Slashdot Mirror

Actually AMD bought a company called NexGen which designed the chip which was later rebranded as the AMD K6. Here is the link.

There are quite a few stories out there about companies who were a) negotiating technology deals with Microsoft, b) the deal fell through, and c) Microsoft introduced a technology of their own with striking similarities.

The most conspicuous was probably the Stac patent-infringement suit.

What these stories have in common is: first, it is probably not so much a case of Microsoft outright stealing technical details, but more "we like that general approach, this company has shown us that it works, let's do it that way ourselves." For every case of outright infringement, there are probably a dozen more of moral, but not legal theft of ideas.

Second, and more ominous, even in those cases like Stac where Microsoft was challenged in court and lost, in the long run it didn't matter. Stacker is a distant memory, Stac Electronics is all but forgotten, and their website isn't responding right now...

The April 1998 Byte cover story has a graphic Why PCs Crash, and Mainframes Don't. It's interesting to see how little has changed in almost 5 years.

The April 1998 Byte cover story has a graphic Why PCs Crash, and Mainframes Don't. It's interesting to see how little has changed in almost 5 years.

Brooks, in the "No Silver Bullet" essay referenced above, stated that there is both essential and accidential complexity in software development, and because of that there never would be a "silver bullet" to slay the software "monster". However, there are fundamental practices that increase the likelihood of success and fundamental pitfalls that every project faces. And, in the end, the root causes of most failed IT projects are human factors; in fact, you could just cite the "seven deadly sins"--pride, envy, gluttony, lust, anger, greed, sloth--and probably hit the nail on the head.

In conjuction with that, far, far too many practitioners in the IT field lack one or more of the following:

• Talent
  
• Sufficient (or any) education in software engineering (or even computer science)
  
• Any familiarity with the literature from the past 30+ years. I'm not talking about IEEE/ACM Transactions, I'm talking about standard classic works such as _The Mythical Man-Month_ (Fred Brooks), _The Psychology of Computer Programming_ (and everything else by Gerry Weinberg), _Principles of Software Engineering Management_ (Tom Gilb), _Peopleware_ (and everything else by Timothy Lister and/or Tom DeMarco), _Assessment and Control of Software Risks (and anything else by Capers Jones), _Death March_ (and anything else by Ed Yourdon), _Journey of the Software Professional_ (Luke Hohmann), and any of the 100 or so texts on software engineering on the bookshelf behind me.
  

To quote George Santayana (who is often misquoted):

Progress, far from consisting of change, depends upon retentiveness...Those who cannot remember the past are condemned to fulfil it.

Software engineering is hard enough--with all the human issues--without further handicapping ourselves with ignorance of all that has been already discovered and documented. Yet that is exactly what most IT workers do. Until we find a way to solve _that_ problem, the failure rate for IT projects will remain high indeed. ..bruce..

Check out Byte's current situation.

I personally have had problems with every Linux distro and Windows version I have ever tried....

This is off-topic, but the act of installing "modern" operating systems has almost never been completely without hitch (okay, the BeOS install has come the closest in my experience...). Windows is already installed on most computers that people buy for their homes, thus removing one frustrating step. The problem is that Microsoft does not allow OEMs to offer anything different.

Could you have exaggerated any more? It's a good thing we don't hear stuff like this from anyone else, because it's false.

On equal hardware FreeBSD and Linux perform similarly. (FreeBSD has a slight edge on Linux.) I haven't seen the latest stable Linux vs FreeBSD 5, but here is a benchmark of FreeBSD 4.1.1 vs the 2.4 kernel: http://www.loureiro.eng.br/artigos/Linux_Vs_FreeBS D.pdf
Here's another more recent comparison: http://www.byte.com/documents/s=1794/byt20011107s0 001/
Makes a 733Mhz p4 feel like a 3Ghz p3? Right.

Of course, even though the comment is bullshit, it gets modded up because FreeBSD has about .01% of the market, and therefore qualifies as 'l337', which means any pro-BSD comment will automatically be modded up. Nice.

It is available from their archives. Not too bad actually.

I think this is the link; a good read:

Why Mainframes Rarely Crash

Probably not.

Moshe Bar (Of open Mosix fame) recetly wrote a byte column on how he got his GPRS phone talking to his laptop while on his honeymoon in the Italian alps.

What his new wife thought about it was not recorded.

I hope this helps.

I like sites like this ... but isn't their already an effort to define and tie blogging communities via the The Weblog MetaData Initiative?

I mean, Waypath is at one level convenient, but no more so than well established weblog communities such as
blo.gs, the Eaton WebPortal and blogs4God. Moreover, when it comes to gleaning headline news via a blog, I would suspect the real weapon of cohice would be our personal aggregators such as Amphetadesk and HotSheet?

Which is where the WMDI comes in. It helps me identify sites via xml-ish mechanisms such as the Dublin Core Initiative ... which is why I would think someone who's blogging their brains out for the hottest headlines might not be better served by the WMDI.

Then again, your mileage may vary.

I didn't see it here on Slashdot, but here's the Byte.com article that was probably the basis of the /. article.

Ultra-basic summary:

There was no particular tuning of either OS, the daemons were set up equivalently on each OS using precompiled binaries, and Linux had KDE loaded and running in order to make a fair comparison. Linux came out marginally faster.

Russ

My favorite columnist, Moshe Bar, has a current article up on Byte that discusses this very topic. Short answer: Right now, Linux is faster than OS X on same hardware.

There's an essay at byte.com written by Moshe bar comparing the performance of Linux and MacOSX from a server point of view. Very interesting read.

Moshe Bar says: "The fact that OS X needs to improve in VM and I/O handling is understandable given its relatively young age." That is his opinion from testing XServe. (Note there was things he could have done to improve the test, but on a whole it was a good test.)

Moshe Bar has written an article at Byte in which he benchmarks and compares performance between Mac OS X and Linux at various tasks on the same hardware.

Moshe Bar has written an article at Byte in which he benchmarks and compares performance between Mac OS X and Linux at various tasks on the same hardware.

His point is to slap some real useable software on top of any OS and live there, not at the OS level with files, folders, permissions, etc.

Wasn't that Netscape Constellation?

Were you aware that not too long ago, Microsoft infused a couple of million into Corel to keep them afloat? As such they are a partially owned subsidiary of Microsoft, NOT a competitor

These are the definitions that I was using in my post:

What definition YOU choose to use is pretty irrelevant. A court has ruled that Microsoft has a Monopoly in it's market. It That ruling has been uphead on appeal. In a court of law it is considered to be a fact.

How come millions of people use Linux, BSD, Mac OS X and every other operating system on the planet if they "NOT free to buy someone else's product"?

A monopoly does not have to be 100% to be a monopoly, and not all of those alternatives are in the same market. The court ruling precisely defines the relevant market. For example Mac is not part of that market.

By your argument power companies are not a monopoly because you get put up solar panels and AT&T was not a monopoly because people could use mail and ham radio. Standard Oil was not a monopoly because you could use coal.

A company's motivation is never "to get more of the market." It is always a company's goal to make a profit;

Yes, a company's goal is to make more profit, and the size of the market they capture a primary factor in that and you know it. You gave an example of a company selling a product at a loss, that rarely happens unless a company is abusing short-term losses to kill competition to harm the market to enable it to impose higher profits.

The more they try to "exploit", the quicker they will loose their monopoly position.

You are assuming there are no barriers to do so, and that the better product always wins. Even if Microsoft did not abbuse its monopoly in any way there would be tremendous barriers to entering and caputuring the market. There is a tremendous amout of software written for the Windows operating system. What good is an OS without software that canm run on it? On top of that there are the barriers that Microsoft has created using its monopoly position. Just look at the list of prohitited activities listed in the Microsoft settlement for a sample, and those are far from the only barriers Microsoft has created.

All of your arguements about the great way capitalism works are absolutely true - so long as you have competition and minimal barriers to entering a market. Microsoft has a monopoly, there is insignifigant competition in its market, and there are massive barriers to entering the market. When that happens the market forces vital to the proper functioning of capitalism break down. When the rules break down in one market, that market can reach out and disrupt the proper functioning of other markets.

the Wall Street Journal printed Microsoft Windows project leader David Cole's 1991 e-mail. In it, he discussed purposely putting a bug inside an early version of Windows that would cause a competitor's software to crash.

Read the entire article that that link points to, it describes EXACTLY one of the ways that the breakdown in competition in Window's market can harm other markets.

-

We have this price list from 1998, which is the second or third generation of machines. I have what I think is a first generation SPARCbook (a 3XP w/ 85 mHz MicroSPARC II) I got on eBay for $110 after shipping. Here's an idea of what this might run you. Prices are in Australian Dollars (1998 dollars. I ran 36,990 AUD through a converter and it spit back $20,669.64, to give you an idea).

SPARCBook price List

Start selling your children and vital organs now to avoid the market flood in Dec when these things come out. What I'd like to know is if it's running a stock Solaris, or a special Sparcbook version like they had for mine. It had a different kernel and I believe a different version of CDE (I don't know, I don't have it, or any OS on the machine for that matter). Kernel was to deal with all the whacky hardware packed into the magnesium monster, especially the microntroller that prevented ungraceful heat death and power management, that I don't think any BSD or Linux deals with 100% correctly. Moot point as the 3XP doesn't have any good way of getting an OS onto the machine (no floppy or CD-rom included, whacky network port [that you can get a cable to convert to AUI, which is less whacky]).

It's nice to see that they are still doing new hardware, but these definately aren't meant to take market share away from Apple or x86 in consumer laptops.

See also Tadpole AlphaBook, Tadpole 7007/IBM N40 RS/6000 laptop with a neat Byte writeup here. (portable AIX is a steal at $12,000. Now there is NO escape!)

#include <stdio.h>
void main( void ){
int x;
long y;
y = 28.2839281;
x = 339829;
y = x / y;
printf("Content-Type: text/html\n\n");
printf("Hello, world ");
}

Byte released a comparision of linux and OS X at here

Easy mistake to make, but I suspect you're thinking of Jerry Pournelle Maybe international terrorism will end now that he has DSL.

Cheers, Paul

Write ONCE media.
If cameras and PDA's were the target, this was the first fatal flaw.
(multisession sure, but that just eats up even more precious space)

EXPENSIVE MEDIA
($11.66 each, $0.46/MB. Compare that to MD- DATA2 650MB discs at $2.00 (And they're rewritable)

Digital Rights Managment, included FREE! for those who can't govern themselves.

Trying to fill a niche that the might of Sony couldn't for 10 years

Proprietary transport. With no S/PDIF digital out. In other words, NO FAIR USE.

UGLY and EXPENSIVE players.

Media -begging- to be lost.
How long do you think you could keep from losing this $11 gem??

I could add many more reasons but I have to get to work! ;)

Maybe we could keep this list growing and send it to the (out of work) CEO...but then again, I'm sure they already knew what a horrible format it was before unit 1 shipped. This is a blatant attempt to copy Minidisc's design and burden it down with DRM (as well as package and market it to kids) ---NEXT!

-=Chud-Wretch

You already forgot?

> Can you prove that Apple's campaign is a success?

Not to be glib, but that's what I did.

Advertising can rarely be translated directly into sales, since there are so many factors involved in making a sale. A campaign can be a success if it brings more people into the store, even if they don't buy (that's a promotion success, but a sales failure). Thus, the direct goal of the advertisement is to get people to think about your product and act (hopefully buy, but not necessarily) upon it. From the example criteria I listed in my first message, I'd say Apple's campaign has been a success.

Note, you used the term "switch" even though you didn't buy a Mac. That meets criterion 1. If the ad prompted you at any time to go to Apple's web site and price out a new G4, then the campaign worked. (If you didn't do that, then I call into question your claim that the Mac would be $3,000 more)

To give you another example, the company I work for made Business 2.0's list of top 100 stupidest business decisions of 2001. Although most people wouldn't be happy about that, our promotions people responded with, "Great! There's no such thing as bad publicity!"

I'd say that your one lost sale hardly calls Apple's products into question. The buzz has been strong since Apple released Mac OS X about how there's finally a desktop Unix. Several respected Unix forums are now paying attention to Apple's products in a way that never would have happened with Mac OS 9: Slashdot's created its own Apple section, O'reilly has an entire series of books on Mac OS X, and former Linux users are gravitating to Apple's new offerings. For example, in a recent Byte article, the author talks about how he saw at a recent Linux conference, he saw "maybe 20 or 25 people running Mac OS X". Plus, in that article, the word "switch" was used three times in the context of someone moving from one computer platform to another. So the former is a sales/product success and the latter is an ad campaign success.

Still, it does have some pretty impressive specs when it comes to memory bandwidth. See?

This is a bit of information I have been spouting off about for the last several months.

In order for any company to compete, and perhaps destroy the Intel/Microsoft cartel they have to be able to offer 2 things.

Thing 1: Easy to use operating system...

Many Linux fanatics view Linux as the silver bullet for restoring competition back into the computer industry. I say BAH! Linux, in order to offer more acceptance into the computer community needs to work on it's GUI in order to be something every Joe Blow user is going to want to use, and feel comfortable using. Apple is light years ahead in research for usability. Granted some usability experts will tell you there are some, but few, inherent flaws to MacOS X's usability, but it still reigns supreme over others nonetheless. Here is a good article that is somewhat related to this MacOS X Article

Thingy Numbero 2:

Apple needs to get into bed with a chip manufacturer with more production facilities. This is a good idea since this will cheapen the hardware needed to make an iMac or a PowerMac. If I recall correct, Apple recently has decided to get IBM to produce the PowerPC chips that will be in their boxes. IBM has more facilities to manufacture said chips, which makes them a wiser choice over Motorola. I personally don't think this will last. I think that Apple should hook up with AMD in very much the same way that Microsoft has with Intel. If memory serves, AMD has even more facilities than IBM. Additionally AMDs architecture isn't too dissimilar from the PowerPC architecture. This sort of alliance would only serve to make both companies better.

I personally believe that this is the only way anyone is going to give Intel/Microsoft a run for their money.

FYI if you think that Microsoft is the only problem think again. Intel is just as bad as Microsoft in the scheme of things. There is a need for reestablished competition in not just the software industry, but the processor industry as well.

Remember the SoftCard, a Z80 card that let you run CP/M on an Apple II?

Microsoft mice have a great reputation and they're actually pretty innovative about them.

Child psychologists use the term "magical thinking" to describe how young children understand the world. In this worldview, effects and their causes are not objectively determined, but mediated by the child's own desires; for instance, "It is raining because I am sad," or "My friend broke his leg because I was mad at him." Of course, even adults sometimes retreat into magical thinking. Take, for example, the U.S. Commerce Department Technology Adminstration workshop on "Digital Content and Rights Management," where a room full of grown men and women discussed, very seriously, the possibility that wanting a thing badly might be enough to make it happen.

. . .

Legal Bits
Stall Beg I
Gab Sell It
Beats Gill
A Bill Gets

And finally...

All Bis Get

Did I post this twice? You Decide!

Legal Bits
Stall Beg I
Gab Sell It
Beats Gill
A Bill Gets

And finally...

All Bis Get

>though Intel is slightly more underhanded about theirs. Ie, people are encouraged to run 64 bit stuff, because the 32 bit stuff on Itanium will run as slow as a wet week.

Man, I've never heard that one before...

It's not the first time someone's mentioned this. Check out this Byte article from Feb 1996, as found on an advocating OpenBSD page. Many of the other tips on that advocacy page are relevant to the linux camp as well.

Do you have a link/biblio-reference to the info about the Russians buying the pinball machines?

One of my favorite Russian-CS-is-screwed is the story about the metric chips... This Byte article alludes to the original story... In short, the Russians stole western-technology and produced knock-off copies using "the metric inch" -- except when their poor-quality copied failed, they couldn't use real (stolen?) chips to repair their machines.

This is one of the reasons I'm reluctant to pony up money for any web site.

was a firmware bug in the built-in fax modem (based on a Sierra chip) in the old Momenta pen computer. Back in 1992/93, I was at Alien Computing, pioneers in PC-based faxing. We had an existing Windows 3.x product we were porting to Pen Windows under an OEM agreement with Momenta, and my responsibilities included pen-ifying the existing Windows UI and adapting the TSR-based communications module to the custom modem chip.

In the process of testing the comms module, I noticed that many of the received faxes were a bit garbled, as though bytes were being corrupted or dropped. After almost a week's analysis of the modem's raw datastream, including decoding the CCITT data, I discovered that the modem would at random intervals repeat a single nibble of the raw data, corrupting the image.

Once I found the bug, I then spent over a week convincing the Momenta engineers of the bug's existence -- they were conviced their modem firmware was bulletproof. I finally got through to them, but as I recall, the bug was never fixed, probably because Momenta wasn't doing well financially -- the pen computer, it turns out, was about a decade ahead of its time.

See now you are just trolling. You say "mozilla is ___so____ much better, etc..." have you actually tried to write an HTTP daemen?

I have no intention to troll. None whatsoever. And please don't put words in my mouth. I never said that quote you have above.
No. I have never tried to write an http-daemon. What does that have to do with anything?

I tried to download the zipfile you posted and as you said, it works fine in IE but not with Mozilla. I have no idea why this happens but I'm sure the mozilla team does. You have filed a bug-report so that they can fix it, right?

Why does this happen from your webserver and not from other servers that I've downloaded zipfiles from? Is it really a Mozilla error? This is an honest question, because it really seems to work elsewhere.

About the bastardizaton of html. It's no secret that Microsoft has added new tags and stuff that only works in IE. Pick up any book about web-design. In most of them it sais right there [Only works in IE].
Although I can't prove it, it wouldn't surprise me if they did the same to the http. They have some authentication-scheme that isn't a part of http don't they?

Just because I haven't done exactly the same things like You it doesn't disqualify me from having an opinion. Especially if its backed up by some facts.
Here they state that Mozilla "may be the most compliant of all current browsers."


Here they say something similar. Written in 99 they haven't tested the 1.0 release and aren't all positive. But they agree that it complies with standards.

Another link about how well Mozilla follows standards.

So, in conclusion I'd like to repeat what I said before. Mozilla isn't perfect, but it's my browser of choice until IE implements the features that I find useful in Mozilla [Gestures, Image/popup-blocking, tabs among others] and prove to be faster and better (subjectively decided by me) and can be run on the platform [OS] I run at the time.

Best regards

.haeger

gravity alone could be used to generate electricity

That's what we do now.

- Fern from 1M years ago dies and washes down a sandy stream.
- It's covered in layers and layers of dirt and sand.
- It gets pressed and pressed by the material on top of it.
- We dig it up as oil and put it in our cars and power plants.

Tada! Gravity as a power source.

Or, you could wear these!

Serving With Linux is interesting.

I'm looking at an aluminum case, probably the Lian-Li that everyone is so fond of. Anyway, I read Jerry Pournelle's Chaos Manor Column up at Byte religously, and noted that he had a bad experience with a flimsy motherboard tray at one point (see this article). Anyone have any opinions? Did he just suffer because he bought some "el cheapo" knockoff?

I'm looking at an aluminum case, probably the Lian-Li that everyone is so fond of. Anyway, I read Jerry Pournelle's Chaos Manor Column up at Byte religously, and noted that he had a bad experience with a flimsy motherboard tray at one point (see this article). Anyone have any opinions? Did he just suffer because he bought some "el cheapo" knockoff?

Independent of whether or not you're trolling, this article needs someone to link to Advocating OpenBSD, and especially to a link off of that page, The Sound And The Fury.

and (hopefully) security conscientiousness continues past the end of the month

Hmmm interesting those corporate buzzword generators win again. Trouble is there first has to be a great awakening like WTC. This has not happened in the Internet world yet, after all bin Laden didn't recruit any script-kiddies. Anyways here's a better article that's been posted before on /. that's relevant here about how Linus likes lots of kernel trees and why he thinks this helps development.

I've heard the new VM is really good. RedHat 7.2 has kernel 2.4.9-31 - I read in this article and elsewhere that 2.4.10+ has the new VM is much nicer. I wanna have it!

I am not as pessimistic as you that a good quality portable decoder cannot be built. I agree that it probably won't be, but headphones should encourage better fidelity, not worse.

I can hear the Rio One decoding errors even with the cheap earbuds because they're mid-high-frequency errors which the earbuds reproduce especially well. If they were over 16kHz or under 200Hz it would be a different story. The warbling of too low a bitrate or a poor encoder is also easy to hear.

The earbuds don't provide as much isolation as over-the-ear headphones, so listening in an airplane or subway might hide the errors, but I listen on a couch or in bed in a quiet room after my son has gone to sleep. I can't use speakers at the volume I like & don't have whole house audio from my changer (though I'm thinking of changing that). I'd rather rip the discs to an archive on my PC & grab a few to an mp3 player that I can jack into my car, carry around the house & take to work.

----

When we say 24 bit audio, let's clarify what we're discussing. Junk like the Rio One use 16-bit *fixed point* DSPs. I'd assume that most PC decoders use single-precision (32 bit?) floating point. I don't know what the "24 bit" decoder plug-in for WinAmp is all about, but it is supposed to sound better than the built-in one.

The StrongArm can retire 12 bits of [multiplier] result per cycle and the adder can retire 32 bits. This yields a total latency of two to four cycles for a 32-bit result and three to five cycles for a 64-bit result. Maybe softfloat can handle quality mp3 or ogg vorbis decoding, but it seems that if inner loops are coded in assembly it might be possible.

A better approach might be to use a heftier DSP. I know the Motorola 56k series & this would seem to be a good candidate: 24 bit datapath with a 56 bit accumulator for 24x24=48 results accumulated without rounding until the end (so a 256-tap FIR is possible).

Sure it's more power hungry than a C54, but hell, skip the hard drive (use flash) but keep the lithium-ion battery and you're all set.

It'll be a lot of work to write the code, I'll grant you that, but fixed-point mp3 decoders have been written (by xaudio for example). Ogg Vorbis will be a chore too.

Actually, there already is a portable player that uses the 56k series, the PJB 100

If you're willing to deal with even worse power requirements, a CPU with FPU is possible. We can trade off things like a nice big backlight for a nice big color LCD and use a Pentium or PowerPC. I'm not going jogging with the thing, so if it weighs a pound or 2, I'm OK with that.

My old 120Mhz Pentium laptop can run WinAmp. An embedded Pentium 166MHz draws 2.35A at 1.8V or 4.23 Watts. Modern NiMH AA cells give 1800mAH each, so four might give you 3/4 to 1 hour of use. That's not enough, but it's within an order of (10) magnitude.

The "portable supercomputer" PPC 440GP also draws around 4Watts, but it runs up to 400MHz.

The TMS320C32 32-bit floating point DSPs draw nominally .675 and max 1.4 Watts @ 60MHz.

-M

Why not do what Moshe Bar did and run Linux S/390 on a PC using Hercules? Sure the IBM site certainly has more resources, but if it's to just periodically recompile/test a few programs, it may not be too bad to do it on your own hardware. I suppose it could also be a good reason if you are developing a closed source program and are paranoid about having the source on another company's site.

Did you even read the excellent article by Scot Hacker?

Be was offering its OS for FREE to manufacturers. Having BeOS as well as Windows on the computers you sell clearly would have added value to your product, because everyone who spent time getting to know BeOS loved it. Clearly, the reason manufacturers didn't take Be up on their offer to preinstall the OS for free is pressure from Micro$oft.

Good luck with the lawsuit, Be!