If I ever need to go back in time and buy a processor, I'll keep that in mind. Meanwhile here in the present Sun is pretty much dead last in performance and they have no advantage in features either.
Jonathan Shapiro wrote a great article analyzing the Windows Common Criteria certification; much of it applies to Linux as well. Among other things, it explains why Windows can get certified even with its remote root exploits: "An EAL4 rating means that you did a lot of paperwork related to the software process, but says absolutely nothing about the quality of the software itself. There are no quantifiable measurements made of the software, and essentially none of the code is inspected."
If you separate that calling sequence into an address calculation and a call, then you can insert a prepare to branch immediately after the address calculation. Then branch prediction doesn't matter. I don't see any problem with objects being moved; obviously they don't get moved while the program is trying to access them.
I imagine that modern garbage collection, which moves objects around, might actually also make this worse, as the branch target buffer relies on real addresses.
I don't think GCs move code around very often, and code addresses are all that matters for branch prediction.
UltraSPARC III has a "prepare to branch" instruction that can allow the processor to correctly prefetch past an unpredictable unconditional branch like those found in virtual method calls.
Can someone familiar with the decision making process post a summary of why the LSB group simply didn't choose to implement POSIX rather than creating their own standard?
The LSB group can't choose those things, because LSB only documents the existing behavior of Linux/glibc. The decisions are actually made by Linus and Ulrich Drepper, and they usually have good reasons when they deviate from POSIX.
Unfortunately that means LSB can't deprecate anything in POSIX, and POSIX is probably not willing to deprecate anything at all, resulting in decades of cruft.
Can you imagine the dark cave we would still be living in if TCP was somehow specifically bound to ethernet?... As a result, this will wind up like previous CE standards effort and fail to provide a consistent interoperable system.
Can you imagine the dark cave we would be living in if HAVi was not bound to any link layer in particular? HAVi/Ethernet devices couldn't connect to HAVi/FireWire devices, and it would fail to provide a consistent interoperable system.
Re:Reiser4? Competition?
on
Reiser4 Benchmarks
·
· Score: 2, Informative
The port of the Plan9 filing system is said to be one of the fastest filing systems out there - enough so that it's a part of a Government research program called "Pink", run by some mad scientists at Los Alamos.
I think they are using the Plan 9 network filesystem protocol (9P2000), not the Plan 9 local filesystem (which isn't anything special AFAIK).
Re:To bad Evolution probably wont support it
on
Kroupware Komplete
·
· Score: 1
The Steltor, er Oracle calendar server is working on CAP support. Last I checked the RFC was about to be issued, so it's no surprise that apps are just starting to support it.
Re:To bad Evolution probably wont support it
on
Kroupware Komplete
·
· Score: 4, Informative
If both Kroupware and Evolution supported standards, then they wouldn't need any extra support to work together. Unfortunately, Kroupware is taking a weird approach of accessing contacts and calendar entries using IMAP instead of LDAP and CAP. Evolution doesn't support CAP either.
A2. IIRC InterTrust has patents on pretty much anything related to DRM. Since MS advertises that their products have DRM, they are obviously in violation. (I leave it to you to decide whether the patents are valid.)
Re:Before all the flamers get in.
on
Qt On DirectFB
·
· Score: 1
The software is designed to help network operators get more value from on-demand and other digital TV services.
Is there any benefit to cable customers? Is this basic program guide any different from the basic program guide that's built in to all digital cable boxes?
No shit. But when vendors call something a "4-way server", they mean that model can be configured to hold 4 CPUs. Most of the time it can also be configured to hold less. You will never get a 4-processor PowerPC 970 system for $3500, mark my words.
Slashdot Mirror
And don't forget JMF.
If I ever need to go back in time and buy a processor, I'll keep that in mind. Meanwhile here in the present Sun is pretty much dead last in performance and they have no advantage in features either.
Jonathan Shapiro wrote a great article analyzing the Windows Common Criteria certification; much of it applies to Linux as well. Among other things, it explains why Windows can get certified even with its remote root exploits: "An EAL4 rating means that you did a lot of paperwork related to the software process, but says absolutely nothing about the quality of the software itself. There are no quantifiable measurements made of the software, and essentially none of the code is inspected."
Component is so last year. DVI is the way to go for HD.
Don't you mean flash is getting cheap these days? Not long ago it cost over $1/MB.
If you separate that calling sequence into an address calculation and a call, then you can insert a prepare to branch immediately after the address calculation. Then branch prediction doesn't matter. I don't see any problem with objects being moved; obviously they don't get moved while the program is trying to access them.
I imagine that modern garbage collection, which moves objects around, might actually also make this worse, as the branch target buffer relies on real addresses.
I don't think GCs move code around very often, and code addresses are all that matters for branch prediction.
UltraSPARC III has a "prepare to branch" instruction that can allow the processor to correctly prefetch past an unpredictable unconditional branch like those found in virtual method calls.
As long as the universities process those thousands of DMCA takedown notices promptly, they aren't liable. So this is a solved problem.
Set up 6to4; a little Googling will find a tutorial for your OS. You don't need any cooperation from your ISP.
BTW, don't use a tunnel broker; it's much less efficient than 6to4.
All of the Java platform is available on a free license for open-source developments, including the test suites. This is what the Kaffe people use.
Can you provide a link? I did some Googling and only found messages saying "we wish we had the JCK".
That FAQ is from 2002, which predates JCP 2.5.
Can someone familiar with the decision making process post a summary of why the LSB group simply didn't choose to implement POSIX rather than creating their own standard?
The LSB group can't choose those things, because LSB only documents the existing behavior of Linux/glibc. The decisions are actually made by Linus and Ulrich Drepper, and they usually have good reasons when they deviate from POSIX.
Posix is(shuld be) a subset of LSB...
Unfortunately that means LSB can't deprecate anything in POSIX, and POSIX is probably not willing to deprecate anything at all, resulting in decades of cruft.
Can you imagine the dark cave we would still be living in if TCP was somehow specifically bound to ethernet? ...
As a result, this will wind up like previous CE standards effort and fail to provide a consistent interoperable system.
Can you imagine the dark cave we would be living in if HAVi was not bound to any link layer in particular? HAVi/Ethernet devices couldn't connect to HAVi/FireWire devices, and it would fail to provide a consistent interoperable system.
The port of the Plan9 filing system is said to be one of the fastest filing systems out there - enough so that it's a part of a Government research program called "Pink", run by some mad scientists at Los Alamos.
I think they are using the Plan 9 network filesystem protocol (9P2000), not the Plan 9 local filesystem (which isn't anything special AFAIK).
The Steltor, er Oracle calendar server is working on CAP support. Last I checked the RFC was about to be issued, so it's no surprise that apps are just starting to support it.
If both Kroupware and Evolution supported standards, then they wouldn't need any extra support to work together. Unfortunately, Kroupware is taking a weird approach of accessing contacts and calendar entries using IMAP instead of LDAP and CAP. Evolution doesn't support CAP either.
A1. It doesn't matter. :-)
A2. IIRC InterTrust has patents on pretty much anything related to DRM. Since MS advertises that their products have DRM, they are obviously in violation. (I leave it to you to decide whether the patents are valid.)
See NeWS. It was complex.
The software is designed to help network operators get more value from on-demand and other digital TV services.
Is there any benefit to cable customers? Is this basic program guide any different from the basic program guide that's built in to all digital cable boxes?
If you look at the Web page, Red Hat Linux X (Cambridge) will not include 2.6. The first version with 2.6 is Cambridge++.
No shit. But when vendors call something a "4-way server", they mean that model can be configured to hold 4 CPUs. Most of the time it can also be configured to hold less. You will never get a 4-processor PowerPC 970 system for $3500, mark my words.
Mac OS X does not rely on any kind of "Mac boot ROM"; all it needs is OpenFirmware. Guess what kind of firmware pSeries uses...
VisualAge C++ (xlc) for Linux
The base configuration of a 4-way server is probably 1 CPU.