There was a huge hullabaloo on rpg.net recently on the desirability of rulings versus rules. To create a synopsis on this strange controversy: there are some people who object to the idea of the GM making rulings, because that is too much like the GM making things up as he goes along based on his subjective judgements, and is consequently unpredictable and open to abuse. There are others who like the idea of the GM making rulings, because it is quick and easy, masses of rules covering all eventualities do not have to be memorised, and it allows the GM to be flexible and respond using his own understanding of his own game and how he wants it to be.
I fall into the latter camp. But what rarely gets mentioned in these sort of debates is that all it is really is fiddling around the margins: even in the most rules-heavy games, the vast majority of the GM's work is composed of making rulings.
Firstly, outside of the specialized and strictured arena of combat, even if the GM is applying a rule mechanic, the means by which he applies that rule mechanic is almost always discretionary. For example, in most skill-based games, the GM provides a numerical difficulty rating which the player has to beat - but since the rule book will not be able to give anything more than a few examples to provide guidance in what that difficulty rating should be for a given task, that rating will ultimately come down to the GM's own judgement: it is, in effect, a ruling, albeit a guided one. Many mechanics in RPGs have this character, when you think about it.
Secondly, once you get away from skill checks and similar, much of what happens in a typical RPG session is effectively at the GM's discretion, though it is not commonly given the title "ruling". For instance, when the adventurers go and visit the Frog King in his throne room and enter into negotiations with him, how does he react to what they say? If the game includes a reaction-dice style mechanic and his reaction is "unfriendly", how does that unfriendliness manifest itself? What does the Frog King think of the PCs - will he try to manipulate them? Threaten them? How cunning is he? These decisions that the GM has to make are all functionally identical to rulings, even though though he probably does not think of them in those terms.
Likewise, in a fight with a group of kobolds, how do the kobolds strategise? How do they choose their targets for ranged attacks? What does Kobold 3 do this round? Does Kobold 4 run to get help, or throw a javelin at the party's magic user? Again, these decisions are functionally identical to rulings: they are down to the individual GM's judgement.
It will ever be thus. In most modern legal systems, especially in common law jurisdictions, judges have a great deal of discretion in how they apply the law. Terms of art such as "reasonableness", "proportionality", and "balance of probabilities" are everywhere; when a provision in a piece of legislation requires that an action be "reasonable", it is down to individual judges to determine the meaning of reasonableness in a given context - to make a ruling - albeit often with guidance. Legal systems recognise the importance of this approach, because it is not possible to write laws which will cover every eventuality and always be so reliable and accurate that they can simply be applied robotically whatever the situation.
The same is true of rules and rulings in RPGs: no matter how comprehensive a set of rules can be, the number of eventualities they can actually cover will be vanishingly small. It is not possible, for instance, to draw up a set of rules to cover precise niceties of social interaction - "If the players say x, the monster says y" - and this is why social interaction is basically an exercise in the making of rulings. Similarly, it is not possible to draw up a set of rules to cover all potential combat tactics - "If x, a kobold will do y" - and this is why the decision-making of NPCs in combat is an exercise in the making of rulings.
The great majority of what a GM does in any given session, then, is ruling. It seems quite plain that the only course of action, in light of these circumstances, is not to attempt to develop comprehensive rule systems (a quixotic goal), but to create games which teach novice GMs how to make rulings effectively.