And, of course, we're not adding new generating capacity quickly enough to keep up with demand. Should I buy a generator?

For individuals, a small generator might not be a bad idea. But what about a small business? Have they thought about backup?

It doesn't take a power grid failure, either, to turn out the lights: Ask anybody in Florida what they have to prepare for every year during hurricane season.

Shockingly, Emerson Network Power has done some research that finds that, despite years of headlines, most small businesses haven't gotten the message:

For small businesses, power outages can be disastrous if they are not prepared. Technologies that make it easy to quickly transfer to back-up sources can keep small companies in a struggling economy not only out of the dark but also, quite possibly, out of the red.

Unfortunately, not all businesses — especially small ones — have prepared themselves for outages. In fact, a 2008 survey of 451 U.S. small businesses commissioned by Emerson Network Power showed that 61 percent do not have back-up power supplies.

Think about the small businesses you know. How many could survive five days or a week in the dark without a catastrophic impact on operations?

Steve Weis, on Google's security blog, writes:

Cryptography is notoriously hard to get right and if improperly used, can create serious security holes. Common mistakes include using the wrong cipher modes or obsolete algorithms, composing primitives in an unsafe manner, hard-coding keys in source code, or failing to anticipate the need for future key rotation.

Google's code addresses that "by choosing safe defaults, tagging outputs with key version information, and providing a simple application programming interface," Weis says.

Keyczar, which is a free download from Google Code, is being released under an Apache 2.0 license.

Despite being "hard to get right," Google is prepared to tackle crypto - - with all of its potential for headaches. On a Keyczar discussion thread, commenter Scott Markwell raised one potentially thorny hypothetical:

what are the plans to handle migrations between (algorithm)? Say a better then brute force attack is developed for AES that makes it a significant risk? Internally are things setup to allow a replacement with a new algo that is more trusted?

Google's Weis answered that it should be difficult to migrate from one crypto algorithm to another, and if new keys are needed, developers could just push one out.

(Update: Weis emailed to say that his point was that migration should not be difficult under this crypto technology.

The social web continues to mature at a rate and pace that truly makes it difficult to keep up. Take, for example, Twitter.

For those of you who aren't familiar, Twitter is the micro-blogging web site that lets you follow thousands of people and their 140-character comments - - one at a time - - from issues ranging from lines at the local bank to earthquakes in California. It also gives you a platform from which to provide your thoughts, daily updates or other information you want to share.

It doesn't sound cool, but, once you use it, Twitter quickly becomes a social, cultural and news ticker from the world's open-source news room.

Six or seven months ago, Twitter appeared to be more novelty than useful service. But, like Metcalfe's Law, Twitter has grown in value (and usefulness) by the day as more and more people sign up and start to use the service. A search function that Twitter recently bought from a third party, Summize, allows anyone to quickly look up comments and conversations on the universe of topics - - baseball to recipes, technology to carpentry, parenting to dating.

Now add to it a service like Phweet, which allows users to invite other Twitter users into voice conversations or conference calls. (Twitter users are known as "tweeters" whose comments are "tweets." Hence "Phweet.")

So how, might you ask, is this different from a chat room or chat line?

Say you're a technology company executive and you find, through Twitter search, that a number of customers are all complaining about one of your products. You invite them into an impromptu focus group "Phweet," and quickly pull out of them specifics on what they think went wrong with the product. Or you're in tech support and you encounter a problem for which there's no documentation or ready solution; quickly call together a meeting of the minds of other tech professionals via Tweet-Phweet and it might be possible to brainstorm it out - - even at 2 a.m. on a Sunday morning.

Maybe one of those impromptu brainstormers can even do a live demonstration in that solution via Qik. (The live-streaming web video service that operates via smart phone.)

It might be the better part of valor for you to tablecloth out ways you can start leveraging all of this stuff for your business before your competitor does. It sounds like the last boarding call for Web 3.0 is sounding.

(You can interact with me and read my impromptu thoughts, or catch links to Channelweb stories, via this Twitter link.)

In a takeoff of the "blind taste tests" from the '80s, Microsoft videotapes a bunch of people -- seemingly from off the street -- and lets them view a PC operating system thinking its named "Mojave." When they rave about it, the Microsoft folks tell them it's actually visit. Surprise!

The point Microsoft is trying to make is that the more you know about Vista, the more you will like it.

The reality is that it's not about the market getting a better understanding of Vista. The reality is the market just needs a lot more RAM to make it work well. The market's experiences with Vista, for the most part, have been on dual-core PCs running one or two GB of RAM. That's like driving a performance vehicle on a bumpy dirt road. Uphill. In the rain. In reverse.

Since Day 1, Microsoft has downplayed the hardware requirements for Vista -- to this day even listing its minimum requirement as 1 GB of memory and a system with a CPU with a 1 GHz clockspeed. While tech professionals will roll their eyes and say, "of course that's the minimum requirement, but only a fool would use that," not everybody is a tech professional. People and businesses on budgets often have a way of trying to make due with minimum requirements.

(Although, Kevin Hoffman, who writes at the .NET Addict's Blog, contributed a couple of chucklers by saying that the " 'Mojave Experiment' participant selection committee was probably borrowed from the Jerry Springer show" and, "They must have tried VERY hard to find that many computer illiterate people.")

It would be interesting to know whether Microsoft presented Vista to its Mojave audience with PCs running the OS's minimum hardware requirements.

In the Test Center lab, last week we built a PC with standard components -- but spec'd it out with an Intel Core 2 Quad Q6700 and 4 GB of DDR2 RAM at 800 MHz. It scored 4326 on Primate Labs' Geekbench 2.0 benchmarking software. It's not surprising that's about double the score of systems with dual-core CPUs and 2 GB of RAM that run Vista. But forget the numbers. What's surprising is the human experience with Vista on a system with that horsepower.

The quad-core PC boots in about 30 seconds, can simultaneously handle Skype calls and other office tasks and run an anti-virus scan in the background with no noticeable performance degradation.

But remember, systems with that kind of hardware weren't available to the masses when Vista launched a year and a half ago.

So for Microsoft to say to PC makers, system builders, resellers and users, "Hey, Vista is great if you're using a $1,600 desktop but only so-so at best if you're using a $600 one" isn't going to fly. Instead they're fighting back with the Mojave Experiment.

Just remember, without the extra RAM and CPU capacity, Vista is much more likely to leave you high and dry.

"We've known it for a long time: the web is big. The first Google index in 1998 already had 26 million pages, and by 2000 the Google index reached the one billion mark," wrote Jesse Alpert and Nissan Hajaj, software engineers for Google's Web Search Infrastructure Team.

"Over the last eight years, we've seen a lot of big numbers about how much content is really out there. Recently, even our search engineers stopped in awe about just how big the web is these days - - when our systems that process links on the web to find new content hit a milestone: 1 trillion (as in 1,000,000,000,000) unique URLs on the web at once!," they wrote.

Google doesn't index every page, the duo write, because some URLs are redundant or are developed by auto-generated content. The company's goal, they said, remains to "index all of the world's data." The process of counting web pages includes, essentially, starting with established, trusted sites and following the URLs that are linked out from those pages.

In addition, Google continues to spend to keep its own infrastructure strong enough to contend with that growth. During a conference call with financial analysts earlier this quarter, Google CFO George Reyes said the company's capital expenditures for the most recent quarter hit almost $700 million - - most of that for IT infrastructure.