How to learn Java-android?

In graduate school, is it more important to learn to do things, or, to learn how to learn how to do things?

• By learning to learn how to do things, I mean that the information you are trying to learn or the skill you are trying to acquire, would be less valuable than the process of trying to acquire those skills. Is the academic knowledge that you have directly learn in graduate school, more important or less important than grasping the know-how to acquire these knowledge? After all, the knowledge you have gathered in graduate study may be so narrow in its scope that it might cease to be applicable out of your department or group once you have graduated with a PhD. On the other hand, the skill of acquiring knowledge can always be utilized to pick up knowledge in another field new to you. How many percent of effort should a PhD student invest in learning new things, and how many in learning how to learn new things?

• Answer:

  I'd definitely say HOW. Doing something is actually quite easy, if you think about it. Anything is easy once you know the recipe behind it. Even the hardest theories can be broken down to trivial equations if it has been thoroughly researched beforehand in the past. In terms of transferrable skills, I can learn about Organic Chemistry, or Einstein's field equations. But are they transferrable? All of the topics ultimately boils down to how well you can solve a problem. That doesn't change whether you are studying Physics, Chemistry, or Mathematics. Sure, there are some trivial differences but the mindset is the same. Individually, they are not very transferrable, unless you plan to go into that SPECIFIC field. This is rare in academia. Now, knowing how to solve a problem? Since graduate school and PhDs in particular is all about solving problems that are NOT known, knowing how to approach a problem is critical. You have to look at a project in a analytical manner, and break down whatever complexities there are into something more manageable. You often *need* to learn new stuff on the fly since you will be looking at papers that are not strictly related to your topic. For example, I research Nanoparticles using Molecular Dynamics. Does that mean I can get by just learning about those two things? NO. A bit of background; I'm (was) a Organic Chemist. Which means that I had little to no experience of Nanoparticles or anything related beforehand. I've had to learn on the fly (with a bit of blood and tears involved..) Programming (Honestly, my supervisor literally threw the B&K C programming book in front of me and asked me to code) Monte Carlo and Markov Chains Nanoparticles and their research done in the lab (I hate working with them so much) Learning to use LAMMPS without guidance from supervisors (they all use GROMACS - those *******s) Using Linux and so and so. I didn't have time to divulge in huge detail about the above topics; I needed to get to a working standard so I can *produce* something. I still struggle, but what I did was translate whatever gobblty-gook that was in the textbook/paper of so-and-so and break it down into its nuts and bolts. That  is the skill that is required and the reason why many PhD students are often depressed; they never feel like they are good at anything since they are always struggling to learn something new very quickly. So I suggest not worrying about the topic at hand too much. Because no matter how complicated a paper or textbook comes down to, what you *do* need to find is: What's the problem How did they solve it How can I use this in my work Thats it. One can be all fancy and talk about the philosophy behind all of it blah blah.. thats utter BS as far as I'm concerned. Science is all about knowing how to approach a problem. The approach I take with unfamiliar stuff is: e.g. A extremely mathematical notation heavy equation, we're talking about hamiltonians, PDES, integrals etc.. Sigh (loudly) and go on a cursing spree What is this (goddamn) equation telling me behind all this mathematical crap? How can I code this without tearing my hair out? I don't care about the mathematics behind the theory and the so-called 'elegance'. All I want to do is make it work for my problem.