I know some general ideas about the sicilian, but I've got a question about the purpose of the move 6.Be3.

What is the purpose of this move besides obviously developing the bishop. Does if support the kingside pawn advance somehow? Or maybe take advantage of the weakened a7-g1 diagonal. I now this move has become popular but I am just curious as to the idea behind the move.

http://en.wikipedia.org/wiki/Sicilian_Defence,_Najdorf_Variation This site provises some general ideas of the Najdorf and even the typical move ideas and even some plans but it doesn't reveal the purpose of this move.

So what is the idea behind this move?

I don't think you need so much justification for moves like Be3. You have to develop the bishop somewhere, right? All other squares look funny to me like Be2 (blocking the queen and possibly a rook in the future), Bf4 (loses to ...e5 fork), Bg5 looks like too committing and unnecessary, and well, you've left with Be3 I guess. which controls the dark squares around the marching kingside pawns after f3 and g4. 

I like Bg5 because it has a clear reason. The e4 pawn can be easily created into a target pawn by black through moves such as ...Bb7 after the b5 push ...Nbd7-Nc5, ...e6 followed by ...d5. Not to mention the f6 knight already aimed at it so white is trying to reduce an attacker on this pawn. I don't see how you this move is too commiting as it has proved to be a fine move in many games

I get that other squares are are worse,but what makes e3 which is more popular today, but what makes e3 better than g5?

Bg5 is overly commital relative to Be3.  You don't necessarily want to give the bishop for the knight on f6.  e3 is a nice place for the bishop to stand behind the e4 pawn where it wont be disturbed.  Developing the bishop from c1 makes queenside castling happen faster.  ( I realize that is also true of Bg5. )  It probes the b6 weakness that is almost always created in the Sicilian by the moves ...c5 and ...a6.  I guess Be3 is one of those moves that doesn't have an absolutely concrete justification.  It is just the right place for the piece.

One thing is that Bg5 doesn't work so well with a e4-f3-g4 pawn attack, since it's in the way of g4-g5.

But in the actual variations it's also sometimes helpful that it controls c5 and b6 (think of the knight on c3, chased away to a4 by b5-b4, that also controls those squares).

Edit: That said, saying that Bg5 is "overly committal" or whatever is a bit silly as that has been the most important move in the Najdorf for much of the last four decades... It's just that Bg5 goes along with f4 better, Be3 with f3 and g4.

6.Be3 is a very subtle draw offer which can be accepted by 6...Ng4 7.Bc1! Nf6 8.Be3.

Be3 also helps control some of black's dark squares on the queenside, which can become weak and subject to attack if black rushes his queenside pawn advance.

i dont think so rigamagician because 7.Qxg4

In the dragon variation, white put their bishop on e3, their queen on d2, and moved the bishop to h6 to trade it off for blacks. The attack is very effective, and people decided to go for the same setup against the Najdorf.

i think white loses after 7 Qxg4 Bxg4 but im not sure

Yeah, after Bxg4, White might have chances to hold a draw.

6.Be3 Ng4 7.Qxg4?? Bxg4 is a very bold and original queen sac indeed.  White might find it hard to wriggle out with a draw though I fear.  Those darn queens can come in handy at times.

 Although! That setup wasn't believed to be effective against the Najdorf for a long time since white won't really prepare to exchange bishops and it was therefore believed that 6. Bg5 is the better move. However it was later discovered that the setup with Be3, Qd2 promotes a plan with f3, g4 better than the setup with Bg5 (because the bishop is blocking the crucial g5 point).

When this was discovered 6. Be3 also became a big part of the enormous theoretical discussion.

Exactly