Before I say anything, I know that I sound extremely foolish by starting this topic or even talking here at all. But I believe I have genuine questions and I hope that you people out there will be kind enough to me to give me some advice.

My epic vision/goal/dream that all of you would probably laugh at would be a game based off of a serious of books I used to read, about some feral cats (sounds pretty dumb already, right?). In my head I would like it to be something like a big epic Legend of Zelda Twilight Princess-style thing, that follows a fixed story line with certain worlds, and those types of 3D graphics and animations. It would not have weapons (cats don't have weapons!) but it would likely have to include other functions such as that type of battle and action system; several buttons do several things or create some sort of particular action. There would be things that I can't even think about programming that would have to be implemented for the game to be created as I see fit. There would also be many, many, characters that would need to run well on AI. And, foolishly, I can imagine movie-like cutscenes to help advance the action. Slightly differently, however, the characters would be confined to a quite finite world with no bosses or the like, and would have to perform actions that even I can't comprehend right now. Not to mention that it would all take place outdoors with no human characters whatsoever. It would be the most enormous project that any one person could actually try to embark on, only as a pathetic hobby to recreate some dumb books.

Yes, I know that I sound pretty dumb and all right now, but I also know that this isn't something I can start right now even with the best game engine. I don't have very much significant experience either, only with some minor coding and hexing in an old computer game of mine and a couple pathetic attempts at C++ (think "hello world!") when I was too young to know what in the world I was doing, anyways. Not to mention that I forgot basically all of what I have learned of C++, and that it now appears that it's pretty dumb to start learning with that language anyways. I know I won't be anywhere near my epic dream anytime soon (haha) but I would like some advice.

So if I need to start with simple games with pingpong and other clones like that, do I need just a compiler or do I need a full-on engine? I honestly wouldn't know.

As it appears to me, Javascript and C# are the two best languages I could start learning, or at least for my purposes. Would this be correct? And isn't C# a bit similar to C++?

However, when I get into more complicated 2D games, I have the sense that any and all work would be divided into two different fields: animations and graphics, and then the actual programming. When I get to this point, would it be a good idea to get an engine, such as Unreal Development Kit or something? As it appears to me, these engines can take a good load off, like with the whole Kismet thing.

And whenever I am actually awesome enough to embark on my dream project, I'm pretty sure I would need something like the Unreal Development Kit. For those purposes I would like support for large worlds and terrains, and natural stuff (like speedtree for trees, it sounds very useful as far as I can tell). What engine would have a not-so-horribly-steep learning curve that could make my aspirations possible?

Well, thanks to anyone that responds. I must repeat again that I totally have the idea that it will take me a lot to get to where I want to go, but I appreciate any help, whether I will use it now or later.

Personally I highly advocate C#, XNA is a good library to start with, it does a lot of the complex work for you in learning basic ways games work, but not enough to where you won't learn. There's a lot of material around for it too.

As for whether or not C# is similar to C++ or not, it has some elements but I would peg it much closer to Java than anything else(although a much more sensible version IMO.) I really do not recommend starting with C++, it's just counterproductive to learning basic skills.

Before I say anything, I know that I sound extremely foolish by starting this topic or even talking here at all. But I believe I have genuine questions and I hope that you people out there will be kind enough to me to give me some advice.
...
My epic vision/goal/dream that all of you would probably laugh at would be a game based off of a serious of books I used to read, about some feral cats (sounds pretty dumb already, right?).

...

And, foolishly, I can imagine movie-like cutscenes to help advance the action.

...It would be the most enormous project that any one person could actually try to embark on, only as a pathetic hobby to recreate some dumb books.

Yes, I know that I sound pretty dumb and all right now, ...

Not to mention that I forgot basically all of what I have learned of C++, and that it now appears that it's pretty dumb to start learning with that language anyways. I know I won't be anywhere near my epic dream anytime soon (haha) but I would like some advice.

Don't preemptively disqualify yourself, regardless how many times others may have disqualified you. How can we possibly form a good opinion of your ideas, if you tell us what opinion to form before we have a chance to review it? You said: dumb x4, foolish x2, pathetic x1, sarcastic 'epic' x2, laugh/haha x2. If you let me form my own opinion, I assure you it differs from the pre-packaged opinion you want me to have of you.

If you want to make a game inspired by the Warrior Cats books, then go for it! (As long as you don't violate any copyrights, by not cloning the series but rather letting yourself be inspired by it to create something new)

I've never read the books myself, but my younger sister has and greatly enjoyed them, and she's better read than I am (for example, having read Shakespeare and Dante's Inferno and the like, on her own initiative).
For additional inspiration, I suggest reading Watership Down (Which I have read. It's a great book! - normal rabbits that only are capable of actions real rabbits could do - similar to Warrior Cats, but even more realistic), Redwall Abbey (I only read the first one - many animals interacting in human-like ways (writing, swordfighting, etc...)),and maybe the Chronicles of Narnia (talking animals with human personalities interacting with humans in often non-animal ways).
Oh, and maybe hit up the Aristocats movie (Cats, but it's a disney cartoon, so very cartoon-esqe jumping and walking and such), which is 'play instantly' on Netflix - ask me how I know.

Also, 101 Dalmations, the book, is worth a read as well (dogs and cats, behaving in dog- and cat-like ways, interacting with humans non-verbally, but with human-esqe personalities).

Yes, alot of these are "children's books" (except for Watership Down), but I don't pass up reading a good child's story every now and then, and get alot of ideas from them.

So if I need to start with simple games with pingpong and other clones like that, do I need just a compiler or do I need a full-on engine?

Just a compiler, and whatever APIs (libraries) you'll use. No engines.

As it appears to me, Javascript and C# are the two best languages I could start learning, or at least for my purposes.

I think you mean 'Java' and C#. Javascript isn't the same as Java.

I'd also toss in Python as a highly recommended option.

So to narrow the 10 or so valid choices to 3: [Java], [Python], [C#]

And isn't C# a bit similar to C++?

Yep. C# is similar to C++. Java is also. Python is slightly similar to C++... and it's built ontop of C++, and interacts with it.

Many languages are similar to each other, and it just so happens C++, C#, and Java are all descended from the 'C' family of languages, and Python is directly built ontop of C++, iirc. That doesn't mean they are as hard to learn as C++; all three of those languages are 'higher level' (in abstraction, not in difficulty) than C++. Python is probably the highest level of abstraction of the three, and probably the easiest to get started with.

However, when I get into more complicated 2D games, I have the sense that any and all work would be divided into two different fields: animations and graphics, and then the actual programming. When I get to this point, would it be a good idea to get an engine, such as Unreal Development Kit or something?

Nope. Any 2D game you basically build on APIs and libraries, most of the time - an engine is almost a nuisance with 2D. Only well you get to 3D do you even bother with engines - and then it's almost a necessity. The exception are tools directed towards 2D like GameMaker, RPG Maker, Torque 2D, and Construct.

As it appears to me, these engines can take a good load off, like with the whole Kismet thing.

Yes, for 3D they are a great benefit. Not so much for 2D.

And whenever I am actually awesome enough to embark on my dream project, I'm pretty sure I would need something like the Unreal Development Kit. For those purposes I would like support for large worlds and terrains, and natural stuff (like speedtree for trees, it sounds very useful as far as I can tell). What engine would have a not-so-horribly-steep learning curve that could make my aspirations possible?

When you get there, then the landscape of available software will change (technology shifts rapidly). When it's time to go 3D, several years from now, then come and ask!