Reducing Electricity Bills For Buildings With XML

Roland Piquepaille writes "Even if new buildings are connected to Internet, they usually don't communicate between themselves. And when it comes to electricity, these buildings are selfish and consume what they want without any coordination. Now, an XML-based system developed at Lawrence Berkeley National Laboratory is using Web services to collectively adjust power usage to variations in price. In 'Internet ups power grid IQ,' Technology Research News reports that the system was successfully tested for two weeks on five commercial buildings. 'Beyond price, systems could be programmed to respond to changes in air quality or to tap into sustainable energy sources.' You'll find more details, pictures and references in this overview. [Additional note: The system described here is completely different to the one mentioned in Slashdot last March in Building the Energy Internet.]"

Comments

Problem with ambiguity

[Clipper]

Totally off topic, but....

upon first reading the subject, I thought to myself "They make buildings out of XML now? But why do those buildings get lower electricity bills?!"

Oh, the loveliness that is ambiguity

Re:Enough with the XML

[Moraelin]

Well, yes, but the innovative part is the whole monitoring and calculations that happens there. XML is nothing more than one of the many possible data formats used, and only for a tiny little part of the system. It's just one of the many possible "glues" between the systems which do the actual work. Maybe a cleverly selected glue, but nothing more nevertheless.

Basically: saying that XML is the important part there, is like saying that tin is the most important part in your computer. I mean, hey, it's what keeps those components on the motherboard. It's got to be the most important part, right? If we used more tin, we'd have a much faster computer. (Just in case anyone doesn't get the joke, no, it's other parts made of silicon that are far more important there.)

Same with this. Whether the system works well or not, and its advantages are because of the whole idea and design behind it, not just because of the XML buzzword. If you just slapped an XML-based server on a building, it would do exactly _nothing_ by itself. It's the rest of the system that actually accounts for any power savings.

And IMHO the sooner we get past this mindless fetish about buzzwords, the better we'll all be in the long run. "Woo! It has XML, XSLT, EJB, DRM, JMS, and a dozen other buzzwords! It must be great! Dunno what it does, but all those buzzwords _must_ make it work great!"

Makes me think of the stereotypical horde of lemmings jumping off a cliff. Everyone else does it, so it _must_ be great. Jumping off cliffs is the way of tomorrow!

Well, no. If it works well, it's because someone actually did some real analysis and design. If they actually studied the problem and decided which technologies to use, but much more importantly: which _not_ to use.

Lemmings which can just make a list of all the fashionable buzzwords... and then need man-decades to make a dysfunctional product that wastes 6 seconds per request just parsing XML data again and again, those are a dime a dozen. (BTW, that's not a joke. I actually benchmarked 6 seconds of pure CPU time just in XML parsing per request in an application here.) It's those who know when _not_ to use a fashionable buzzword that are the real architects.