Opening Principles

I was taught this:

1. King Safety

2. Center Control

3. Development

4. put the e or d pawn in the middle 

5. Knights before Bishops

6. Keep the Queen back

7. Do not move the same piece twice

8. Castle

9. Connect Rooks

-----------------------------------------------------------------------

I am cool with the first 3 principles, however the order of development ideas bother me quite a bit.

If both players followed this strictly we would basically play 4 Knights game all the time.

So what is the justification for breaking the order in say, Bishops Opening or Italian game for example? Center Game? Counter-Center? Gambits?

Then Hyper-Modern Openings seam to play with a completely diferent idea of development.

So why is this order taught in so many chess books?

Is it time to play by a diferent set of opening rules, or perhaps strip this back to just the first 3?

These are different kind of principles, and most of them just concrete ways about how you should realize a given principle. 

For example, first you wrote King safety. Yes, it is very important. And usually you place your king into safe position by castling. So the 1. and 8. principle are related.

Then there is the 2. principle you wrote, Center Control. It is very important in the whole game, not just in the opening. The 4. is not really principle, rather just advice is one way to realize a part of the goal, to control the center.  The pawn on e4 or d4 (e5 or d5) controls the center well.

5. Knights before bishops means that for example if you develop your king side pieces you should develop the knight before the bishop.
This can be seen for example in Ruy lopez, Italian game, Petrov defence, and in most of the queen's gambit lines as well.
There are other openings that don't follow it, but it is the usual way. The bishop is active on its initial square, the knights are passive.

But it does not mean that you should play 4knights variations.

Don't develop the queen early and don't move the same piece twice are just another rules, these are slowing down your development, they give lots of tempo for your opponent.

"Then Hyper-Modern Openings seam to play with a completely diferent idea of development."

In the opening it is important to control the center. There are two ways how this can be done. The first is the classical way, placing your pawns in the center.

The Hyper-Modern way is to control the center with your pieces, and later attack it with your pawns.

Both are good ways.

Basically what you need to know about principles is this:

Control the center.
Develop your pieces.
Castle. 
Develop the queen to connect the rooks.

There are many rules on how to realize this, (do not develope queen early, do not move same piece, etc) but this is what you need to keep in mind. 

hihi, silly hypermoderns, always ruining everything

i'd say that it's better to learn the rules before learning the exceptions, though it's good to know that these are in no way absolute. :P
anyway, you are kinda right, 1-3 are the actually important things that no respected opening ignores, hypermodern or not (if one seems to, it's either just on the surface or there is good compensation) and 4-9 just possible consequences of 1-3.

like, if you waste too much time with the same piece, it's likely that you'll fall behind in development.
and, if you bring the queen out early, that might also lead to waste of time if it's too exposed to the opponent's minor pieces.
"knights before bishops" is a funny one, the reason behind it is because the knight often only has one square that is clearly best, while the bishops usually has more choices. i think it's "wrong" though, like the next poster stated. ;P

Every rule has its exceptions.

1. King Safety - nice and vague. For instance the 1. e4 e5 2. f4 exf4 3. Bc4 Qh4+ 4. Kf1 - the king is "safe" i.e. it's not going to be mated with fairly good play on white's part, while black's queen is exploitable.

2. Centre Control - this can be achieved directly via pawns and classical development, or via hypermodern ideas with fianchettoes and flank pawn breaks, or even via indirect control in e.g. Trompowsky (Bg5) or Chigorin QGD (...Bb4, ...Bg4 and ...Bxf3.)

3. Development - This is the big one. It can mean absolutely anything, and therefore is the best principle out of the lot. Basically get your pieces doing something.

4. Push central pawns - either immediately or as a freeing break. This is played in almost every opening, or as part of a middlegame plan.

5. Knights before Bishops - It's not a good principle. The correct way of stating this would be "play the obvious before the optional". The reasoning behind this is after, say 1. e4 e5, It's fairly obvious where the knight should go (hint: 2. Ne2 is not best) but not at all where the bishop belongs. After 2. Nf3 Nc6 follows, now it's apparent the bishop can go to c4 (aiming at f7, controlling d5) or b5 (indirect centre control eyeing the c6-knight and d-pawn.) 3. Nc3 on the other hand would restrict options with the c-pawn (c3-d4 is a common plan in Italian/Spanish games) while keeping bishop options open. 3. Bc4/Bb5 keep options with c3 or Nc3 open, so all options are fine. "Knights before bishops" is a specific case of the maxim "keep your options open" or retaining flexibility.

6. Keep the Queen back - Unless it's a good move! In the Scandinavian 1. e4 d5 2. exd5, the "principled" move 2...Nf6 leads to the Icelandic Gambit, which gives a pawn for fast development and hopefully an attack. 2...Qxd5 instead exposes the queen, but for white to take advantage has to play 3. Nc3, blocking the c-pawn, and limiting his central control as a result. Black gets a solid pawn structure and immediately simplifies the centre for equality (one central pawn each, black has no weaknesses.) In other openings, 1. e4 e5 2. Nf3 Nc6 3. c3 d5 4. Qa4! (Ponziani, pressuring e5 via the pin) and 1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Bg5 Nbd7 5. Nf3 c6 6. e3 Qa5! (QGD Cambridge Springs) are cases of early queen moves that are possibly best.

7. Do not move the same piece twice - Unless it's attacked, or needs to be manoeuvered to a better spot. Like ...Ne7-g6/f5 (French), ...Na6-c7 (some KIDs) or even...Bb7-c8-e6 (some QGD Tarrasches.)

8. Castle - not always, but in general yes.

9. Connect Rooks - Sometimes it's better to leave rooks and king at home, like in some Caro-Kann classical lines, where an Rh1 supports a pawn on h5 for instance.

Gambits - 1 pawn ~ 3 tempi in the opening. Gambiteers sac pawns for a lead in development and with it, quick dashing attacks if the defending side plays too sloppily. The key idea is that pitching pawns accelerates 3. Development which gives a temporary advantage for that side, which in turn can hopefully later be converted to equality or a permanent advantage with accurate play.