I would recommend you start with something you're the most curious about and/or find the most useful in your day job. It's like this, you're hired to bring valu

I would recommend you start with something you're the most curious about and/or find the most useful in your day job. It's like this, you're hired to bring value to a company so it should make you productive if you know it better. Get to know that well, seek out edge cases and new techniques. If you put your time into it, you will gain that knowledge fast. The important part is not to stop when you're finished. Pick up the next thing and do the same. After you go through a few phases you will have enough general knowledge to apply in many areas. In this way you start a specialist and become a generalist. You'll know fully well what your tools and technologies will be capable of and you will be able to give reasonably accurate estimates. As far as I can tell, unless you have an eye for design, it's generally easier to start on the back end and move to the front end. This way you can be productive and are able to move between companies. Front-end is still changing a lot. But there are a lot of promising releases/tools in the works that are making getting started a lot easier.