How do you start learning physics by yourself? I think this question has its place here because I am sure some of you are "self-taught experts" and can guide me a little through this process.
Considering that :


*

*I don't have any physics scholar background at all.

*I have a little math background but nothing too complicated like calculus

*I am a fast learner and am willing to put many efforts into learning physics

*I am a computer programmer and analyst with a passion for physics laws, theories, studies and everything that helps me understand how things work or making me change the way I see things around me.


Where do I start ? I mean.. Is there even a starting point ? Do I absolutly have to choose a type of physic ? Is it possible to learn physics on your own ?
I know it's a general question but i'm sure you understand that i'm a little bit in the dark here.
 A: Like you, I want to self-teach myself physics, yet here I am still at around the same stage I was a few years back. Why?
Because to learn physics effectively, I need to be immersed in it for days, weeks, months even years at a time. I also need to be coached by good teachers and peers that can steer me in the right direction, and prevent me from being lead astray by wrong ideas and bad problems. It's therefore essential people are in the right social environment.
So my answer would be to find the right social environment to make things easier for yourself, such as studying part time at night school, joining a physics club, forum--as you've done here.
I really do think people underestimate the importance of the correct social environment when studying physics because of the support it provides and the warding off of depression from social isolation.
A: I would recommend starting with calculus and maybe linear algebra. A basic understanding of the properties of functions, derivatives, integrals and especially differential equations is vital. Vectors, matrices, tensors, different coordinate systems and their metrics, vector calculus are equally important. I would also take up a book on general (classical) physics, a good reference for that is Physics for scientists and engineers by Randall D. Knight. It's a big book, but it starts from the very basics and it goes all the way up to Special Relativity and QM (although you'll probably want to study those topics more in-depth in other, more specialized books).
From then on, analytical mechanics would probably be a good choice of topic. Learning about the lagrangian, Hamilton formalism,... That should give you a nice idea of the beauty in theoretical physics as well. And from then on I think you can go pretty much any way you want. For QM, a good book to start with is Introduction to Quantum Mechanics by David J. Griffiths. But you're not there yet, the maths comes before the physics. Good luck!
A: 1.) Find something that interests you. The secret to learning is to do something you can be passionate about. For one person it may be building metal detectors (cicuits etc.) and another may be more interested in string theory or crystal physics. Explore your local library's physics section. 
2.) Become competent in the area that interests you. Thomas Edison was home-schooled and taught himself everything, but he also was given the freedom to learn whatever interested him. You will have to use relevant books. The internet alone will not suffice. 
3.) Learn the Math. Focus on the theory and the equations should come naturally. Albert Einstein had to learn a lot of math before he could express his ideas in equations. Calculus 1, 2, and 3 are commonly used in much of physics. Khan academy may be useful here. 
4.) Find another topic that interests you within physics. That should be easy since you will doubtlessly have stumbled across many fascinating concepts while investigating the first.  
5.) Repeat steps 2-4.
There is nothing we can't do if we work hard, never sleep, and shirk all other responsibility.
A: The learning time problem affects everybody, physics is intimidating because to learn it, you have to recapitulate the history, there's no underlying axiomatic system to deduce from. On 't Hooft's website, you will find a self-study guide put together for this purpose. It should get you started, and I don't think I can improve on 't Hooft.
But if you know mathematics and programming, there is a simple way to study the field--- just go over the famous problems yourself. You can simulate the Ising model in a very short time:
Make a 2d grid of bits with value 0 or 1. Choose a bit at random, and calculate the energy it feels with its neighbors, by counting the number of neighbors that have a different bit value. The energy is this number.
Now flip the bit, and if the energy goes down, keep the flipped value. If the energy goes up, keep it with a probability $e^{-\beta\Delta E}$ where you are free to change $\beta$. Do this a very large number of times, and you get an equilibrium configuration. Then you can draw a picture of what it looks like.
As you change $\beta$, you will see the phase transition appear. The Ising model will settle to be either mostly 0's or mostly 1's. Studying this transition and the related problems is a good introduction to most modern physics (past 1940). It contains the seed for everything except string theory.
A: Just a note in addition to the advice being given here is this:
ACTUALLY DO THE PROBLEMS. Like on pen and paper. Do not under any circumstances look at a solution and go "Oh yeah, I get this. Next!" That is absolute bull and what many, many people who attempt to self-study physics end up trying and why a lot of them fail. It is very easy to skip on the grinding, difficult work associated with trying to actually solve problems and just read examples and theories but at that point you may as well pick up a popular science book and save yourself some heartbreak.
I am not kidding about this. If you take one piece of advice from this thread, let it be this. I am sure other people who have been formally educated as physicists will echo my sentiment.
A: You start by starting the process.  It really doesn't matter all that much how you start, only that you start.  Go to the library, look through some books, etc.
At first, you'll find much of what you read opaque.  But, in a short time, you'll start connecting some dots and then more and then more still.  You'll revisit material that was initially incomprehensible to you and find that bits and pieces now make some sense.  This non-linear process will continue with great strides followed by slow spells.
At some point, you'll determine that you must make the effort to become familiar with the language of calculus if you're to make any more meaningful progress.
Once you've done that, it's like you've opened your eyes to an entirely new world.
A: First and the most important thing to do is to study calculus . I recommend the 2 textbooks by Apostol.Then, study classical mechanics(Taylor is good) and electrodynamics (Griffith) .They are basically calculus .If you understand linear algebra, read Griffith book in Quantum mechanics(You don't need anything outside calculus and linear algebra to read this book) Classical mechanics ,Electrodynamics and Quantum mechanics constitute the foundation of physics . After that , you can learn whatever you want in physics or math.
Be sure ,to solve all the problems contained in the textbooks you read.
A: First and the most important thing to do is: don't get Apostol's Calculus. Unless you want to go into math as well, I don't see the purpose of getting that book when you want to self-study physics. Second, I am shocked that the previous commenter said that you only need calculus and linear algebra for Griffiths' QM when he said in his book's preface that you also need to be familiar with the complex variables and fourier series. You need more math than just calculus and linear algebra.
So what do you do? I say that you can do this free. Here's my list with websites.


*

*precalculus 
(http://www.opentextbookstore.com/precalc/)

*Strang's calculus 
(http://ocw.mit.edu/resources/res-18-001-calculus-online-textbook-spring-2005/)

*Nearing's undergraduate math methods
(http://www.physics.miami.edu/~nearing/mathmethods/)

*Stone and Goldbart's graduate math methods 
(http://webusers.physics.illinois.edu/~goldbart/PG_MS_MfP.htm)


And for the physics? It's just one website: (http://farside.ph.utexas.edu/teaching.html)
A: It strongly depends on what you want to do with your physics knowledge. 
If you want to make something practical or experiment then don't agonize over the math or you'll end up learning math alone and you won't understand the whole point of doing physics. Learn math on an as-needed basis and stick to the basic undergraduate physics literature. 
If you want to do some theoretical/simulation modeling or just curious about the universe, like for example string theory/astrophysics/field theory etc, you need to start from basic calculus and work your way up to multi-variable calculus, differential geometry, abstract algebra and group theory. Then study just a little about introductory classical mechanics, E&M and go straight into QM/QFT or whatever. Don't be scared. Solving problems as opposed to just reading stuff will greatly quicken the learning process. 
Hope this helps.
A: I've begun a self-study schedule on my own.  I began with a review of algebra and geometry and have now begun calculus.  I'm basically trying to follow a typical college physics curriculum.  There are alot of resources on the internet: MIT online, cheap used books on amazon and barnes & noble, etc.  I think most of the above comments are on target.  I want to get the math down first, then on to classical physics....like college.
