One of my most cherished maxims is that anything that's worthwhile to learn or implement, you should be able to do within 24 hours. I've found it's true for everything computer-related. If you get an idea in the morning, you should have a working prototype that evening. If you start researching a subject at lunchtime, by lunchtime the next day you should have developed some expertise in it.
Conversely, if you can't get something working within 24 hours, whatever system you're working with is too user-unfriendly, or the idea is a dead end, or the methodology doesn't match up well with your mindset.