Long before we picked on George Bush for his befuddling phrases in speeches, we had then-VP Dan Quayle, who besides not knowing how to spell ‘potato’ when judging a spelling bee, said things like “We don’t want to go back to tomorrow ....” Are we moving towards innovation in technology or reinventing the wheel?

Long before we picked on George Bush for his befuddling phrases in speeches, we had then-VP Dan Quayle, who besides not knowing how to spell ‘potato’ when judging a spelling bee, said things like “We don’t want to go back to tomorrow ....” Are we moving towards innovation in technology or reinventing the wheel?

Well, although it is humorous when one of our important leaders stumbles when picking words and phrases in public speaking, I can’t but think how dumb I’d probably sound in the same situation. Fortunately people seldom have paid to hear me speak in public, and I’m cut considerable slack.

My eyebrows really go up, though, at some of the befuddling ideas I hear that are getting substantial funding. Here’s a recent “Huh?” moment … Pano Logic.

The founders have an idea. If we’re moving to cloud computing and virtualization, why not virtualize the desktop? After all, the desktop or laptop computer is running some operating system and has a CPU and local storage, all of which have little purpose if you’re just going to use it to connect to a network and have the real processing done on a remote server. If your application isn’t running on your desktop computer and your data is saved on a network store somewhere, then why do you need a 2 GHz CPU, 4 GB of RAM, 320 GB of disk space? Why not go for a thin client approach? So the Pano Logic founders do have a good point.

But the NCD was a good idea, too … an X-Windows thin-client terminal, it had a keyboard, display, network interface, and knew the X-Windows protocols, little else. Remember those? They worked well for lightweight applications. We threw the last NCD terminal out in about 2001 as they were slow (strained servers needed additional cycles to handle the growing number of thin clients and thin client applications), costly to maintain, and not really upgradeable. Of course what we replaced them with was the ubiquitous desktop or laptop computer, running terminal emulators when we needed to connect to the servers, but also using local applications when we didn’t have to compute on the overloaded servers. And we moved from that to doing as much as we could on our faster, dedicated personal computers because they were, well, dedicated for our personal use, not shared with some resource hog.

For those of you who followed the NCD link, a quick quiz ... (a) where are they today? (b) how many times did you see the word "legacy"? (c) what is the date of the newest document you found in a web search of the term "NCD", besides this one?

Before the NCD, we did distributed computing, with departmental servers and text-oriented terminals, no colorful graphics. We were trying to move away from the centralized mainframe, which was slow, costly to maintain, and upgraded every few years, towards departmental servers that were more flexible and maintainable.

So the idea of returning to the dumb terminal just doesn’t jump right out at me as “Wow! Great idea!”  There are, of course, some distinct advantages to the ‘cloud computing’ concept. In one of my previous roles, we supplemented engineering servers with farms of inexpensive rack-mounted generic PC’s, set up a plan for rotating 1/3 of the computers out each year and replacing them with the inevitable faster and cheaper computers on a regular basis, and carefully monitoring utilization to see when we needed to add more compute power. Cloud computing, properly implemented, should offer the same advantages. But budgets are tight this year, and I’d place a wager that cycling out this year’s 1/3 might just get postponed. And then we’re back to overloaded servers and wanting to have more control over our own destiny, such as by having our own real computer rather than a dumb terminal.

Incidental question: suppose I sign up for a service with ACME Cloud Server Company, and they go off the air due to their own financial crisis. Where’s my data? I took advantage a number of years ago of a free online storage service, and put some things there for clients to pick up. Funny, that free service didn’t last long. Luckily I had local copies of all my data on my own disk drives.

Takeaway: when doing strategic planning of where you’d like to be 5 years from now, don’t forget to take history into account, and the mindset of people like me that are influenced by history.