I heard a fantastic quote once from actor Rutger Hauer. He was being interviewed on his tiny role in some big blockbuster movie and the interviewer asked him if he thought his role was great. He replied "Great... I got a chance to be great once. I took it, and I was great. I won't ever have that chance again, but I was, once." A wonderful bittersweet vision.

I suppose when you've created arguably the most significant microprocessor that ever existed, you get to feel that rules don't apply to you. Although the Pentium Pro team could give him a run for his money.

Anyways, I certainly don't begrudge these important old-timers the credit for their work well done.

Also worth mentioning: Jack Tramiel was the only person who ever won a business deal over Bill Gates. When Jack Tramiel was looking for a BASIC for his computers - the Commodore PET specifically - he called in Bill Gates and wrung the worst deal out of him that anyone has ever produced. It's documented in the fantastic "Commodore" book by Brian Bagnall (http://www.amazon.com/Commodore-Company-Edge-Brian-Bagnall/dp/0973864966/ref=sr_1_3?ie=UTF8&qid=1334012789&sr=8-3).

Every Commodore computer used Bill Gates' BASIC code and Bill got a pittance.

Bill Gates has never since let anyone get the best of him. I suspect the experience of getting Tramieled directly led to his success in negotiating the rights to PC-DOS and winning the IBM PC contract.

Here's to you, Jack. You gave Chuck Peddle the chance to be great, and you scared Bill Gates into building modern computers. That's a pretty damn good run.

Your comment is truly awesome. I remember binary coding back in those days; frankly I'm amazed at the amount of work you and your team produced in the mid-80s. It took me literally a year after reading Petzold's book to realize that a handle was actually a memory pointer within a heap.

Kudos on you ;)

Yep. Demos sure look clean if you remove A) all the necessary controls that allow you to do useful work, and B) the context that helps the human eye figure out how data point relates to the overall picture.

This is much more a "dashboard demo" than an "application demo". But dashboards are hot right now; everyone wants one. Certainly no harm in offering appealing dashboards except that they obscure how much work is required in order to make a dashboard show something useful in a relevant context.

They made three transformers movies???

I only saw the first one; it was awesome and had Orson Welles in it. Who knew they could still afford Japanese animators and Nelson Shin after all these years?

/sign

Thank you for reminding us that we don't have to suffer fools forever. We can, and should, insist on a slightly higher standard.

This sounds like the scheme we used to use when I was at Vivendi Universal. It was very successful even if the names weren't quite memorable.

I definitely like the idea of using CNAMEs to give department servers fun names for their groups; but I wouldn't go overboard. Most end users I know don't really care to name their servers, they just need a page they can look up to find the right server name. Which brings me to my next point...

No matter what naming you choose, create and maintain an IT department wiki. The instant you create a server, open up a new wiki page for it and make notes on whatever you did to the server. If stuff fails or if it bombs out or needs to be rebuilt, keep a running log on the wiki of what happened to it. Makes disaster recovery much easier and gives everyone a solid place to go to for documentation.

At my startup company, we named servers after notable videogame characters. It was quite nifty when we had three servers; it stayed fun for years. But when we reached 30 servers, gradually problems crept in. One machine needed to be rebuilt and the name kept getting reassigned. Similar names were confusing.

Server naming schemes are cute until you outgrow them. Hint: Determine for yourself when you outgrow them. We now name servers by their function and their sequence number.

Some executives tasked a junior guy with forecasting how much usage their new network would get back in 1997.

He wrote a spreadsheet that multiplied the "expected number of users" by the "expected data amount per user."

He produced three forecasts for each: high, medium, and low. The end result was a tic-tac-toe board of "here's how much network we'd have to build for each of these nine forecasts".

The resulting 3x3 grid was tossed into a board meeting where uninformed executives argued "this is too high" or "this is too low" mostly because of how much money they wanted to spend, rather than how realistic the numbers were.

Then they published whatever number they settled on in their contract. And as soon as their competition thought they could get a leg up, the competitors issued a press release "We're 1GB/month more than the other company!" until rapidly everyone settled on the same number. ... Or am I too cynical about business?

I concur with what others have said. There are numerous services, being provided by Azure, that are completely unreachable, and have been so for longer than seven hours.

AT&T is scheming that they will be able to trick customers into paying more for their data. Their "choice of plans" approach penalizes everyone except those who know what their usage will be in advance. This shouldn't be permitted; it's a form of taking advantage of unsophisticated consumers.

A good regulation would say "If you offer a choice of multiple billing plans to your customers, you must automatically switch the customer at the end of each month to the plan that would charge them the least." Then AT&T can charge whatever covers costs most effectively for their business, but consumers don't have to expend effort every month figuring out when to change their plans.

Or, heck, just insist that data providers move to packet-based billing like a water meter or an electricity meter. That seems fair.

This is the key bit. Very few companies nowadays spend money on a separate work phone for executives. Over the past decade, virtually all companies I know have abandoned the "Company provides you with a separate phone" policy and instead gone towards "key executives get reimbursed for their personal phone expense," or "company provides an iOS phone".

When people buy iOS devices and Android devices for their personal use, and when an iOS / Android devices provide more functionality than a blackberry device, a company would be stupid to pay for a second phone for each person when there's no need to double up.

Thank you for this post. I always enjoy reading Slashdot when I get a stray comment from someone who is knowledgeable about a topic.

It did crush my little 10-year old spirit a bit to realize that the Space Shuttle couldn't go to the moon and didn't have the thrust to get much beyond LEO. But every time I'm lucky enough to visit JPL and see the amazing things people can do with carefully tuned science and technology, I am glad to be able to watch some of the things people like you do.

Netflix subscribed to a management theory called "eating your own lunch." The idea is that any business, if you wait around long enough, will get mummified as you keep trying to protect the revenue generated by your ancient business model. The theory says that, as the big company keeps struggling to keep its moribund business alive, a younger, hungrier competitor with a slightly different business model will steal your lunch. So, the theory goes, you should eat your own lunch and embark on your own variant business models. That way, when the business world shifts, you'll still be in business.

The theory points to such past projects as the CD industry, Blockbuster, and others. The idea is that such industries failed because they were too wedded to their ideas to change.

The trouble is, Netflix went overboard. They had two different business models running perfectly smoothly side by side. There was no mummification, nothing preventing them from being innovative or seizing on the new streaming business. In fact, their DVD-by-mail business was helping them wield great power in the movie industry, and helping them to get deals for streaming content.

So if they were paying attention clearly, the only reason to kill off the DVD-by-mail business is if it was scaring off the customers, starving the company of funds, or somehow preventing innovation. None of those were true. I'm glad to see they came to their senses.

This guy got caught up in his metaphor and the article doesn't impart much useful information. There's probably a few nuggets of worthwhile advice there about documenting or specifications or vendor lock in. Next time, focus on the IT part and less on the "Dante's Inferno" part.