Monday, October 02, 2017

Berlin Day 2

The hotel corner is a hub for several trams, and the sBahn is half a block away. The rooms face the courtyard, so the noise doesn't bother people--or at least doesn't bother me. On the other hand, the room we met in parallels the tracks, so every subway or commuter train distracts a bit.

My talk went well, the others were interesting, though I wonder about the Markov Chain Monte Carlo(*) approach to scattering. As one of the attendees pointed out, there is no theoretical reason why the approximation he was testing should work, and in the end you have to rely on "Well, the distributions look OK in this case"--and the only way to know that they "look ok in this case" is . . . to do the usual estimation, with all the CPU time expense that entails.

I forgot to mention one nice feature of the airport: when we got off the plane, the baggage pickup was right there. No heading down to a common pool of baggage carousels and scouting about for the right one--or what will be the right one when the current deposit is done. Security is probably much better, too.

The Grand Bar served up food very quickly, and with much nicer presentation and ambience than you expect from a "bar." If you eat equal volumes of broccoli and potatoes, they cancel out, right?

I was curious if the simple pre-pay cards would let me make calls from here to the US, but O2 says no. Oh well.

I've never seen Cabaret. I suppose I should, one of these days. There are plenty of posters and graffiti up condemning the AfD as nazis: I don't think nazis will make any kind of comeback in my lifetime, or my children's lifetimes. Not with that name anyway, whether chosen or imposed.

Tour boats on the Spree at night are lit up with bright lights, and they sometimes use searchlights on the clouds as well. Tried to take a picture, but the contrast was terrible. There's nothing like the dynamic range of the human eye.

(*) No, I'd never heard of it before, but Jakob said he and some others had brainstormed it for a while and concluded that it was semi-impossible. I really really want a Journal of Stuff That Doesn't Work.

No comments: