That is correct, but not what the GP meant. If you can model the distribution (e.g. you 'know' that B is 90%) then you can weigh your random guessing such that it is correct in >50% of the cases, even without looking at the case itself (it is still 'random' in that sense)

Extreme case: I can predict whether someone has Ebola without even looking at them with >99.99% accuracy by just guessing "no" every time, since the prevalence of Ebola is >.001%.

Suppose the supreme court has 70% chance of overturning (e.g. because they choose to hear cases that have 'merit'), then an algorithm that guesses 'overturn' 100% will have a 70% accuracy. A random guess that follows the marginal of the target distribution (e.g. guess 70% overturn) also scores >50% (58% to be precise).

(to those downvoting this: Hotel Charges Guests $500 For Bad Online Reviews)

Why don't you impose a $500 fine?

I'm not sure you understand what 'digital' means...

It was meant neither as an excuse nor as an explanation.

Let's just say that apparently Toady has the kind of skills and character that enabled him to implement DF in the current fashion. And that is both a big compliment and a big gripe...

embarrassingly parallel has a specific meaning, namely that the task is composed of a (relatively large) number of sub tasks that can each be performed completely independent of each of the other sub tasks. So, any sane attempt to pathfind (say an A* search) is not embarrassingly parallel since whether a path can be pruned depends on the best paths found so far in other branches, and there is an optimal ordering for which branches to descend into first which is also dependent on what's happening in the other branches. I'm sure you can make a parallel version by e.g. forking out N possible branches to some depth, gathering state, and then pruning and ordering centrally and branching out again on the most promising branches, but this is not "embarrassingly" parallel.

s/van/can/... I want an edit button on /., even if only for the first 5 minutes after posting...

For one unit, sure. The CPU problem in DF (as far as I understand) is that there are 200 dwarves, 100 goblins and 400 kittens all trying to pathfind at the same time. Unless I miss something, each of these units van pathfind in parallel since they don't "know" about the other's paths.

Yes... From what I gather, the developer ("Toady") is an autodidact that doesn't use any sort of version control and no multithreading. Although the simulation might be difficult to run in multiple threads, I think the path finding is one of the biggest CPU drains and that should be embarrassingly parallel. Also, he is really giving the community a tough time by having a monolithic game engine + GUI instead of some sort of modular system, which would allow the many programming-savvy fans to build tools much more easily (tools such as dfhack and therapist now use direct memory hacking, which is annoying (therapist needs root access, dfhack encapsulates df itself) but also unstable.

Also, if you've never played DF, it is probably best to either wait a couple months until the worst bugs are fixed and the wiki etc are updated, or just grab the previous versions. This new version will still be quite rough at the edges and some of the info on the wiki, youtube etc will be outdated.

I'm pretty excited about the new release, I've been playing for a couple years and I hope that especially the AI behaviour has been improved although it is not listed in the change log as such...

I had to read it twice, but that's actually quite funny :-)

(no mod points today...)

This is an important part of the story. Public decency laws, and nude beaches as an official exception to them, are not there to protect the nude people, they're there to protect the prude from the nude.

The sad truth is, however, that while being nude at a nude beach is OK, having a picture taken of you and distributed outside that context is not OK. For one thing, it violates my feeling of privacy more than a picture of me walking in the park (I guess there is still a remnant of prudishness there), but it can also damage my reputation and social standing among people who dislike nudity. Thus, it makes perfect sense to be stricter about taking and distributing pictures from nude beaches, just like there is a distinction between taking a picture of me in my front garden (maybe ok?), sunbathing in my back garden (less ok), watching television in my living room (bad) and having fun in the bedroom or bathroom (really bad).

(Note also that most people don't go to nude beaches because they're exhibitionist: they go there because it is much nicer to sunbathe and swim without swimming gear. If you've never swum naked, you should really try it one day, it's a world of difference)

For most* languages out there you can automatically create a binding to a C api, if 1:1 API use is what you want. It is true that C is the default goto language for libraries, but if you actually want to do something like parse plain text (or json/xml), communicate over a socket, use SCP, etc etc, it is by no means the easiest language to do something in.

For the record: perl was king more like 20 years ago. I started around 1998 and it was all perl. 5 years later I had switched to python and I think most people started realizing that perl was not the future. If you look at e.g. tiobe, you'll see that python went mainstream around 2004 and has stayed more or less constant since.

Yeah, if you want to have fun tinkering with computers, haskell (or prolog, lisp, erlang) are extremely fun to toy around with.

If you want to get something done.... well let's say it ends on ython ;-)

Yeah, although I understand the reasoning for the whitespace and don't object on principled grounds, it can be quite annoying practically if you are copy/pasting code, need to (de)indent a large block, and especially if you are forced to develop somewhere where your favourite editor/IDE that actually handles these cases well is unavailable and you have to work with something that actually inserts a tab when you press tab... *shudders*

It's also a shame that some of the things corrected in python3 were not corrected earlier, but at least they did have the courage to make some breaking changes, instead of waiting for the next language to come around and start without the excess baggage but also without the built-up community and design experience...