Standard of journalism at INQUIRER sinks to all time low

Where's the picture of the Linksys rep?
Of PC Celebs and Snacks

I have to say I was seriously disappointed today. I find that the Inq has always had the highest standards of Journalism. Yet reading the discussion about CeBIT America I was thoroughly and total disappointed.

The Inquirer has always made a point about giving its readers what they want. Yet this article was seriously lacking in one particular area.

And I quote: "I suppose the big thrill of the night was gawking at the Linksys rep in her black dress with plunging neckline."

Right. So where are the pictures then eh?

It was long ago said by some drunken Journo somewhere "give the people what they want" Well we want pictures of the leggy linksys rep, and we want them now.

Most of us out here are a sad lot of geeks who couldn't get a date in a whorehose. To make an inflammatory statement like that and not following up with a couple of pictures of said neckline, is in my opinion grossly incompetent.

Name supplied

Got USB? Confused? Surely some mistake
USB Forum creates USB consumer confusion

My understanding is as follows:

USB 2.0 is the first revision of the USB standard that supports 'high speed' 480Mbps data transfer. It's backwards compatible with earlier USB standards - the 'low' and 'full' speeds specified in USB 1.x - in the sense that you can plug 'low' and 'full' speed devices into a USB 2.0-compatible port.

It appears that you may be able to plug a 'high' speed device into a USB 1.x port and it will still function, but only at 'full' speed (I bet they regret that name now...)

Therefore they're saying that devices should be marked simply 'USB', not 'USB 1.1' or 'USB 2.0'.

Of course this confuses things like my new Asus P4T533 motherboard, which has both USB 1.1 and 2.0 ports on it (two of each on the back panel, plus headers for two more of each on the board). Oh, I'm sorry, I meant to say "two 'Hi-Speed' compatible and two non-'Hi-Speed' compatible" ports. Doing things like that is bound to confuse people...

Mike Dimmick

Hi,

Thanks for the article about the USB.

A good example of this can be found with the Canon scanners; the Lide series 20 and 50 were openly advertised as USB 2.0 (at least here in France) and it turned out that they were far from reaching the 480 mbps that we were supposed to expect.

It's really a shame because they are excellent scanners and I am sure people would be ready to pay the little extra difference to get real USB 2.0

Keep on the good work,

Sebastien

There is nothing wrong calling a USB full-speed device "USB 2.0" as the device meets the USB 2.0 spec. What you need to look for is the "high-speed" classification (and logo - which you have in your article) which indicates the 480Mb/s speed support.

Companies who sell product as "USB 2.0 full-speed" are not mislabeling - in fact, that is exactly how it should be labeled. Part of the job of being a consumer is to be informed - don't blame the USB developers if you fail to follow (or understand) the technology.

Chris

[Being a consumer is a job. Now you know, folk. Ed.]

Where's the confusion?

USB 1.1 Supported 12MBit (full speed) and 1.5MBit (lo speed). USB 2 Supports 3 speeds of device 480MBit (hi speed) and the two USB1.1 speeds.

Saying something is USB2 doesn't say how fast the link is in exactly the same way that USB1.1 didn't specify the speed.

Who needs a 480MBit link to a mouse or keyboard anyway?

Andrew Ansell

Pot calls kettle black
AMD's Opteron: does it measure up?

Dear Mike,

[This] is an incredably misleading article. Just considering the Xeon MP vs. the Opteron (similar arguments exist with regard to the dual Xeon and the Itanium 2), it is preposterous to ignore the 2.8 GHz CPU which will be released in fewer than two weeks and which will murder the Opteron, even if AMD raises its speed to 2.0 GHz simultaneously. Furthermore in this regard, the observation that when the shrink of the Opteron to 90 nanometers occurs the secondary cache size is likely to be increased to 2MB ignores the fact that the Xeon MP already has a 2MB cache, will have this increased to 4MB in Q1 2004, well before a 90 nm Opteron is produced, and will likely have an 8MB cache when 90 nm Potomic comes out, probably before the 90 nm Opteron emerges.

In an article which justifiably criticizes abundant media sources for unbalanced reporting, it seems the pot is calling the kettle black.

Best regards my friend,

Stuart B. Arm

Mike,

I've always enjoyed INQUIRER articles, even when occasionally I didn't agree with them. And I truly enjoyed Mario's recent article on "AMD's Opteron: Does it measure up?" but found one minor or major flaw in it, depending on how you view it.

Your assertion that the extra two stages of Opteron over Athlon would enable higher clock frequencies was incorrect in this case. While extra stages CAN be used for splitting the workload out among the additional stages if designed to, this isn't the case with Opteron. The 2 added stages to Opteron are not there to share any other stages workload, but are there for the purpose of encode/decode scheduling and will increase IPC (I know I don't have to explain IPC to you) only, not clock speed. If anything, they (the extra two stages) may prove to increase the difficulty of clock speed ramping of Opteron and this may be a part of the reason why AMD is having difficulties hitting higher clocks, though certainly not the only reason (SOI difficulties come to mind here).

Don't get me wrong, I'm not gloom and dooming Opteron. I respect the massive increase in overall performance per clock the Opteron touts, but I fear that clock speed itself may be sacrificed for higher IPC to the point of ending up like PowerPC, under-appreciated unfairly due to low clock speeds (which, by the way, could be part of the IT industry's reluctance to blow the Opteron horn, beyond the strong-arming that has occured from a certain company).

I found the rest of your article's reasoning to be solid overall and share your feelings that Opteron could squash Itanic's hopes for survival if implemented correctly. Lets see what AMD has up their sleeves next, eh?

Greg Manning

Hyperthreading. Will Lotus bring out
1-2-3 V4 faster with 800 programmers?
HP warns hyperthreading might not work

The views expressed herein are not those of my employer, etc.

Before I launch into this, I love your site and most of your articles - but sometimes your urge for cynicism (which I normally appreciate) makes you put a downer on something for the most trivial of reasons. Your article "HP warns hyperthreading might not work" is a great example. You jump on a true statement by HP, contrast it with a non-contradictory quote from Intel, and make a story which heavily implies lack of performance improvements with Hyperthreading. A hell of a lot of discussions and articles involving Hyperthreading miss out one very important item, including this one.

The facts are these -

1. Hyperthreading can not improve performance of a single-threaded application running on it's own (like during benchmarking a single app). Intel have never said this to be the case - they consistently mention 'multi-threaded applications', including in your article. In reality, a single thread may run a little slower than on a non-hyperthreaded CPU of the same speed. I have seen this myself, but it's not to a degree that's a problem in any way.

An analogy is two office workers being available to do a job. If a job cannot be split up, it's better for one worker to stick with the job, getting it done, than two workers swapping over all the time, taking turns at doing parts of the job.

2. Hyperthreading can improve performance of multi-threaded applications, at times when multiple threads are trying to run at the same instant. In our imaginary office, two (or more) jobs now need doing. Clearly it's better for each worker to do a job - even if they have to share bits of office equipment, it will get done faster than with one worker constantly swapping between the two jobs, doing little chunks of each till it's done.

But there's a third fact which has consistently not been mentioned in articles I have read.

3. A PC and user using multiple single-threaded applications can benefit from Hyperthreading. If I'm running Outlook, Track-It, Phoenix, Exchange Administrator, VNC, IE, and MMC (as I am now), Hyperthreading would allow two of the active processes (including all the background things XP is doing, the virus checker, etc) to run at the same time as opposed to one. With our office workers again, if they are very busy with many things needing doing ASAP, obviously it's better for two of them to be there than one.

Hyperthreading is not a panacea, in the same way dual-CPU desktops never were. However, in the modern world PCs are only ever running one app when someone is benchmarking them. Hyperthreading does help when lots of apps are vying for CPU attention, and I speak from experience. Hyperthreaded PCs feel much smoother to use when under load than non-hyperthreaded ones - not as smooth as a true dual CPU box, but noticeably better than an old-school one-thread-at-a-time one. Intel know this, HP know this, and so I imagine do you boys. It'd be nice if you reflected that once in a while. Intel haven't given us world peace and the abolition of poverty, but they have given (OK, sold) us CPUs that improve my working experience by helping the OS respond to me faster without costing the earth that a dual-Xeon box would have.

Your title "HP warns hyperthreading might not work" isn't even true. HP warns that not all software will benefit from it. Very true, but 'might not work' is a hell of a stretch from there. I personally wish to hell my brain could do hyperthreading.

Regards (publish if you want),

Dan

Dan Lawrance
Technical Consultant
Communisis

Shooting, hanging, burning
is too good for them

Dear Inquirer,

I must agree wholeheartedly with Randy "Kill The Pirates" Hubbard.

The laughable "due process of law" provisions set down in the United States Constitution's Fifth Amendment are a complete waste of taxpayer resources. They allow the guilty to go free just so the innocent may avoid unjust imprisonment.

A far more efficient system would authorize law enforcement agents and private citizens simply to execute any suspected criminal or "lowlife thief" on sight. Any citizen foolish enough to open an e-mail attachment containing a trojan that sets up a covert warez server deserves to have all their personal data -- as well as their expensive PC hardware -- destroyed by RIAA hackers.

This sort of punishment could justifiably extend to those whose computers secretly participate in DDoS attacks or spam dispersal, and corporations whose employees engage in off-shore Internet gambling.

This would be a VERY practical means to make an individual whose trust is abused pay a real penalty for their crime. Kill 'em all and let God sort 'em out! RIGHT ON, Randy!!!

With much love and sarcasm,

Dylan P.

Has Randy considered the issues involved if a mistake is made?

I run a licenced copy of Windows 2000 Professional as well as a legitimate copy of Gentoo Linux. I also have a file called "Office.iso" on this machine, which I share publically; it is actually an ISO image of OpenOffice materials, not of MS Office, but again it may convince someone I'm involved in copyright violation.

If someone acts to damage my PC remotely, in a manner authorized by this bill, who is liable for the resulting mess? The copyright holder who did the damage presumably will not be liable; if they are, what stops a copyright violator claiming after his PC was rendered unusable that he wasn't engaged in illegal acts? After all, there is now no evidence to show that he wasn't sharing OpenOffice, wxWindows, or other similarly named files that are legitimate to share.

On the other hand, if I have to pay to clean up the damage, what stops me releasing a file called "XPSP1", and then exploiting the bill to damage Microsoft's systems for their role in apparently illegally distributing my file? Surely if they are not liable for the damage to my machine, since they acted in good faith, I cannot be liable for the damage to their systems? Or is there going to be one rule for rich copyright holders, and one rule for poor ones?

Points to think about,

Simon Farnsworth

Yeah!! What a great idea!!

While we're at it, if you see someone parked illegally, why wait for the traffic police to do something about it? Just get some large, heavy object and smash up their car. That'll learn them!! And this applies if someone cuts you off too.

Also if you have any suspicion that your neighbour may be doing anything illegal, how about setting fire to their house? This will curtail most illegal activities they may be involved in and teach them a good lesson too. No proof? No problem! Judge, jury and executioner!

The best thing of all is.. this works just great even on technicalities. It doesn't matter whether there's anything IMMORAL involved... hell, forget morals altogether! The law knows best! Now get out there and BREAK and BURN!!!

Nicholas

In response to Randy Hubbard's letter regarding the destruction of people's PCs as a proper response to MP3 "piracy", I ask what we should do with all those "lowlifes" who violate the speed limit? Should we have the "criminals" cars blow up when they are trying to pass, or keep up with normal traffic? How about someone going 80MPH in a school zone?

Should all the people who jaywalk have their legs amputated, or just the (hypothetical, in this case) people who purposefully dart in front of cars to cause accidents? "Theft is theft" is a very naive statement. Lessig is right: there must be leeway built into these DRM systems. After all, if a friend wants to lend you the tape he made of your children's school performance, so you can make a copy, should the FBI shouldn't rappell in through the roof?

If Mr. Hubbard has his way, of course, they wouldn't. Your PC would just stop working, seal your hard drive for "evidence purposes", and send home a brief notice of infringement to, say, Microsoft and the Ashcroft Justice Department. "Shurely shome mishtake" (as you are fond of putting it) "World Controller Gates, sir..."

Abe J

PS As a side note, those spammers may be annoying, but some of them are just sending a lot of e-mail, and not breaking any laws. Does a modern - fully legal - annoyance really modern breaking, entering, and willful destruction? Well, I don't like SUVs anymore than the next person, but that is too far...

Hmmm, not as bad as calling it terrorism; of course, downloading mp3's causes massive amounts of terror to the music companines. Causing widespread panic to society? I don't think so.

That's nice Randy. Oh we accidently destroyed the wrong computer, sorry to have destroyed irreplacable data on it with all your families pictures and cause days of downtime as I reinstall everything.

Keep supporting the RIAA and it's music companies extortion of music prices, and the price fixing involved which otherwise is illegal for others to do. Keep supporting the labels who make all the money off of the artist and try to take ownership of the music from the maker of it.

Keep supporting Disney who will lobby for copyright laws to last for eternity, or just keep making it longer everytime it's about to expire, even though the acutal maker of the material is long dead.

Oh sure you can download music legally, just the kind that you never own or if you can it's not cd quality. And keep supporting legislation like this because the DMCA is not enough; it's a crime to get music to play on your computer by bypassing the so called piracy detection on it. F.Y.I. remember back in 1993 when minidisc came out with SCMS?

That was due to the good old music compaines making sure that multiple digital copies couldn't be made. In order to pass that law, computers HAD TO BE EXEMPT due to the inconveniences it would cause to put in protection (oh that's why I can use normal data discs in computer to burn music...). Nowadays that doesn't seem to matter.

Force is not the answer. "The more laws, the more offenders." (hmmm... where did I get that from?) On one end of the spectrum is the blatent pirates, at the other end is the copyright owners who won't let go of the material or sell it overpriced.

There will never be a balance between people who want stuff for free vs. those who want it to be high priced and that you will never own the material (Microsoft anyone?). What is the logic of selling a brand new cd and one that came out 40 years ago for the same price? Or the fact that I have to pay for a replacement cd everytime it's scratched beyond repair?

Of course for those of use who run Linux/Mozilla they wouldn't even know how to destroy the system.

Name supplied

"The laughable "due process of law" provisions set down in the United States Constitution's Fifth Amendment are a complete waste of taxpayer resources. They allow the guilty to go free just so the innocent may avoid unjust imprisonment.

"A far more efficient system would authorize law enforcement agents and private citizens simply to execute any suspected criminal or "lowlife thief" on sight."

Actually, with passage of Patriot Act I (and soon Patriot Act II), secret detentions/arrests/trials/executions are now the law of the land. DOJ has even given itself the power to execute even those who are not even suspected of having committed a crime. Israel has also publicly declared its intention of assassinating on American soil Americans and others that the Israeli government considers 'terrorists'.

This policy has actually been in effect since Waco, where all 80-odd inhabitants of the now infamous 'compound' were executed by members of the U.S.Delta Force at the direction of the White House with members of special forces from several foreign nations looking on as observers.

See

http://www.dabney.com/wacomuseum/
http://www.infowars.com
http://www.infowars.net
http://www.prisonplanet.com
http://www.propgandamatrix.com

Dave F

FUD's thud bud
HP slings mud with FUD about Sun's future

Mr. Magee

Your article states that HP sales reps are telling potential customers that Sun's "UltraSPARC III has reliability problems and "glitches" that result in customer systems going down."

As a UNIX admin for a Fortune 5 company (whose name I prefer not to reveal) I can tell you that HP's claim is not FUD but fact. We run scores of Sun Fire systems, and there is definitely a problem with the quality of the 900MHz processors on the Sun Fire uniboards. They are subject to unpredictable failures that can (and do) lead to system unavailability. The problem is sufficiently acute that Sun has agreed to replace all the 900MHz processor uniboards in all our Sun Fire systems with the 1200MHz processor version at Sun cost (not expense, but cost). This effectively gives us a very (well, relatively) inexpensive upgrade path, and can certainly do nothing to help Sun's bottom line!

Please keep my name, and that of my company (should you discover it) confidential.

Name withheld. [Company not yet discovered. Shouldn't be too hard though, Ed.]