BeOS and Transmeta Crusoe: Two Sad Stories
January 9th, 2009
While playing around on my new Acer Aspire One (story upcoming), I was thinking about probably my favorite time in my history with computers. This would have been back around 1999-2000. I was young and idealistic.
This was the time of uncapped cable modems, and all the music you cared to download via Napster. TechTV was in full swing, and I lived and breathed all things computing. The magic of online gaming with ever improving performance consumed much of my time. The internet, while not something new, became more than a slow novelty with the addition of broadband. I was young and idealistic, and foolishly thought it would never change. Alas, this is all for another post. For now, we will look at two bits of technology whose promise seemed almost endless to me, and who both were complete disappointments that totally failed in their goals.
First, let’s look at the now defunct BeOS. BeOS started as a PowerPC operating system that was to be run on Mac hardware. The company wanted their software to be the new OS for Macs. Unfortunately for them, this did not happen, so in 1998 they released BeOS R3 for x86. It was a very neat OS, and I ran it on one of my machines for quite a while. It was a fairly thin OS, and as a result was very fast booting. Beyond that, I don’t know why it was so attractive. Most likely its underdog status really appealed to me at the time. Not many people ran it, so it felt like I was part of a special club.
By this point Be was doomed already. They gained no traction on the desktop market, and the tech just was not there for the little internet devices and tablets they aimed at once they knew the desktop market was out of reach. We all remember the endless promises of neato tablet PCs, and small net appliances of that time. The magic devices never materialized, and by the time they did become viable Be was long dead.
This leads us to my second disappointment of the era. The Transmeta Crusoe processor. The President of Transmeta was on every tech show constantly regaling us with the magic of his new chip. It was supposed to use a lot less power than anything else out there, while still having great performance.
This played in well to the promise of the tiny web appliance running Be of course in my mind, and a tablet PC prototype is usually what the Crusoe was demonstrated on. Throw in the fact that the firm hired Linus Torvalds to write software, and you had a geek dream. At the time, I bought into the stupidity that big companies were evil, so anything that challenged them was a good thing in my mind.
Transmeta’s hiring of Linus was probably their high point however. Once the chip became available we all found out the sad truth. It sucked. Crusoe turned out to be a slow performing, not all that energy efficient chip. In the real world you saved about fifteen percent on power consumption, at a lesser level of performance as a comparable Intel chip, and the Crusoe cost more. This truth came out once people tested the chip with real-world software. Their weird software x86 emulation just did not work all that well. Their promised results had come from benchmarking software only, which turned out to be completely unrealistic.
So it goes. Two companies that one could be excited about, that made poor business decisions. They are certainly not alone, as there were hundreds of such examples when the dot com bubble bursted.
If only TechTV had survived, this would not be such a sad story.
