Dreamt last night that I died and went to Hell.
Which turned out not to be as bad as all that (except for the weather - there was a pretty bad thunderstorm at one point), mostly due to the fact that I hadn't done anything utterly reprehensible. I met a few interesting people and had an odd time, but largely it was reminiscent of my childhood camping holidays.
Firstly, realise that you _are_ going to make mistakes, and that if you don't make the mistakes you will never learn anything. Growing up I made a lot of mistakes with computers, and very few with people (by avoiding interaction most of the time). This has meant that at the age of 32 I've learnt how to avoid most major mistakes with computers, and am still capable of making absolutely huge ones on the people side of things. On the other hand, I've made some _huge_ mistakes in the last couple of years, so hopefully I'll have learnt from them.
Secondly, realise that the problem is more complex than it looks. You can sit and plan out your exact solution in advance, and preparation is almost never harmful, but once you actually get stuck in you'll find that some of the things you've assumed aren't quite right (or possibly even actively wrong) that things wok differently to how you (or, sometimes, anyone) thought they did, that you wanted something quite different to what you thought you did, and that the best solution is only obvious once you're immersed in the problem. Which isn't to say that you should leap straight in, but be aware that spending all your time in planning means less time doing.
Thirdly, realise that there is no such thing as a perfect solution to any real life problem. Real life is complex and difficult and full of contradictions. People frequently want both X and Y without even realising that to have either one would mean that the other is impossible. Finding the compromise or realising that compromise is impossible is the tricky bit.
Fourthly, when you're trying an idea, try it firmly. Going in half-heartedly will just mean you stop at the first tiny obstacle. For a period of time, assume that you're going about things the right way - if you hit a problem, assume there's a solution if you look for it. There's nothing more likely to succeed than someone who knows that they are right.
However, Fifthly, don't get attached to your mistakes. The point of trying something is not to get attached to the path you took - whatever you invested in that path is in the past - you may well have got something out of it, it may well have been the right answer, but if it wasn't then continuing down that path isn't going to suddenly make it the right one. This is the one I've found the hardest to learn over the years - I've created prototypes at work on several occasions that were just plain _wrong_, but because I'd become attached to them I found it almost impossible to throw them away and do things the right way. When you find yourself defending what you know is a mistake, you're well on the way to the dark side.
Which leads nicely onto Sixthly - get feedback from other people. At work I have a Gordon - when I think something is pretty much there, I get him to take a look at it and tell me what's wrong with it. I've yet to have a solution so perfect that he didn't find _something_ wrong with it - and frequently some good pointers as to what would make it work even better. Sometimes you don't even need a person - just writing down your plan can trigger your internal editor to point out the blatant goof you've made - but if you can find your own Gordon (or Ed, my equivalent at home) then use them as a sense-check.
Seventhly, decide as late as possible. Which isn't contradicting number two - trying something and making a firm decision that it's 'the answer' are two different things. Be aware that making certain decisions preclude you from other options, and therefore it makes sense not to make those decisions until necessary. Necessary can, of course, mean different things in different situations, but being aware that some decisions don't actually need to be made right now can make it easier to focus on the ones that do.
As an aside - I have no idea how I know the word 'ineluctable' - I just knew it was the right word to use, even though I didn't actually know what it really meant - I had to go and look it up. Most of my writing is, in fact, done by my subconscious.