Liberty version 1 is contingent on trust relationships negotiated out of band between identity provider and service provider.
Liberty version 1 doesn't make provisions for sharing personal information -- it only defines protocols for federation, single sign-on, federation termination, and logout.
IPL, the source code release that is linked to, is at version 0.1, not version 6.0. The original posting is not very clear. Note that IPL is not the same thing as Sun ONE Identity Server.
Buy the base station from Apple for $300 and then buy the cards from Lucent -- the WaveLAN (now Orinoco) cards are supported on lots of hardware and operating systems.
We've done this here at the office (used the Lucent cards with the WaveLAN) and it works great. You may need to use a Mac to configure the base station the way you want it -- it uses SNMP though so other software should be usable. The author mentions that the Karlbridge software for Windows does everything right. Else you can figure out what its default IP address is and use that (it's listed in the documentation somewhere, or maybe on http://til.info.apple.com/). Or get friendly with someone who owns a PowerBook... or bring the base station with you to some sort of public computing facility that has Macs on ethernet and use the Airport Admin Utility which is a free download from Apple's web site (you don't have to install the AirPort software; you can just unpack the archive and there is a copy of the admin utility there that you can run).
The other good thing to do to your base station is to get one of the Lucent range extenders and then drill another little hole in the case to run the wire in... you can get stronger range this way.
The things really do work awfully well. Very handy when setting up new headless machines for me to telnet into the console server from a laptop over the wireless net and then be able to carry around the console terminal if I need to go back to fiddle with the hardware... without losing my console session and having to continually connect back in.
One thing the author didn't quite explain is that there are more reasons than the lack of an integrated antenna that the AirPort cards will not work in machines other than late model Apple machines. The bus is weird. The AirPort card actually sits on an ATA bus!
One final interesting AirPort hint is that if you have an older Mac and want to use the Lucent card with it under MacOS, you can use the AirPort 1.1 software with it instead of the drivers Lucent provides, which seem to be shoddy in our experience here (they really mess up the system on MacOS 9 since it somehow manages to install the 68k version of the driver by mistake).
only active, needed chunks of the code are loaded into RAM.
I think the technical term for this is "demand paging." Except isn't it "code fragments" instead of pages for MacOS? Or am I confused. I'm not really a Mac person . . .
Huh? I think you may be confused about your terms.
Virtual memory (VM) is how you implement processes with separate memory spaces; with a page table, and a page fault handler in the kernel. It's how your swap space gets used on Linux.
Certainly you could have protection without paging; there were machines in the 60's that did that if I recall correctly. It's kind of silly, though, because you have to either always use position-independent code, or have a loader which relocates your code on the fly. Or you can do weird segment register things. A page table (and hardware assist via a TLB) is a much more flexible solution, considering it lets you do things like paging to disk (instead of swapping entire processes to disk, which you had to do before you had VM).
Are there any mainstream OSes out there which don't do virtual memory? I think MacOS is weird this way somehow, but I'm not sure exactly what it does and doesn't do (just wait for MacOS X.)
@Home seems to do a pretty good job on the receiving end (or maybe this is just due to cancelbots). But the amount of spam I see on their news servers is pretty tolerable.
They're really fast news servers as well, and they seem to keep several thousand articles in each newsgroup. I was really going to be bummed if I had to read news elsewhere in the mean time...
Email spam, on the other hand, is horrible; I get more spam at @home than anywhere else. One last week looked like someone was actually iterating through all possible eight-character usernames... not sure why @home is such a spam target.
It's been a year and a half since any of these pages have been updated, and when they were left in the state they are now, the guy hadn't figured out much at a software level at all; this project looks much more like a hardware hack, and an old one at that.
It is neat that Linux is the type of OS that can be easily ported to lots of different platforms, and is rather well suited for embedded applications (easy to pull out stuff you don't need, among other things). But it doesn't look like anyone is actually working on Linux on the N64, which is kind of pointless except as an academic exercise (which is what this project apparently was when it was alive).
IMHO it's not worth getting worked up about. Personally, I think Palm is a much spiffier target...
Slashdot Mirror
IPL is available from http://liberty.sunsource.net/ and the code is under the SISSL (you can see the license on opensource.org).
The license is not Apache style--it's fairly unique. (I should know--I am the IPL author.)
Liberty version 1 doesn't make provisions for sharing personal information -- it only defines protocols for federation, single sign-on, federation termination, and logout.
See the Liberty architecture overview (in the specs section on the Liberty web site) for more information.
Not quite. See the Liberty Alliance Project web site for more information.
IPL, the source code release that is linked to, is at version 0.1, not version 6.0. The original posting is not very clear. Note that IPL is not the same thing as Sun ONE Identity Server.
This is the first open source implementation that we know of. There are other (non-open-source) implementations.
The license, SISSL, is OSI approved, and the FSF considers it to be a free software license.
We've done this here at the office (used the Lucent cards with the WaveLAN) and it works great. You may need to use a Mac to configure the base station the way you want it -- it uses SNMP though so other software should be usable. The author mentions that the Karlbridge software for Windows does everything right. Else you can figure out what its default IP address is and use that (it's listed in the documentation somewhere, or maybe on http://til.info.apple.com/). Or get friendly with someone who owns a PowerBook... or bring the base station with you to some sort of public computing facility that has Macs on ethernet and use the Airport Admin Utility which is a free download from Apple's web site (you don't have to install the AirPort software; you can just unpack the archive and there is a copy of the admin utility there that you can run).
The things really do work awfully well. Very handy when setting up new headless machines for me to telnet into the console server from a laptop over the wireless net and then be able to carry around the console terminal if I need to go back to fiddle with the hardware... without losing my console session and having to continually connect back in.
One thing the author didn't quite explain is that there are more reasons than the lack of an integrated antenna that the AirPort cards will not work in machines other than late model Apple machines. The bus is weird. The AirPort card actually sits on an ATA bus!
One final interesting AirPort hint is that if you have an older Mac and want to use the Lucent card with it under MacOS, you can use the AirPort 1.1 software with it instead of the drivers Lucent provides, which seem to be shoddy in our experience here (they really mess up the system on MacOS 9 since it somehow manages to install the 68k version of the driver by mistake).
Virtual memory (VM) is how you implement processes with separate memory spaces; with a page table, and a page fault handler in the kernel. It's how your swap space gets used on Linux.
Certainly you could have protection without paging; there were machines in the 60's that did that if I recall correctly. It's kind of silly, though, because you have to either always use position-independent code, or have a loader which relocates your code on the fly. Or you can do weird segment register things. A page table (and hardware assist via a TLB) is a much more flexible solution, considering it lets you do things like paging to disk (instead of swapping entire processes to disk, which you had to do before you had VM).
Are there any mainstream OSes out there which don't do virtual memory? I think MacOS is weird this way somehow, but I'm not sure exactly what it does and doesn't do (just wait for MacOS X.)
There are some interesting ones that I had not heard about before; several large companies (several of their customers, actually). Rather interesting.
BTW: Who are the underwriters for this? Do we know yet?
--jss.
They're really fast news servers as well, and they seem to keep several thousand articles in each newsgroup. I was really going to be bummed if I had to read news elsewhere in the mean time...
Email spam, on the other hand, is horrible; I get more spam at @home than anywhere else. One last week looked like someone was actually iterating through all possible eight-character usernames... not sure why @home is such a spam target.
It is neat that Linux is the type of OS that can be easily ported to lots of different platforms, and is rather well suited for embedded applications (easy to pull out stuff you don't need, among other things). But it doesn't look like anyone is actually working on Linux on the N64, which is kind of pointless except as an academic exercise (which is what this project apparently was when it was alive).
IMHO it's not worth getting worked up about. Personally, I think Palm is a much spiffier target...