to code anything. He did an awsome thing (something you can't claim) and has proven himself both as a coder and as a manager (something you can't claim). Any two bit asshole idiot like yourself can post on slashdot saying "what has a done recently". I tell you what he has done. He got off his butt and made a difference in the lives of thousands of people. He is till working to shape the future in his image, he is still working his ass off trying to come up with better ideas for how computing should be. He doesn't have to he is rich as hell. He does not have to work another day in his life. He could lie on some private island and have nubile young women bring him drinks. But he works anyway.
What the hell did you do today (besides whine on slashdot that is)?
War is necrophilia.
[It] is now an open source effort, with a vision to enable developers to create innovative services and applications
Great! As a programmer, I've always longed for something like that. No matter how hard I or my colleagues have tried, we have not been able to make either services or applications that were at all innovative. I'm glad that these guys are finally doing something about it. No more thinking about these program things that I could write! Now, programmers everywhere will be able to turn the designs that they've had in their heads for the last 60 years and turn them into real programs! I've had this idea for a while about a program that displayes remotely stored pages of what I call 'hypertext'. Maybe I should patent it...
Honestly, these guys are the best people to enable me to proactively leverage my synergies and grow my productivity so that the whole is more than the sum of the parts!
--
SecretAsianMan (54.5% Slashdot pure)
Washington, DC: It's like Hollywood for ugly people.
...we're gonna have problems. Let's hope one of the directions for the project is to add some self-organizing behavior so that we can set up a hierarchy dynamically. That would eliminate the problem of search propagation. I realize it's a difficult problem, but it would be very cool. Imagine an OpenNap or Gnutella network that was like a Hydra - you shut down one directory node, and the network re-organizes to anoint another one. Just try to shut down that kind of network! Too bad we can't do it like SMB browser elections - it's too easy to hijack the process. Besides, the packet waste with a browser election (or just the NMB protocal in general) is phenomenal.
Right...
And Joy, come on, he's one of the reasons you have *BSD.
Without an application to see SOME functionality, even if it's just an IRC clone, I don't get it.
--Mike--
I just picked up working with RPC in C after several years working with CORBA, RMI, and other more recent protocols, and I have to agree completely. RPC really is a very nice and simple distributed invocation protocol. See my sig to check out how I am using it.
In case you are interested, I've also written an XML stream for XDR (the data translation layer of RPC) so that structures can be deserialized and reserialized easily to XML, but without giving up XML's loose coupling. I wrote up an article on the idea here.
LibBT: BitTorrent for C - small - fast - clean (Now Versio
Looks like there is a trend of inventing half-baked complex "protocol based" distributed systems, ever struggling to find a decent basic abstraction for generalized communication between application processes in different machines: proprietary RPC mechanisms, myriad in-compatible application spesific XML languages, JXTA "pipes", Linda/Jini tuples, CORBA "objects" and finally a hodgepodge of classical text line based TCP/IP application protocols with a mix of transport and presentation layer issues unsolved.
Plan 9 "only" has distibuted files. With dynamic, on-the fly synthesized algorithmic response. With authentication, security and distributed user management with group rights. With naming and access categorization bits.With unified interface for intra-node and inter-node tarffic. With straihtforward binding to classical channel stuctures for multiprogramming intra-process messaging and synchronization with a flexible threading concept. With compatibility with all classical local static data file access code and algorithms. With distributed inheritance and with stackable half-transparent union directories.
Download Rob Pike's and friends' Plan 9 and Inferno from Vita Nuova pages and onwards to Bell Labs pages.
Anssi Porttikivi / app@iki.fi
There is more to JXTA than just p2p file transfer. I think the really interesting application of this technology is as a software agent space.
The JXTA services layer includes features that allow the agent to propergate it's self, it includes services allowing searching and indexing.
I think, though, P2P and distributed networking is trying to realize the dream of "The network is the computer" which I think is a quote attributed to Scott McNealy, if I'm not mistaken. hmm....
So to reiterate, JXTA is just a protocol, it is not a Java package.
So no, it cannot be serialized anymore than HTTP could be.
"Mind, as manifested by the capacity to make choices, is to some extent present in every electron." -Freeman Dyson
However, Java 1.4 (Merlin) is nearing completion and they have a new I/O API coming designed specifically for scalibility.. I guess we will have to wait and see.
"Mind, as manifested by the capacity to make choices, is to some extent present in every electron." -Freeman Dyson
JXTA is not a Java package, it is a language independant protocol spec. Most of the work done so far is on the lowest Core layer, which involves services such as peer discovery and grouping. They have written an early IMPLEMENTATION of this in Java.. and why not? Java was built for platform independance and networking.
But anyone can write an implementation in any language they want. It uses XML for communication.
My company has recently become involved in JXTA as we are developing a pure Java P2P file sharing app called File Rogue. (I am lead coder)
"Mind, as manifested by the capacity to make choices, is to some extent present in every electron." -Freeman Dyson
That's a good point. What I meant was that Gnutella and Napster are high-level application protocols that are designed to do something specific (share files, in this case). JXTA is more low-level; it's totally general and doesn't actually do anything.
Honestly, I don't think it's all that similar to Beowulf clustering (which tends to be focused on high performance or high availability, while JXTA trades off performance for portability/implementability) and the only similarity to Gnutella is that it's peer-to-peer.
The most obvious difference between JXTA and the popular P2P systems (like Napster, Gnutella, Freenet, Mojo Nation, Jabber, etc.) is that JXTA isn't an application; it's a toolkit for building P2P apps. Unfortunately, that means in the short term it doesn't really do anything. But it appears (from a very cursory inspection of the docs) to handle mundane details like finding other peers, sending messages (over a variety of protocols apparently including Bluetooth, not just TCP), relaying through NATs and firewalls, etc.
If you want to play with JXTA, the Getting Started PDF has a tutorial for using the shell.
BTW, here's a convenient Mac OS X package of the JXTA Shell since Sun didn't build one.
No, it's like defining TCP without defining IP
first. There's nothing under it. So it's
impossible to implement it. The only way
to interoperate is to reverse engineer the
code. And Sun is holding out, I know, because
Applied Metacomputing is interoperating C JXTA
with the Java JXTA, so there has to be more doc
internally.
-I like my women like I like my tea: green-
Hey Sun! Get lost!
Ok, I know this ones a troll but the sentiment of other comments doesn't sound good. What's the problem? So Sun is a Mega-Corp. But the're giving us stuff! They did the whole Open Office thing. And now they give us JXTA which is not some worthless bone thrown out of Bill Joy's garbage bin but something that could potentially make Gnutella, Napster, and Freenet look like tin cans with some string. They've got specs. Thats the kind of organization OSS projects need. When was the last time you saw an OSS project with a spec. Never. Invariably, you just start typing and half the time it turns out to be crap. Whatever the truth is, you should look at the software before posting fooling comments. I suspect the normal people are off looking at the white-paper.
The code for all of these is clean? Mozilla works well? You wanna tell me where the documentation for Samba is please? Guess which company John Ousterhout worked for while doing some of his most significant work on the Tcl language? You guessed it, Sun Microsystems.
It's released under the apache (read: bsd style) license.
pathetic trolls bother me.
Mooniacs for iOS and Android
Did you even read the info???
The JXTA protocol(s) etc are going to be STANDARDS that can be worked into ANY application, regardless of programing language.
Furthermore, I believe, that all comunication (Between Peers, etc) is using XML anyway.
This project is AWESOME!!!!! Finally, a standard for Peer-to-Peer! yay!!!
I am become Troll, destroyer of threads
No Joy from P2P vets for Sun's Jxta
By: Andrew Orlowski in San Francisco
Posted: 26/04/2001 at 00:52 GMT
Sun wheeled out its Mount Rushmore of cerebral greats - Gage, Joy, Gosling - to herald the unveiling of its Jxta peer-to-peer project today.
Announced by Bill Joy at the O'Reilly P2P conference in February, Jxta (pronounced "Juxta") is now live and we're awash with positioning papers, technical documentation and real downloadable code. But the instant reaction from the peer-to-peer community - who've been at this for a little while longer - was cool.
"It's no good for FreeNet, next to no use for MojoNation or Gnutella, and no good for SETI at home," FreeNet developer Adam Langely told us. "It is buzzword compliant, though."
And Jxta's reliance on XML brought an "Oh my god," from the developer - a contributor to the excellent O'Reilly P2P book, Disruptive Technologies - who's juggling a rewrite of the FreeNet core in C++ whilst studying for his GCSEs.
It's not as if the guest of honour has marched in to the P2P party, wolfed down the free booze and fondled the hostess. Almost, but not quite.
This party doesn't really need a guest of honour it seems, even if it is Sun itself in best-behaviour mode. Bill Joy modestly described Jxta as a project that attempts to define protocols, that's all. Within a year he told us today, we might have enough usable protocols to embed in some real devices.
But watching these billionaire new frontiersmen earnestly describe the problems that P2P networks need to overcome, after we've watched 18 months of very public sweat and anguish from the Gnutella, FreeNet et al networks as they tackle these problems, strikes as the definition of redundancy.
"These networks develop in vertical silos, and they don't interoperate," said Gage in his introduction today. Which is true: "The P2P projects have nothing in common except TCP/IP", agrees Langely. But far from being their weakness, it's really their strength. Gnutella began life as a brute force, quick-and-dirty mechanism for file sharing, and FreeNet as a long term project to build a secure space free from surveillance. To adopt Sun's Jxta plumbing would not only entail throwing away these hard-won lessons, but it would compromise what each network was created to do. For example, FreeNet is inundated with offers of help to turn it into a platform for instant messaging, a global anonymous email gateway, or the new Napster. Take your pick. But as FreeNet luminary Brandon Wiley unfailingly points out - FreeNet is uniquely useful for dissidents in China (it was inspired by Ross Anderson's Eternity service meme) - so please don't mess it up.
The road to hell is paved with good intentions, and indeed, well-though-out but pointlessly blue-sky RFCs, and Sun's error is really in mistaking social spaces for technical problems. This conundrum was best illustrated at the O'Reilly conference when a panel moderator (forgive us, we can't remember which one, and we paraphrase liberally here) asked: "Is there a P2P? Is there a P2P business model? Or will it be like client/server? Will we be sitting around at a client/server conference in a year's time?"
So Sun's Jxta is a technology project looking for social uses, and the P2P networks are social projects looking for technology solutions, and the two seem to be passing each other like the proverbial ships in the night.
But let's get some perspective: it's a benign adventure, and doesn't deserve the rancour that say, a Microsoft P2P 'solution' - let your imaginations run riot here, folks - would attract. We've seen so many such pogroms in the past (Pen Windows, anyone?) that trample over not only optimistic start-ups, but entire business models, and with Jxta being the hesitant Apache-licensed venture that is, comparisons don't stand up to scrutiny.
As if P2P had never happened
We'll go into the technical details when we've had time to digest them (comments welcome), but Jxta's a layered set of protocols tackling not just interoperability but monitoring and performance too.
If you were starting from scratch, then Jxta would be an obvious place to go. The monitoring stuff is nice, as plenty of fringe edge networking gets proscribed by vigilant BOFHs, fearful of congestion at network and disk choke-points. And not just BOFHs, either - any local ISP worth its salt should by now have recognised that P2P is a loyalty/community trump card, too.
Interestingly, Joy is thinking small with Jxta. It could be, he suggested, a way of steering users between the mess of access networks that we'll be faced with pretty soon - between 2.5G GPRS/EDGE packet data, 802.11 networks, and our local LAN or dial-up connections. "Devices are too small to carry ten protocol stacks," said Joy optimistically, without quite convincing us that a Jxta-enabled device would solve the problem. But give the man credit, he's looking for an answer to a problem most folk haven't even recognised yet. Unfortunately, the conversation took a turn into the utterly surreal, as Joy began to explain how embedded IP devices in schoolkids' sneakers could cause havoc for teachers, and how Jxta-enabled sneakers would solve the problem, because of their device recovery and monitoring characteristics. Sensibly, and abruptly, Gage drove the conversation back on to dry land before anyone had time to notice.
Let's kill the geeks
But if the distress in the people's P2P community wasn't enough, the opprobrium unleashed on the P2P meme by a lordly tech press is nothing short of astonishing.
"Bill Joy is catching the tail end of a euphoria that never came into existence," declares the New York Times, grandly.
The CNet/ZDNet conglomerate has outsourced its opinion to Gartner Group analysts, who opine:
"Sun was careful to avoid the term P2P, not wanting to be associated with a technology that appears to be going out of fashion." A fashion created by ... analysts such as Daryl Plummer and David Smith as recently as last
August, we seem to remember.
Ouch! Since when were the NYT and CNet such pernickety style mongers, we wonder?
Ever since they had the P2P meme foisted upon them, we suspect, and there's more than a hint of snobbishness at attempting to bury an idea that the geek press had the temerity to name before they did. O'Reilly might not have named P2P - we're not absolutely sure who did, and we really couldn't care less - but the meme left the industry elite gasping for air, and without an industry elite to follow, the industry-led tech press was left experiencing a kind of zero-gravity for the first time. The Fourth Estate marches to a well regulated beat. OK, we'll give you 'Open Source' as a rebranding excercise, you can hear them think, but P2P, that's just too much weird shit...
P2P networks, or whatever they'll be called now, are about to be touted as the saviour of Europe's 3G crisis, for the very simple reason that they're communication rather than content based. And while we don't claim to predict the future, that it's a model that's as sane as anything else on the table.
Check out the Vinny the Vampire comic strip
"It is a greater offense to steal men's labor, than their clothes"
JXTA does not mandate how messages are propogated
Its sort of like the IP protocol, which needs TCP built on top of it to guarantee that all packets are delivered properly. This make their protocol very flexible and lets others build on top of it just like TCP/IP was built.
It's a bit strange that Sun goes from being the great advocate of fat servers and thin clients to this peer-to-peer protocol. But I guess the popularity of Napster gave them a vision of what might come, so why not embrace it.
I'm sure we'll see alot more about Net vs. One in the news from now on...
Oh, I can't help quoting you because everything that you said rings true
Wise up buddy. All companies want your money, even your pal IBM. (or do you work for them)
This is the company that weasled out of anti-trust by snowballing 10 years, oh and if you want to see what IBM really think of Linux and open source read on...
IBMs view on open source
"You can't hold an open-source community accountable," says Dick Sullivan, a VP of marketing in IBM's software group. "Customers say, 'How long do you think I'd last if I told my CIO, I've registered this problem on the Internet and I'm waiting for the open-source community to get it fixed.' It's not in the business culture today."
IBM on Linux
"Customers currently running Linux apps might have concerns about management and scalability. So now their Linux app can run with the AIX back end," says Mike Kerr, IBM's VP of marketing for the pSeries.
"a notion that caused their sales to drop 73 percent, "
...
Linux installations tend to replace older , lower end Sun boxes used for DNS, ftp etc
System running Oracle or Sybase still run Solaris and , as far as I can tell, people are not talking about replacing them with Linux.
In short, this guy from Sun was rigth.
The only way liberals win national elections is by pretending they're not liberals.