News

When were the first real "blogs?" Nostalgia calls!

"Why are you doing a blog? You're a bit of a late starter, aren't you?"

No, actually, I'm not. If you honestly think you know an older blogger, let's see it. But in the days when I first did a blog, it was so pre-blogosphere, we didn't actually call it a blog. It was called "Guy Kewney's Online Diary" and here's a cut-and-paste for sceptics:

Thursday 12th October 2006.

Guy Kewney

ZDNet UK
March 07, 1998, 07:00 BST

An illiterate press release from a company claiming to "becomea" leading supplier of enterprise application "developmenttechnology". They are clearly running out of spaces; there are other new English words like "ofvisigenic" and and "thetransaction" -- and to make my day complete, it comes with a complete supply of last week's unopened paper press releases. All of which, of course, have already reached us via email.I'm in a poor mood to deal with this sort of thing; I've been ill AGAIN, and I'm well ticked off with it. A sore throat all last week has succumbed to the magic antibionic concoction of the local pharmacist, leaving me with the discovery that the bacteria seem to have scarred my larynx. So no sleep for Guy; cough, cough, cough. "Well, you're an old hack, you deserve it," is the sympathetic message from a colleague.

A mad frenzy of rage results: I note just how much utter crap is issued as "Urgent Media Announcement" through my letter box. Here's last week's pile:

Prestel discovers 'new man' on the internet! (27% of Internet users are now female).

Zergo to acquire Security Domain Pty! (Zergo's 'entrance into the Asia Pacific market).

Going to CeBIT? (Book interviews with Psion top management)

Will you come to our Windows Show Party? (Adobe needs to spend money on champagne)

No, I can't stand it. Into the bin.

A nice evening thrash: four of the more experienced public relations freelances around the country have amalgamated themselves into a "virtual" PR company. To celebrate it, very definitely non-virtual gatherings take place in the Cheshire Cheese in Fleet Street. Time was (I'm going nostalgic again) when such a pub would have been handy for 90% of the journalists in London. Tonight, not a single Editor works within a mile of the place.

Tuesday

Career-limiting moves # 27: "Praise the boss's strength of character, in being able to live with his apalling wife."

A close simulation of this is given, at a ceremony in Gloucester Road, where Intel unveils the "secrets" of the Pentium II derivatives, including the "Celeron" -- a Pentium II without second level cache.

You might think that a P6 without L2 cache might not run particularly quickly; I can't possibly comment. "This information is all embargoed," warned the executives from Swindon. "Well, actually, it's all public, by now, as you'll have noticed. But there are some genuine NDA terms in it: specifically, you can't talk about performance benchmarks, and you can't talk about prices, until April 15th."

Performance a no-no, eh? This isn't as daft as it sounds, even though we're about a fortnight away from the Hannover computer fair, CeBIT -- where every damn new chip will be on display, built into a PC, and working and testable. The reason it's not stupid, is that Intel has a problem with the "low cost" Pentium II chip.

Here's the logic, made a bit tricky because I'm not allowed to tell you why it's logical!

Your older relatives will recall a chip called the Intel 80286, known affectionately as 'the 286' for short. The 286 was the standard chip of the last years of the 80s, and Windows version 3.0 would work on a 286, as well as the new, exciting, super-powerful 386.

And it was clearly apparent that despite the fact that the 386 was 32-bit, and the 286 "only 16 bit", Windows was faster on the 286.

Similarly, you may recall that when Intel released the P6 and called it the Pentium Pro, it turned out to be slower than the Pentium, clock for clock, at Windows 95.

The name of the new processor, Celeron, uses the Latin root "celer" meaning "swift" -- from which we get both celerity, and celery (which supposedly grows fast). (I made that up). Without giving away any secrets, Intel now faces a similar problem. A cacheless Pentium II is not going to be as fast as a Pentium II with half a megabyte of closely coupled second level cache; and until Intel can crank the clock up, the only thing you can do to make it seem fast, is to give it a fast-sounding name.

The other precaution Intel is taking, is to make sure nobody mistakes this for a Pentium II, which is genuinely fast. So fast, in fact, that the presenter, Sheryl Rigby, gets rather carried away.

"This high end Deschutes processor in Slot 2 is really fast. To give you an idea of how fast, let me list some applications which it is suitable for... and look at the benchmarks..." [I have to leave the figures out].

There was no sign of her audience becoming infected with her enthusiasm. Reaching deep into her store of simile, she said: "Look, this is really a breakthrough. With this chip, for the first time, you can actually load Active Desktop. It really runs, without virtually stopping your PC. You know, how if you normally load Active Desktop, your PC just slows right down, and there are these long gaps between when you click, and anything happens, and...."

She chatters gaily on about how awful Active Desktop is, and how only a Deschutes at 350 Mhz with a 100 MHz system bus and 2 meg of L2 cache can make it even executable at all... mean while, her colleague are turning pale, and preparing their excuses for the day when all this gets printed.

It wouldn't matter, but the "boss's wife" really is as appalling as the joke makes her seem.

Wednesday

There are days when it would be better not even to get out of bed.

The one thing I hoped to retrieve from the ruins of Wednesday was this: that though I couldn't get to the office, and though every tiny piece of technology I touched turned to dung, I would at least be able to run down the road in the car and pick up the broken telly. In the end, I only had five minutes spare between other disasters, what with the telephone engineer coming to install my Wireplay line, and leaving the door open so that the dog got out and ran away and.... oh forget it. You truly don't want to know.

And the PC threw my data into the ether, and for absolutely no reason whatever, deleted it before checking that it reached the other end.

A friend rings, with news about his crashed hard disk. It seems that he and a colleague were sharing a flight to California on an Airbus. They wanted leg room, so they chose to sit at the door. After takeoff, they produce their notebooks, and begin work. A few minutes later, and errors start appearing. The errors increase. The programs begin malfunctioning. Finally, the machines cease operation.

Back at base, the notebooks are analysed: it turns out the whole disk is scrambled. Both disks. Totally scrambled. "Have you had this in a very strong magnetic field?" asks the service engineer?

Phone calls to the airline, elicit no data. In despair, one of them contacts an engineer at Airbus Industries. "Oh, you'll have been sitting in the seats by the bulkhead!" he responds.

Those seats, of course, have no way of pulling tables down from the seats in front. So they're built into the arms of the chairs. You lift the arm, pull out the table, unfold it, and you're trapped until the stewardesses * choose to free you. And apparently, the airline operators got complaints from passengers: "I booked that seat to sleep, and all I got was rattle, rattle, rattle -- sounded like the aircraft was falling to pieces!"

So, to prevent the rattling, Airbus decided to heavily magnetise all the table components.

A whole day wasted. Well, one redeeming feature. About eight o'clock, giving up all hope of getting my PC Mag column re-constructed from the digital ruins of my hard disk, I decide to take the dogs for a walk at the lake.

It's raining, gently but insistently. We slide gently through the oak thicket, down to the lake-side; where there is a pond. And it is Spring! No mistaking this; the ground is covered with leaves. "Look, a frog!" says my white dog. "No, I'm NOT," says the frog. "I'm a small, lightish coloured leaf, blowing along the grass inthe strong wind."

Amazing, though, how often these leaves appear to have got stuck together...

Thursday

An urgent phone call on voice mail: it is Seagate, wanting to invite me to a press meeting at a hotel. Venue, subject, speakers; and what time? Not worth mentioning in vmail, huh.

So the first thing to do is to call the hotel. It's the Lanesborough; they are prompt with the information, and I rush down to Hyde Park corner -- late, but not too late.

The announcement is an odd one. Quinta is a subsidiary of Seagate, and it's a startup which was taken over a couple of years back. And Seagate took over Quinta because Quinta, quite simply, have solved the problem of "the superparamagnetic Limit."

Disks are magnets, we all know (see above). Normal disks have a magnetic powder stuck to their surfaces. A very, very tiny magnet magnetises very, very tiny bits of this magnetic powder, and can read the pulses back again. Storage!

The trouble is, nobody can see a way of storing data once these tiny magnetic domains get down to something like 10 gigabits per square inch. Quinta, in short, has cracked it.

What they do: they don't even try. They use an "amorphous, rare-earth, transition metal alloy" which normally, you can't magnetise, or not easily. What you can do, is to heat it up to a certain temperature (known as the "Curie Point") -- and then, suddenly, it is terribly easy to magnetise it - and it holds that field for some time. Densities of 40 gigabits per square inch are just the beginning... El Dorado!

Because what you did before, was to make your write head VERY small indeed, and hold it VERY close to the surface of the disk -- one microinch, in fact. That way, you can magnetise a very small area of the surface -- without upsetting the neighbouring bits.

With the aid of a powerful laser, however, you can use the brute force, massive ignorance method. You shine your laser on the tiniest spot you can, and wait till it gets to the Curie Point (about a nanosecond). Then you magnetise the whole damn area. You saturate it with magnetism. It doesn't matter: only the tiny speck of Praseodymium alloy (Samariam? Neodymium? Dysprosium? -- I'm guessing!) which is hot, will change polarity.

So the head can be 50 microinches above the surface; no need for super clean-room manufacturing. There is, alas, a small flaw! [this was reported in full elsewhere - sadly lost]

Friday

A mad colleague lends me a music CD -- Hans Reichel; the title "Shanghaied on Tor Road." It is (says the sleeve) "The world's first operetta entirely performed on the Daxophone." You want to know what a Daxophone is? "It is small. Years ago, when I started using it in concert," says the sleeve note, "lots of people thought I wanted to make a fire on stage..."

The noise is that of flatulent badgers.

Reichel apparently contributed many fonts in the Adobe Type Library, including one called Dax (German for badger). I think the reality check probably can't cope with it. It's a Sign, that &deity& wishes me to go down the pub to celebrate Rupert's "departure."

Exeunt, pursued by a badger...

* Stewardesses: longest authentic English word which a touch typist would enter into a keyboard with the left hand only. And the answer to your next question is: polyphony.