It is silly advice. Moving the bishop on the second move is perfectly good but the standard advice is to move the knight to f3 first because it extremely rare that the knight is better on any other square it can move to, if at all.
Why 2. c4, and not 2. Nf3, for beginners?

For chess openings, I would recommend Fundamental Chess Openings by Van der Sterran. It's more of a talky book (instead of just tables of moves) that explains the ideas behind the main openings and the main lines. I'd also recommend checking out Heisman's Novice Nook articles (they're online).

the elephant gambit is refuatable - just go learn it, and it depending on your opponents level the latvian gambit can be declined with Nc3... that doesnt lead to a very complex position...

I'd also recommend checking out Heisman's Novice Nook articles (they're online).
at chesscafe.com. They're great, but I haven't read them in a while. I'll go check them out again.

the elephant gambit is refuatable - just go learn it, and it depending on your opponents level the latvian gambit can be declined with Nc3... that doesnt lead to a very complex position...
What exactly does it mean that it's "refutable"??? In know what the word means, but not in this context. Does it mean that Black will be in serious trouble, if White knows a certain series of replies?

Well, that doesn't mean Schiller's other books are "utter crap". It doesn't even mean the one Miles reviewed is.

"utter crap" is probably harsh, but I have Schiller's book, and I would recommend finding something else. You have to sift through a lot of unnecessary things and, yes, occasionally silly advice, like moving the bishop to avoid gambits. There are definitely better books out there.
I can't recommend opening books (I mostly use online resources), but I *would* recommend that you check out some of Jeremy Silman's books when you're ready to move beyond the opening, such as "The Amateurs Mind" and "Reassess your Chess". They are adaptable to any level of play, and they are some of the best reference material for positional play. It's just as important to understand the positions you get from an opening as it is to know the moves to make an opening. In fact, I think it's more important!
Hope that helps.

Thanks, fran.... I'm actually reading Silman's "AM" at the moment. Find the positional aspect of chess much easier to understand.
That's not actually what it means......in the context of chess, "refuted" simply means there is a counter to something that prevents whatever move or opening from allowing a player to gain a huge advantage. For example, the formerly dangerous Maroczy Bind variation in the Accelerated Dragon has been "refuted", but that doesn't the Accelerated Dragon is unplayable.

I think Schiller is just trying to help a new player, so you don't have to learn much theory. Certainly the Elephant gambit and the Latvian are not good, but you need to learn it. After 2.Bc4 f5 you can play the simple d3 if you want to, and you don't need to worry about anyone dislodging your f3 knight.
I cannot vouch for the book, just take it in the spirit it is written in. It's probably meant to allow you to play with a minimal of theory and/or complications. If you have more time to study, then take on more openings. But, no one can make a full opening reportoire at once, and it isn't even interesting for a new player to do.

The only problem I've found with Van der Sterren's Fundamental Chess Openings is that now I want to play all possible openings. They're all interesting!

Hahaha. I profoundly dislike all Dutch opening books that try to tell you every opening, half. The only general opening books I liked were from John Watson, ''Mastering the chess openings.''
If you really want to know an opening with all the variations, buy a book on it.

That's all good. But is it true? I cannot take things for granted, so I looked this up in the Game Explorer. It turns out that after 2. Nf3, the Latvian Gambit and the Elephant Gambit will be Black's reply in less than 1 % of all cases, added together. So, that shouldn't really be something to worry much about. Further, it appears that after 2. Bc4 (as Schiller recommends), there is the Calabrese Countergambit (2... f5). It's also very rare, but according the the stats, it's much more dangerous for White. It actually favors Black, whereas the two others gambits strongly favor White.
At first, you should stop evaluating rare moves according to the game explorer stats because the sample is made out of few games from strange-playing players (to have chosen that moves). Otherwise, you should deduce that 1.Na3 is the best move for White by far (60% wins, 40% draws... on a sample on five games).
Most of the theory recommends on the contrary 2.Nf3 first, because (by increasing subtelty, not by importance) :
1- The knight covers on f3 important squares against an attack from the black queen, while the bishop on c4 loses the protection of g2 ;
2- The knight has no other good square to go in the king pawn's opening, so it will anyway go to f3, while the bishop has at least two squares to go to (c4 and b5) ;
3- It pressures the pawn on e5, which should be the focus of White's strategy (according to Reti, and my opinion joins his) and takes initiative while the bishop hopes for tactical skirmishes with no real threats, after 2.Bc4 Nf6 3.Nc3 c6 and eventually d5 Black has taken initiative ;
4- It creates no target for Black while the bishop on c4 is undefended and attackable by a future ...d5 (see e.g. the line mentioned in 3-).
I am not saying 2.Bc4 is not playable. It is just inferior to 2.Nf3. If you are so afraid of the (unsound) gambits that may occur, play the queen's pawn (1.d4) with 2.Nf3 to avoid the Albin (d4 d5 c4 e5) and the Budapest (d4 Nf6 c4 e5) gambits, and you can't get in any tactical mess within the first moves.
Being a patzer, but not a pawn, I've now taken two steps back, to finally acquaint myself with the basic beginners openings I've been neglecting until now. I'm reading Eric Schiller's "First Chess Openings", and on page 40, he writes:
"After you have planted a pawn in the center, it is time to turn your attention to castling. In order to get castled, you must bring out both the bishop from f1 and the knight from g1. Your life will be a lot easier if you bring the bishop out first, because this reduces the number of options available to Black. If you bring the knight out first, you have to contend with the tricky and dangerous Latvian Gambit (2...f7-f5), and Elephant Gambit (2...d7-d5). Those strategies are not available when you move the bishop to c4 instead."
That's all good. But is it true? I cannot take things for granted, so I looked this up in the Game Explorer. It turns out that after 2. Nf3, the Latvian Gambit and the Elephant Gambit will be Black's reply in less than 1 % of all cases, added together. So, that shouldn't really be something to worry much about. Further, it appears that after 2. Bc4 (as Schiller recommends), there is the Calabrese Countergambit (2... f5). It's also very rare, but according the the stats, it's much more dangerous for White. It actually favors Black, whereas the two others gambits strongly favor White.
So, what's the deal here? Is Schiller giving silly advice, or am I missing something?