Apple would probably be as happy selling boxes destined to run Yellow Dog Linux as OS X.
I doubt it; then Steve couldn't lock you into his idea of the digital lifestyle. After all, if you're running Linux, you won't buy a copy of Final Cut Pro ($999) or DVD Studio Pro (also $999), nor would you be very likely to shell out over $1,000 to attend their developer conference (see ad banners on/. today).
The case you cite sounds like a separate issue; the PGA Tour was spending a lot of effort to collect scores in a central system, thus adding some proprietary value to them. A more interesting question to me is: If someone keeps score for a sports event on their own, can they be prevented from publishing that information?
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.
After reading Carnage4Life's excellent post, you might also want to dig up the papers about Spring from SunLabs. Spring wasn't just a microkernel; it was a complete multi-server OS with some neat features like Plan 9-style extensible namespaces and single-system-image clustering.
It looks like JXTA is so low-level that stuff like scaling isn't even defined; as the docs say "JXTA does not mandate how messages are propogated."
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!
This sounds a lot like the Chord protocol from MIT. When one node leaves the network, it moves all the metadata it is holding to a nearby node. Since each Chord node holds (more or less) the same amount of metadata, there are no large targets to go after (and shutting down nodes doesn't hurt the network anyway).
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.
The CAD folks will get their SMP workstations in Q3, along with a ServerWorks DDR chipset with PCI-X and 6.4 GB/s of memory bandwidth. It'll cost an arm and a leg, but they'll get it.
But if you don't want to pay > $5K for an SMP system, AMD looks like the way to go.
No, I've heard from Apple engineers that the BSD code runs in kernel space and isn't a separate server. Mac OS X isn't a microkernel; it doesn't use a microkernel; a microkernel architecture is not involved.
Time Password system. You can either generate a list of passwords and carry them around in your pocket or run a password-generating utility on a PDA. If someone snags one of those "disposable" passwords, it will do them no good.
2. Don't use passwords; switch to public keys and store your private keys on a smart card or iButton (which someone else already mentioned).
All the chips are buyable on the open market, the core OS source code is open, and the ROM is no longer used in OS X.
You certainly can't buy a Uni-N (northbridge), KeyLargo (southbridge), or Pangea (combined Uni-N/KeyLargo) on the open market. So what northbridge could you use in a PPC motherboard? I haven't heard of a single one that supports recent interfaces like AGP 4X.
Speaking of ROMs, you'd need OpenFirmware on your motherboard if you want to boot Darwin/OS X; that would be quite a lot to develop yourself and probably isn't cheap to license, but at least there are a couple of companies selling OF implementations.
I haven't seen any indication that Intel is planning to use InfiniBand as their system bus. Even 12x InfiniBand (the fastest currently-specified version) is only 3 GB/s, while the P4 bus is 3.2 GB/s (granted, IB is not a bus; it's switched).
However, I agree that LDT is likely to beat the pants off the P4 bus.
Speaking of DivXNetworks, what kind of business plan do they think is going to work? MPEG-4 codecs are a commodity and their competition (like PacketVideo, Philips, Sorenson and probably every other codec company out there) has a huge head start.
I disagree; CPUs are cheap these days, and I think the CPU time needed to compress outgoing pages is likely to cost less than the bandwidth saved, thus saving money overall.
Slashdot Mirror
Apple would probably be as happy selling boxes destined to run Yellow Dog Linux as OS X.
/. today).
I doubt it; then Steve couldn't lock you into his idea of the digital lifestyle. After all, if you're running Linux, you won't buy a copy of Final Cut Pro ($999) or DVD Studio Pro (also $999), nor would you be very likely to shell out over $1,000 to attend their developer conference (see ad banners on
The case you cite sounds like a separate issue; the PGA Tour was spending a lot of effort to collect scores in a central system, thus adding some proprietary value to them. A more interesting question to me is: If someone keeps score for a sports event on their own, can they be prevented from publishing that information?
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.
After reading Carnage4Life's excellent post, you might also want to dig up the papers about Spring from SunLabs. Spring wasn't just a microkernel; it was a complete multi-server OS with some neat features like Plan 9-style extensible namespaces and single-system-image clustering.
It looks like JXTA is so low-level that stuff like scaling isn't even defined; as the docs say "JXTA does not mandate how messages are propogated."
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!
This sounds a lot like the Chord protocol from MIT. When one node leaves the network, it moves all the metadata it is holding to a nearby node. Since each Chord node holds (more or less) the same amount of metadata, there are no large targets to go after (and shutting down nodes doesn't hurt the network anyway).
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.
That's why there's a USB model.
It's a service; you can't download it. (Unless you want to cut a licensing deal.)
It looks like Compaq is doing the hardware and MIT is doing the software. They have some extremely ambitious, yet vague, goals for Oxygen.
The CAD folks will get their SMP workstations in Q3, along with a ServerWorks DDR chipset with PCI-X and 6.4 GB/s of memory bandwidth. It'll cost an arm and a leg, but they'll get it.
But if you don't want to pay > $5K for an SMP system, AMD looks like the way to go.
Did you post this to the decentralization list? I don't think I've seen it there.
No, I've heard from Apple engineers that the BSD code runs in kernel space and isn't a separate server. Mac OS X isn't a microkernel; it doesn't use a microkernel; a microkernel architecture is not involved.
Hey, if developers pf proprietary software want to use the same tools as developers of Open Source, that's fine with me.
SourceCast sounds similar to VA's SourceForge OnSite. Occasionally Larry McVoy talks about starting a similar "BitForge" project, too.
I guess we're either supposed to go to .name (whenever it opens) or fight the cybersquatters in .com.
My browser munged the first part of my post. It should have said:
1. Use a One Time Password system.
Time Password system. You can either generate a list of passwords and carry them around in your pocket or run a password-generating utility on a PDA. If someone snags one of those "disposable" passwords, it will do them no good.
2. Don't use passwords; switch to public keys and store your private keys on a smart card or iButton (which someone else already mentioned).
Steve Jobs in talks with Be? LOL! At the time Apple was looking to purchase an OS company, Steve didn't work for Apple.
I think the Aqua PLAF calls into the Appearance Manager (native code), so it probably isn't portable.
All the chips are buyable on the open market, the core OS source code is open, and the ROM is no longer used in OS X.
You certainly can't buy a Uni-N (northbridge), KeyLargo (southbridge), or Pangea (combined Uni-N/KeyLargo) on the open market. So what northbridge could you use in a PPC motherboard? I haven't heard of a single one that supports recent interfaces like AGP 4X.
Speaking of ROMs, you'd need OpenFirmware on your motherboard if you want to boot Darwin/OS X; that would be quite a lot to develop yourself and probably isn't cheap to license, but at least there are a couple of companies selling OF implementations.
I haven't seen any indication that Intel is planning to use InfiniBand as their system bus. Even 12x InfiniBand (the fastest currently-specified version) is only 3 GB/s, while the P4 bus is 3.2 GB/s (granted, IB is not a bus; it's switched).
However, I agree that LDT is likely to beat the pants off the P4 bus.
Here's another project:
http://mpeg4ip.sourceforge.net/
Instead of a codec, they're doing the other necessary stuff like a player and streaming server.
(/. rejected a story about this, too.)
Speaking of DivXNetworks, what kind of business plan do they think is going to work? MPEG-4 codecs are a commodity and their competition (like PacketVideo, Philips, Sorenson and probably every other codec company out there) has a huge head start.
I disagree; CPUs are cheap these days, and I think the CPU time needed to compress outgoing pages is likely to cost less than the bandwidth saved, thus saving money overall.
Only certified games will run on an Indrema console.
Indrema games won't run on any non-Indrema hardware.
Indrema will sell the console at a loss and make money by taking a cut from every game sold.
Indrema's Linux distribution is full of copy protection/DRM stuff.
What's your point in mentioning Indrema? It's not open at all.