Slashdot Mirror

After seeing what the standardization process did to VRML, I can see how this will be good not only for Sun but for Java also.

The standardization process for VRML allowed input from all kinds of people. Some of these people were 3d engine writers who found that they had a hard time implementing some of VRML-1 features. The result? VRML97 cannot do some things that VRML-1 could. Obviously the later version can do more, but there was no rational reason for them to go backwards in some areas and forwards in others, except the fact that it was put together by a committee.

Sun is not stupid. They realize that great ideas and innovations seldom come out of committees.

Suppose they decide to come out with a later version of Java that's not a byte code, but a compressed 32bit word code (maybe this would have some advantage in processing Java VM instructions, maybe it wouldn't, it's just a hypothetical.) Now, if this were in a standards committee, you can be sure that a lot of people would push to specify that the VM be little-endian, because that's how intel is. But then Sun would have to kludge Java to make it run better on its own big-endian architecture.

How would such a kludged, hodge-podge language benefit the Java community?

Besides, the standard for the Java VM is Already Open

AC as always... you know why.

Can someone tell me why I have to pick a platform target for downloading the JDK javadocs or extended APIs??

Oh, for cryin' out loud. Normally I ignore the mass ignorance on /., but this is just too much. Below is a quote from Sun's download page for the Java 2 documentation.

"The download choices have identical documentation content and differ only in the compression format, so "theoretically" any choice should work on any OS. However, due to tool imcompatibilites, this is not entirely true. Do not download the COMPRESS tar or GZIP tar formats and try to install them with WinZip or a non-Solaris version of tar (such as GNU tar). About 10 of the included files have paths that are 100 characters or longer, and these programs will not install them correctly."

As you can tell, it's OS and tool problems. Sun didn't write your freakin' decompression tools, so don't blame them. (And ironicly enough, it seems to suggest that a GNU tool is at fault here.) At least they went to the trouble of packaging it multiple times and documenting things which is more then I get from most companies or open source projects.

In my view (and I am not speaking for my employer) this should assuage the developer's fears of not having the standards body oversee the process.

This URL has more information.

Congratualtions Sun. You have _finally_ done what more than 1500 Java developers have clearly been asking you to do for two years, TO THE DAY.

Bug #4097810 was posted on December 8, 1997. The number one bugfix request (now request for enhancement) for the entire period, indeed the request that has had more votes than the rest of the top 25 requests COMBINED has been to treat Linux as a first tier platform, to handle the releases for Linux in-house, instead of the arm's length (although increasingly more direct) support that, to date, has been given to Blackdown.

Today's announcement is great. This is what many hundreds of Java developers have been looking forward to, and what Blackdown and Sun have been working towards. I think that Sun and Blackdown deserve kudos for achieving this excellent result.

The focus of clueless slashdotters, who literally don't know how Blackdown came to have access to the code in the first place, how much direct assistance Sun has given to Blackdown, and who haven't noticed that Blackdown is actively involved in what's happened today is on an equally uninformed opinion piece in LinuxToday.

Wake up people, this announcement is what Sun, Blackdown and hundreds of Java developers have looked forward to for years.

This story is not about open-source, it never was, Blackdown was working with confidential code from day one, and it is apparent from looking at Blackdown's site that today's announcement isn't a problem, it's just part of what they've been working towards.

• Sun reverses plan for Java standard
• Sun feels the Java heat from rivals
• Sun to drop some Java royalties

Part of what is not mentioned in any of these articles is that the former head of Java development at IBM (Patricia Sueltz) was lured to Sun towards the end of the summer and is now in charge of the Java effort there. While the article which announced the change emphasizes her wide view of Java and desire to keep Java unified, it remained to be seen whether this was lip service or not.

My thoughts are that Pat Sueltz did enough good for Java while at IBM that she ought to be amenable and keep a listening ear for response from those of us who count ourselves as "individual" (vs. corporate) members of the Java Community. Feel free to use this link to the java.sun.com feedback page, but please, be polite. Turning the wrath of /. on any company via their provided feedback mechanism(s) has bad karma, and a few well thought out, reasonable responses will get more attention that a thousand flame messages.

...[Opera] would handle everything except Java...

Java works fine on Opera with the Java plug-in (after all, it can use Netscape plug-ins, as you mentioned)

Anyway, Bochs is not free software, in either the beer or speech sense. The "you can look at our code and send us suggested changes" reminds me more of the SCSL.

I think that Sun (which few people know stands for S tanford U niversity N etworks.) is one of the coolest names/acronyms, not to mention logos out there.

--

Get the damn language spec and virtual machine specification and get coding, don't expect to get source handed to you like bloody social security. Or trink a leetle Kaffe to get better.

(I assume e.g. C++ isn't open either, then, since AT&T didn't release their source, and you'd have to pay ISO a hefty sum for the standard. But then you'll probably start yabbing about GCC, and I'll cover my ears just like you.)

Get the damn language spec and virtual machine specification and get coding, don't expect to get source handed to you like bloody social security. Or trink a leetle Kaffe to get better.

(I assume e.g. C++ isn't open either, then, since AT&T didn't release their source, and you'd have to pay ISO a hefty sum for the standard. But then you'll probably start yabbing about GCC, and I'll cover my ears just like you.)

Well, partly because they've given their source code to the Blackdown group who is doing a great job. If you don't like their port, you can always use IBM's.

I'll tell you why - they aren't interested in seeing the language shine on [Linux et al].

That's funny. They're hiring Linux developers.

OO ... is not a productivity enhancing technology.

Any technology can suck if you implement it badly. Done right, OO is extremely productive. Don't blame OO because you can't get it right.

I realize that as a Slashdot reader, this information may be new to you. Maybe you should check other sources before spewing such ignorant crap.

I'm not associated with Sun and would rather not have another discussion on whether the SCSL is either a Good Thing or Evil Incarnate, but recently Sun announced the availability of their high performance clustering toolkit as source under the SCSL that has been used by many, many machines in the top 500. Pulling that announcement out of my mailbox, it looks like:

QUOTE ON

Sun is pleased to announce that on November 15th Sun HPC ClusterTools[tm] software was made available through Sun Community Source Licensing (SCSL).

We appreciate the interest you have already expressed in this offering and are excited about the opportunities that the Community Source Licensing model presents. Since our initial June announcement we have completed the work needed to make HPC ClusterTools available through SCSL, and we are very encouraged by the large number of individuals who have registered their interest at our Web site. We look forward to development by the community and wish to extend our thanks to all who will be joining this community effort.

You can access HPC ClusterTools SCSL source code via the Web at:
http://www.sun.com/hpc/communitysource

This site contain a licensing overview, FAQ's, technical information, download information, support information, and related links. This provides an excellent introduction to the Sun HPC ClusterTools SCSL product and should make it easy for you to download, build, and start developing with the product.

Again, we thank you for your interest and look forward to your active participation in the ongoing development of the community.

QUOTE OFF

All that stated, the most useful info is actually on the Technical Description page or on the previous ClusterTools 3.0 documentation..

I'm not associated with Sun and would rather not have another discussion on whether the SCSL is either a Good Thing or Evil Incarnate, but recently Sun announced the availability of their high performance clustering toolkit as source under the SCSL that has been used by many, many machines in the top 500. Pulling that announcement out of my mailbox, it looks like:

QUOTE ON

Sun is pleased to announce that on November 15th Sun HPC ClusterTools[tm] software was made available through Sun Community Source Licensing (SCSL).

We appreciate the interest you have already expressed in this offering and are excited about the opportunities that the Community Source Licensing model presents. Since our initial June announcement we have completed the work needed to make HPC ClusterTools available through SCSL, and we are very encouraged by the large number of individuals who have registered their interest at our Web site. We look forward to development by the community and wish to extend our thanks to all who will be joining this community effort.

You can access HPC ClusterTools SCSL source code via the Web at:
http://www.sun.com/hpc/communitysource

This site contain a licensing overview, FAQ's, technical information, download information, support information, and related links. This provides an excellent introduction to the Sun HPC ClusterTools SCSL product and should make it easy for you to download, build, and start developing with the product.

Again, we thank you for your interest and look forward to your active participation in the ongoing development of the community.

QUOTE OFF

All that stated, the most useful info is actually on the Technical Description page or on the previous ClusterTools 3.0 documentation..

I'm not associated with Sun and would rather not have another discussion on whether the SCSL is either a Good Thing or Evil Incarnate, but recently Sun announced the availability of their high performance clustering toolkit as source under the SCSL that has been used by many, many machines in the top 500. Pulling that announcement out of my mailbox, it looks like:

QUOTE ON

Sun is pleased to announce that on November 15th Sun HPC ClusterTools[tm] software was made available through Sun Community Source Licensing (SCSL).

We appreciate the interest you have already expressed in this offering and are excited about the opportunities that the Community Source Licensing model presents. Since our initial June announcement we have completed the work needed to make HPC ClusterTools available through SCSL, and we are very encouraged by the large number of individuals who have registered their interest at our Web site. We look forward to development by the community and wish to extend our thanks to all who will be joining this community effort.

You can access HPC ClusterTools SCSL source code via the Web at:
http://www.sun.com/hpc/communitysource

This site contain a licensing overview, FAQ's, technical information, download information, support information, and related links. This provides an excellent introduction to the Sun HPC ClusterTools SCSL product and should make it easy for you to download, build, and start developing with the product.

Again, we thank you for your interest and look forward to your active participation in the ongoing development of the community.

QUOTE OFF

All that stated, the most useful info is actually on the Technical Description page or on the previous ClusterTools 3.0 documentation..

See Sun press release. For Toy Story 2, they used 120 E4500 with 14 UltraSparc-II's each - total of 1680 CPUs, along with 4.5terabytes of storage. List price of around $30M I guess, though I presume they got some kind of discount ^-^. btw, one of the requirements was for the render-farm to be pretty compact. Performance Computing magazine have a review of the E4500 here. Pixar used Sun kit for their previous stuff too. If they do a Toy Story 3, by then the UltraSparc-IV should be out, which'll be about 5x faster in FP than current top-end UltraSparc-II's.

...that Penguin Computing still sells AMD systems; Compaq sells Alphas running Linux; Apple still sells PowerPC systems that are Linux-compatible; and that you can still by SPARC stations from Sun or for cheap from many auction houses and hardware resellers.


The Kulturwehrmacht

The best has yet to come because with current technologies and knowledge, we can reach 1.6 GHz. But after that, the chip would melt anyway due to high temperature. 0.1 micron is the limit. Then we could multiply the numbers of processors but it's just a short term solution.

I've heard that HAL computer systems (working for Fujitsu) is working a prediction method to find data before it is calculated to increase speed. A little like the machine O (the Oracle) of Alan Turing.

There is also the MAJC computer of Sun that could be used to build a neuron network (like a machine B) to give more "smart" data processing.

Finally there are also the Quantum Computers that could change everything but that's almost science-fiction

All this information has been mainly taken from an article in LOGIN:

I would recommend a Sun MultiPack with Solstice DiskSuite for management.

I would recommend a Sun MultiPack with Solstice DiskSuite for management.

Definitely use XML over HTML. With XML, you can make up your own tag-set that accurately represents the structure of your documents. It would then be trivial for you to write an XSLT (see http://www.w3c.org/TR/xslt) stylesheet to transform your document into HTML (which has very little structure beyond lists and nested 'div' containers) for delivery, complete with auto-generated TOCs, indices, etc.

Then, if you decide to change your HTML style, you can just re-generate it by changing your stylesheet - without touching your content. It's sort of like generating HTML forms out of content in a database.

In terms of internationalization support, XML documents can contain just about any Unicode character. So basically you can write an XML document in practically any language.

Your XML source can capture things like:

12345-67

Whereas in HTML it would be more like:

12345-67, or at best 12345-67

In HTML, the only way to reproduce the 'build-your-own-vocabulary' capability of XML would be to have your whole document be a sea of div> and elements, with their 'class' attribute set to different values. But processing (and reading) such documents would be a real bitch.

A plethora of XML tools are available and tons more are on the way...

I recommend using James Clark's "Xt" (http://www.jclark.com/xml/xt.html) XSLT engine in conjunction with Sun's "Project X" (http://java.sun.com/xml) XML parser.

James was editor of the XSLT spec, and an outstanding programmer. The Sun XML parser has been shown to be the most conformant, and is quite fast.

Both applications are written in Java.

More XML info:

• http://www.oasis-open.org
• http://www.xml.org
• http://www.xml.com

I notice Sun has just 404ed the top25 bug page without explanation. :)

These are the little bits I know:

• Built-in session management.
• Customizeable session swapping (session info is swapped to the hard drive if session is inactive for a specified time period.
• Modular.
• Cross-platform.
• Compiled.
• All the features of Java, such as inheritance.

--

I found HotJava working quite OK.
It's kind of slow on my machine, but it's very usable.
It renders tables very fast... Just try it on Freshmeat. I just love it :)

Thanks for your time.

There are a lot of officetools around... If U goto sun or corel (WordPerfect) U can get some office warez allready.

The SEC filing specifically states that the delay will be until 2nd half of Sun's fiscal 2000 year. However, Sun's fiscal 2000 year is 1st July 1999 to 30th June 2000. In other words, this means the final product will be out and shipping in volume by the 1st half of calender 2000. The beta will be coming out by the end of this year.

Still, it does mean StarPortal will be 3-6 months later than previously stated...

on a happier note, Sun have had over 1,000,000 StarOffice 5.1 downloads in the last 2 months. See Press release. They have a download counter on their homepage, or go here for a direct link to the image. They also have a download counter for Java 2 SDK, and Solaris - though in the latter case, this is the number of "Free Solaris" orders... not downloads...

The SEC filing specifically states that the delay will be until 2nd half of Sun's fiscal 2000 year. However, Sun's fiscal 2000 year is 1st July 1999 to 30th June 2000. In other words, this means the final product will be out and shipping in volume by the 1st half of calender 2000. The beta will be coming out by the end of this year.

Still, it does mean StarPortal will be 3-6 months later than previously stated...

on a happier note, Sun have had over 1,000,000 StarOffice 5.1 downloads in the last 2 months. See Press release. They have a download counter on their homepage, or go here for a direct link to the image. They also have a download counter for Java 2 SDK, and Solaris - though in the latter case, this is the number of "Free Solaris" orders... not downloads...

The SEC filing specifically states that the delay will be until 2nd half of Sun's fiscal 2000 year. However, Sun's fiscal 2000 year is 1st July 1999 to 30th June 2000. In other words, this means the final product will be out and shipping in volume by the 1st half of calender 2000. The beta will be coming out by the end of this year.

Still, it does mean StarPortal will be 3-6 months later than previously stated...

on a happier note, Sun have had over 1,000,000 StarOffice 5.1 downloads in the last 2 months. See Press release. They have a download counter on their homepage, or go here for a direct link to the image. They also have a download counter for Java 2 SDK, and Solaris - though in the latter case, this is the number of "Free Solaris" orders... not downloads...

...I can't find anything on Specialized's or Sun's websites that mention a word about Sun-purple bikes. Very disapointing. They had a picture of McNealy being presented with a custom-made bike that was sun purple. But, according to the press release, the co-branding ACTUALLY involves putting a logo saying 'Desgined on Sun' on all the bikes and helmets. Oh well. Too good to be true, I guess.

...I can't find anything on Specialized's or Sun's websites that mention a word about Sun-purple bikes. Very disapointing. They had a picture of McNealy being presented with a custom-made bike that was sun purple. But, according to the press release, the co-branding ACTUALLY involves putting a logo saying 'Desgined on Sun' on all the bikes and helmets. Oh well. Too good to be true, I guess.

...I can't find anything on Specialized's or Sun's websites that mention a word about Sun-purple bikes. Very disapointing. They had a picture of McNealy being presented with a custom-made bike that was sun purple. But, according to the press release, the co-branding ACTUALLY involves putting a logo saying 'Desgined on Sun' on all the bikes and helmets. Oh well. Too good to be true, I guess.

I do not need to be inundated with the problems of the netherworld of MS Windows users. Since I run Solaris, the most reliable, scalable OS on the planet, I need not be bothered with such mindless drivel! Out with the post!

Yes, this is old news. The only thing I can find at the moment is on Wired but I remember Judge White saying (on techweb or cnet, probably) he would use it. Read the link: ATT used it, and it would get around those Appeals judges.
I keep hearing everyone say stuff about this case that's just wrong. Another thing is "What's the worst that can happen? Break them up?" NO. Answer: All of the above. Break up the company AND impose behavioral remedies. Joel Klein said "Let me make clear, we are not looking for any financial penalties. We're concerned with competition. This is not a penal action, and we're not going to seek monies.'' So a fine is out of the question. They probably will not send Bill Gates to jail, either. :-( But Microsoft will be subject to lawsuits from other companies, like the one currently filed by Caldera for potenially vast sums of money. And what about the Penal part - you can't let lawbreakers go unpenalized. Maybe another DOJ case?

John

MAJC developers connection

A chat with the lead designer for the MAJC architecture

MAJC 5200 chip press release From MAJC docs page: First MAJC implimentation presentation.

Some comments... Firstly, I don't think it was made clear that the MAJC architecture can execute compiled C/C++ as easily as Java. You'd also probably be using a dynamic compiler like HotSpot rather than a JIT compiler. I think the guy got it wrong about the functional units being data agnostic - the registers definately are though. Still, you can (pretty much) execute 4 of any type of instruction at once - the first block is slightly different compared to 2-4 (which are indentical), though I don't know how. The very interesting STC concept is much easier to do with the Java programming model (because of certain issues with data) compared to C/C++. They don't say how easy it would be to apply STC to C/C++ though - might be impossible in the general case, though possible in some "limited" cases. The MAJC does also have scoreboarding for instructions for dynamic execution times eg loads, though I'm not sure if this applies to things like FP multi/div/sqrt etc. There's a couple of other interesting things the guy missed - the cross-bar data switch and the steaming data ports, for example. Though apart from that, I think it was a pretty decent review.

With regards to the first chip - the MAJC 5200, it's supposed to "tape out" (get first physical implimentation) by the end of the year and sample in Q2 next year. The 5200 has 2 CPU units on chip (with a shared L1 data-cache and a seperate L1 instruction caches), will be made on a .22 micron process, run at 500Mhz and consume 15W.

Here are some "here's how fast it is" stuff from the PR:

• This chip is expected to be able to handle over one hundred voice-over-IP channels while enabling encryption and decompression of the packets over a 10 gigabit-per-second Ethernet connection. This means that a very large number of simultaneous phone conversations could be supported in a small-footprint gateway server device. In the area of image processing, the JPEG 2000 test set is anticipated to run at 78 MB/s while encoding 8-bit sample images. For networked video, two streams of MPEG-2 data representing five Mbps interlaced sequences have the ability to be simultaneously decoded, with additional processor capacity still available. For teleconferencing under the H.263 standard, six decodes and one encode are expected to be handled in real time. In advanced audio applications, an AC-3 decode of 5.1 channels at 384 kps is predicted to use only seven and one-half percent of the capacity of one of the 5200's two processors.

btw, Sun have a 2000 processor array for simulating their UltraSparc-III chip and they do do some pretty accurate simulations, including things like booting Solaris. In the MAJC 5200 PDF/PS file, they also quote some (estimated) speed-ups gained from using the STC technique for the SPECjvm98 suite of programs - they get from 40-60% or so, which I think is very impressive. The MAJC 5200 also has a graphics pre-processor (it's going to be used in Sun's new high-end graphics systems) and they quote some triangle processing figures with different levels of lighting detail. I don't really get what they mean, but they quote from 60 million triangles/s to 90M/s or so, which is in about the same region as the PlayStation2, or about 4x faster than the fastest current mainstream PC graphics card. However, that doesn't mean you can use 60m-90m in real world stuff...

In general, the chip is aimed at "low end" (though for Sun, "low end" equates to less than $100,000 generally) embedded solutions.

MAJC developers connection

A chat with the lead designer for the MAJC architecture

MAJC 5200 chip press release From MAJC docs page: First MAJC implimentation presentation.

Some comments... Firstly, I don't think it was made clear that the MAJC architecture can execute compiled C/C++ as easily as Java. You'd also probably be using a dynamic compiler like HotSpot rather than a JIT compiler. I think the guy got it wrong about the functional units being data agnostic - the registers definately are though. Still, you can (pretty much) execute 4 of any type of instruction at once - the first block is slightly different compared to 2-4 (which are indentical), though I don't know how. The very interesting STC concept is much easier to do with the Java programming model (because of certain issues with data) compared to C/C++. They don't say how easy it would be to apply STC to C/C++ though - might be impossible in the general case, though possible in some "limited" cases. The MAJC does also have scoreboarding for instructions for dynamic execution times eg loads, though I'm not sure if this applies to things like FP multi/div/sqrt etc. There's a couple of other interesting things the guy missed - the cross-bar data switch and the steaming data ports, for example. Though apart from that, I think it was a pretty decent review.

With regards to the first chip - the MAJC 5200, it's supposed to "tape out" (get first physical implimentation) by the end of the year and sample in Q2 next year. The 5200 has 2 CPU units on chip (with a shared L1 data-cache and a seperate L1 instruction caches), will be made on a .22 micron process, run at 500Mhz and consume 15W.

Here are some "here's how fast it is" stuff from the PR:

• This chip is expected to be able to handle over one hundred voice-over-IP channels while enabling encryption and decompression of the packets over a 10 gigabit-per-second Ethernet connection. This means that a very large number of simultaneous phone conversations could be supported in a small-footprint gateway server device. In the area of image processing, the JPEG 2000 test set is anticipated to run at 78 MB/s while encoding 8-bit sample images. For networked video, two streams of MPEG-2 data representing five Mbps interlaced sequences have the ability to be simultaneously decoded, with additional processor capacity still available. For teleconferencing under the H.263 standard, six decodes and one encode are expected to be handled in real time. In advanced audio applications, an AC-3 decode of 5.1 channels at 384 kps is predicted to use only seven and one-half percent of the capacity of one of the 5200's two processors.

btw, Sun have a 2000 processor array for simulating their UltraSparc-III chip and they do do some pretty accurate simulations, including things like booting Solaris. In the MAJC 5200 PDF/PS file, they also quote some (estimated) speed-ups gained from using the STC technique for the SPECjvm98 suite of programs - they get from 40-60% or so, which I think is very impressive. The MAJC 5200 also has a graphics pre-processor (it's going to be used in Sun's new high-end graphics systems) and they quote some triangle processing figures with different levels of lighting detail. I don't really get what they mean, but they quote from 60 million triangles/s to 90M/s or so, which is in about the same region as the PlayStation2, or about 4x faster than the fastest current mainstream PC graphics card. However, that doesn't mean you can use 60m-90m in real world stuff...

In general, the chip is aimed at "low end" (though for Sun, "low end" equates to less than $100,000 generally) embedded solutions.

MAJC developers connection

A chat with the lead designer for the MAJC architecture

MAJC 5200 chip press release From MAJC docs page: First MAJC implimentation presentation.

Some comments... Firstly, I don't think it was made clear that the MAJC architecture can execute compiled C/C++ as easily as Java. You'd also probably be using a dynamic compiler like HotSpot rather than a JIT compiler. I think the guy got it wrong about the functional units being data agnostic - the registers definately are though. Still, you can (pretty much) execute 4 of any type of instruction at once - the first block is slightly different compared to 2-4 (which are indentical), though I don't know how. The very interesting STC concept is much easier to do with the Java programming model (because of certain issues with data) compared to C/C++. They don't say how easy it would be to apply STC to C/C++ though - might be impossible in the general case, though possible in some "limited" cases. The MAJC does also have scoreboarding for instructions for dynamic execution times eg loads, though I'm not sure if this applies to things like FP multi/div/sqrt etc. There's a couple of other interesting things the guy missed - the cross-bar data switch and the steaming data ports, for example. Though apart from that, I think it was a pretty decent review.

With regards to the first chip - the MAJC 5200, it's supposed to "tape out" (get first physical implimentation) by the end of the year and sample in Q2 next year. The 5200 has 2 CPU units on chip (with a shared L1 data-cache and a seperate L1 instruction caches), will be made on a .22 micron process, run at 500Mhz and consume 15W.

Here are some "here's how fast it is" stuff from the PR:

• This chip is expected to be able to handle over one hundred voice-over-IP channels while enabling encryption and decompression of the packets over a 10 gigabit-per-second Ethernet connection. This means that a very large number of simultaneous phone conversations could be supported in a small-footprint gateway server device. In the area of image processing, the JPEG 2000 test set is anticipated to run at 78 MB/s while encoding 8-bit sample images. For networked video, two streams of MPEG-2 data representing five Mbps interlaced sequences have the ability to be simultaneously decoded, with additional processor capacity still available. For teleconferencing under the H.263 standard, six decodes and one encode are expected to be handled in real time. In advanced audio applications, an AC-3 decode of 5.1 channels at 384 kps is predicted to use only seven and one-half percent of the capacity of one of the 5200's two processors.

btw, Sun have a 2000 processor array for simulating their UltraSparc-III chip and they do do some pretty accurate simulations, including things like booting Solaris. In the MAJC 5200 PDF/PS file, they also quote some (estimated) speed-ups gained from using the STC technique for the SPECjvm98 suite of programs - they get from 40-60% or so, which I think is very impressive. The MAJC 5200 also has a graphics pre-processor (it's going to be used in Sun's new high-end graphics systems) and they quote some triangle processing figures with different levels of lighting detail. I don't really get what they mean, but they quote from 60 million triangles/s to 90M/s or so, which is in about the same region as the PlayStation2, or about 4x faster than the fastest current mainstream PC graphics card. However, that doesn't mean you can use 60m-90m in real world stuff...

In general, the chip is aimed at "low end" (though for Sun, "low end" equates to less than $100,000 generally) embedded solutions.

MAJC developers connection

A chat with the lead designer for the MAJC architecture

MAJC 5200 chip press release From MAJC docs page: First MAJC implimentation presentation.

Some comments... Firstly, I don't think it was made clear that the MAJC architecture can execute compiled C/C++ as easily as Java. You'd also probably be using a dynamic compiler like HotSpot rather than a JIT compiler. I think the guy got it wrong about the functional units being data agnostic - the registers definately are though. Still, you can (pretty much) execute 4 of any type of instruction at once - the first block is slightly different compared to 2-4 (which are indentical), though I don't know how. The very interesting STC concept is much easier to do with the Java programming model (because of certain issues with data) compared to C/C++. They don't say how easy it would be to apply STC to C/C++ though - might be impossible in the general case, though possible in some "limited" cases. The MAJC does also have scoreboarding for instructions for dynamic execution times eg loads, though I'm not sure if this applies to things like FP multi/div/sqrt etc. There's a couple of other interesting things the guy missed - the cross-bar data switch and the steaming data ports, for example. Though apart from that, I think it was a pretty decent review.

With regards to the first chip - the MAJC 5200, it's supposed to "tape out" (get first physical implimentation) by the end of the year and sample in Q2 next year. The 5200 has 2 CPU units on chip (with a shared L1 data-cache and a seperate L1 instruction caches), will be made on a .22 micron process, run at 500Mhz and consume 15W.

Here are some "here's how fast it is" stuff from the PR:

• This chip is expected to be able to handle over one hundred voice-over-IP channels while enabling encryption and decompression of the packets over a 10 gigabit-per-second Ethernet connection. This means that a very large number of simultaneous phone conversations could be supported in a small-footprint gateway server device. In the area of image processing, the JPEG 2000 test set is anticipated to run at 78 MB/s while encoding 8-bit sample images. For networked video, two streams of MPEG-2 data representing five Mbps interlaced sequences have the ability to be simultaneously decoded, with additional processor capacity still available. For teleconferencing under the H.263 standard, six decodes and one encode are expected to be handled in real time. In advanced audio applications, an AC-3 decode of 5.1 channels at 384 kps is predicted to use only seven and one-half percent of the capacity of one of the 5200's two processors.

btw, Sun have a 2000 processor array for simulating their UltraSparc-III chip and they do do some pretty accurate simulations, including things like booting Solaris. In the MAJC 5200 PDF/PS file, they also quote some (estimated) speed-ups gained from using the STC technique for the SPECjvm98 suite of programs - they get from 40-60% or so, which I think is very impressive. The MAJC 5200 also has a graphics pre-processor (it's going to be used in Sun's new high-end graphics systems) and they quote some triangle processing figures with different levels of lighting detail. I don't really get what they mean, but they quote from 60 million triangles/s to 90M/s or so, which is in about the same region as the PlayStation2, or about 4x faster than the fastest current mainstream PC graphics card. However, that doesn't mean you can use 60m-90m in real world stuff...

In general, the chip is aimed at "low end" (though for Sun, "low end" equates to less than $100,000 generally) embedded solutions.

MAJC developers connection

A chat with the lead designer for the MAJC architecture

MAJC 5200 chip press release From MAJC docs page: First MAJC implimentation presentation.

Some comments... Firstly, I don't think it was made clear that the MAJC architecture can execute compiled C/C++ as easily as Java. You'd also probably be using a dynamic compiler like HotSpot rather than a JIT compiler. I think the guy got it wrong about the functional units being data agnostic - the registers definately are though. Still, you can (pretty much) execute 4 of any type of instruction at once - the first block is slightly different compared to 2-4 (which are indentical), though I don't know how. The very interesting STC concept is much easier to do with the Java programming model (because of certain issues with data) compared to C/C++. They don't say how easy it would be to apply STC to C/C++ though - might be impossible in the general case, though possible in some "limited" cases. The MAJC does also have scoreboarding for instructions for dynamic execution times eg loads, though I'm not sure if this applies to things like FP multi/div/sqrt etc. There's a couple of other interesting things the guy missed - the cross-bar data switch and the steaming data ports, for example. Though apart from that, I think it was a pretty decent review.

With regards to the first chip - the MAJC 5200, it's supposed to "tape out" (get first physical implimentation) by the end of the year and sample in Q2 next year. The 5200 has 2 CPU units on chip (with a shared L1 data-cache and a seperate L1 instruction caches), will be made on a .22 micron process, run at 500Mhz and consume 15W.

Here are some "here's how fast it is" stuff from the PR:

• This chip is expected to be able to handle over one hundred voice-over-IP channels while enabling encryption and decompression of the packets over a 10 gigabit-per-second Ethernet connection. This means that a very large number of simultaneous phone conversations could be supported in a small-footprint gateway server device. In the area of image processing, the JPEG 2000 test set is anticipated to run at 78 MB/s while encoding 8-bit sample images. For networked video, two streams of MPEG-2 data representing five Mbps interlaced sequences have the ability to be simultaneously decoded, with additional processor capacity still available. For teleconferencing under the H.263 standard, six decodes and one encode are expected to be handled in real time. In advanced audio applications, an AC-3 decode of 5.1 channels at 384 kps is predicted to use only seven and one-half percent of the capacity of one of the 5200's two processors.

btw, Sun have a 2000 processor array for simulating their UltraSparc-III chip and they do do some pretty accurate simulations, including things like booting Solaris. In the MAJC 5200 PDF/PS file, they also quote some (estimated) speed-ups gained from using the STC technique for the SPECjvm98 suite of programs - they get from 40-60% or so, which I think is very impressive. The MAJC 5200 also has a graphics pre-processor (it's going to be used in Sun's new high-end graphics systems) and they quote some triangle processing figures with different levels of lighting detail. I don't really get what they mean, but they quote from 60 million triangles/s to 90M/s or so, which is in about the same region as the PlayStation2, or about 4x faster than the fastest current mainstream PC graphics card. However, that doesn't mean you can use 60m-90m in real world stuff...

In general, the chip is aimed at "low end" (though for Sun, "low end" equates to less than $100,000 generally) embedded solutions.

Bollocks. Data General have been offering 64 CPU Intel based servers for some time now. See http://www.dg.com/avi ion/html/av_25000_enterprise_server.html. 128 CPU and higher versions are due to follow early next year.

Sun HW offers 1000-way SMP capabale systems(in theory at least)

Nope. The most Sun offer is 64-way with the E10000 StarFire. We have four of them here. http://www.sun.com/servers/highe nd/10000/spec.html. Sure, you can go past 64 CPUs with clustering or server farms, etc., but not SMP.

Linux is slowly getting there but it;'s not there "just yet" Let's be sensible....

Agreed. I've even had Linux running on a Sun Ultra Enterprise server, but it certainly doesn't make as good use of it as Solaris does. It'll get there in time, but it's not there yet.

what you seem to be forgetting (or perhaps everybody is) is the need for a networked infrastructure for all these online devices. simply having a network accessible device is good and all as a toy, but it doesn't do it any good unless it can talk to your other networked devices for planning and coordination needs.

presumeably this infrastructure (whether it be jini-like, or hive-like) has to have the ability to manage all the devices on the local network. and once you have a way to adminstrate all the devices, think of an AutoRPM system or something similar to what RedHat now has for registered users -- let your infrastructure automatically update your applicances.

of course, this all assumes you buy upgradeable devices, whether it be in software or with FPGAs or something. i don't understand why anybody would by non-upgradeable devices if they had the choice anyway. in the worst case, if there is a bug in your "coffee machine", let your infrastructure take it off the network until a fix has become available.

networking everything is the way that we are moving. networking coffee machines, or microwaves is an interesting concept as toys, but the more interesting things happen when you network your whole house and have it have all these independant devices work together to form some emergent behavior (oh yeah - this had better not have a bug in it -- think of a house that is out of control and out to get you :)

sorry about rambling..........

They've also promised source for the HotJava Browser. They must have retracted that promise because the last time I went to the HotJava page, there was no mention of SCSL...

Gate's big idea was to make sure his company's software and operating systems were distributed so freely and aggressively they were on every desktop.

Sun is offering StarOffice 5.1 for FREE. Sound familiar?

cheers,

At The Register, there is an interesting article: Judge: Linux can't break Windows monopoly. See also, What will happen if Microsoft finally loses?, Judge's ruling opens way for Caldera Win95 suit, Caldera judge finds MS 'grossly misprepresented' facts.

At Reuters Business News, there is this report: Silicon Valley Cheers Microsoft Ruling, I liked the quote from a Sun lawyer "The aura that surrounded Microsoft as this all powerful, inexorable force that always won has now been significantly diminished,"

On Sun's website, on Friday, they put up this page: Sun Responds to Department of Justice vs Microsoft Case, where they give their own ideas about what to do to MS:

• Microsoft should be prohibited from buying the distribution channels of the future (e.g. cable and wireless) and from buying rather than inventing technologies. Microsoft's unfettered use of a cash hoard created out of monopoly profits is a competition killer;

• The government needs to foster competition in the software industry by assuring that the technical interfaces of Microsoft's monopoly products are open;

• Microsoft must be forbidden from entering into exclusive or preclusive agreements;

• Microsoft must be required to make their pricing policies non- discriminatory and public.

Sorry that the link was filtered out. Here's what I meant: JFC Tutorial

Actually, JDK 1.3's version of the JFC is codenamed Kestrel and is supposed to have some significant changes. See this for more information.

http://java.sun.com/docs/books/tutorial/informatio n/download.html

There's lots of other tutorials there too...

I quote from http://www.sun.com/announcement/findings.html

As attention turns to potential remedies, Sun believes that:

• Microsoft should be prohibited from buying the distribution channels of the future (e.g. cable and wireless) and from buying rather than inventing technologies. Microsoft's unfettered use of a cash hoard created out of monopoly profits is a competition killer;
  (Since we are talking about innovation into better technologies and serving the customer's need, not just engulfing all available distribution channels and competitors, right?)
• The government needs to foster competition in the software industry by assuring that the technical interfaces of Microsoft's monopoly products are open;
  (C'mon, you don't need to keep your protocols and API's private to protect innovation, do you?)
• Microsoft must be forbidden from entering into exclusive or preclusive agreements;
  (...And I'm sure you don't need to bully OEMs into not including your competitor's products, since yours are so much more innovative!)
• Microsoft must be required to make their pricing policies non-discriminatory and public. Sun believes it is important that steps be taken to ensure that one company is not allowed to stifle true competition in an increasingly networked and dynamic industry.
  (No need to manipulate the market and distribution channels with secret backroom deals -- your software is so good it shouldn't need that!)

Here's Sun's equally predictable response

The existence of the MS monopoly has encouraged a lot of good things to happen lately. Sun with Java, IBM with their eBusiness initiatives, Oracle and Linux have risen to the challenge. Of course, who knows how these companies and technologies would have fared in an environment where MS didn't have to be concerned with their business practices?

The increased scrutiny that MS is getting seems to be encouraging even more good things. MS seems committed to doing a lot of their future work using XML, which would be good. It's a good thing when anyone commits to an open standard. I'm satisfied that if MS implements a lot of technology based in XML, that others will be able to easily develop compatible products. XML file formats for Office documents would be a boon to those who wish to import and interoperate, for example. An object protocol based on XML (such as SOAP) would allow competitors to interoperate well.

I'm fairly comfortable with the marketplace today. I think you can, now, compete with MS broadly without too much fear of reprisal. As I said, I do feel that MS needs close scrutiny.

What makes this all so difficult is that Microsoft signed a consent decree having to do with bundling and OEM contracts back in 1995 and then proceeded to continue with business as usual. The law, and respect for the law, are incredibly important and no company should be allowed to continue to ignore their responsibilities without consequence.

I agree with Scott McNealy, no breakup of MS is necessary. I think the ideas that Sun is publishing here are good.

Of course, if MS gets a slap on the wrist (again) and they just continue with their unfair practices, then it's time they get serious consequences. I guess I'm surprised a little that the Judge didn't hold the MS lawyers in contempt for the rigged video demonstration. If MS ignored court orders in the future, there could and perhaps should be criminal penalties to MS executives.

I have reason to believe that MS "gets it" now. MS is making conciliatory sounding statements that would lead you to believe that they are ready for a settlement. Ballmer castigated his people in highly publicized meetings earlier this year for becoming out of touch and not focused on value. We'll have to see if there are real changes in MS corporate behavior now. They need to focus on value and away from how to leverage their money and power into even greater market dominance.

First I must warn all that I have long since dropped out of my business analysis train of thought as I steadily receded back into loving electronics and OS's - *NIX being my favorite.

Businesses and managers are very impressionable folks. In that I mean they truly believe in commercialism. I have worked with government and commercial sector and I have seen the way salesmen can appeal to these goons. With the government it is - well kinda perverse in a way but it works, with business it's easy -- we offer you this and get that.

So let's break off and follow that last statement for a second, we offer this and we are saying you get that.

Now let's apply the late 80's early 90's MS edition of that very same statement:

If you want that you must buy this in order to survive and have that, oh yeah we forgot to mention you have to also buy this other pile of sh*t as well but we can get inot that later . . .

Now come back to the present.

Joe Blow IT Manager, the very same schmuck whose light up when the dancing paperclip "finally does voice recognition" (f*n idiot) sees this. He now says - sh*t, maybe sinking all of that $$$$ into M$ could kick back - - QUICK, what do the other InfoWorld Ad say? Who should I turn to?

Yes this will happen. Even some delusional idiots will actually turn to Apple to solve their "hi end needs". This will be the first fallout.

But hey, all society advancement comes at great pains and cost, this will be one of them. By the same token just as many smart IT Managers will sit back and "see where things are going" and probably apt for a conservative UNIX Company or quite possibly FreeBSD or Slakware (despite it's "freeness" Slakware still boasts the best uptime and stability in distro - not neccessarily kernel).

IT Managers, believe it or not, think very short term as within 2-3 years or so. They are not forward thinkers (well not most of them I am sure the gang at Red Hat might take me up on that) so many saw NT as a good thing for the next year or maybe 2 years. This makes sense when the byline of your job is "keep from gettin' fired".

So, first, businesses will begin to lose confidence in MS, some will stick with them but many (as is already happening) will turn to companies like VALINUX and Sun.

The home computing market gradually lags anyway due to out of pocket expense limits so breaching the MS infiltration will take several years, but I believe as more becomes available and usable, more home users will seek alternatives. They may or may not turn to Linux, that is hard to tell (but I can say, the time is ripe for FreeBSD to make it's move).

Ultimately home users will not change their hbits for a long long time, I think everyone already sees this. I expect a shift to take place in 2-3 years (given the current buying trends).

I mention them in the same context here because I mean the same thing - both GNU / Linux and OSI now will see something that has been slowly and quietly happening for a long time.

Converts

Developers and geeks alike will also lose confidence in the almighty M$. We now will be given an opportunity to bring in and cultivate untapped talent that simply "didn't know any better" but have now awoken. This will most likely happen (in large numbers) after a ruling which is pretty sure to go against M$. It is already happening, as a matter of fact, I wouldn't be surprised if Linuxnewbie.org's mailing list jumps a little next week.

This isn't much different from other breakups as far as business is concerned. Geeks, however, will reap great reward which as all /. readers know is what matters.