Log in

No account? Create an account

Previous Entry | Next Entry

Jerry Pournelle comments briefly on Google Chrome, which he correctly sees as the third try for network computing. He further notes (in his column) that less than 30% of the US has access to broadband. Add the two together, and it says to both of us that much like the paperless office, we're not in the future yet where network computing actually works well enough to risk your business on it.

And until you can convince America's businesses that network computing is safe and secure, they're not going to sign up for the program, because having nice cheap workstations isn't even half the battle. Take a bank, for example. Are you going to go ahead and put the personal and financial information of thousands, maybe millions of customers out there on the Internet, on servers you don't own and can't guarantee the security on? Oh, sure, Google might promise you those servers are secure, but if somebody cracks in and hoovers up all that data, your customers' lawyers aren't going to sue Google. They're going to sue YOU, and at that point you can kiss your job (along with your freedom, quite possibly) goodbye. We haven't even addressed the issue of a server crash denying you access to data or wiping out that data. Nope. It's just simpler and easier to keep all that stuff in-house, where you can run tape backups every day to make sure that if the worst case happens, you can restore the data and boot up tomorrow, good as new.

I recall that Wells screwed around with the Compaq iPaq for a bit while I was there, and finally tossed them in favor of more conventional Evo workstations. Aside from Compaq's fondness for proprietary hardware, I think what killed the iPaq for Wells was the plain and simple fact that it was just easier and cheaper to buy a hundred thousand conventional workstations that could function without network connectivity in all our far-flung branches, insurance offices, etc., rather than split up the purchase to get a bunch of iPaqs that would work fine within the regional headquarters in Minneapolis where you were hooked into the corporate network 24x7 but not so much in the hills and prairies of the Dakotas, Nebraska, etc., because once you got a few miles outside of Fargo, you're still talking dial-up connections.

Jerry also comments on this article by John Derbyshire about the Apollo program. He has some interesting things to say about the program, and I agree: we should have done it Delos D. Harriman's way, taking a step farther out until going to the Moon was just another step from our L4 and L5 stations. Instead, it got tangled up with the Cold War and national pride, and we did it in one mighty rush. RTWT.

UPDATE AND BUMP: Megan McArdle also has some thoughts on the Apollo landing's anniversary.



( 2 comments — Leave a comment )
Jul. 18th, 2009 06:00 pm (UTC)
As concerns business entities and distributed computing via the Internet
One cannot underestimate the chilling effect of various failed attempts to achieve newer models of computing by large businesses. One also cannot underestimate the stupidity of the average MBA when it comes to *how things work*.

The Internet as we now know it was implemented by engineers and computer geeks, not men in suits (except for the occasional suit-wearing engineer or computer geek, there being no limits to the variety of taste in personal accouterment...). I know someone who worked for several years for a corporation which was trying to achieve a distributed computing network that really worked as a whole, rather than as a collection of disparate systems which all had to have their own high priests and acolytes. The corporation was unwilling to (a) spend the necessary money to acquire the needed hardware and bandwidth, and (b) unable to force their little computing satrapies to give up their power.

(b) above was probably because all of the high priests of those little computing kingdoms feared for their employment, since the modern corporate approach is to flush "used" people and get bright, new shiny ones that they can pay less whenever a paradigm shift occurs.

And yes, a permanent presence on the Moon by 1980 would have been a better goal than a quick landing and photo shoot by 1969, but what can we say? Millions for defense of our national pride, not one penny for actual development of new scientific opportunities. When it comes to spending on science, we (collectively) only want the short-term science which leads to quick new technological developments.

If GWB had spent as much money on new science as he did on going to war, he probably would have been impeached for "wasting money", by both sides of our political spectrum.

Jul. 19th, 2009 09:27 am (UTC)
Re: As concerns business entities and distributed computing via the Internet
If GWB had spent as much money on new science as he did on going to war, he probably would have been impeached for "wasting money", by both sides of our political spectrum.

I can't actually see him doing that in any case; the political tendency he represented was more interested in having scientific research done in the private/academic sectors without Federal funding.

I can believe that NC failed in other companies due to opposition from internal IT sections, but at Wells I'm pretty sure it was more a matter of cost. We already had huge mainframes to serve as, well, the server half of the NC equation, after all.

Edited at 2009-07-19 09:29 am (UTC)
( 2 comments — Leave a comment )



Latest Month

April 2019
Powered by LiveJournal.com
Designed by Lilia Ahner