INQUIRER flooded with RIAA, Apple letters

First, Apple
Intel, AMD accused of cooking the books and Senior chip analyst says Apple lying about PowerMac G5

'He claims that Apple is the "first and so far only company" to market a 64-bit RISC machine as a desktop with prices to match. And, he adds, the Athlon 64 is a "dumbed down" relative of the Opteron. Intel has zip".'

Isn't the PowerPC 970 a dumbed down version of the IBM POWER 4? The Athlon 64 appears to be quite similar to the Opteron, perhaps even identical. Apple has the gall to say otherwise, especially when the PPC 970 is far from the POWER 4? Apple needs to get a grip on reality.

Sincerely,

Micah B. Haber

'And, he adds, the Athlon 64 is a "dumbed down" relative of the Opteron. Intel has zip".'

Isn't the processor in a G5 also a bastardized version of a server chip?

Andrew Wittbrodt

Opteron? Don't make me laugh. I've got an old Alphastation 400/233 that was a 64-bit desktop in 1994, and it was second or third generation... Apple was still selling 68000s back then.

One can laugh at Apple for confusing "desktop" and "personal computer", maybe, but anyone who thinks a dual opteron counts as anything other than the first dual opteron is just as confused.

Name supplied

Dear Mike,

Kudos to the inQrew for great coverage of all things geek.

Many people seem to miss an essential point or two about Apples 32/64 bit strategy... In the WWDC Steevenote, Jobs practised common PR lingo, transmogrifying "workstation" into "desktop computer" (that should've raised eyebrows: maybe Apple is prepping monster workstations to back their claims), which is on par with Microsoft "innovating" (squirming) their way out of bloat-ware categorisation, and maybe less of a sin. What characterises the media's bias is total silence on Ballmer's dubious "Wild Orang-Utan vs. Howler Monkey" impersonations... OK, that was a cheap shot, but boy did it feel good!

First of all: efficiently rewriting all of OS X.3 in 64 bit code (which was a brainless rumour in the first place), would take a great deal of time and would not necessarily turn it into a much, much faster OS (only applications wielding massive amounts of RAM and number crunching truly benefit from 64-bit goodness); however, key portions of Panther DO take in account the 64-bit architecture of the new G5 mobo, but as neither are shipping products, we can only speculate about how the whole widget will perform.

Secondly: Apple's current hardware is mainly powered by 32-bit G3/G4 chips, and will remain that way for the foreseeable future, therefore OS X has to remain compatible as an all 64-bit Panther wouldn't work on the majority of Apple's non-64-bit catalogue. That wouldn't bode well for OS X's slow acceptation rate in the existing user base... As a fervent Jaguar/PowerBook G3 user, I hope that Panther's added value will accelerate the migration to X, and frankly, I think it'll fare quite well.

The big picture here is Apple's rather spotless record of bridging paths for technology transitions: witness the black-and-white to colour graphics with the Mac II, CISC to RISC with the PowerPC chips, clunky legacy ports to practical USB and FireWire with the original iMac, proprietary OS 8/9 to open-source (well, some of it) 'NIX flavoured OS X with built-in Classic and Carbon support, and now, a 64-bit platform that supports 32-bit applications without the shortcomings of other platforms. No rebooting, no slow-downs, nary a hiccup. Would be nice to see the mainstream media run a piece along that tune, "To Stability and Beyond!".

A lot of people should get rid of their trifocal SPEC glasses and wait for the official release of the Panther/G5 combination to prove it's real-world performance. There's still place for optimisation, awe and wonderment from us Macheads, gnashing of teeth and wailing from the WinTel crowd..

Either way, it's Apple's call.

Best regards,

Robert Emslie

If Applelust.com claims this:

"No one publicly questions how Intel arrives at their claimed scores. It is a poorly kept secret that everyone in the industry knows that Intel uses special 'hot box' motherboards and 'reference compilers' to get the scores", he said."

Then the site authors can't read. From the SPEC web-site, we find this:

http://www.spec.org/cpu2000/results/res2003q3/cpu2000-20030630-02332.html

entry submitted by Dell on a Dell machine, not an Intel 'hot box' special motherboard. Dell benchmarked using the "Intel C++ Compiler 7.1" compiler, which is a commercially supported compiler and can be purchased (or downloaded for free personal use!) from here:

http://developer.intel.com/software/products/compilers/cwin/pricelist.htm

It is true that the Intel compilers give better results than gcc or msvc++, but since they are commercially available, I don't see what the big deal is (yes, most developers *don't* use them ... but that doesn't mean that the CPU/chipset is any slower).

Regards,

Mark Roulo

With all the debate going on apple's 64 bit desktop claim, I'm almost amazed that I have yet to see Sun's SunBlade 100 or 150 mentioned. I may be partial since I own one, but has Sun really become the Rodney Dangerfield of the industry to the point where no one even things about the systems they offer in this debate? For starters, the blade runs Solaris 8 & 9, which ARE true 64-bit OS's. On top of that, the entry level blade I purchased a year and a half ago cost $1,000. So, am I missing something here as to why this machine hasn't been brought up in this debate?!? It isn't just you, no other news site that I've seen has even mentioned the Blade, which does indeed have a small, but loyal following.

James

Sigh. (directed at the person quoted in the article, not you, Dear Editor) Take a moment to research the DEC Multia/UDB line of computers, marketed starting about 1994. The UDB stands for (wait for it) Universal Desktop Box, and yes, every member of that line of mid-1990's desktop computers was fully 64-bit.

Google the web or usenet on udb and multia, and pick most any return to put these BS claims to lie.

Name supplied

Mike,

You must acknowledge that although Panther is 32bit, it can leverage the 64bitness of the G5 in every sense of the word. Therefore it is possible have have 32bit 10.3 and a tweaked Photoshop (to take advantage of the G5) to get at least a 100% increase in performance. In AMd land, this is not possible - as far as I know.

Regards

Andrew Sheridan

Hi Mike,

Can you please put me on record as saying at Joe C. Carson of Applelust is the "dumbest @$£$"$$"$"!" I've come across this year? It's time for my 15 minutes of fame damn it (can't wait to hear the laughter at the office as people reload the inq!)

Name supplied

The title of this article is horrendous!

Nowhere in this article do I see where Apple has lied about anything. Let's start with the first 64-bit desktop issue.

Your article claims that there were 64-bit consumer model desktops in the 90's, but names none of them. Where are these desktops? You claim that they ran Windows NT, but Windows NT was never marketed as a home-user solution for computing. That was always the business model, which is geared towards workstations, not desktops...

Now, your article claims that Apple lied about the benchmarks, in that this Peter fellow was able to get higher marks. Yet nowhere in your article does it say if he used the exact same benchmarks, under the exact same conditions, with the exact same compilation flags. That's pretty irresponsible journalism.

Look, I wouldn't be getting so up-tight about an article like this, if not for the fact that you guys blatantly called Apple a bunch of liars. Saying that they flat out lied about the G5. That's a huge accusation to make, and then "back" it with only the claim of this Peter fellow to which you guys give no link to the original article, no mention of who this Peter guy is, nor any contact information for him.

I read The Inquirer religiously. I think you guys do great things, and I love your content; however, this story is over the top. It's drive-by-journalism. Make a couple of accusations, and then "back" it with three obscure sentences. Damn guys.

Jesse Thompson

CS Student

Hi,

Just a comment about today's article "Senior chip analyst says Apple lying about PowerMac G5".

In it, you quote Peter Glaskowsky accusing Apple of irrelevance, saying: "Apple knows perfectly well that clock frequency per se is essentially meaningless".

This is a very odd statement to make. Does Glaskowsky think that Apple (not Intel) has been pushing it's marketing based on processor MHz? Does Glaskowsky think that Intel came up with the marketing phrase "The Megahertz Myth"? [Wasn't that us? Ed.]

In the article you also mention "benchmarks Apple published for the Pentium 4 and the Xeon are "considerably higher than those that Apple reported". "

Apple has quite clearly stated that the SPECfp_base2000 and SPECint_base2000 benchmarks used compilers that were not tweaked specifically for a given processor. The published SPECfp_base2000 and SPECint_base2000 benchmarks on Intel's site are based on their own optimised compilers. Apple hired Veritest to do these benchmarks, and they were done on a level playing field.

But, ultimately SPEC benchmarks are not important to the end user. What is important is system performance using real-world applications. Apple has demonstrated the G5 against the Pentium 4 and the Xeon, using real-world applications, and the G5 shines!

Harvey Lubin

And now the RIAA
Abit throws gauntlet down to RIAA, governments and RIAA will take 2191.78 years to sue everyone

[Abit article, Ed.]

A bold statement it is, and a stupid one.

All SecureIDE does is, to some degree, prevent the RIAA from accessing your data if the HDD is removed from the computer. The key here is that the encryption is transparent from the operating system's perspective, no device drivers or things of that sort are required to use SecureIDE. This means that the encryption, while being real and indeed securing to a point, doesn't do anything to elevate the general level of security in a PC system. For example, all flaws inherent in the operating system are carried over the minute a user starts the computer with a valid key. As long as you start the computer with the valid key, hackers and scriptkiddies have exactly the same opportunity to get at your data as they had before the SecureIDE was installed, since the operating system doesn't know there is encryption done at all and no special disk access control is used either.

What does this mean for the average user?

If someone buys the SecureIDE hoping to use Kazaa or any other P2P file sharing program under the RIAA's nose without them noticing they're terribly mistaken. The electronic footprints left by them on the Internet will be the same, SecureIDE doesn't protect them there, meaning the RIAA will still have as good a lead on them as without the SecureIDE. And if the RIAA should be interested in the data on a user's HDD to such a point that they will actually confiscate the HDD in question, they will surely see the SecureIDE in the computer and come to the conclusion that 40-bit DES encryption is used. If so, they would probably not mind spending a week cracking it using "government supercomputers". If they feel up for it, they may even acquire a specialized DES cracker and crack it all in under 72 hours (http://www.eff.org/descracker/), although getting good plaintext candidates may be harder than with written language which may prolong the procedure somewhat. After all, $250,000 is small potatoes when you're handing out subpoenas left, right and center to anyone who even remotely looks like they're using Kazaa.

This is nothing more than a novelty piece of equipment.

I've written some on the subject here too: http://www.abxzone.com/forums/showthread.php?s=&threadid=51265&perpage=30&hi ghlight=secureide&pagenumber=2

I'm afraid this is not actually true.

Taking Michaela Stephens numbers, then it will take an infinite time.

If you take 60 million people and assume that the same % of the US population share files.

According to the CDC 1.02 % will die per year and 1.45% new ones will be born.

Thus they must sue 60,000,000 * (1.45-1.02) =258,000 people per year, merely to keep the number constant. Thus they must sue 706 people per day, just to keep up with the steady state. At a mere 75 per day, they are walking very slowly up a very fast down escalator at one tenth the speed of its acceleration. I could add that population growth is exponential, so it is actually worse than this.

But that is not the end of their problems.

There are approximately 300 convicted serial murderers in the USA. Their acts are hard to predict, but assuming that a file sharer is as likely to be a homicidal maniac as anyone else, something like 10% of those currently free within the USA will be sued by the RIAA. I of course would not condone the murder of RIAA executives or lawyers.

However I would not personally spend vast sums in antagonising them either. Each year in the USA there are about 30,000 "one off" murders. Again assuming that file sharers are a typical cross section of 10% of Americans, some 3,000 people who will choose to kill someone will be sued by the RIAA every year. Many Americans have guns, so over a 5 year period the RIAA will antagonise an army larger than the front line forces available to a country like Britain or Germany, Although their armed forces are larger, most are engineers, cooks, admin etc, only a tiny % of a modern army ever gets to hurt anyone.

However the violent criminals of the USA are people who have first hand experience of violence, and since the police are reasonably competent, those at large are those smart enough to avoid capture.

Of course some more peace loving Americans, will merely break the limbs of RIAA lawyers and staff, given the assault rates in the USA, group will number in the hundreds of thousands, a force larger than the US Army, including its cooks and cleaners.

The US Secret Service is highly professional, armed to the teeth, and given free reign to do anything it likes. However even then presidents under its care routinely get shot. Same principle applies, does not matter if you are Reagan, Kennedy or a mythical Bartlett, if you are hated by millions of people, one will get to you, doesn't matter if you deserved it.

I would again emphasise that I don't condone violence, nor wish to incite it, however if you upset a large number of violent people you will suffer far worse consequences than a denial of service attack on your website.

Name supplied

A minor caveat about these figures:

I should point out that we'll be having quite a few leap years in the next two millennia, and Michaela has failed to account for this.

By working every February 29, the Recording Industry Association of America will be able to shorten their litigation orgy down to a mere 2190.28 years, a savings of almost 18 months! Since we're assuming that they will work all day long without weekends, holidays, or possibly sleep, they should be able to use this time to get a few cups of coffee at Starbucks or something. I imagine that it's going to take a lot caffeine to finish this project of theirs, and a double-shot espresso or two (or point-eight million) might help.

James Ayres
Litigant no. 1927254

I was intrigued by Michaela Stephens' proof-like observations regarding RIAA's Music Swap Combat Tactics and her conclusion that they would be suing our great grandchildren before they were done pestering the present generation. It seems the Congressmen geezers in Washington, D.C. need a math lesson as well.

Recently, Michigan congressman John Conyers introduced in the House the `Author, Consumer, and Computer Owner Protection and Security (ACCOPS) Act of 2003'. This bill would make sharing copyrighted material on any publicly accessible network illegal, punishable by prison time and fines, regardless of whether the file has been downloaded.

Before I talk prison time or fines, lets do a little math. Sec. 101 of the bill states that the Feds get $15,000,000 to investigate AND prosecute these cases. What's that Michaela? 60 Million P2P users swapping music? Interesting. Even if only 30% of P2P users are actually sharing 'their' music, that is 83 cents per case for the Department of Justice to investigate and prosecute. I can't even get a 20-minute long distance call (or 7c/minute after that) with that.

All joking aside, if they hire some Ex-Enron accountants, and find that 15M is enough to start prosecuting, you might want to know the bill calls for more than the taking of your internet privileges. The guts of the bill go thusly: "`The placing of a copyrighted work, without the authorization of the copyright owner, on a computer network accessible to members of the public who are able to copy the work through such access shall be considered to be the distribution, during a 180-day period, of at least 10 copies of that work with a retail value of more than $2,500.'" I had to read that 27 times before I fully understood what this guy was trying to say.

Basically it says "If you share 1 file on 1 network, Big Brother would treat you AS THOUGH you have distributed (at least) 10 works worth $2500 in 180 days". The real reason it is phrased as such is the copyright laws on the books state the 10work/$2500/180day minimums need to be met in order to be considered a crime. It seems that they have hired some really great mathematicians to demonstrate that one electronic file sitting on one public network, regardless of value, is greater than or equal to 10 physical copies of the same work and $2500.

Bottom line, this buys them the right to put you in jail for 3 years because of 1 file on 1 network (even if it is never accessed). As a side note, I noticed the law would not apply on a "member's only" network, as it would no longer be a 'public network' as the law requires. Very Interesting :)

Will Ackerly

Here is a link to the Bill: here

The US Code it Leverages is http://www4.law.cornell.edu/uscode/18/2319.html

(P.S. Conyer's office is in the Rayburn House Office Building)

Screw the length of time to file the RIAA lawsuits, look at the COST!

A friend of mine assumed 25 million P2P users, and a jail sentence of 1 year each.

25 million American P2P users X 1-year jail terms each X $19,075 per year = 4.76875 X 10^11.

That's 476,875,000,000. Half a TRILLION dollars.

Where is the money for this going to come from?

Using Michaela Stephens' estimate of 60 million P2P users, that cost shoots up to over a trillion dollars.

Scott Bland