Dumb Terminals

I make a point of staying on top of trends in technology. It's sort of my "thing" and it's more fun than say, the hair colors and marital statuses of the rich and famous.

So like most geeks, I've been hearing more and more about "cloud computing," which is supposedly an evolution of Web 2.0 technologies (Web 2.1? 2.5?) and this whole "internet thing," where software and data is something that runs on a server somewhere else, and your computer (via your browser) is a window on the "cloud." Let me back up:

Lets start of with the model traditional personal computing. People have computers, they run software, and they store data. If they have network connection, the network is primarily a tool for pulling new data to be stored and processed using the software and hardware that's sitting on the users desk.

Where as on the desktop you might use a program like Word or Open Office, "in the cloud" a program like "Google Documents," is probably the app of choice. And web/cloud apps have been replacing desktop email clients for years.

And this is important and noteworthy because it's a fundamental change in the way that we use computers, and largely we are all accustomed to this mode of operation. The interesting thing, is that the underlying technologies that support cloud computer: mySQL, PHP, python, ruby-on-rails, and even AJAX are really nothing particularly new. I suspect that the largest contributing factor to the emergence of cloud computing is the fact that network connectivity in the last year or two has improved dramatically.

Having a connection to the internet isn't something that you do for a few moments or even hours a day anymore, but is practically a requisite part of computer usage: the "internet" is always on. And networks are pretty darn fast for most things.

The geeky and historically astute among you, given the title, can probably see where this is going...

The personal computing modality (run applications and store data locally) came about when computing power and storage finally got to be small enough (it could fit on your desk!) and powerful enough (whole kilobytes!) that it became reasonable for non-specialists to run and operate them.

Before this, computers were pretty large, too powerful by the standards of the day for one person to run themselves, very expensive and very finicky to run, so they ran in secure/controlled locations operated by specialists, and users had "dumb terminals," which included some sort of connectivity interface (RJ-11 or coax likely), a monitor a keyboard, and a chip board that was just enough to tie it all together and send the single back to the real computer where all the processors and data lived. [1]

And then computers got smaller and faster than the network connections could keep up with. Hence desktop computing. I'm just saying that things cycle through a bit, and everything that's old is new again.

Thinking about cloud computing as an old modality rather than as a new modality makes it a much more exciting problem, because a lot of the nitty gritty problems/interface questions were solved in the 70s. For instance, X11, the windowing system that most *NIX systems use, is designed to run this way and in fact sort of acts as if instances where windows appear on a screen attached to the computers running the applications is an interesting coincidence. Which is pretty logical and makes a lot of very cool things possible, but is admittedly kind of backward in the contemporary perspective.

Anyway, cool stuff. Have a good weekend, and if you have any thoughts on this subject, I'd love to hear them.

[1]In fairness these connections were, I believe almost always over intranets, rather than over some sort of public internet, though as I think about it, there were probably some leftovers of this in the BBS-days, with regards to terminals and what not.
comments powered by Disqus