Interesting Links for 13-02-2017

[andrewducker]

Anglia Ruskin University sets up sci-fi and fantasy centre (tags: scifi academia uk ) Commonwealth citizens should have UK visas fast-tracked after Brexit, MPs argue (tags: immigration uk visa politics ) West Coast states form a wall against Trump’s reactionary agenda (tags: usa politics ) Cheating Frenchman sues Uber for tipping off wife about

Comments

[gonzo21]

Oh for fucks sake, so Labours immigration plan would either require border control checkpoints at every region to make sure immigrants aren't moving into London and then moving into the rest of the country, or... data-chipping and real time monitoring or something? Insane.

[danieldwilliam]

You could probably get most of the benefit by tying National Insurance numbers to regional immigration status and cross referencing that against payroll tax returns.

[gonzo21]

Corbyn is finished isn't he? He's alienated too many of the people who rejoined the Labour party now with this anti-immigration pro-Brexit stance.

[danieldwilliam]

PoliticalBetting seems to think he is. The odds on the next PoliticalBetting article being about how Corbyn is finished have been slashed from 2/1 to 5/1on. Lots of talk about timing, process, changes to nomination rules and succession planning.

I'm not sure it's entirely to do with Brexit and immigration. Lots of Labour supporters and voters are lukewarm on the EU and immigration. I think it's more to do with his poor planning and execution of practical opposition and the very, very unfavourable view voters have of him.

I, personally, am particularly cross with him (and the Labour Party) for their half-hearted support for the EU but I think I'm in a minority there.

[momentsmusicaux]

> There are 2^2^2=16 distinct boolean functions of two operators.

How do they get that?
I don't mean how they get 16, of course it's that. But how that formula?
I would have said it's 2^4, because each function can be represented as a 4-digit binary number that's the RH column of the truth-table.

[khoth]

> I would have said it's 2^4, because each function can be represented as a 4-digit binary number that's the RH column of the truth-table.

It is, and the reason the binary number is 4-digit is that there are two inputs with two possiblities each, ie 2^2.

So overall, 2^(2^2) ( = 2^4 )

[danieldwilliam]

I was thinking on the way in to work today that if the UK wanted a proactive renewables (or low carbon) policy then backing cost reductions in offshore wind was probably our best option. We don't have the sun to make best use of solar PV, we seem grumpy about onshore wind and we're probably so far behind China and others in terms of building our own nuclear fleet that we've already missed the boat but we've got a lot of wind potential offshore, perhaps as much as 675 GW of capacity in UK waters. We're also not bad at building sticky up energy structures in the sea.

[nojay]

Offshore wind is expensive to build and maintain, just like any offshore structures like oil rigs and platforms. At the moment the ROC (Renewables Obligation Certificates) system guarantees offshore wind electricity operators £149.50 per MWhr for any electricity they generate. Even at that inflated sum, three times the cost of onshore gas, coal or nuclear power several promising offshore schemes have been abandoned as too expensive to build given the expected returns. I don't see much slack in the costs for offshore operations, they'll still require designing and overbuilding the equipment for operation in a hostile environment and requiring ships and helicopters for installation and maintenance.

The report you link to is written by someone who now works for the offshore wind industry, talking up their prospects and not mentioning the downsides.

"We're also not bad at building sticky up energy structures in the sea. " Brent Alpha, Alexander Kielland and Deepwater Horizon would beg to differ.

You've got to accentuate the postive, eliminate the neg

[danieldwilliam]

Yes, offshore wind is expensive to build and maintain. It is a long way from people and in a pretty hostile environment. There are several design requirements differences between onshore wind and offshore wind which mean that you can't just take onshore kit and place it offshore

[nojay]

I mentioned this before in another blog entry but developing the experience to build nuclear reactors only really helps the country building out its own generating capacity while they're building it. Once the reactor fleet has been built it doesn't need to be replaced for at least two generations and more (new Russian VVER-1200 reactors have an expected operational life of a century) and by that time the technology has moved on and the folks who built that fleet of older reactors are long gone from the industry

[danieldwilliam]

Who'd have thought that States' Rights would become a rallying call for the left in the USA.

I also love the idea of California launching its own sattelites. You'd almost be tempted to start up that programme just to watch Trump try and shut it down.

[simont]

The thing that always strikes me about XOR is that mathematicians - by which I mostly mean the serious academic ones who publish papers in maths journals - have so many different ways to describe a function that is basically bitwise XOR, and some of them are so amazingly long-winded, and I always feel they could save so much time if they just learned our nice short word for it.

Some of the words and phrases I've encountered include:

The symmetric difference operator on sets, used when there's no particular ordering on the 'bits' of the 'number' (and, in particular, still applicable if there are infinitely many bits).

Nim-addition, as defined by John Conway in the course of completely solving the problem of optimal strategy in the game of nim. Coming from a conceptually very different starting point, in that in this context it's critically important that it isn't being applied to a bag of essentially unordered independent bits but rather to positive integers, and the rest of the analysis depends vitally on the ways that XOR

[andrewducker]

Cheers, that was fascinating (even with my lack of maths).

I'm a bit surprised that Boolean logic isn't fundamental that most mathematicians wouldn't be aware of the basics.

[simont]

Yes, it is a bit surprising - and even when mathematicians do deal directly in Boolean logic, they often seem to content themselves with AND, OR and NOT, and build up any more interesting things they need via combinations, so you might find that XOR to a logician was just some cumbersome expression along the lines of (a & ~b) | (b & ~a) which sheds no light whatsoever on all its elegant and useful properties like commutativity, associativity, self-inverseness etc.

In fact, another thing I've seen mathematicians do in Boolean logic is to narrow down to just a single operation, namely the ⇒ connective, i.e. (A ⇒ B) meaning ~(A & ~B). You can build the usual operators from that and the constant FALSE, because ~A is just (A ⇒ FALSE), then A | B is (~A ⇒ B) and finally A & B is ~(~A | ~B). (And if you expand out each of those by repeated application of the previous rules then you get some really cumbersome expressions.)

I think the emphasis tends to be on wanting fewer primitive operations, because you keep having to prove theorems by