MicroSoft talks about making their code available in some sort of shared arrangement. The point that worries me is what conditions will they place on people who are allowed to look at their code? The last time I was in a position to get access to NT code as a non-MicroSoft employee there was a clause in the agreement that I could not work on any other OS for 2 years after I stopped having access to the code. This kind of restriction would make MicroSoft's talk of sharing completely worthless.
Now, has anyone ever made actual USE of such an agreement? That I can't tell you.
Oh yeah, they have. I was one of the founders of
a company called Netwise. We made a cross-architecture RPC compiler and some of our
customers required a software escrow clause. When
MicroSoft bought out Netwise the escrow clauses
went into force and the customers got their own
private copy of the source to our product. Whether or not they
could make heads or tails of it I don't know but
they got it.
PS: Yes, MicroSoft bought Netwise. No, I didn't
make a penny off the deal. I'd quit a year before
and the 22,000 to 1 reverse stock split kind of
destroyed my holdings. I'm still a wage slaev.
Intel's
strategy for their upcoming CPU is to not make it backwards compatable. This is a *serious* flaw in their
plans and I don't think we're going to see them succeed with this venture.
If I have to respond to this kind of comment one
more time I will SCREAM.
Intel is, repeat is, retaining backwards
compatibility with IA32 application code on their IA64 processor.
Read the literature that is out there right now
before you make this completely erroneous assertion.
I have an IA64 machine in my office running Linux
right now. I run Quake 3 Arena on it. I guarantee
there is not an IA64 version of Q3A, I'm running
IA32 Q3A binaries, I run IA32 StarOffice binaries,
I run IA32 Oracle binaries. I don't
know of any IA32 Linux application that
doesn't run. If fact, if you know of one, please
let me know, one of my jobs in life is to make
sure that IA32 applications run on IA64 Linux.
Disclaimer: I work for VA Linux, I'm a Unix bigot and I abhor MicroSoft products and practices. Keep that in mind as I try to rebut a few of the points in this article.
why
wasn't Microsoft justified in integrating Windows with its much-hyped OS?
Wouldn't doing otherwise prove corporate suicide? Was Microsoft really
supposed to sit back and allow competitors to dominate this critical market,
surely threatening Windows in the process?
They are fully justified in creating a browser that competes with Netscape. They are not justified in tying their browser to their OS and then using the monopoly status of their OS to turn their browser into a monopoly also. This is the heart of the case - you can't use a current monopoly to create for yourself a new monopoly. This is exactly what MicroSoft has done.
Of course, the ironic thing is MicroSoft probably didn't have to do this. Their browser now is arguably better than Netscape. If they had just competed fairly it is quite conceivable that they would have won the browser wars on technical merit alone. As it turned out, they got scared, they cheated, and I sincerely hope they get punished for it.
Judge Jackson... who made critical
comments about Microsoft and its founder to reporters while the appeal
process is still underway.
I think it was The Register that pointed out that Judge Jackson didn't start making his provocative comments until 4 days after MicroSoft filed it's appeal. At that point in time he no longer had control over the case, it was now in the hands of the appeals court. Technically, Judge Jackson did nothing wrong, I think the appeals court is mainly having a personality conflict with Judge Jackson. (An old legal joke is that the way to increase the average intelligence of the appeals courts and the trial courts, at the same time, is to appoint a trial judge to the appeals court.)
Are other giant theme park operators really free to create new versions of
Disney World
Actually, yes they are free to do so and are doing so successfully. The Six Flags and SeaWorld franchises are two competitors that immediately come to mind. All over America you will find local amuzement parks that, although not necessarily of the same size as Disney, still provide viable competition.
There are a lot of really fishy things going on in the details of this agreement, and quite a
few of us suspect that, in this end game, the 3dfx management may be trying to benefit themselves at the
expense of their own shareholder
Remember, the board of directors of 3DFX does indeed have a fiduciary duty to the owners of the company, the stock holders. Much as I hate to say it but the term shareholder lawsuit immediately comes to mind. This is the USA, you always have the right to sue someone (and in this case it might even be the proper thing to do).
even whole cities are divided in parts where there is ADSL available and parts where there isn't yet.
Don't expect this to change any time soon. This is still the case in the overly wired USA where parts of a city have ADSL and other parts don't. The problem is the wires from your home to the Central Office have to be within a certain distance, I think around 2 miles. If you exceed this length then you're out of luck, no ADSL service. Also, the speed falls off as you get near this limit so you can be in the situation where you can get ADSL but its no faster than ISDN.
There can be other problems too. I have a friend who lives in a part of the city that will probably never get ADSL service. The telephone installer told him that the ADSL circuit cards take up two slots in the junction box, reducing the number of lines the telco can provide, and are very expensive. Combine the two and the local telco will probably never upgrade his area.
Bad thing about hardware is that you can't just issue a bugfix patch
Actually, this is exactly what the Microprocessor Update feature of the current Pentium's is all about.
Intel releases firmware updates as binary files that can be loaded into the processors firmware memory and can fix processor errata.
Sorry, Intel won't tell you what the firmware instruction format is and they also encrypt the update files that they release so you can't change the instruction set willy nilly but this is a nifty way of correcting processor mistakes.
I should preface these comments by noting that they were created on an Itanium machine running Linux.
Yes, the Itanium release has been delayed many times but if you think about it it is understandable. Unlike the PIV the Itanium is a completely new architecture, not just a bigger and faster version of a current CPU. We've had to create completely new software generation tools, port OS's & applications, debug errata's in the silicon... The astonishing thing is that we have anything working at all, not that it's late.
McKinley on the other hand will be more like a PIV release, it is basically a bigger & faster Itanium. I would expect that the McKinley project will adhere to its schedules much better.
We've had something like 4 different beta releases of TurboLinux for the IA64. Linux will be ready for Itanium as soon as Intel decides to start selling them, which should be soon. I don't see any chance that McKinley will come out before Itanium.
The issue that is really twisting my shorts is what happens to the Samsun et. al. agreements if Micron wins it's case and invalidates the Rambus' DDR SDRM patents (I know, I'm an optimist)? My guess is that Rambus gets to keep collecting money for no good reason at all but I'd be curious if anyone knows for sure.
I agree that AMD's Hammer looks better positioned though (mostly due to being IA-32 compatible
vs IA-64,s software emulation).
If I hear this comment one more time I'm going to SCREAM. You can fault Intel for many things but believe me, they do NOT do software emulation of IA-32 instructions on an Itanium. I know this for a fact because I am working on the Linux kernel for Itanium and, in fact, I'm the lead engineer working on IA-32 support in the kernel. Itanium runs IA32 binaries and runs them in silicon. For proof I've got Quake 3 Arena running on my IA-64 machine (performance sucks but it's a hell of a lot better than software emulation would be).
For good or ill this is one area where the AMD marketing people have succeeded in pulling the wool over the technical community's eyes and for
reasons that escape me Intel is ignoring this issue.
In conclusion, and to repeat myself, the Itanium processor executes IA-32 applications in silicon with complete compatibity with current IA-32 processors.
Is VA Linux seriously going to sue some college kid in Texas because he stole their site design?
Somehow I doubt it. Imagine the great publicity that would generate...
Great, I predict it will take less than 24 hours for the rumor to start that VA is suing a college kid. And it's all your fault:-)
If IBM buys Novell, IBM will have the final piece of the puzzle: a network operating system from which to integrate Linux/AIX compatibility.
If you'll cast your mind back a few years you'll remember that at one point in time Novel bought USL which gave Novel the final piece of the puzzle: the SVR4 version of Unix which they could integrate into their network operating system.
This integration was such an abysmal failure that Novell eventually sold their Unix division to SCO. If IBM is trying to solve a puzzle they need a few more pieces than Novell.
How about a section of "streaming cache" for data that will pass through, but only once?
Please RTFM, in this case the instruction set reference for the PIII. Part of the new Streaming Simd Extensions is a set of instructions that:
1) Prefetch from memory to anywhere in the cache hierarchy. 2) Write to memory, bypassing the cache hierarchy.
Using these instructions I was able to write block bopy routines that achieved transfer rates of up to 600 MBytes/Sec. on a 500Mhz PIII. The same transfer using the GLIBC bcopy routine could get no more that 235 MBytes/Sec.
Because if there isn't, I'm not sure if migrating to a different OS -- even a similar one like Linux -- would be easy, or even possible.
Hint, iBCS is your friend.
I've taken very complex programs written for UnixWare and run them successfully on Linux. I had to copy all of the shared libraries over (we'll ignore any copyright issues associated with that) but it worked quite well. I don't know how your OpenServer programs would work but you might want to look into this, just to prepare for the inevitable:-)
--
-- Don Dugger VA Linux Systems
Re:reports of SCO's demise...
on
Endgame For SCO
·
· Score: 1
I have got to respond to this comment:
One bad quarter != a dead company.
No, but multiple bad quarters followed by minimal prospects for the future do indicate a company in serious trouble.
From what I know SCO's major products are OpenServer, UnixWare and Tarantella. OpenServer - Based upon Xenix => nobody in their right mind would buy it new today. UnixWare - Based upon SRV4, actually a very good enterprise server that SCO has just not been able to market. Tarantella - I have to admit I never understood this, it's just another piece of software that gets between the client and server but it seems to be selling well so it must do something.
Bottom line is that SCO's OS line is severely hurting and, unfortunately, seems to be going nowhere.
At any rate, it *will* be able to run 32-bit apps natively, not through emulation as with Merced--err, Itanium (dumb name).
RANT The specifications for Itanium are available for free from Intel. I really wish you would get a copy and read it before you post blatantly incorrect statements. /RANT
The Itanium does not emulate i386 instructions. A considerable amount of real estate on the die is devoted to executing i386 instructions. I know this for a fact, I execute i386 processes on an Itanium on a daily basis and, believe me, there is no emulator in the system. --
Yes, these are binaries that were copied from an i386 machine onto the IA64 machine and they just worked, including the shared libraries (which were also just copied over from the i386 machine).
Performance is kind of a dirty word right now, the IA64 cpu's and chipsets are just too new to give real performance numbers yet. Having said that I believe that even Intel will tell you that IA32 programs will not run as fast on Itanium as they will on the fastest IA32 processor available at that time. Let's face it, this is a 64-bit machine. If you want 64-bit performance use Itanium, if you want 32-bit performance use Pentium. --
I can't comment on Sledgehammer but I can talk about the IA32 support available on the Itanium. I've been working on adding IA32 support to the IA64 Linux kernel for the last 6 months so I have some knowledge of the subject.
Individual processes can select which instruction set they wish to run in, IA32 or IA64, even though the kernel is executing entirely in the IA64 instruction set. We've added IA32 kernel interfaces to match the system calls available currently on the i386 Linux kernel. This is not vaporware, this is running and has been publicly demonstrated at conferences this year.
Currently I've run IA32 versions of bash, gdb, gcc and netscape. All of these programs are running now with no known problems. I'm sure there are IA32 programs out there that don't work yet but my goal is to make sure that eventually all IA32 programs will run on the Itanium.
I admit to having a bias on this subject as I work for VA Linux and my job is to help create Linux for the IA64 processor. --
Sizes are strictly a convention and the compiler is free to use any sizes it likes. The Cygnus and SGI compilers used by the Trillian project both use the LP64 convention (longs and pointers are 64-bits, all other sizes follow IA32 conventions). MicroSoft decided to use a compiler that uses the P64 model (pointers are 64-bits, all other sizes follow IA32 conventions).
The `advantage' of the P64 model is that you get fewer compiler warnings when you compile programs that were originally written for IA32. They could still be semantically wrong but at least the compiler doesn't complain.
The advantage of the the LP64 model is that there is a native type available for all integer sizes.
From my comments you can guess which model I believe is the correct one to follow, especially given that my experience has been that Linux applications just re-compile and work on IA64. There has been very little changes required by applications to get them to work. Of course, this is undoubtedly helped by the fact that Linux supports the Alpha.
You must not have seen the demo itself. The `doom' that was running was an x86 binary that was created on an x86 system and copied to the IA-64 machine. One of the Trillian project goals is the ability to run current x86 binaries with no modifications required. The `doom' demo was the proof of concept that this goal is achievable.
Remember, the IA-64 chip comes from the same company that produces CPU's that can still run 8086 binaries - compatibility and continuity are indeed important.
Slashdot Mirror
--
Don Dugger
VA Linux Systems
--
Don Dugger
VA Linux Systems
Oh yeah, they have. I was one of the founders of a company called Netwise. We made a cross-architecture RPC compiler and some of our customers required a software escrow clause. When MicroSoft bought out Netwise the escrow clauses went into force and the customers got their own private copy of the source to our product. Whether or not they could make heads or tails of it I don't know but they got it.
PS: Yes, MicroSoft bought Netwise. No, I didn't make a penny off the deal. I'd quit a year before and the 22,000 to 1 reverse stock split kind of destroyed my holdings. I'm still a wage slaev.
--
Don Dugger
VA Linux Systems
If I have to respond to this kind of comment one more time I will SCREAM. Intel is, repeat is, retaining backwards compatibility with IA32 application code on their IA64 processor. Read the literature that is out there right now before you make this completely erroneous assertion.
I have an IA64 machine in my office running Linux right now. I run Quake 3 Arena on it. I guarantee there is not an IA64 version of Q3A, I'm running IA32 Q3A binaries, I run IA32 StarOffice binaries, I run IA32 Oracle binaries. I don't know of any IA32 Linux application that doesn't run. If fact, if you know of one, please let me know, one of my jobs in life is to make sure that IA32 applications run on IA64 Linux.
--
Don Dugger
VA Linux Systems
--
Don Dugger
VA Linux Systems
OK, then can you give me performance numbers for SQL7/NT4.0 vs. SQL2000/Win2K?
--
Don Dugger
VA Linux Systems
why wasn't Microsoft justified in integrating Windows with its much-hyped OS? Wouldn't doing otherwise prove corporate suicide? Was Microsoft really supposed to sit back and allow competitors to dominate this critical market, surely threatening Windows in the process?
They are fully justified in creating a browser that competes with Netscape. They are not justified in tying their browser to their OS and then using the monopoly status of their OS to turn their browser into a monopoly also. This is the heart of the case - you can't use a current monopoly to create for yourself a new monopoly. This is exactly what MicroSoft has done.
Of course, the ironic thing is MicroSoft probably didn't have to do this. Their browser now is arguably better than Netscape. If they had just competed fairly it is quite conceivable that they would have won the browser wars on technical merit alone. As it turned out, they got scared, they cheated, and I sincerely hope they get punished for it.
Judge Jackson ... who made critical
comments about Microsoft and its founder to reporters while the appeal
process is still underway.
I think it was The Register that pointed out that Judge Jackson didn't start making his provocative comments until 4 days after MicroSoft filed it's appeal. At that point in time he no longer had control over the case, it was now in the hands of the appeals court. Technically, Judge Jackson did nothing wrong, I think the appeals court is mainly having a personality conflict with Judge Jackson. (An old legal joke is that the way to increase the average intelligence of the appeals courts and the trial courts, at the same time, is to appoint a trial judge to the appeals court.)
Are other giant theme park operators really free to create new versions of Disney World
Actually, yes they are free to do so and are doing so successfully. The Six Flags and SeaWorld franchises are two competitors that immediately come to mind. All over America you will find local amuzement parks that, although not necessarily of the same size as Disney, still provide viable competition.
--
Don Dugger
VA Linux Systems
Remember, the board of directors of 3DFX does indeed have a fiduciary duty to the owners of the company, the stock holders. Much as I hate to say it but the term shareholder lawsuit immediately comes to mind. This is the USA, you always have the right to sue someone (and in this case it might even be the proper thing to do).
--
Don Dugger
VA Linux Systems
Don't expect this to change any time soon. This is still the case in the overly wired USA where parts of a city have ADSL and other parts don't. The problem is the wires from your home to the Central Office have to be within a certain distance, I think around 2 miles. If you exceed this length then you're out of luck, no ADSL service. Also, the speed falls off as you get near this limit so you can be in the situation where you can get ADSL but its no faster than ISDN.
There can be other problems too. I have a friend who lives in a part of the city that will probably never get ADSL service. The telephone installer told him that the ADSL circuit cards take up two slots in the junction box, reducing the number of lines the telco can provide, and are very expensive. Combine the two and the local telco will probably never upgrade his area.
--
Don Dugger
VA Linux Systems
Actually, this is exactly what the Microprocessor Update feature of the current Pentium's is all about. Intel releases firmware updates as binary files that can be loaded into the processors firmware memory and can fix processor errata. Sorry, Intel won't tell you what the firmware instruction format is and they also encrypt the update files that they release so you can't change the instruction set willy nilly but this is a nifty way of correcting processor mistakes.
--
Don Dugger
VA Linux Systems
Yes, the Itanium release has been delayed many times but if you think about it it is understandable. Unlike the PIV the Itanium is a completely new architecture, not just a bigger and faster version of a current CPU. We've had to create completely new software generation tools, port OS's & applications, debug errata's in the silicon ... The astonishing thing is that we have anything working at all, not that it's late.
McKinley on the other hand will be more like a PIV release, it is basically a bigger & faster Itanium. I would expect that the McKinley project will adhere to its schedules much better.
We've had something like 4 different beta releases of TurboLinux for the IA64. Linux will be ready for Itanium as soon as Intel decides to start selling them, which should be soon. I don't see any chance that McKinley will come out before Itanium.
--
Don Dugger
VA Linux Systems
--
Don Dugger
VA Linux Systems
If I hear this comment one more time I'm going to SCREAM. You can fault Intel for many things but believe me, they do NOT do software emulation of IA-32 instructions on an Itanium. I know this for a fact because I am working on the Linux kernel for Itanium and, in fact, I'm the lead engineer working on IA-32 support in the kernel. Itanium runs IA32 binaries and runs them in silicon. For proof I've got Quake 3 Arena running on my IA-64 machine (performance sucks but it's a hell of a lot better than software emulation would be).
For good or ill this is one area where the AMD marketing people have succeeded in pulling the wool over the technical community's eyes and for reasons that escape me Intel is ignoring this issue.
In conclusion, and to repeat myself, the Itanium processor executes IA-32 applications in silicon with complete compatibity with current IA-32 processors.
--
Don Dugger
VA Linux Systems
Great, I predict it will take less than 24 hours for the rumor to start that VA is suing a college kid. And it's all your fault :-)
--
Don Dugger
VA Linux Systems
If you'll cast your mind back a few years you'll remember that at one point in time Novel bought USL which gave Novel the final piece of the puzzle: the SVR4 version of Unix which they could integrate into their network operating system.
This integration was such an abysmal failure that Novell eventually sold their Unix division to SCO. If IBM is trying to solve a puzzle they need a few more pieces than Novell.
--
Don Dugger
VA Linux Systems
Please RTFM, in this case the instruction set reference for the PIII. Part of the new Streaming Simd Extensions is a set of instructions that:
1) Prefetch from memory to anywhere in the cache hierarchy.
2) Write to memory, bypassing the cache hierarchy.
Using these instructions I was able to write block bopy routines that achieved transfer rates of up to 600 MBytes/Sec. on a 500Mhz PIII. The same transfer using the GLIBC bcopy routine could get no more that 235 MBytes/Sec.
--
Don Dugger
VA Linux Systems
Hint, iBCS is your friend.
I've taken very complex programs written for UnixWare and run them successfully on Linux. I had to copy all of the shared libraries over (we'll ignore any copyright issues associated with that) but it worked quite well. I don't know how your OpenServer programs would work but you might want to look into this, just to prepare for the inevitable :-)
--
--
Don Dugger
VA Linux Systems
One bad quarter != a dead company.
No, but multiple bad quarters followed by minimal prospects for the future do indicate a company in serious trouble.
From what I know SCO's major products are OpenServer, UnixWare and Tarantella.
OpenServer - Based upon Xenix => nobody in their right mind would buy it new today.
UnixWare - Based upon SRV4, actually a very good enterprise server that SCO has just not been able to market.
Tarantella - I have to admit I never understood this, it's just another piece of software that gets between the client and server but it seems to be selling well so it must do something.
Bottom line is that SCO's OS line is severely hurting and, unfortunately, seems to be going nowhere.
--
Don Dugger
VA Linux Systems
RANT
The specifications for Itanium are available for free from Intel. I really wish you would get a copy and read it before you post blatantly incorrect statements.
/RANT
The Itanium does not emulate i386 instructions. A considerable amount of real estate on the die is devoted to executing i386 instructions. I know this for a fact, I execute i386 processes on an Itanium on a daily basis and, believe me, there is no emulator in the system.
--
--
Performance is kind of a dirty word right now, the IA64 cpu's and chipsets are just too new to give real performance numbers yet. Having said that I believe that even Intel will tell you that IA32 programs will not run as fast on Itanium as they will on the fastest IA32 processor available at that time. Let's face it, this is a 64-bit machine. If you want 64-bit performance use Itanium, if you want 32-bit performance use Pentium.
--
--
Individual processes can select which instruction set they wish to run in, IA32 or IA64, even though the kernel is executing entirely in the IA64 instruction set. We've added IA32 kernel interfaces to match the system calls available currently on the i386 Linux kernel. This is not vaporware, this is running and has been publicly demonstrated at conferences this year.
Currently I've run IA32 versions of bash, gdb, gcc and netscape. All of these programs are running now with no known problems. I'm sure there are IA32 programs out there that don't work yet but my goal is to make sure that eventually all IA32 programs will run on the Itanium.
I admit to having a bias on this subject as I work for VA Linux and my job is to help create Linux for the IA64 processor.
--
--
The `advantage' of the P64 model is that you get fewer compiler warnings when you compile programs that were originally written for IA32. They could still be semantically wrong but at least the compiler doesn't complain.
The advantage of the the LP64 model is that there is a native type available for all integer sizes.
From my comments you can guess which model I believe is the correct one to follow, especially given that my experience has been that Linux applications just re-compile and work on IA64. There has been very little changes required by applications to get them to work. Of course, this is undoubtedly helped by the fact that Linux supports the Alpha.
--
You must not have seen the demo itself. The `doom' that was running was an x86 binary that was created on an x86 system and copied to the IA-64 machine. One of the Trillian project goals is the ability to run current x86 binaries with no modifications required. The `doom' demo was the proof of concept that this goal is achievable.
Remember, the IA-64 chip comes from the same company that produces CPU's that can still run 8086 binaries - compatibility and continuity are indeed important.