[There's been a lot of correspondence this week, some down to struck nerves, others the usual insanity. Here's a digest of just a few that make us laugh/cry]

One always gets through the net

Spin, spin, spin !! Aren't you getting dizzy yet? Rambus too expensive - Intel .... duh !! You look like a deformed version of Hugh Grant's "flatmate" in the movie Notting Hill... except HE looks intelligent! What the fuck's up with you boneheads? Flatmates butt crack signifies your IQ. What credibility will you have when this is over? Rambus is NEXT idiots !!!!

[I picked up a nasty cold this week - as the other Reg hacks and the bin of used tissues will testify. So it seemed only logical to do a story on the Net's advice on colds and flu]

Greetings from a fellow sufferer thousands of miles away (Sydney, Australia).

The principles behind homeopathy are insane. Which wouldn't matter, if it worked, but there is no evidence to suggest that it works better than placebo for anything. Just because the Royal Family use it doesn't mean it's good. If you're going to take Prince Charles' advice on matters of science, you're beyond help.

Large doses of C can give you the runs. Which is all you bloody need when you'd much rather spend the whole day in bed. You do need quite large doses to risk a case of the liquishits, though. A hundred honkin' milligrams of codeine will make the problem go a LOT further away. For that, though, you need codeine without paracetamol or other annoying contaminants, as are present in the codeine pills available over the counter in most countries where ANY codeine is available over the counter.

My own, personal recommendation? Herbal medicine. A number of reputable establishments sell a variety of healthful infusions of juniper berries, anise, caraway seeds and other life-enhancing botanicals. The clear beverage in question is given a long shelf life by a 40% alcohol content, and goes nicely with tonic and a slice of lemon.

Daniel Rutter

I read somewhere recently that large doses of Vitamin C dramatically increases your chances of heart disease. It seems that we just can't win.

Tom Condon

I'm only mailing because I found this out the other day myself, but you use zinc tablets (or whatever format you've got, bar bits of metal) in conjunction with vitamin C because one makes the other get absorbed by your body much better, and therefore helps against your cold. However, the guys telling me about it then started arguing about whether it was the Vitamin C that fought the cold and was speeded up by zinc, or whether it was the zinc that fought the cold with Vitamin C as it's speedy sidekick (the idea there being that we've always got some zinc in our bodies so the vitamin C tablets help us break it down). I've seen some tablets in chemists across Europe which contain a mixture of both. I gave up at that point and went home for some Lemsip.

Regards,
Mario Miniaci

[We covered BT announcement that it was to delay its rollout of ADSL again. This is about the fourth or fifth time. Unsurprisingly, we received an enormous number of letters - some angry, some very balanced. In fact, there were so many that we won't put them here or the letters bit will be too long. We'll set up a special ADSL letters bit next week. Okay?]

[Mike covered a new publication by PR firm Miller Freeman which promised to tell IT like it is in terms of journalists working in the industry. Needless to say, we tore it to pieces. We also received an entertaining email from Computer Weekly's news ed.]

Hey you,

As a regular reader, I've come to expect slip-shod foppery from The Register, but your story on the "Miller Freeman PR Audit" takes the whole f@cking cookie jar.

Where do I start?

1) its Clark - no "e".
2) the most defamatory mistake, I work for the UK's leading business IT mag Computer Weekly, not VNU fly-by-night cowboys Computing.

How dare you! And no, you can't just say, "Oh well, it's Miller Freeman's fault". Only limp-wristed journo ammeters hide behind that kind of excuse. Check your facts before you put your stuff on the Web.

Lindsay Clark
Bad-to-the-bone News Editor
Computer Weekly

PS Everyone on The Register smells of wee and Mike, I hear you drink like a girl.

Disclaimer: Everything in this email is true and I mean it. Bite me.

[The one-liner is an art form as we all know. It is also useful for our readers who tend to be busy bunnies (and who don't want to give us too much ammunition). Here's one or two]

Slowlaris? Good Journalism. Thanks to the Net I get to read articles by 3 year olds.
Bryan

Subject: What a crap story.

Do you work for AMD?

Christopher J King

[Andrew was also severely disturbed when he received the following email]

From: tizzo [mailto:tizzo@****.it]
Sent: 29 June 2000 20:53
To: andrew.thomas@theregister.co.uk
Subject:

[And that was it. "What does he want from me?" cried Andrew, a broken man. So he hid in the pub]

[We pointed out the possible hypocrisy of branding anyone who tried to sell a domain name a cybersquatter, but applauding giant Procter & Gamble when it did the same. Oh, and we made mention of the American mentality]

I think you have been a little unfair on P&G regarding the selloff of domain names.

These names are all generic, non-branded names - and anyone who had the nous to register such names can hardly be accused of cybersquatting, since the names do not infringe on anyone else's trademarks or identities. For example, had you or I thought to register flu.com, no large organisation, except perhaps a hypothetical FLU Inc. could expect to acquire it by accusing the owner of cybersquatting. [goes on like this for a bit]

In the meantime I think you owe P&G an apology for implying that their conduct in selling off these URLs or indeed acquiring them in the first place, was in any way unethical. On the contrary, it shows considerable business acumen. They simply took the old adage 'buy land, they ain't making any more of it' and applied this to URLs.

rgds.
Andrew Mayo

[Of course, we believe that this letter simply proves the point we were making]

Although I sadly can't quarrel much with your conclusions about my republic's national adoration of the rich & powerful, I don't think your analysis of "cybersquatting" is quite fair [similar sort of argument]

[However, someone could quarrel with the US' national adoration]

The difference is a matter of perceived intent, not a matter of wealth in and of itself, and I resent your publication's condemnation of Americans without considering the issue more fully. As a citizen of the nation you attacked (and someone who enjoys the vast majority of The Register's articles, by the way), I'd like to request an apology. [You can request all you like, rich or poor]

[Mike asked How the hell - do I countersue Rambus? Then he asked it again]

Being a young and naive techy, I'm curious. Do all legal proceedings look like this? No wonder this sort of litigation takes forever, if it takes over two pages to say, "Me Hitatchi, You Rambus"

I get the feeling all of this is just a smoke screen. I want to see the third part of this document, the real stuff with entries such as:

173. Rambus has, on no less than six known occasions, used internal memoranda between senior executives to level allegations of sexual impropriety at Hitatchi's mum. Document HCNF245 is a copy of one memorandum, stating: "I've had Hitatchi's mum, and she were crap". Hitatchi refutes all allegations of this nature and would make it clear, at this point, that it believes Rambus smells of sick and wears spacky glasses.

Hitatchi only dropped the counter-suit when Intel, Rambus' dad, issued a statement prior a fight between the two parties in the car park after school. "You lay one finger on Rambus and I'll kick yer teeth in.".

The truth is out there. Apparently.

[In last week's letters, we ran a seemingly convincing denouement of Kenwood's claims to have created an MP3 reading machine that created files as good as the CD they were taken off. Not all agreed.]

This letter is bollocks. Firstly this sentence is clearly rubbish:

Secondly it is not data compress that removes the information, (else how would WINZIP work ?), but the special techniques and algorithms designed to reduce the complexity of the sound and hence increase the data compress rates. What this means is that the data ...

Errr.... I think your reader is confused between "lossy" and "non-lossy" data compression techniques. Winzip is non-lossy (you cannot afford to lose even a single bit of data).

Something like JPEG or MP3 compression IS lossy. It's how the compression routines work."the special techniques and algorithms designed to reduce the complexity of the sound". Yep. That sounds like compression to me.

Secondly, you CAN retrieve the lost data. What Kenwood are claiming is possible. Every "musical" tone consists of a fundamental, plus certain harmonics (typically the odd harmonics dominate (i.e. thirds, fifths, sevenths)). By examining the various tones, you can guess which are fundamentals, or low order harmonics, and then calculate which higher-order harmonics should be there. You then use a tone generator, and stick them back in.

Simple? Actually no. It's really hard, because most music (apart from anything by "Aqua") is harmonically complex, and you get lots and lots and lots of different tones, harmonics, and inter-modulations. The clever bit is deciding which tones should be "regenerated", and which are just noise. So while it is technically possible, this technology probably won't work :)

(Having said that, this technology is not new... it's been used for a fair long while taking old vinyl LPs, and remastering them. It returns some of the crispness lost from old, bad quality and worn pressings.)

Bye
Roger
(who may work in PR, but was once an electrical engineer too).

Just to be awkward and counter Mark Newman's claim that Kenwood's "better than MP3" technology is nothing more than a BOFH-style scam: In the music biz there's a piece of equipment called an "exciter" that does exactly what Kenwood's post processing does. It really does work, and for much, much lower frequencies than 20kHz.

In it's most basic (and still useful) form, it takes the high frequencies, clips them adding the other harmonics, then mixes these back with the original signal. Every recording studio worth its salt has an exciter or three.

cheers,
Richard F. Smit

It seems Mark Newman made the mistake. MP3 compression is a lossy compression, quite unlike the compression used in the zip programs. Only the highest power frequency peaks are stored in the compressed file, and later recreated in the output. The Kenwood breakthrough is electronically replacing higher order harmonics that originated from the original peaks.

Brian Peltzer

[So there you have it]

[Two examples of where it can go wrong]

EXAMPLE ONE

Sorry Andrew, but this story is bollocks!

You've got it back to front. Intel chips are multiplier-locked BEFORE they are tested. They are locked when the core is packaged. Intel admited precisely this in that Taiwan legal case a while back (last year?).

This means that they have to decide how many of the cores are going to be each speed before testing them, based on the relative demand for each speed. If a P3-1000 MHz fails testing it CANNOT be sold as anything else: it has to be thrown away. Intel decided that it is cheaper to do this than speed-test the cores before packaging them.

The same applies to deciding if the core will be P3 or Celeron. They both use the same core; you can bet that the Coppermine Celerons actually have 256KB L2 cache physically on board, but half is disabled, just like the old 486SX actually had the FPU on board (like the DX), but it was disabled in the packaging process.

On the other hand, the Thunderbird Athlons and Durons are NOT multiplier-locked when they are packaged. They are all made unlocked, then tested, and finally the multipliers are locked AFTER testing by burning the "golden bridges" on the top of the package. This means that the AMD chips CAN be sorted for speed and the ones that fail at one speed CAN be sold at a lower speed.

Are the T-Bird Athlons and Durons "completely different processors" ? Hardly! The extra L2 cache on the T-Bird is the only difference - if the Duron doesn't just have it disabled. The rest of the core is identical.

So Andrew, you've got it arse-over-tit!

Yours etc.,
Orca
Mike writes in support:

The other guy said this was bollocks, this guy said your story is bollocks, Chris Tom from AMD Zone says your story is bollocks, and I say your story is bollocks too (all for different reasons, which amounts to a vote of no-confidence in the first four paragraphs) :)

Still, that's never stopped you in the past, so why break with tradition now?

EXAMPLE TWO

Andrew,

Were you drunk when you wrote your "Duron shortages looming?" article?

First off, how many 1 GHz P3's have you seen? Even the 933's are rumored to have thermal problems, presumably because they are in effect "overclocked". The 5% die-shrink due in late Q3 is supposed to deliver 933's in volume (finally) but there are no volume 1 GHz claims being made. This is 3 going on 4 months after the March paper announcement.

Secondly, what's with this off-the-cuff remark about Tbird being *twice* the die size of the Duron? Tbird comes in at 120 mm2, and Duron at 100mm2. That's 20% larger, not 100% larger. Even the most rabid Intel partisans on the SI thread aren't prone to such blatant misstatements. And the Austin facility is on a tried-and-true aluminium process that's running with very high yields, so I see little problem in cranking out more and more Durons as the K6 line is retired. Dresden is yielding *even higher* per AMD management, so I doubt we have that many Tbird wafers hitting the dumpster.

Thirdly, as far as your claim that there is little reason to buy an Athlon classic anymore, THEY'VE ALREADY STOPPED PRODUCING THEM. Is your whole article some sort of strange joke? The channel is supposedly completely depleted of classic Athlons as of about now.

There would be concern about Duron cannabalizing Tbird sales, but for the fact that AMD separates them by clock frequency. As they release higher clocked Tbirds, presumably they'll replace the low-end Tbirds with higher-end Durons.

Sober up, before you write again!

Doug Barnum

Mike writes in support: Thomas confirms he was drunk.

Hi,

I feel like worshipping you, or your reader who found out this information.

I searched/installed the firmware upgrade on my HP+8100 and made a 690 MB CD on my "new" Sony CDX100E. I couldn't believe it. I even verified the size using Explorer. Still can't believe it. Maybe it's all a dream.

Thanks for writing the most useful news I've ever read! :)

Cheers,
Alex

Hi.

Finally someone who tells it like it is. I have always been wondering why nobody ever wrote down what you did. It feels so good to have the 'world + dog' being able to read that article. Most of all I hope the Novell Sales People will read it and start acting upon it.

I really like your site. All the jokes all of you come up with make the articles all the more readable. Keep on writing like this and I can prophecy big success on your side. Anyway, just wanted to say thanks for a very well written piece. Love to read stuff like that.

Yours,

Kuno Weiss

Ed Ross wants us to tell you all that he's now finished his A-levels, so you can start adding your sites to his Register Readers' Webring. ®