In any case, I would imagine there are likely to be other such brief bits that are not large enough to warrant a separate article, or perhaps are so general that they don’t fit well into more specific categories.

]]>Let’s suppose I hit the Tricki wanting to find the normality of a constant. Any proof so far is on a highly “manufactured” constant (Copeland–Erdős, say, which has 0 followed by all the prime numbers in base 10) and a general proof method is unknown. Should the entry just give known properties and the known proofs and leave it at that? Or will it give suggestions for further research? If so, what format will these suggestions take? Will it mention, say, continued fractions, even though while there seems a connection intutively (and continued fractions are useful in irrationality proofs) no particular method has been established and the connection may just be illusionary? Where’s the dividing line between established math and rampant speculation, in other words?

]]>1. You seem to have re-invented or re-discovered the idea of an expert system (see http://en.wikipedia.org/wiki/Expert_System). This is particularly clear when you speak about an interactive dialog for using it. I believe similar ideas have been proposed to assist doctors in diagnoses, which I suppose more closely resembles problem solving, rather than theorem proving.

There may, therefore, be some literature on expert systems that could be helpful.

2. On the idea of repeating some items to add redundancy and robustness: there are pros and cons to doing this in software systems. I would recommend at least keeping an index somewhere so if there’s a development pertinent to a field, you can easily find all the pages needing an update.

3. On the point about making adding links easy and making deleting links hard: this reminds me of the perpetual question about how many features a piece of software has. It’s easy to miss that too many features can be bewildering. Similarly, too many links from a page could be bewildering. I would recommend that at least you consider sorting the links based on popularity from early on in the use of the Tricki.

I think, by the way, that this sounds like a splendid tool, and I wish you good luck with it!

]]>I’m not sure whether counterexamples count as tricks in quite the same way, but ‘What goes wrong when I try to construct a counterexample?’ is one question to ask if stuck.

]]>In a similar vein, Leo Corry has studied the changes to the classification of topics in algebra over 50 years around 1900. The changes were profound. Possibly, though, we may find greater stability now.

]]>Let’s suppose I’m trying to solve a combinatorics problem and I realize after enough searching that I want to use the pigeonhole principle. The pigeonhole principle itself is fairly well documented; at that point should the page simply point elsewhere, or should it go through examples like one might see in a textbook?

]]>COngratulations on the initiative. I hope I can lean a lot of Mathematics from Tricki. One note though, personally I would have picked another name that “smelt” a little bit more mathematics.

Thanks

]]>