Slashdot Mirror

Eh? What's this then?
IBM XL Fortran (PowerPC)

No, they won't. At all. Sony does provide intricate DMA libraries and have several methodologies they "suggest". But this will not automate the development of code or execution of code in any way.
Actually, yes, it does. However, as the previous AC pointed out it is somewhat slow. http://www-128.ibm.com/developerworks/power/librar y/pa-cellstartsim/

This is partially correct. "Engines", or at least middleware is a pretty big deal these days. However, approximately 95% of games out there aren't first person shooters, and there aren't many viable engines which aren't fps engines. But the bigger issue here is that the "engine" code traditionally is only a small volume of the overall code. And those bits and pieces are the ones which will actually be fairly straightforward to move to the individual units (character deformation, misc vertex operations, physics).
This makes no sense. First of all, there is nothing to suggest that Unreal Engine 3 is limited to FPS games. For example Bioware is using it for Mass Effect, an RPG. Second, the engine code is a small volume of game code, it's also where 90% of processing time is spent. Anything else can be handled on the PPE.

Probably :) Regardless it's a reasonable assumption that the two will be pretty same - they're from the Big Two. In fact, I would go so far as to say if one or the other had, say, 2x-4x times the pixel throughput, that would still be considered comparable, in terms of realizable performance.
So by this logic you must believe that PS2, Gamecube, and Xbox had identical graphical performance.

ThinkPlus USB Travel Keyboard with UltraNav http://www-131.ibm.com/webapp/wcs/stores/servlet/P roductDisplay?catalogId=-840&storeId=10000001&lang Id=-1&dualCurrId=1000073&categoryId=2581898&produc tId=8717673 This is basically the ThinkPad keyboard. It's great--pointing stick and trackpad.

I like the IBM Trackpoint SpaceSaver. In addition to being a nice compact keyboard, the Trackpoint is great for people who don't like to reach for the mouse all the time.

3.3. MFC DMA Commands
This section describes functions that implement the various MFC DMA commands. See the Cell Broadband Engine Architecture for a description of the DMA commands, including restrictions on the size of the supported operations. MFC DMA command mnemonics are listed in Table 3-98.
 
Table 3-98: MFC DMA Command Mnemonics1
Mnemonic Opcode Command
MFC_PUT_CMD 0x0020 put
MFC_PUTB_CMD 0x0021 putb
MFC_PUTF_CMD 0x0022 putf
MFC_GET_CMD 0x0040 get
MFC_GETB_CMD 0x0041 getb
MFC_GETF_CMD 0x0042 getf
 
1 MFC command enumerants are defined in spu_mfcio.h.
mfc_put: move data from local storage to effective address
(void) mfc_put(volatile void *ls, uint64_t ea, uint32_t size, uint32_t tag,
uint32_t tid, uint32_t rid)
 
Data is moved from local storage to system memory. The arguments to this function correspond to the arguments of the spu_mfcdma64 command: ls is the local-storage address, ea is the effective address in system memory, size is the DMA transfer size, tag is the DMA tag, tid is the transfer class identifier, and rid is the replacement class identifier.
 
Implementation:
spu_mfcdma64(ls, mfc_ea2h(ea), mfc_ea2l(ea), size, tag, ((tid<<24)|(rid<<16)|MFC_PUT_CMD))

Apparently you were caught in a class loading problem you didn't understand. You need to have a look at the article series Demystifying class loading problems, on the IBM DevelopersWorks site.

More, and more recent, information can be found at another IBM site, developerWorks, at http://www.ibm.com/developerworks/power/cell. You can even download (for free) your very own Cell simulator, tool chain, and Linux from there. So go test the claims yourself...

Sony and IBM clearly have continued to contribute to GCC for Power. Cell is based on Power and the GCC compiler (and a whole tool chain) for Cell is available from http://www.ibm.com/developerworks/power/cell
(also note that URL -- /power/cell -- gives away IBM's view on the matter)

Could we all just remember what IBM had in mind when they designed Cell? If you have a read of the Introduction [Pdf warning], you can see they identified the primary bottlenecks to performance, back in 2000, one of the most important problems being memory latency. Now, if you've done some work with assembler, you should know that every time you touch main memory, you loose about 20-30 clock cycles through your memory's low speed. If you want an example, I have a 3GHz computer, but the memory goes at 400Mhz. Just think how much time it must spend waiting for that memory?

Cell counters this problem by using SIMD in combination with what they call "Local Storage". Instead of having to wait for every single memory transfer, threads can read blocks of memory into storage actually on the SPE, process it, and then read it back. All with a couple of instructions, and execution continues even while the memory is been read/written.

The closest that present-day multi-processor computers can get to that is by caching the data. However, that still means that a cache miss will halt execution for many cycles, and each processor / core has to constantly check what other processors / cores have in their caches, ocassionally invalidating them.

What this all adds up to, is a level of efficiency that hasn't been seen before. However, I don't think it's gonna be anyway near "movie quality" graphics, you'd need a farm of Cells for that.

http://www-128.ibm.com/developerworks/power/librar y/pa-fpfxbox/?ca=dgr-lnxw09XBoxDesign

If you want to read more about the CELL heres a link for you...
http://www.research.ibm.com/cell/home.html

More information about the Cell processor directly from the source : The Cell project at IBM Research

I really should stop replying to my own postings... but leak detection tools are available for UNIX... (or at least for some platforms)

Why not check out Ruby On Rails?

Websites
RubyOnRails - http://www.rubyonrails.org/

Ruby general
http://www.ruby-lang.org/

Programming Ruby book - 1st edition online.
http://www.rubycentral.com/book/

Ruby Code and Style Online Mag
http://www.artima.com/rubycs/index.html

Small article on how to program in ruby.
http://www-128.ibm.com/developerworks/linux/librar y/l-ruby1.html

RubyOnRails vs Java for web development efforts
http://www.relevancellc.com/blogs/?p=92#comments

Another comparing Java and Ruby for Web Efforts
http://article.gmane.org/gmane.comp.lang.ruby.rail s/24863

No way, no how. The cheap solution is always going to leave something out - quality.

No way, no how. The cheap solution is always going to leave something out - quality.

A quick search on Google would suggests that there is increasing interest in this field. How aboutIBM , as well as a start up company called D-Wave Systems located in Vancouver, for a start.

As for my two cents, don't bet on an up-and-comer quantum-computer-making-business "knock them [the processor giants] them of their perch". The article (in addition to previous stories) doesn't predict a quantum computer that you'll be able to buy off the shelf and use on your desktop. Perhaps a look at the current prospects for implementations of quantum computers, and a miniscule amount of common sense would convince you of this unliklihood.

That's quite true: Haskell's notion of pattern matching is much more powerful and extensible than mere regular expressions.

Jim Clark's Haskell implementation of the derivative algorithm for validating Relax NG is a wonderful example of how powerful, elegant and concise Haskell is. Relax NG is all about tree structured regular expressions over "hedges" (mixed trees of XML elements and text). It's based on the same automata theory as regular expressions, extended to describe "hedges" (XML documents).

An Algorithm for RELAX NG Validation By James Clark, January 07, 2001. Author's note to XML-DEV: 'I have written a paper describing one possible algorithm for implementing RELAX NG validation. This is the algorithm used by Jing, which I believe has also been adopted by MSV... If you try to use this to implement RELAX NG and something isn't clear, let me know and I'll try to improve the description.' From the introduction: "This document describes an algorithm for validating an XML document against a RELAX NG schema. This algorithm is based on the idea of what's called a derivative (sometimes called a residual). It is not the only possible algorithm for RELAX NG validation. This document does not describe any algorithms for transforming a RELAX NG schema into simplified form, nor for determining whether a RELAX NG schema is correct. We use Haskell to describe the algorithm. Do not worry if you don't know Haskell; we use only a tiny subset which should be easily understandable." Jing is a validator for RELAX NG implemented in Java;

Another interesting Haskell library for XML is HaXml:

Consider Haskell in lieu of DOM, SAX, or XSLT for processing XML data. The library HaXml creates representations of XML documents as native recursive data structures in the functional language Haskell. HaXml brings with it a set of powerful higher order functions for operating on these "datafied" XML documents. Many of the HaXml techniques are far more elegant, compact, and powerful than the ones found in familiar techniques like DOM, SAX, or XSLT. Code samples demonstrate the techniques.

Here's some more stuff about Relax NG, comparing the Haskell implementation to the Java implementation (jing): Maximizing Composability and Relax NG Trivia:

Here's some interesting stuff about the design and development of Relax NG:

James Clark wrote about maximizing composability:

First, a little digression. In general, I have made it a design principle in TREX to maximize "composability". It's a little bit hard to describe. The idea is that a language provides a number of different kinds of atomic thing, and a number different ways to compose new things out of other things. Maximizing composability means minimizing restrictions on which ways to compose things can be applied to which kinds of thing. Maximizing composability tends to improve the ratio between functionality on the one hand and simplicity/ease of use/ease of learning on the other.

Clark describes the derivative algorithm's lazy approach to automaton construction:

I don't agree that <interleave> makes automation-based implementations impossible; it just means you have to construct automatons lazily. (In fact, you can view the

I feel your pain about bugs in libraries that you must use without the source code. I had an arrangement like that for nearly two years with extremely buggy code. Just relinking the static library with changes to my code, changing where in memory the library would reside, would often cause huge problems. Let's just say I got really good at debugging in assembly with gdb. I got where I could call them up and say something like, "you have some code at the end of function foo that looks like 'a[2] = b', but a was never allocated." They'd always reply with something like, "Yes it is ... oh wait..."

If I had to guess, I'd say most likely his problem is caused by fence-post type errors in dynamically allocated arrays (malloc and friends). But on the off chance he's got problems with buffer overruns caused by user input (gets, etc.) there's also stack protectors like ProPolice in addition to libsafe.

Using a complete hardened Linux distro is not necessary for "normal" development work, but it may be a good idea to verify that your application actually works there too.

In addition to the run-time checkers you can also look for static checkers like Splint. It can provide you with some extra information about potential problems that only occurs under certain conditions that maybe aren't met during runtime.

If the effort of trying to track down stack corruption is worth it? - YES! Absolutely! Catching bugs in an early stage is essential to keep down the lifecycle cost of a system. Also consider the risk of badwill if your product is prone to strange behavior and crashes.

You can now take pictures of atoms with a scanning tunneling microscope.

Researchers at IBM even move individual atoms around to create artwork.

More here: http://www.almaden.ibm.com/vis/stm/corral.html

I know IBM can do it, if they're paid enough.

My company (130,000 people) outsourced it's hardware support to IBM. Just at my location, depending on severity, we've had response times of less than a half-hour (when our IBM 3174 failed to reboot after a power failure, cutting off half our building), to days (when a single monitor released it's magic smoke).

You laugh but I bet IBM's Millipede would outlast present technology (including tape).

...just because they claim to have invented magtape and have a big stake in the market.

My memory was a bit foggy.

http://www-128.ibm.com/developerworks/linux/librar y/l-boot.html

It's an article on how to optimise the Linux boot-process by parallelising the init script. This is done using make.

more importantly, MICROSOFT would prefer that you don't know anything about partitioning, as that might let some moron (er, user) install a dual boot with some other OS than Windows, and who would EVER do that? For that matter, Microsoft has made no effort to allow multiple boots off the same hard disk(s), intentionally overwriting the MBR with their single boot loader and never creating a multiboot loader like GRUB or LILO (read this article).

Microsoft uses the KISS BOW IBM method - Keep It Simple, Stupid But Only When It Benefits Microsoft.

My God, you are truly a pathetic nerd. He "stole" money, huh? From people who purchased his product?

Steal. v. To take (the property of another) without right or permission.

If I go to your house, take away your money, but give back a "gift" in exchange, would you consider it stealing? Specially if that "gift" cannot replace the money I took away?

To put it in other words, if I FORCE YOU to buy my products, either by threatens or by AN IMPOSED MONOPOLY, I am, in fact, stealing.

Alright, you don't like the term? How about UNFAIR TAXES? To breathe, you must pay a tax. Or in this case, to run your business, you MUST purchase Microsoft Windows(TM). Of course, everything has its price...

The only difference between a monopoly and taxes, is that, unlike taxes, the owner of a monopoly does not have the legal obligation to return that money to the people. So we are, in fact, under a technological dictatorship.

See where I'm getting at?

Straining there aren't we chuck. MS has compition, there are alternatives. People are just not into learning how to use the competition no matter how easy it is. Just as VHS was successful because it was "good enough" MS is #1 because it is "good enough". They are just about as decietful and evil as any other company. Us geeks dislike them because the the biggest most obvious target of what we think the tech industry is doing wrong.

Response time. Experiments by IBM in the 1970s and 1980s showed that sub-second response times improved programmer productivity. See this paper, for instance:
http://www.research.ibm.com/journal/sj/231/ibmsj23 01E.pdf

Where are we today, 30 years later? Sub-second response times from a PC are certainly possible, but not guaranteed. Sub-second response times from a Web server are rare, and can in practice be tens of seconds.

The remote server is to blame for the lousy response time, but I still feel like smashing my keyboard in frustration. If computers are to feel our mood, they need to react on the timescale of 10ms, never mind 100ms. 1 second, 10 seconds, 30 seconds, are just out of order.

You might want to take a look at http://www-306.ibm.com/software/tivoli/products/st orage-mgr-space/

Or come up with your own scheme of migrating data to tape... if you've got stuff you cant delete for years, you probably need to back it up. So if you can migrate those files off on to two sets of tapes, you're safe and have a scalable solution.

IBM has been pushing Java pretty hard for a few years now and has contributed a lot to the community. They may very well contribute to Apache, but I'm not sure on that. The article above was written for IBM's developerworks site which has frequent articles on all of the technology and development tools that IBM have their fingers in, including Linux, Java, Open-Source, and quite a bit more. It's a cool site to visit every so often.

Yes.. The dinosaur is moving.

And check out their "games on demand" division.
(there are a few interesting intranet links, but here's one for the web. )

I just wished their "On Demand" philosophy applied also to the job I'm doing. *sigh*

I need a new manager.

Wired LAN?? It wasn't all that difficult. Even if it had no NIC, I'd pop in a PCMCIA nic card and on XP it would work out the box, drivers included in the OS. Ever tried setting up PCMCIA on linux? Then the myrad of third party NICs and accessories? Look, these are things that I've done, and they weren't easy or intuative a lot of the time.

I've never owned a laptop that did not come with a floppy drive, so as a last resort I could have gotten the drivers on that way.

It's the base, bog-standard medium that all i386 PCs have.

I've seen plenty of business machines with floppies and no-cd rom, but not the other way around.

And no, Linux is still harder to set up than windows.

Provide me with one where setting up Linux is easier, if you can...it's a great OS, but saying it's easier to admin than windows is a tad off-the-mark.

Wired LAN?? It wasn't all that difficult. Even if it had no NIC, I'd pop in a PCMCIA nic card and on XP it would work out the box, drivers included in the OS. Ever tried setting up PCMCIA on linux? Then the myrad of third party NICs and accessories? Look, these are things that I've done, and they weren't easy or intuative a lot of the time.

I've never owned a laptop that did not come with a floppy drive, so as a last resort I could have gotten the drivers on that way.

It's the base, bog-standard medium that all i386 PCs have.

I've seen plenty of business machines with floppies and no-cd rom, but not the other way around.

And no, Linux is still harder to set up than windows.

Provide me with one where setting up Linux is easier, if you can...it's a great OS, but saying it's easier to admin than windows is a tad off-the-mark.

iSeries will run Windows. You can keep the reliable hardware and use multiple operating systems, including windows, on it. If Windows doesn't meet your expectations, you can change the OS but keep the hardware. http://www-03.ibm.com/servers/eserver/iseries/soft ware/os/

Nice try, except with a logitech "mac comptaible" USB keyboard ***NONE*** of the keyboard shortcuts work with the mac-mini. Not "hold C" or "alt-option-o-f" or the many others. Not a single one. You probably need a blessed official Apple keyboard or something.

OTOH I may have found a solution to your problem: before you reboot to boot from the non-MacOS X install medium (i.e. Debian, *BSD, whatever), open a shell in OS X and type:

% nvram auto-boot?="false"

This will automatically drop you into the OF console after power-up or reboot. (See excerpt of dW article below.)

Frankly I'm not impressed by a computer lacking a BIOS [or monitor of some sort]. Even my 8051 board has startup software to help use it...

From An embedded view of the Mac Mini, Part 1:

Many people assume that, since the Mac doesn't display a prompt to enter a PC-style BIOS, it's not configurable. In fact, it's quite the opposite: the Mac's boot firmware is years more advanced than PC boot firmware. It has a command prompt which is a full-featured Forth interpreter. Open Firmware needs some kind of console. On older Macs, you could use a serial port, but the Mini doesn't have serial ports, so Open Firmware connects to the keyboard and display by default. The display has to be some kind of VGA display. The video output won't work properly until the system's fully booted, because the boot firmware doesn't know about video refresh rates. If you went ahead and got into Open Firmware without knowing this, just type mac-boot to cause it to boot up. Open Firmware is also useful for ejecting stuck CDs; the Mini has no eject button. The trick is to get into Open Firmware and type eject cd.

In the long run, you might want a way not to use a display. Displays chew up power, and if you're using the machine in an embedded project, you may not have any actual use for one. You're in luck: Open Firmware can be configured to run over the network! All recent Macs, including the Mini, support this. Apple's Tech Note 2004 (see Resources) discusses the procedure for setting this up. This does require an ethernet connection: you can't use the wireless network to do it.

A few other likely things to do at the Open Firmware console are disabling auto-booting (setenv auto-boot? false), or setting the machine to verbose boot mode (setenv boot-args -v). Verbose boot mode will be familiar to experienced UNIX users, who have probably been wondering where the familiar and comforting string of console messages have been going all this time.

Verbose booting is most useful if you're trying to debug driver problems that occur early in the boot. You can also tell the machine to go into verbose boot mode by holding down command-V.

Subversion or CVS as version control is about the same when it comes to a small project. Subversion is the successor to CVS and should normally be the choice unless there are other factors that makes it hard to implement.

Obviously - the boss in question is probably also rejecting ISO 9000. And I even wonder about what the accounting looks like. So it may be a good idea to silently look for another job and change before the shit hits the fan.

I'm personally considering the following rules when coding:

• No compiler warnings - unless there are very strong reasons for leaving one alone.
• Any kind of simple version control. (cvs, svn, cms, rcs or whatever is convenient/available)
• Some kind of bug tracker - like Mantis. Excel isn't really good enough.
• Check C code with Splint.
• Write well-structured code - not necessarily with a lot of comments - but the code shall be easy to read.
• If self-healing can be handled in the code - use it. (like checking if a file was closed when the finalize() method is called in Java and then close the file if it wasn't)
• If possible - run the code through Purify or a similar tool.
• ALWAYS braces (or what the language used dictates) around the body of an IF-statement.
• Each method/function shall have a comment describing that comment. (OK, I'm not always doing this)
• Try to keep variables fairly short - too long variable names cuts down on the readability.
• Use of single letter variables are permitted - using the old-fashioned variables i, j and k as index variables in loops.
• Indent code properly - using SPACES to get compatibility with most editors. Eclipse is fairly good at this.
• Align the braces so that an opening brace is in the same column as a closing brace - this makes the code a lot easier to read than if an opening brace is over the right edge somewhere.
• Keep opening/closing braces on their own line. - Yet another readability issue.

You know, it's interesting. We (humans) are quite good at indexing and sorting things. When most of us were toddlers, we learned how to put square pegs through square holes...how to organize things (toys) by color, shape, or function. When we learned how to read, we started small, but worked our way up to understanding the sections of a newspaper and what was fun (Comics) and what was not (Business). As our skills developed, some of us found journalists that we liked and some that we didn't - this helped us further refine our "searches" through newspapers.

Sections, headlines, story titles, author, location - all metadata that is used by us to index the info in a newspaper. I don't think we have the capacity to use many more pieces of metadata to index a newspaper - there's a reason the newspaper is in the form it is today...we have hundreds of years of refinement. Newspapers that sucked weren't bought and "natural selection" resulted in industry norms etc that present the user with a very consistent interface. If I read the NY Times, the interface is similar to the Wall St Journal, USA Today, etc. It's predictable, easy-to-use, and well-defined. My choices are limited, but the format is similar.

The web is a lot bigger than a newspaper - and the web presents the user with a number of different sources for info - all in a very inconsistent format. when I was growing up, there were 2 daily newspapers that we subscribed to - the Hartford Courant and the Journal Inquirer. Neither one was perfect, but they worked the same way for me. Today, if I don't like something I see on CNN.com, I've got a gazillion other choices out there. The challenge to the user is to find a mix of news sources that meet their appetites for knowledge. They also need to wade through the mountains of crap website designs that mean no 2 news sites are the same. I'm all for individual expression and I love clever unique designs - I like to think I've come up with a few in my time - but it does present the user with an interesting problem to solve. Some of us get it, some of us don't. Remember that article about Google users having a richer online experience? Some of that is because of Google - but some of that is because of the people as well. People who use Google are more likely to be able to cope with information overload and quickly parse out the important bits.

CNN has another take on it here http://www.cnn.com/2005/TECH/12/26/information.eve rywhere.ap/index.html. They're not talking about online newspapers, but I believe the issue they describe is the same one we face with online newspapers.

How do we make sense out of petabytes of data? This is why I think the work Google is doing w/Google Print and the work IBM is doing with UIMA http://www.alphaworks.ibm.com/tech/uima is so critical. We've long since outgrown the day that a file cabinet was capable of organizing all the info that's important to us. We've outgrown the filesystem as well. And the web has outgrown us.

Link to all the articles in order.

For me, the first thing I would want is a decent open source IDE that would allow multi-threaded debugging of Ruby apps.

Ask, and ye shall receive.

Seriously, though, half the problem with some of these new environments is over-complexity for a task (I'm looking at you, Java). The fact that you think you need an IDE and eg., multithreaded debugging to make the environment work show that you're in that over-complex world.

When you see an old-timer validate his code because it itself allows for inspection and dumping state, *and* has such a simple internal structure that you can understand the working parts, you realize that the IDE crowd missed something.

Ibm's blade center takes dual opteron cards. 14 blades (28(x2 for dual core) Procs) in a 7u space.

http://www.ibm.com/servers/eserver/bladecenter/

IIRC, (What the salesman told me), you've paid for the chassis at around the 5th blade purchase if you're going up against some x336s. Of course, you need a architecture where you can live without local disks on each machine.

that's pretty sad, and in all seriousness i'd take it as an indication that you should be looking around the industry a bit more, or significantly re-evaluating your knowledge

AIX is IBM's Unix system, and has been a commercial offering for nearly twenty years (next year). it's got a very large installed base (as server OSs go), impressive enterprise-grade functionality, and enviable stability. it's also a very large part of the fuss around the SCO v. IBM lawsuit, perhaps the biggest legal issue in IT at the moment. check out AIX's Wikipedia page, which gives a much better overview of history and such than IBM's official AIX page.

I don't know what you uconsider low cost workstations, but something starting at 5,575 doesn't seem to qualify as low cost. If you want to run a real Unix on a supported platform Sun's workstations start out a lot cheaper and similarly loaded workstations are cheaper. http://store.sun.com/CMTemplate/CEServlet?process= SunStore&cmdViewProduct_CP&catid=48612 AMD workstations start out as little as 895 and UltraSparc based workstations go for as little as 1,395. They can run Solaris 10 as well as RedHat and Suse. Similarly, you can get Sparc and AMD based servers from Sun for under a grand. Being able to run multiple lpars may be cost effective if you want to run multiple lpars but if you want to have one affordable server for development.

> It will be interesting to watch them duke it out - done right, both may benefit. If IBM really wants to compete,
> they need to start selling low cost workstations to encourage further development on AIX.

They've got them, and as low as $5,575. Though I think they are primarily intended for unix graphics packages like Catia:
      http://www-03.ibm.com/servers/intellistation/power /

They've also got very low cost power5 servers that start at $3700:
      http://www-03.ibm.com/systems/p/hardware/entry

There's not much of a price difference between these and intel boxes. And they can run either AIX or Linux. In fact, with the P510 (around $5k) you should be able to run multiple lpars, including linux lpars. Pretty cool stuff.

> It will be interesting to watch them duke it out - done right, both may benefit. If IBM really wants to compete,
> they need to start selling low cost workstations to encourage further development on AIX.

They've got them, and as low as $5,575. Though I think they are primarily intended for unix graphics packages like Catia:
      http://www-03.ibm.com/servers/intellistation/power /

They've also got very low cost power5 servers that start at $3700:
      http://www-03.ibm.com/systems/p/hardware/entry

There's not much of a price difference between these and intel boxes. And they can run either AIX or Linux. In fact, with the P510 (around $5k) you should be able to run multiple lpars, including linux lpars. Pretty cool stuff.

Allow me to expand on this. People might question Intel's motives with "performance per watt" instead of pure megahertz. In the consumer market, it is a bit more confusing (although I'll bet it helps silent PC's move into the mainstream). In the server market, it's critical.

Right now, the best datacenters being built have power and cooling capacity for 110W/sq ft, maximum. A standard 19" rack with proper allowances for cooling, access, etc takes anywhere from 20 to 25 square feet (let's use the 25 sq ft number, since that's what Qwest quoted me last week).

This means in a given rack, you can have at most 2750W of power. One BladeCenter requires two 2000W power supplies, and four if you use more than 6 blades, all in a 7U form factor. Yes, that means you need to use about 35-40 square feet of floor space to support the power and cooling for a 7U box (albeit, an extremely dense one).

Likewise, look at your standard 4U/4 processor server (now more frequently 3U). Five years ago they were using dual 300W power supplies. Now they use dual 1300W power supplies. What's taking all the power? The main board isn't, the PCI adapters aren't, and the disks may even be taking less (faster disks, but 2.5" SCSI is making its way in - see the xSeries 366, for instance). Even if some of the other components are taking more power, the lion's share is the CPU.

Allow me to expand on this. People might question Intel's motives with "performance per watt" instead of pure megahertz. In the consumer market, it is a bit more confusing (although I'll bet it helps silent PC's move into the mainstream). In the server market, it's critical.

Right now, the best datacenters being built have power and cooling capacity for 110W/sq ft, maximum. A standard 19" rack with proper allowances for cooling, access, etc takes anywhere from 20 to 25 square feet (let's use the 25 sq ft number, since that's what Qwest quoted me last week).

This means in a given rack, you can have at most 2750W of power. One BladeCenter requires two 2000W power supplies, and four if you use more than 6 blades, all in a 7U form factor. Yes, that means you need to use about 35-40 square feet of floor space to support the power and cooling for a 7U box (albeit, an extremely dense one).

Likewise, look at your standard 4U/4 processor server (now more frequently 3U). Five years ago they were using dual 300W power supplies. Now they use dual 1300W power supplies. What's taking all the power? The main board isn't, the PCI adapters aren't, and the disks may even be taking less (faster disks, but 2.5" SCSI is making its way in - see the xSeries 366, for instance). Even if some of the other components are taking more power, the lion's share is the CPU.

NASA uses Ruby also (and no doubt a few dozen other languages)

see here about NASA Ruby (and Python and Perl and other open source) use and also here for NASA and other companies

Python runs on an amazing number of OS , from embedded ones to mainframe (maybe a couple java can't?). Ruby can run on Windows, Unix/Linux/BSD, OpenVMS, BeOS, amiga. Ruby even runs with threads on MS-DOS.

Vlissides' generation gap pattern can help solve a lot of the maintainability issues with code auto-generated from a DB's structure. The idea is that you use some tool to auto-generate stubs from your model, and then you extend (in the formal OO sense) those stubs. When your model changes, you can regenerate your stubs without trying to do some ugly merge between your customized templates and those that reflect the new structure.

While we're on the topic of keyboards, does anyone have experience with the Lenovo/IBM ThinkPlus USB Travel Keyboard with UltraNav?

Does it work with Linux and BSD systems?