powerless

The power went out at Imperial yesterday. I wasn't there, but I was notified, which didn't help one bit. My molecular dynamics simulation that was setting my boss's graphics card on fire had unceremoniously crashed. While I have restart files, I don't exactly have the time to restart things. But it could be worse. Colleagues' protein purification runs hung halfway, wasting their prep and days of work, and the computer farm that calculates three-dimensional reconstructions from thousands of electron-microscopic images went down.

It wasn't the first time. In November of last year, there were three instances when the lights went out and the computers down - it seemed to happen once a week. It got so bad that I sent an irate email to one of the building managers asking whether they had considered putting a damn emergency generator in some under-utilized basement to mitigate the situation.

I wouldn't have gone so far as to request a reliable power supply. This is England, after all, the nation that started the industrial revolution and still lives off it. London got electric street lights in the 1870s and, aside from a few coats of paint to keep it from completely falling apart, not much has changed since then.

The building manager pointed out that the cost of installing effective generators would be astronomical. Fair enough, but it seems to me that it is more of a question of priorities than of money. At the same time the power went down regularly last year, the central library at Imperial got a new door.

This would normally require a trip down to the DIY store and an hour of a handyman's time, total cost a few hundred quid, all done in an afternoon. What happened instead was a month-long operation during which a kind of aquarium was grafted to the library entrance, a glassy tumor the protrudes out onto the walkway in front of the building and encroaches onto the Queen's Lawn.

The lawn, in preparation of the elaborate door change, was clipped on the side of the library and a set of wide stairs was built that seemed designed to connect the building with the lawn. Nice, except the new door doesn't open onto the lawn. When you exit, you see the grass in front of you through floor-to-ceiling glass panels, but you have to turn left to find the sliding doors. You might be momentarily confused because on your right is a similar set of doors, but they don't open. I dare not guess whether that's a design feature or a flaw.

In either case, deceiving people exiting the building with clear glass panels and an inviting set of steps onto a leisurely lawn seems to me a particularly idiotic idea, stupid and dangerous. I was reminded of the entrance block at the Citadel of Aleppo that boasts a daunting bent entrance where potential invaders would see their progress slowed by six turns up a vaulted ramp with concealed defensive features. In the early 13th century, this was the state of the art of fortress architecture.

It is beyond me why anyone would adopt this design for a library entrance. The disadvantages are too obvious. Let me point out just one, relevant in this health-and-safety obsessed country. In case of a fire or other accident, you want people to evacuate quickly. Seeing them get pushed against a glass wall by the panicking masses behind won't be pretty.

Had the door stayed as it was, by the way, everything would be all right. The door was working fine. After an investment of millions and an effort of four weeks during which students chilled to the bone in the Library Café because side doors needed to remain wide open when it was freezing outside, it still does, but less so.

Meanwhile, the idea of assuring a reliable power supply for a university that likes to see itself near the top of the world is unthinkable. But I insist; finding the money can't be that hard. How about negotiating a third-world discount with the power supplier or taking out an insurance policy against power cuts and enjoying frequent pay-outs?

The best solution would of course be to start at the root of the problem and bring the infrastructure into the 20th century, but even I see that it would take more than a few interrupted simulations and wrecked experiments to bring that about.