theorem

If someone proves a new theorem, it takes some work by the reader to decide whether or not to upvote it. An amusing cartoon takes less.

Ha ha, theorem proved, bitches!" [image of him doggy-style dry-humping a woman he assures you in the comments 'he has no idea who that girl is'] ...

When I failed calculus because I didn't prove a theorem the way it was in the lecture notes and then was insane enough to argue the point, I thought it was pretty stupid. Here, I don't think the word even begins to describe the situation.

In fact, Bayes' theorem is never even mentioned. There's just a hand-wavy collection of statements like "The Bayesian approach is to rely on past knowledge and then adjust accordingly".

At work last week, we were working on project euler at lunch, and had the one CS PhD in our midst not jumped up and explained the chinese remainder theorem to us, we wouldn't have had a chance.

In philosophy, real answers are difficult to find and prove, but in math and science, even though every answer brings 10 new questions, you can look back and say: I proved that theorem, I empirically verified the acceleration of the earth. To quote my favorite xkcd T-shirt "Science, it works, bitches!"

Com/977/ I wonder if there is some sort of theorem that describes which fidelities you can get out of a flat projection of the surface of a sphere. For example, a projection could have accurate area ratios or accurately reflect point to point distances but not both.

To mathematicians as a whole, this result is largely meaningless; to set theorists, it's a minor theorem of some interest, though even they, I think, probably would look at it more as an interesting puzzle with a beautiful solution. And that, by itself, is nothing to sneer at.

They know what maths is, and they know what programming is, and they'll never want to visit a forum and tell everyone how awesome it was when they elegantly solved some mathematical problems or came up with a new proof for some random theorem. They will on the contrary yearn to talk about shoes, interior design, literature, art, perfume, TV shows, movies, diets and so on.

I have a theorem which I refer to as the Fuzzwah theory, it goes like this: "The more you know about a topic, the more it will annoy you when it is covered by the main stream media." Once you realize this with regards to topics you're knowledgeable about you start taking coverage of ALL topics with a grain of salt, knowing that someone who knows all about it is probably cringing and / or yelling at how poorly it is being reported.

This doesn't beat the CAP theorem. The approach can be quickly described as a log-based append-only database with read-repair code based on last-write-wins. The system will have inconsistent views for the whole time a netsplit lasts. If I have two nodes, A and B, and that they both hold a piece of data related to a meeting M0 at 11 on Monday, then a netsplit happens, the strategy suggested won't help with the following case: Jim and Peter have access to node A, and Mary and Julie have access to the node B. Because there is a netsplit between A and B, communication is impossible there. However, Jim and Peter agree to move the meeting M0 to 3PM on Monday, and we call it {A,M1}. Mary and Julie, however, agree to move the meeting to 1PM on Tuesday. We call it {B,M1}. Now, the M1 record on A and B is no longer the same. If the netsplit isn't resolved before Monday, then inconsistent data will have caused the loss of our team meeting! When the netsplit is resolved and we consult the meeting, the approach of the blog post will pick either B's version of M1 or A's version of M1 based on the map-reduce we've had. To the reader, we'll have discarded either of them. This is why the last write wins -- it's the latest state in the log of events, the other one is ignored. Last-write-wins is a default mode to be used on read-repair, and nothing solving the problem of losing consistency. During the netsplit, views of the data remain inconsistent. The way to keep them consistent would have been to have A and B blocking writes. As such, we would have kept Jim, Peter, Julie and Mary from writing new meeting times and made sure that they all had the same hour presented. An easy way to test whether you beat the CAP theorem or not is to imagine your system functional during a week-long netsplit. Either it keeps on working with inconsistent data, or it has a way to keep it consistent. The CAP theorem tells us it's impossible to do both at once, and the blog post didn't disprove that, but merely provided a way to resolve conflicts in data in an automatic way. I would expect someone to 'beat' the CAP theorem as much as I would expect someone to 'beat' the Pythagorean theorem. The blog is interesting in its own approach to make the CAP theorem simpler to handle, but misses its own point of beating it.