Wikipedia and TV Tropes are some of my favorite websites. Sadly, every time you visit one of them, you risk being sucked in for hours. You look up one definition, and the next thing you know it is 3 hours later and you are reading about something completely unrelated to your original query:
Comic © xkcd: http://xkcd.com/214/
The other problem with Wikipedia and similar sites is that they are not designed for quick lookup scenarios – they were made to be thorough and densely packed with knowledge. So if you are trying to use wikipedia to quickly figure out why some mathematical theorem mentioned off-hand in an article you are reading is significant, chances are you will have to do some serious skimming. Math and science articles are notorious for hitting you with complexity and jargon in the very first paragraph. Let me give you some examples – try looking up: Poincaré conjecture, Riemann hypothesis or the infamous P vs NP problem. If you just wanted to know “in layman terms” what these things are, and why they are important, Wikipedia probably will not be much help. Of course there is nothing wrong with that. Last thing I would want is to see my favorite online resource being dumbed down.
That said, I see a niche here for another site: a TLDR wiki site which takes Wikipedia (or other encyclopedic) articles and boils them down to one or two plain English sentences. Think of it as encyclopedic golf – how many words and sentences can your remove and still retain the meaning of an article and let the reader know why it is notable. And no, I didn’t just reinvent a dictionary because most dictionaries do not include complex topics like the ones I listed above.
The reason why this ought to be a community project is the fact that one person couldn’t possibly do this on their own. I feel that such distillation of meaning would take a collaborative effort of hundreds of users and it would be a gradual process. For example, I have no clue how to explain any of the concepts linked above in one paragraph or less. That’s basically the problem – these things are difficult to explain – which is the more reason why we should have a TLDR wiki. I guess this would be a good place to start when working on a definition for Poincare Conjecture, but it still seems too long.
What do you guys think? Is this a good idea? Horrible idea? Do you think it would take off?
This is just a random idea at this point. I have not set anything up for it because I don’t really have the marketing skills to give it that starting push, no place to host it, and hardly any time to babysit it and moderate it. I have seen hundreds of promising public wikis go down the drain because their small community could not deal with avalanche after an avalanche of spam. So for now I’m just tossing this idea around.
Have you seen the Ten Word Wiki?
@ Chris:
Nope, I have not, up until now. I should have known better than to assume that I had an original idea for a wiki type service. :P
Actually, I suspected something like this might exist, I just didn’t google hard enough.
On the other hand the ten word wiki seems to be more of a joke site even if it was probably not intended to. It is overrun by memes and urban dictionary type random joke definitions.
My idea was to make this a serious resource.
This is indeed an awesome idea. I can’t even begin to count the number of times I’ve had a discussion on IRC and turned to Wikipedia to look smart, only to get stuck there and entirely miss the discussion. I’d love to make something like that happen, but just like you I hardly have the marketing skills to pull it off. I could totally figure out hosting and domain name though (I quite like tldrwiki.com).
I agree that Ten Word Wiki isn’t quite serious enough for this, maybe because of lack of moderation. Exactly ten words is a pretty silly rule though if you ask me. It’s just asking for spam. One paragraph would be a better guideline.
I love this idea. Just a quick ‘tell me what it is, nothing else’ in 1-2 paragraphs would be absolutely perfect for a lot of the searches I do every day, and it could then just link to Wikipedia and such FMI. As you say though, getting it started would be the trick.
While I like the idea, it would be very hard to do. Explaining the three topics in your post without using any technical words that’d need lookups or requiring previous knowledge in the field would take… a lot more than 2 paragraphs. Especially if you want to explain why they matter. I think wikipedia is doing a fine job as it is, after all if you don’t know enough about math to understand the poincaré conjecture article, it’s highly unlikely you need the information for other purposes than showing off/passing time.
Also, people say math is an international language… they’re wrong! I can barely understand a thing on the english pages you mention, but when I switch to french it’s much clearer. Maybe the french pages are just better :).
actually, there is also http://simple.wikipedia.org/, which is “wikipedia in simple words”, which kinda does what you want.
(I wanted to post links to “Poincaré Conjecture”, “Riemann hypothesis” and “P versus NP”, but your system called me a spammer :( )
Rats, ido beat me to it. Simple.wikipedia is actually pretty good, e.g. p versus np. Unfortunately, a lot of these simple articles have been getting longer & more detailed. In fact, many articles are tagged “This short article about mathematics or a similar topic can be made longer. You can help Wikipedia by adding to it.” – so the two may have some convergence.
P = NP wouldnt be that difficult to explain in a paragraph or two, for something that specific you could makea a resonable assumption that the person is from a programming, math or similar field. With P=NP, I would assume its fine to put things along the lines of sets P is the set…
But would avoid Non-determistic polynomial time like the plauge
This is reminding me, of a quote by a haskell programmer (cant remember who, and is paraphrased). Monads are not difficutl they sound difficult, we should have called them fluffy bunnies. Another example being Generics doesnt sound to scary, but parametric polymorphisim does.
At any rate, sounds like a good idea.
This page seems a great example (the descriptions of the problems) well the P=NP example, Poincare and Riemann are still pretty dense
http://en.wikipedia.org/wiki/Millennium_problem
Honestly, I think the Summary at the top could be revamped, rather than creating an entire other database… that would be ridiculous. IMHO