May 31st

NVidia, AMD and Micron vs. Rambus

One thing in life is certain.  You do better in general if you have lots of friends, than you do if you have lots of enemies, and no friends.  The latter situation is where Rambus finds itself in it's anti-trust case with the Federal Trade Commission (FTC).  There is an unfriendly lineup of companies waiting to testify against Rambus, which probably is enough to make even the most hardened Ramboid worry.  Bill Teel comes to mind, but that's another tale.  

As things stand now, it's impossible to tell what is going to happen.  Rambus could lose it's patent rights, or it could come out the winner.  It's too soon to tell, mainly because it will depend on the judges involved, and how they view anti-trust law.  Rambus clearly violated JEDEC memory standards rules by participating in meetings while at the same time applying for patents to the open standards being discussed.  The question is, will the judges see it that way or not? But with NVidia, AMD, Micron and Infineon all waiting anxiously for their day in court, it's not going to be fun and games for the Rambus crowd. I expect the silence from ex-friend, Intel, to be almost deafening. 

                                               Dr. John

Tom's Guide to Digitizing Your Old Tapes and Movies!

Tom's Hardware has a great new article posted for folks wanting to dabble in digitizing their old VHS and 8mm movies.  This is something that lots of people are getting interested in doing, and Tom's got all the tips you'll need to get started, here.

How Businesses Make Themselves Hated

Have you ever noticed that there are some companies out there that most people love, or like, (Sony comes to mind), while others seem to have PR and management teams from Hell?  NVidia is a company who's products I generally like, but I don't like the company because of their dirty tricks, and heavy-handed tactics.  Of course, Intel is also in that category, and Microsoft practically invented the category.  Indeed, Microsoft is a funny bird, with about 60% of people loving it (the blissfully ignorant, or those employed by MS), and the remainder, which despise the company.  Microsoft has single handedly turned the IT industry into a truly world-class, cutthroat environment.

And yet there is another class of companies, without the customer loyalty base that Microsoft can count on, that have created a remarkably  horrific reputation for themselves, and are suffering greatly as a result.  Rambus, and now SCO, both come to mind in this category.  Rambus could have been the PC enthusiast's dream memory, now it is the subject of constant jokes and disdain. In fact, Rambus has worked tirelessly for years to destroy their own image among PC users.  It's as close to a corporate death-wish as any company has managed to date. 

Now comes SCO, a small company with a small range of purchased products (they didn't make them, they bought the rights to them), that has decided to take on the entire Linux community, probably the most iconoclastic, in-the-know, independent-thinking, tight-knit group among computer users. This isn't David and Goliath, this isn't Iraq vs. the USA, it's sardine vs. Killer Whale, and it's pretty obvious where this affair is headed. The fight isn't even fair, as you can tell by the amount of bad news SCO received during the first week of the assault. If this drags on much longer, expect SCO to become one of the most despised companies in the computer industry, and expect their fortunes (and stock price) to continue declining.  

                                               Dr. John

PS, Microsoft probably put SCO up to this, which makes a lot of sense. It's an end run on Linux, which Microsoft would love to be successful. 

May 29th

SCO Owns Linux, And You Will All Pay!

An interview with SCO's Chris Sontag might seem like it's right out of the Twilight Zone, but the fun part is, it's real! Chris says that everybody better stop working on Linux immediately, or else!  Oooooo. Now everybody is scared. Chris kind of reminds me of the Iraqi Foreign Minister, but that's another story. 

When asked if everybody would do better to get rid of their copy of Linux, and maybe burn it to eliminate the evidence, Chris says that he can't make any recommendations just yet.  When asked if SCO would begin demanding licensing payments from Linux users, Chris said he couldn't rule anything out.  OUCH! Rambus hasn't got anything on these guys!

I wonder if anyone, anywhere, will ask for an investigation into Microsoft's recent large contract deal with SCO. This topic is screaming for scrutiny. There is no one on Earth that is more interested in stifling Linux development than Bill Gates, and Mr. Gates just signed a large contract with SCO, around the time that SCO decided to threaten Linux developers. I see a major scandal brewing here, and it will be interesting to see if the government is up to investigating Microsoft again. Why do I doubt that Ashcroft et al. would even consider it?

You can read the incredible interview with Chris Sontag here.

                                              Dr. John

May 28th

SCO Leadership Has Gone Insane

If the latest press release from Novell is accurate (and it seems very convincing), SCO does not  own the patent rights to Unix as they have stated publicly.  Novell owns the rights.  Indeed, Novell says that SCO has been trying to secure those rights for Unix from Novell for months, making SCO's claim of ownership and infringement seem almost insane.  If you are going to take a page from Rambus Inc's play book, at least make sure you own the patent rights first. Here is what SCO has to say about the Novell statement.

                                              Dr. John

Microsoft Patch Needs Patching

Microsoft's latest Windows XP patch does not work with other companies firewalls, and prevents Internet access. Considering that tens of thousands of loyal Microsoft customers have been convinced to leave the "automatic update" feature on, I expect that many XP users are learning the meaning of frustration. Many of these users probably didn't  even know that Windows had automatically updated itself with the onerous code, and as such couldn't figure out why they had lost access to the Internet. 

Microsoft is the largest software company in history, by far, and this has me pondering how they could have failed to test a major patch for their flagship operating system with other companies firewalls.  Either they are incompetent, or they are playing games here with third party security software.  Either way, it makes me even more leery of Windows XP. Time for those of you who still have the 'auto update' feature enabled to disable it.

                                              Dr. John

Novell Calls SCO's Bluff

Novell claims that it owns the rights to the Unix operating system, and it will fight SCO's attempt to control Linux by means of litigation. The whole affair is getting very murky, and I imagine this will take the courts years to sort out.  

Background? SCO has decided that Linux is a threat to their Unix business, and declared Linux in violation of their acquired Unix patents. Now, with Novell challenging the standing of those patents, I expect SCO is starting to wonder if they bit off more than they can chew. SCO has to prove both that they own all the patent rights, and that Linux is in violation of those patents.

                                              Dr. John

May 27th

Microsoft Can't Give Windows Away

Two stories out today have me wondering if Microsoft is losing its grip on their operating system monopoly. First is the story that Microsoft CEO Steve Ballmer cut a ski trip short when His Billness called and said that Germany was quitting Windows.  Ballmer rushed to Munich to offer the German government a 90% discount on MS software (to get them hooked), but the Germans were too smart for that ploy. This has become a standard operating procedure for Microsoft. As soon as any large organization threatens to dump Windows, they get offered free, or nearly free software. His Billness knows that he needs to keep them hooked... a few days off the stuff, and they might stay clean. 

The second story concerns Microsoft's decision to increase its philanthropic donations of Windows to various US organizations (get them hooked early) to nearly 1 billion dollars.  Many critics immediately accused Microsoft of trying to bolster its monopoly on operating systems for PCs, which is of course just what Microsoft was intending to do. But if you think about it, as Microsoft gives more and more copies of Windows away for free, it deflates the fair market value of the remaining copies for sale. Eventually, if MS tries to out-freeware Linux, they will go out of business.  So what's a software giant to do?  Keep giving it away to get everyone hooked, or let Linux slowly eat away at their paying customer base?

                                              Dr. John

May 25th

ATI Caught Cheating Too

It's a sad day in 3D land.  The two most popular video card makers have been tarnished, tainted, and exposed as cheaters.  NVidia claims the prize for blatantness and gall, ATI gets the consolation prize for trying to keep up with NV by similar, unsavory, means.  Extremetech, the guys who couldn't afford the Futuremark "beta partners" program but got it anyway, have finished their exhaustive testing of both new driver releases from the two graphics card companies.  Both have unfair optimizations for one of the toughest tests in the benchmark suite, "Game 4", or Troll's Lair. NVidia had additional cheats in their drivers, which in one case boosted performance artificially by 44%!!  That's a cheat to write home about.

We can hope that the results from this affair include things like, honest driver releases from both companies, and a change in Futuremark's "beta partners" program that includes all major video card manufacturers.

                                            Dr. John

May 24th

NVidia Cheating Questions

As the news about NVidia getting caught at cheating with their new driver release reverberates around the web, several folks have made note of the two "partners" that fingered NV. Both Beyond3D and Extreme Tech have joined the Futuremark "beta program", which means they get all the tools to examine 3DMark benchmarks on different platforms in great detail. The question is, how did these web sites become beta partners?  Neither web site can afford to pay the tens of thousands of dollars required to join by the normal method (i.e., pay big bucks and you're on "the inside").  So what other method did these two web sites use to become beta partners with Futuremark?  Obviously, they worked out a deal behind the scenes.  Until we know how this happened, a growing cloud of uncertainty hovers over Futuremark's head. NVidia was not a "beta partner" because they refused to pay the absurdly high "entrance fee".  Something is rotten in Futuremark land, and I want to know what.

                                            Dr. John

Beyond 3D

ExtremeTech

May 23rd

NVidia Caught Cheating

Ouch! That's all I can say. Futuremark, the folk who bring you the 3D video benchmark "3D Mark" have released a scathing report on NVidia's latest driver release. The title of the audit is: "Alleged NVIDIA Driver Cheating on 3D Mark03". It details how NVidia altered their drivers to specifically detect 3D Mark 2003, and to implement altered code that was tailor made for the benchmark.  The cheats boosted performance by about 25%, accounting for virtually all of the speed increase offered by the new drivers.

What I find so bizarre about this is the fact that it would have been found out eventually no matter what.  Hardware review web sites would have immediately noticed that benchmark scores only changed in 3D Mark 2003, but not in any game benchmarks.  What NVidia was thinking is beyond comprehension. I expect this revelation will drive even more PC enthusiasts to ATI's Radeon 9800 video card. That may turn out to be more difficult than you'd expect at the moment, considering that the last two NVidia stumbles (debacles) have created a severe shortage of high-end ATI video cards. 

The damning audit is here.

                                            Dr. John

May 21st

Half Life II Illegal for Minors?

Washington State has just passed a law that forbids selling any computer game to minors that depicts violence against law enforcement officials. So both Grand Theft Auto: Vice City and Enter the Matrix would be off limits for those under 18 years old.  This also means that the upcoming Half Life II, wherein everyone is an enemy, including monsters and swat team members, would be verboten for teenagers. Of course laws don't stop kids from doing much of anything, but it does mean that store owners will have to be careful who they sell games to in order to avoid fines. It all seems like a very big waste of tax payer's money, and legislator's time, but I'm sure this will win them brownie points from some parents. We'll have to wait and see how many other states jump into the fray, and write their own laws limiting sales of computer games to minors. If this keeps up, we may only see games like The Sims in the future, which is a very scary thought.

                                            Dr. John

MSI BIOS Update Kills Mobos

MSI must be kicking themselves for releasing a BIOS update for their i875P (Canterwood P4) motherboards. Apparently, some people are finding their motherboards have been rendered useless after the 1.3 flash update. So best to hold off on updating your MSI board until this gets sorted out. 

                                            Dr. John

May 19th

Doom III Fading, Half Life II Rising?

We've been waiting for the sequels to Doom II and Half Life for years.  Years.  Now both sequels are on the horizon.  Of the two, Doom III has been far more hyped and discussed.  The graphics are beyond anything to date, but due to all sorts of considerations, the promise of a "fully deformable" environment seems to be slipping from Doom III's grasp.  So Doom III may win the eye-candy prize, but not the interactive object prize.  Half Life II, on the other hand, is being reported to have robust  environmental interaction.  Indeed, all objects will have several attributes, including strength, mass, buoyancy and momentum.  Half Life II may not look as slick as Doom III, but it sounds like it's going to be lots more fun.  We'll just have to wait and see. 

The good news is, new games with new features are just around the corner... well, maybe around a few more corners.

                                            Dr. John

May 18th

Cheating in Multiplayer Games

Tom's Hardware has an interesting article on cheating in multiplayer online games.  He goes over various common methods of cheating, including the "client hook", the "OpenGL hack", and the "hard-coded hack". There are at least 3 hacks available for my current favorite game, BF1942/Desert Combat.  One is called the accuracy hack, which reduces weapon recoil, making aiming easier. 

It seems to me we need a national directory of cheater's online names, so we will know which servers to avoid. As cheating gets more pervasive (mainly because kids care more about winning than playing), I expect the more mature players to form clans and set up private servers to keep the obnoxious little kiddies out.

                                            Dr. John

May 16th

NVidia Drivers May NOT Boost Performance

When I mentioned the new NVidia FX driver release, and said that it looked like it really boosted performance more than other driver releases, I may have spoken too soon.  Extreme Tech has  an article on their findings with NVidia's newest drivers, and they suspect that the drivers were artificially boosted by static culling of out-of-scene polygons, which basically is a fancy way of saying they cheated.  If true, then NVidia loses some of their lead over ATI, as well as losing some prestige.  We'll have to wait and see how this pans out, as NVidia hasn't officially responded yet. But it looks like another stumble for NVidia in their cut-throat competition with ATI.

                                            Dr. John

No Multiplayer for Doom III

Doom III may be the most talked about game of all time.... hell, it's been talked about for years, and it still isn't out.  Reports from the E3 conference in LA indicate that Doom III is a single player only game.  While that may allow better graphics, it reduces replayability to almost zero.  I typically can only play a single player game two times through before I'm done with it, maybe 3 if it's a great game.  But that doesn't take very long, and that's it for the game.  Good multiplayer games, on the other hand, like Team Fortress, are around and still being played years after their single player versions are dead (Half Life, in the case of Team Fortress). Oh well.   Doom III will still set the standard for graphics in new DX9 games, but it won't be around nearly as long as lesser games that have good multiplayer action. Maybe it will live on for years as a 3D benchmark, but I can't see anyone playing it 10 times or more.

                                            Dr. John

May 15th

SCO Has Death Wish

What exactly is SCO? SCO is a company that owns the rights to UNIX, and which produces very expensive versions of Linux, as you can see on this page here. You may have heard over the last couple of weeks that SCO is suing IBM because they believe that IBM and other Linux developers have lifted UNIX source code in the process of building Linux.  Well now SCO has decided that it likes the way the RIAA is going after music file swappers, and has decided to go after Linux users in similar fashion. Considering the fact that SCO isn't taking on file-swapping college students, but is instead attacking a huge segment of the US business community, I have to think that they will not only fail, but they will hurt themselves badly, perhaps irreparably. It strikes me that SCO has developed a pathological death wish complex, and they need to seek professional help. They have a hard row to hoe to prove that Linux infringes UNIX copyrights, but that's for the courts to decide. Obviously, SCO thinks that it can shut down all free versions of Linux, and thus be the sole vendor of absurdly expensive versions of the same operating system.  Good luck guys, you're going to need it.

                                            Dr. John

Googlewash Update: Shortly after I wrote this, SCO yanked all their Linux products from their web page. But thanks as always to Google's cache, we present the page, sans pictures... here. Anyone up for the $2,199.00 version of SCO's Linux server?

May 14th

Detonator FX Drivers May Actually Boost Performance

The scuttlebutt on the web is that NVidia's new Detonator FX drivers offer a significant speed boost for GeForce FX owners.  We have heard these claims before, and they have almost always turned out to be wrong.  But this time the reports are coming from people testing the cards outside of NVidia, so these reports seem to be founded in some fact.  If the benchmarks pan out, you can expect a 20 to 30% speed boost in some benchmarks. 

                                            Dr. John

Notes: Anand has a nice (LONG) article on the NVidia GeForce FX 5900, with comparisons to the ATI Radeon 9800. In many ways, Anand is fairly hard on the Radeon, and especially critical of the new 256MB version of the card (which is slower), and the newest Catalyst drivers (which are slower, and cause problems).  The good news is, Doom III is very playable with both video cards, even though the FX 5900 definitely wins the day.

May 12th

GeForce FX 5900 Looks Pretty Good

The reviews have shown up on the web already, and the GeForce FX 5900 Ultra looks pretty good. It doesn't clobber the Radeon 9800 Pro, but it beats it soundly in many benchmarks.  The Radeon's antialiasing looks better, but the GeForce gets the speed crown. So either way, you'll have a great video card, meaning that fans of both company's products will be very happy with these new cards. The only question remaining is whether the $499 price tag for the 5900 holds.  It sounds a bit too rich to me.

NVidia is out of the dog house, and back in the game.  Let's hope the competition drives prices down, while increasing availability.  Expect both cards to be more plentiful in June. 

                                            Dr. John

May 10th

GeForce FX 5900 Womps Radeon 9800??

If the leaked benchmarks that hit the web yesterday are an indication of how good the GeForce FX 5900 card will be, then it looks like a winner for NVidia.  Of course we'll need to see many other benchmarks before saying anything one way or the other, but this little taste is tempting, as you can see.

PS, NVidia, if you want me to take this down, send me an NDA to sign and a 5900 to test! :) 

                                            Dr. John

May 9th

New NVidia Card to Debut Next Week

NVidia has officially admitted the failure of the GeForce FX 5800 video card, and says that the next card will be ready for debut this month, most likely at the E3 conference next week. But this doesn't ensure NVidia's success, considering that ATI's Radeon 9800 is going to provide some stiff competition. As it looks now, the NVidia card will be slightly faster than the 9800, but it will also be more expensive. It's my guess that NVidia would do better to reconsider the debut price if they want to regain market share from ATI. There aren't many people who are willing to expend $500 for a video card, which is the suggested retail price that has been floating the web.  Experience suggests that sales drop off fast as you exceed the $400 price range. 

So keep your eyes peeled for debut reviews, probably next Thursday or Friday.  Will the GeForce beat the Radeon again... and if so, by how much?

                                            Dr. John

Make a Robot From an Old Floppy Drive!

Fran Golden from Goodwin College in Connecticut has instructions on how to turn an old floppy drive into a small, functional robot. Very neat.  You may have trouble loading the pages due to excessive activity, but if you've been dying to do something with an old floppy drive, this is for you.

May 8th

No News is Good News

As they say.  Unless you're interested in SCO suing IBM for giving away Unix secrets to the Linux mob, or in the fact that Microsoft is teaming up with Best Buy to scam customers, there just isn't much IT news out there to be had. But that's probably a good thing.  It means Bill Teel isn't spinning the Rambus FTC trial, and that another glaring security flaw hasn't been discovered in Windows.  It means that NVidia's latest video card isn't an underpowered dust-buster, nor that the latest patch to Battlefield 1942 causes more problems than it solves.  It even means that there hasn't been another deadly disease outbreak in China, further crippling the already beleaguered IT industry. Indeed, it means that no more pre-emptive wars were undertaken yesterday, thus adding to the FUD that has helped drive the markets so low over the last year. So all in all, it's a good thing.... as long as this doesn't get to be a habit with newsmakers.  

                                            Dr. John

May 5th

Microsoft "LongHorn", How Good Bad Is It?

Microsoft is working diligently on their next operating system, called Longhorn. Paul Thurrott has a review of the latest Alpha build of the new OS on his web site, and I have to say, I'm not impressed. On top of being slow as molasses, it's still buggy as hell, and most or all of the "enhancements" seem trivial, and non-productive. Check out the article yourself, and see what you think.  To me, it's more eye candy, and no substance.  Are you dying to have "tumbling windows"? I'm not.  I want a solid, secure, and productive operating system.  But that seems beyond Microsoft's capabilities.

Side Note: I had another Windows XP activation nightmare over the weekend.  I was trying to get a new XP system "activated" because the 30-day grace period was running out.  The system had a brand new, shrink-wrapped version of Windows XP Professional installed.  Despite the fact that the system was on the Internet, hooked through a router, XP's activation routine could not find Microsoft's servers.  This went on for 2 days.  Finally, despite hating entering the 100 or so digits manually, I called the phone number and tried to activate the system that way.  The system said it could not activate my copy of XP, and that I needed to talk to an operator.  I waited until the person came on the line, and then read in the long number, 6 digits at a time. After a few seconds the woman said, "this copy has been activated already".  I said, "but it's fresh out of a shrink-wrapped package!" She then had to fiddle with the system for a few more minutes to try to generate a new activation number.  Finally, I entered the huge number into the empty spaces, and the activation worked. But what I'm wondering is, how did that activation code show up as being previously used?  Is it a bug in their number generation system, or have some people figured out how to generate functional activation numbers illegally?

                                            Dr. John

May 2nd, Dr John, Wash. DC

Bill Teel Plays Reporter at Rambus Trial

Bill Teel isn't a reporter, but apparently he wishes he were. Nonetheless, he is "reporting from Washington DC" on the current Rambus trial for The Inquirer. In a developing series of "reports" from the trial proceedings, Bill continues to paint the issue in terms of Rambus being the falsely maligned little IP company that just wants what it is due. But Bill isn't reporting from the DC courthouse, he is spinning the case as hard as he can. Here is an excerpt from his "reporting" yesterday:

"Long before Rambus entered JEDEC, it filed its original invention with the USPTO, known as the '898. All the claims that Rambus sought to add while they were members of JEDEC, and long after they left, are based on this original filing. Rambus' patent activity while it was in JEDEC is often misinterpreted. All the patents Rambus applied for while in JEDEC are merely claims on their original specification, which was filed in 1990. It would be virtually impossible for Rambus to obtain those patents without having a technical description to support them. Had Rambus submitted new claims (without a new technical description) that was not based on its original technical specification, they could not be granted. Likewise, had Rambus filed an entirely new patent that was purely inspired from what they saw in JEDEC, and then enforced that patent, the FTC would be on to something, and Rambus would probably not be entitled to enforce the patent. But, to our knowledge, nobody is accusing Rambus of stealing ideas from JEDEC in this sense (that's like, sooo Richmond 2001). Rambus had the patents (everyone agrees to that), and obtained them fairly, although possibly motivated to do so faster after what they saw going on in JEDEC."

Does that sound like reporting?  I have a couple of questions for Bill to answer.

Did any of the amendments to the original patent add additional details, make technical refinements, or change the specifications or parameters in any way that could be construed as possibly having been influenced in even the slightest way by JEDEC proceedings?

Were any of the additional patent claims filed by Rambus after "898" pertinent to SDRAM and DDR DRAM? If so, there is at least the appearance they were trying to broaden their patent scope to better encompass SDRAM and DDR.

If Rambus had the specifications for SDRAM and DDR DRAM filed with the U.S. Patent Office in 1990, why did they even join JEDEC in the first place? And if these specifications were "patent pending" at the time, and Rambus joined JEDEC while knowingly not disclosing the technical specifications under consideration (that were also pending patent approval at the time), then Rambus was clearly and knowingly in violation of the rules, as well as violating good faith.

Bill can't have it both ways.  He said that Rambus had SDRAM and DDR DRAM specifications in it's original patent filling. If so, then they joined JEDEC in bad faith, while awaiting approval of a patent that was directly pertinent to JEDEC's proceedings. And why did Rambus file other pertinent patents, and patent amendments while still in JEDEC? There is no logical way out of this dilemma for Bill or his pet IP company. Rambus was in JEDEC for one reason, to find out what they were doing, so that Rambus was in a better position to adjust their patents to make sure they encompassed the technical features under discussion at JEDEC. Any participation in JEDEC after the filing of "898" in 1990 (while not disclosing the patent application to the standards body) was in bad faith, as well as being demonstrably anticompetitive, and that's a fact.

Bill Teel owns lots of Rambus stock.

                                             Dr. John