Slashdot Mirror

There actually are at least some exceptions. IBM is one of them: http://www.ibm.com/ibm/responsibility/company/mana gement/policies/politics.shtml

disclaimer: I work for IBM, but this comment is my own and not that of my employer.

No shit. 2-issue and in-order requires hand tuned coding. Yes there is a whollop for a "cache miss" (fetching out to main mem) on the SPE's of the Cell processor. But there are ways to code around that. Split the local store up into smaller chunks and fetch data to fill the smaller chunks while the SPE plugs away on the chunks filled with data. That's why the SPE has TWO pipes. One pipe is for memory loads, the other pipe is for data processing.

http://www-306.ibm.com/chips/techlib/techlib.nsf/t echdocs/E815CC047A60914687256FC000734156/$file/ISS CC-07.4-Cell_SPU.PDF

http://research.scea.com/research/html/CellGDC05/1 5.html

http://research.scea.com/research/html/CellGDC05/1 6.html

If you don't split up the local store, you're going to incurr a 500 cycle penalty while waiting for memory. If you split up the local store, you can fetch to half the mem and process on the other half. This amortizes, if not completely masks the cost of main memory access.

Correct me if I'm wrong.

It's up to the developer to optimize their code and ensure that it is being scheduled properly.

I'd love to hear from a developer that is actually doing everything they can at the low level to optimize data flow. What's their experience with keeping the processors fed with data?

On a related note, IBM has, contrary to your claim of "jumping off", just posted an article detailing the changes they will be contributing to Linux to take full advantage of the astonishing horsepower of the Cell chip.

On a related note, IBM has, contrary to your claim of "jumping off", just posted an article detailing the changes they will be contributing to Linux to take full advantage of the astonishing horsepower of the Cell chip.

Since eclipse.org is struggling, there's a screenshot about halfway down this IBM visual editor tutorial.

Anyone who writes in Groovy, Python, ObjectScript, JavaScript, NetRexx, and many others. There's a scripting JSR as well. I don't think the original poster had this specifically in mind though.

Aside from that, you can also take a look at some pervasive-computing stuff. They're big into "multimodal technology" - XHTML+Voice provides a way to speak to the device or see your input on a screen (or, presumably, to view it in an accessibility-enabled web browser). Its key "secret" is that it can do much better voice recognition because you limit its vocabularies and grammars to the task at hand. See also the IBM multimodal web site, you can find these toys online, even!

Oblig. disclaimer: I work for IBM. Heck, I work just down the hall from the Accessibility Lab and the lab formerly known as "Pervasive Computing". But I'm just a little intern, and I'll be gone at the end of the summer. Just make sure none of this stuff is construed as official IBM sanctioned stuff, promises, announcements, anything-like-that, mmmk? You know the drill.

Aside from that, you can also take a look at some pervasive-computing stuff. They're big into "multimodal technology" - XHTML+Voice provides a way to speak to the device or see your input on a screen (or, presumably, to view it in an accessibility-enabled web browser). Its key "secret" is that it can do much better voice recognition because you limit its vocabularies and grammars to the task at hand. See also the IBM multimodal web site, you can find these toys online, even!

Oblig. disclaimer: I work for IBM. Heck, I work just down the hall from the Accessibility Lab and the lab formerly known as "Pervasive Computing". But I'm just a little intern, and I'll be gone at the end of the summer. Just make sure none of this stuff is construed as official IBM sanctioned stuff, promises, announcements, anything-like-that, mmmk? You know the drill.

The internal mirror is here: http://fullmoon.torolab.ibm.com/downloads/

But to back up those hazy remembrances, I just googled it:
http://www.google.com/search?hl=en&c2coff=1&safe=o ff&q=blue+gene+node+power&spell=1
http://pcworld.com/news/article/0,aid,113418,00.as p
http://www-03.ibm.com/technology/power/splash/0321 05bluegene.html

Java compiles to bytecode which is 'not' native. The bytecode is for the java processor(virtual machine).

Python and Perl seem fast on startup because they don't have to compile to bytecode.

Native code is what c or c++ compilers create.

http://www-106.ibm.com/developerworks/java/library /j-native.html?loc=j

Your other point about GCJ is what I was saying.

I recently discovered some very interesting things about wireless keyboards:

This used to be IBM's wireless keyboard version. Now Lenovo's.
http://www-131.ibm.com/webapp/wcs/stores/servlet/P roductDisplay?catalogId=-840&langId=-1&partNumber= 73P4067&storeId=10000001
Here is a small picture (sorry, finding large pictures of this keyboard is HARD. )
http://www-131.ibm.com/webapp/wcs/stores/servlet/C ategoryDisplay?catalogId=-840&storeId=10000001&lan gId=-1&dualCurrId=1000073&categoryId=2581898

This keyboard has a range of 10 meters ( 30' ). It's neat and cool looking, the same price as those 8' range keyboards and the 7 top keys are fully programmable ( including macro commands )

And as an unexpected bonus, my wireless 8' logitech mouse now works at 20' with both bases plugged.
http://www.logitech.com/index.cfm/products/details /US/EN,CRID=2135,CONTENTID=9340

With two mice and 1 keyboard... this DX1 sure sound nice. But for half the price I got almost as good were functionnality is concerned. ( not including the logitech mouse... )

I recently discovered some very interesting things about wireless keyboards:

This used to be IBM's wireless keyboard version. Now Lenovo's.
http://www-131.ibm.com/webapp/wcs/stores/servlet/P roductDisplay?catalogId=-840&langId=-1&partNumber= 73P4067&storeId=10000001
Here is a small picture (sorry, finding large pictures of this keyboard is HARD. )
http://www-131.ibm.com/webapp/wcs/stores/servlet/C ategoryDisplay?catalogId=-840&storeId=10000001&lan gId=-1&dualCurrId=1000073&categoryId=2581898

This keyboard has a range of 10 meters ( 30' ). It's neat and cool looking, the same price as those 8' range keyboards and the 7 top keys are fully programmable ( including macro commands )

And as an unexpected bonus, my wireless 8' logitech mouse now works at 20' with both bases plugged.
http://www.logitech.com/index.cfm/products/details /US/EN,CRID=2135,CONTENTID=9340

With two mice and 1 keyboard... this DX1 sure sound nice. But for half the price I got almost as good were functionnality is concerned. ( not including the logitech mouse... )

How many recent patents have you read? IBM alone was granted over 3000 last year...

Slashdot only ever reports on the patents that it sees as beinog frivolous, and even then that's generally only judged from the abstract, which is a dangerous thing to do.

Depends on the threat level.

Nope. Well, unless you think the attacker may be able to defeat modern ciphers. If your enemy is that capable, though, just give up now.

If there's an encryption scheme, there's a key. And depending on who's looking for it, they may resort to coercive measures to get that key.

This is no different from someone resorting to coercive measures to acquire your non-destroyed hard drive.

The reason that encrypting the data is better, from a security perspective, is that it's *easier* to destroy the key than it is to destroy all of the data. If you want really high security, use a strong master key that is generated inside a tamper-reactive device which will (a) never give up the key and (b) destroy it instantly if anyone attempts to open the device, or upon command from a cell-phone call, or if an attached GPS receiver either goes away or reports the wrong location... whatever. Any idea you can come up with for triggering the drive to self-destruct you can also use to destroy the key. And the key goes from fully operational to fully zeroed in microseconds, unlike the mist-in-the-drive, which takes time to really wipe all the data.

Further, with crypto-based security, you can require the presence of multiple keys. For example, you could use a high security FIPS 140-2 level 4 crypto processor like an IBM 4758, which stays inside the computer with the data, plus a key stored in a smart card, plus a (shorter) key that you memorize. If all of the elements are required to successfully decrypt any of the data, then destroying any one of those elements is sufficient to effectively destroy all of the data.

Not only that, but an IBM 4758 can be had for about $1500, and it'll encrypt data for as many drives as you want. So it's not only more secure, it's cheaper, too.

At Almaden they have a zillion Linux boxes spidering the web building data mining databases. They then sell the ability to mine to companies for marketing research. For example, conceptually mapping Slashdot, you find that users who post on ./ also frequent OSNews, Groklaw, and pr0n.

Uh, wait... Hmm. Maybe that isn't such a good thing...

My university had one of these...

Check it out: http://www-1.ibm.com/press/PressServletForm.wss?Me nuChoice=pressreleases&TemplateName=ShowPressRelea seTemplate&SelectString=t1.docunid=930&TableName=D ataheadApplicationClass&SESSIONKEY=any&WindowTitle =Press+Release

IBM already offers this, but I don't think you could afford to put one in your closet...

http://www-1.ibm.com/servers/eserver/zseries/os/li nux/

Linux has played a large role on IBM mainframes since 2000. Many large corporations who own mainframes are now running dozens or even hundreds of Linux virtual machines under z/VM.

http://www-1.ibm.com/servers/eserver/zseries/os/li nux/

Although the hardware by isn't cheap and certainly wouldn't run very well in a closet. :)

HP donates in some way to about 55 projects, but only a few of them are linux specific. IBM on the other hand contributes to at least 162 projects, 44 of them specific to Linux. So IBM's Linux specific contributions alone number nearly as much as ALL those of HP put together. When you count in the projects that IBM works on that are not Linux specific, they stand head and shoulders ahead of HP in the most important measure: code.

Big Blue has spent R&D money thinking about this for us :)

http://www.alphaworks.ibm.com/tech/atomik is a layout for tapping. I used it on my m105, and it's pretty good. Definitely better than graffiti if you're standing still and don,t haveto pay attention to your surroundings, and somewhat better than qwerty or alphabetic (I used paper overlays on the graffiti area).

http://www.alphaworks.ibm.com/tech/sharktext uses the same layout in an ingenious way, where you trace a line between the letters of the word. You quickly remember the shape of frequent words, and it works surprisingly well (I'm using the java demo almost as often as the built-in TIP on my tablet PC).

Big Blue has spent R&D money thinking about this for us :)

http://www.alphaworks.ibm.com/tech/atomik is a layout for tapping. I used it on my m105, and it's pretty good. Definitely better than graffiti if you're standing still and don,t haveto pay attention to your surroundings, and somewhat better than qwerty or alphabetic (I used paper overlays on the graffiti area).

http://www.alphaworks.ibm.com/tech/sharktext uses the same layout in an ingenious way, where you trace a line between the letters of the word. You quickly remember the shape of frequent words, and it works surprisingly well (I'm using the java demo almost as often as the built-in TIP on my tablet PC).

IBM's prices aren't as ridiculous as you make out either. Here's an example. It's expensive at around $5,600, but over $10,000?

I'm not going to comment on SGI or HP, as they live in their own worlds. But anyway, the point is you're wrong. Totally wrong. I hope you're embarassed, I mean, I'm embarassed for you. My advice? Change your name. Seriously. That's about the only way you're going to get your reputation back.

look at the memory prices!

Total memory [6] [Help me decide]
Note: Total memory includes the base memory that comes with the system.
512 MB included in base system
768 MB (2 NonParity DDR2 SDRAM SoDIMM PC4200) [add $125.00]
1 GB (2 NonParity DDR2 SDRAM SoDIMM PC4200) [add $250.00]
1.5 GB (2 NonParity DDR2 SDRAM SoDIMM PC4200) [add $750.00]

http://www-131.ibm.com/webapp/wcs/stores/servlet/X ICLaunchConfigCmd?&configServer=eCimSERVER1&pageTy pe=6&site_type=public&base=18695CU&storeId=1000000 1&catalogId=-840&cntry=840&lang=en_US&ratetype=D&d ualCurrId=1000073&launch_type=newconfig&real_time_ c1=USD&real_time_t1=false

$750 for 1.5 GB ram!

Crucial quotes 1 GB sticks at $148.39

TFA isn't a review -- it's just a rehash of the IBM announcement. Plus it's on a site with zero Slashdot compatibility. Wny not link the IBM/Lenovo page directly? Probably because the submitter wanted to promote his site.

Power refers to the Power line of chips which includes PowerPC Power4 and Power5.

Not sure where you got that number, but the cheapest openpower machine I could find is the base 710, for $3500 with a 1.6Ghz POWER5/32MB L2 cache, 74GB of UltraSCSI and 1GB of RAM. I challenge you to find a Intel, AMD or even PowerPC (Apple G5-type) based box for that price that can match it performance wise.

Hate the name. Running linux doesn't make the architecture "Open". For an affordable power based workstation look here. It runs a few different distros of linux and other os's and more are in the works.

I don't really agree with all the hype around IBM's open source moves. Seems like a lot of what they are doing is using open source as a way to get salespeople in the door to sell proprietary IBM solutions. That's not necessarily a bad thing depending on how it's done. Look at some of the Apache projects, things like Jetty where IBM was a big supporter until they decided it was good enough to fork it internally for WebSphere Portal. That's not so bad, the license allows it but it's still a bit messed up in my opinion.

But other things like Derby and Geronimo are different all together. Take a look at the IBM Cloudscape pages. Cloudscape is not open source. It's "open source-based". They released the code as the Apache Derby project, but Cloudscape is not an open source project. The download link says "trials and betas". It may be free now but it has the potential of not being free when IBM senses it can charge for more than just support. Why would IBM release this? It's not because there weren't other open source java based RDBMS's out there. Hypersonic and Mckoi were already quite popular in that space and there are more. The difference is, Cloudscape has been designed to be a stepping stone to DB2. The java rdbms's have a niche market. One segment of which is users that want an easy portable system to develop on with little administrative requirements. These types generally outgrow the capabilities of such database systems and the user needs to migrate. If you're already using Cloudscape, the logical migration path is DB2. Which definately isn't free.

Also look at their recent acquisition of Gluecode. This gives them the developers of the Geronimo project. It's going to turn into Websphere-Lite. Same type of deal with Cloudscape. Both of these will be at the expense of open source software that is trying to be real open source, not just a carrot on a stick. Hopefully people see this and projects like Jonas and JBoss continue to grow.

I feel IBM is abusing the phrase "Open Source" in it's marketting where it's not appropriate. Like a couple of days ago when they announced they were moving to a more open source type of development for all their projects. They didn't say they were open sourcing all their projects, just that they will be using open source methodologies within IBM. I guess you could call it the Bazaar in the Cathedral. Yet so many mistakenly took the news to mean that ibm was doing more with open source and releasing more products via open source.

To me it looks like IBM is turning Open Source into shareware. They have their free versions to tinker around with but when you're ready for the real thing, a non crippled version, you have to pay for it. From Jonathan Schwartz's Weblog we have this quote:

"Jonas (Red Hat's app server) is just a toy, it's just for the low end" said IBM's exec at the Smith Barney Tech Conference I just attended in NYC.

A similar message is presented (though a little less harshly) when IBM compares AIX to Linix

Also look at Eclipse. It's used as an IDE but in many respects it is crippled in that regard, it's real purpose is as a tools platform. To get real productivity out of it you need to purchase plug-ins or upgrade to professional tools based on Eclipse like WebSphere Studio. The Eclipse Foundation is working on a number of open sourc

The real power of the open power concept is that IBM is opening the design of the entire architecture. Has anyone else ever done that? Nothing on this scale that I know about.

The article is about a particular part of this open process - allowing developers free access to an OpenPower 710 server. Have you looked at these things? The price is hard to beat for the performance.

By the way - IBM does sell AIX 5.3 for POWER, just not for the "OpenPower" series of servers.

The real power of the open power concept is that IBM is opening the design of the entire architecture. Has anyone else ever done that? Nothing on this scale that I know about.

The article is about a particular part of this open process - allowing developers free access to an OpenPower 710 server. Have you looked at these things? The price is hard to beat for the performance.

By the way - IBM does sell AIX 5.3 for POWER, just not for the "OpenPower" series of servers.

iseries are the intel based boxes

You can buy power from... IBM. And it's not cheap. And it doesn't run AIX, only Linux. Sort of. Many applications require some porting love, as per the bounties on http://www.linuxonpower.com/

I generally like what IBM does, and use their x86 servers, storage, and software.

But "Open" is pushing it here.

I'd never be able to justify a recommendation to buy Open Power, that is, unless the sales guy left a flashy car in my parking spot...

Jonathan Schwartz (Sun CTO) had it right when he noted that that was as silly as them shipping Open Sparc boxes. Mind you, there are Fujitsu SPARC64 chips, and OEM sparc-based system builders.

Of course, IBM is just loving Solaris, particularly Solaris 10. Some assistance in your Solaris to Linux on Power migration? http://www-128.ibm.com/developerworks/linux/librar y/l-pow-portsolaris/
(Though it is a well written piece - good quick guide to Linux and Solaris system calls, signals.)

for example, there is no real reason why IBM shouldn't be able to build a system roughly a thousand times as fast as the BG/W system if someone paid the necessary 40 billion dollars

For example, somebody like DARPA?

The missing link: TrackPoint Mouse

This has absolutely no tactile feedback for the amount of scrolling. I don't find this to be a problem when scrolling in ordinary applications. There is almost never any need to scroll text with total precision. It's an advantage sometimes because if I need to scroll a lot I just need to keep it pushed in that direction. It's is however useless for selecting weapons in games.

IBM Scrollpoint It's only $21, too.

IBM's Linux Watch enhanced shell with Bluetooth

Some may find this article informative.. http://www-106.ibm.com/developerworks/library/l-em bdev.html
It discusses the various pros and cons of file systems, graphic systems, etc.

Link to Blue Gene

--
So who is hotter? Ali or Ali's Sister?

4758 Secure Coprocessor Driver for Linux
This project is a Linux device driver for the IBM 4758 PCI Cryptographic Coprocessor, which is a tamper-sensing and responding, programmable PCI card. It provides a highly secure subsystem in which data processing and cryptography can be performed.

ATM on Linux
ATM support for Linux is currently in pre-alpha stage. There is an experimental release, which supports raw ATM connections (PVCs and SVCs), IP over ATM, LAN emulation, MPOA, Arequipa, and some other goodies.

Abstract Machine Test Utility (AMTU) for Linux
Abstract Machine Test Utility (AMTU) is an administrative utility that checks whether the underlying protection mechanism of the hardware is being enforced. This is a requirement of the Controlled Access Protection Profile (CAPP) FTP_AMT.1.

Ananas Project: Summary
This is the source for Working XML, a column on developerWorks with companion project code that demonstrates the evolution of full-fledged XML applications. This is distributed under the artistic license.

Apache HTTP Server
The Apache project develops and maintains an open-source HTTP server for various modern desktop and server operating systems.

BlueHoc simulator
BlueHoc is a tool that predicts the performance of Bluetooth wireless hardware technologies. BlueHoc simulates the baseband and link layers of the Bluetooth specification.

COIN (Common Optimization INterface)
Developers can use Common Optimization INterface (COIN) to build optimization solutions. IBM mathematical optimization researchers opened the code they use in finding the optimal allocation of limited resources. The code has many applications in a variety of industries.

Channel Bonding
The Channel Bonding project works on methods to join multiple networks on Linux into a single logical network with higher bandwidth. The project team works with the Beowulf Ethernet Channel Bonding project, where bonding work began.

Consensus prototype
Consensus is a joint European project carried out by six companies. The project is partially funded by the European Commission. The project goal is to provide technology to support single-authoring for mobile devices. developerWorks hosts the open source implementation developed by the Consortium. Detailed information about the project is at the Consensus Project home page (http://www.consensus-online.org./

Content Query System (CQS) Project: Summary
Content Query System (CQS). CQS is a distributed peer-to-peer query system for the purpose of discovering content or data. XML messages are passed between systems and query "engines" are used to access the data that is being made available on the system.

Crypto Accelerator Driver
Device Driver Support for the IBM eServer Cryptographic Accelerator.

Crypto Interface Library
Generalized Interface library for the IBM eServer Cryptographic Accelerator Device Driver. Note, this is a low level api for the Specified adapter, it is not intended to be an interface which is written to by applications. Applications should use the openCryptoki PKCS#11 api for interfacing to the token.

Dynamic Probe Class Library (DPCL)
DPCL is an object-based C++ class library that allows tool developers and sophisticated tool users to build parallel and serial tools using a technology called dynamic instrumentation.

Embedded IBM PowerPC 4xx Linux Support
This project contains packages which enable add

First of all, IBM and HP are two of Sun's biggest competitors. It's not surprising to hear them talk down sun. Sun was started by a couple of true geeks, an mba and a guy with a lot of money and vision. They were always very technology focused and anti microsoft. They worked with open technologies before someone decided to put a capitalize O and S in open source. IBM has a shady reputation, not a lot has changed. IBM will say anything to get their products in the door. I've seen it first hand. This is one of my favorite IBM /Sun stories from a while back.

The Sun/MS thing is a recent occurence. Before that, Sun was doing it's own thing competing against Microsoft. IBM and HP were MS partners the whole time. Who do you think has benefitted more from their relationship with MS, Sun or IBM and HP?

The reason this whole MS/Sun deal came about was because MS was abusing Sun's license on Java. Sun always seemed to be about open standards and level technological playing fields. Their chips are open, they built their OS on open source, they released a lot of open source and published a lot of information to allow compatability and competition. Look at what they're doing with Java. They're trying to keep the process open to all with the JCP so that one vender can't wrestle control. If Sun was as underhanded as you think, they'd at least be the number one or number two java server vendor. They're not. IBM's been trying to get more control over Java for years, not as bad as MS though, but it is a situation that needs to be watched. IBM has recently been championing linux but other than that, how have they supported openness in the past? Have a look at the Compuware suite which was settled for $400 million a couple of months ago.

MS tried to screw that over and take control of the Java market. Sun stood up against MS. Hell, they stood up for all developers that MS has tried to screw. Most of the times MS succeeds. Sun and Oracle really pushed for the anti trust cases here and in Europe. They put a lot of effort into the fight. It really was "Mankind vs Microsoft", not Sun vs Microsoft. MS had screwed over a lot of smaller software companies and now they were getting together to put an end to it, with Sun leading the pack. Oracle was a big player too but none of them were as anti-microsoft as Sun. So after a long, world-wide legal battle, MS loses and gets declared a monopoly. Woohoo! Ok, hold your horses because a new team just came into the Justice Dept and things don't look as rosey as they did before. On top of that IBM and HP have been nipping away at Sun's market share with their own Unix lines as well as using Linux. Now the open source community is starting to boo and hiss at Sun too. The main reason was because Sun liked Solaris better than Linux. Some of Sun's managers might have been a bit colorful but in interviews with developers, including Bill Joy you heard good things about Linux. Sun really didn't say anything about linux that was different from what IBM said about Linux vs AIX back then. What else did they say? Oh yeah, they'd indemnify their customers that use Linux. How is that bad and OSRM not?

With the justice dept dropping the ball, the community that should have been more supportive of Sun being against them and MS having made inroads into Sun's server turf Sun decides to settle. The settlement includes a big fee for infringment as well as an agreement to interoperate better. Sun's customers have been wanting this for a long time. They use Sun and they use Windows. They want them to work together. Not only that, but MS now has to be more cooperative in general. MS has also taken the initiative to end other lawsuits

What kind of geeks do we have here?? :) IBM has published a line of hands-on books for many years that machine room mole-people live and die by.. the IBM Redbooks series. Go to http://www.redbooks.ibm.com/ and click Drafts in the left Navigation pane. There are currently 39 books in progress, all from multiple authors who are actively seeking feedback.

How do you figure that? Maybe it used to be true, and it is certainly one of NetBSD's goals, but it's simply not true that NetBSD runs on more systems than Linux. NetBSD gives this impression by listing multiple "ports" for a single CPU architecture.

In fact, NetBSD supports 17 different types of CPUs, some of which are just variations of the other CPUs. It's difficult to find a complete list, but Linux supports at least 22 different system architectures according to this article, and many more of them are useful than the NetBSD ports. Not to mention the much wider variety of peripherals and interface cards that Linux supports than any of the BSDs support.

You're living in a different reality than the rest of us, friend. There are many, many user applications out there that work only on Linux, some of which will never be ported to BSD because they are commercial products. Like Maya, for instance, the software that is used for most computer animation today. Even some open source software runs so poorly on BSD that it's not worth using -- like MySQL. The fact is that even if these problems are mostly because of Linux's greater popularity and not technical, Linux is much better as a general purpose OS.

I agree, except for the BTX example. I think BTX is being adopted (gradually) by the mainstream. We can't just expect them to drop all of their proven, mature ATX designs right away for newly designed BTX cases. For mainstream BTX that's already been released, see:

• Gateway 9000 series
• IBM ThinkCentre A51p
• Shuttle i Series
• Dell Dimension 9100 and 5000
• Dell Optiplex SX280, GX280, and GX620

Allow me to refute your alternative possible explanation with *facts*. :)

[snip netcraft]


Ha ha. Very funny. You know, it has been nearly a decade since you needed a really good OS to serve websites, right?

No, if you want to be a Linux contender, you have to do this, this, this, this, this, this, this, this, etc.

I don't see the word FreeBSD anywhere, do you? Did you think maybe SGI is "obscuring" FreeBSD because it runs on their 512 CPU SSI Servers so much better than Linux? Or the alternative explanation is that FreeBSD (even the current 6 branch) barely scales to 4 CPUs.

Oh, and don't get me started on clueless idiots.

There are already some simple proposals that go a long way to solving the man in the middle issue already, without resorting to grossly inefficient schemes such as yours (or HTTPS). The problem is in getting them adopted.

My advice, IBM Lotus Domino products. It's the most powerful, flexible, and comprehensive tool out there. Technically, it's a software development platform for collaboration and messaging. But there's a subset of tools called "Quickplace" which includes out-of-the-box features that I think would be perfect for you.

From the IBM Quickplace site:
* Provides anytime, anywhere access to collective knowledge, information-sharing, tasks and team calendar events whether on-line or disconnected.
* Seamlessly establishes a working community with a sense of accountability, whether team members are centralized or geographically dispersed.
* Increases team productivity and efficiency by virtualizing asynchronous collaboration processes, and optionally integrates them with real time.
* Increases responsiveness among colleagues, customers, business partners and suppliers by facilitating instant formation of working teams -- whether team members reside within or beyond the organization.
* Facilitates faster, collective decision making by centralizing timely and accurate information, and granting all team members equal opportunity to review and react.
* Helps make your teams more productive and self-sufficient through easy, instant assembly of collaborative applications using team space templates -- with just a browser.


I've been using it for years, and I still haven't found anything I can't get it to do. Note that it's not for transactional systems (airline ticket systems) and it's not a relational database system (which is a plus!), but it's just perfect for what you described...

My email is shown here at Slashdot (thank you SpamCop), so please feel free to email any questions you might have and I'd be happy to help. I'm not a consultant and I'm busy anyway, so I can't build you a system or anything (this is not a pitch:), but I would be happy to field any questions you have (I have about 10 years experience with Domino).

Don't miss the developerWorks site for Domino (plenty of info), and feel free to ask questions in the forums (it's a friendly group!).

Yes, there is a system which can add capacity at will, without a reboot or other disruption. Without even uncrating a box, in fact.
http://www.ibm.com/eserver/zseries/linux

Bugger... I meant to add the URL where I found the quote as it's a good read...

Unix Utilities Part 4

I'd say Sony would get a fair chunk of profit too, seeing that they developed Cell in partnership with IBM and Toshiba.