In 1st edition you got the multiclass elves that took both classes at the same time. They were balanced by their horribly slow rate of advancement. I seem to recall second edition being similar (which is why few people I played with ever chose elves).
3rd edition tried to standardize the classes in some ways and balance them as well. As anyone who played 3rd edition knows, for the first few levels the melee (non magic) classes do well but after about level 10 they are second class party members to the druid, cleric and wizard. By level 20, they are not much more than cheerleaders. In 3rd edition the wizard had no incentive to be a Gish. He could summon creatures that could out fight the fighter and still have 30+ spells a day to do other reality warping things.
Pathfinder has leveled the playing field somewhat. The Fighters can actually fight better then other classes now. Sure clerics can still buff themselves to a level equal to a fighter but it takes several rounds of buffing, in which time the fighter has probably killed most of the bad guys. The wizard is still bending reality, but he no longer can do it all as easily as before. Having a fighter friend around to deal the the segregated, slowed, and singed enemies keeps the wizard from having to exhaust all of his resources in a single battle. Druids can no longer be a wise but skinny weakling in human form and then turn into world wrecking animal. They are now strong in combat or master spell casters, not both.
Lets look at a few options for a sword and sorcery build. We'll compare them at level 12 which is around where most builds have matured but before the game has started to break down too much.
The first up is the Eldritch Knight. One level of fighter followed by 5 wizard levels seems to be the standard entry method to EK. At low levels if is awful and almost unplayable. The rest of your group will hate you as you are a deficient spell caster and a hopeless melee combatant. By level 12 things are not so bad. You are only 2 caster levels behind a full wizard so you can sling around a couple level 5 spells a day and you are starting to collect a few of the strong mid-game melee feats (although no "Greater" Fighter feats yet). I see the EK as a wizard that can melee a bit. Damage output in melee is quite comparable to the other builds I'm going to show but lacks their tricks for extra damage. My EK build focuses on Str and Int and probably needs some optimizing work. His saves were the worst of the bunch (12,9,12) but his hit and damage bonus respectable. +20 to hit, 1d6 +15 for damage. His AC was the worst at a self-buffed 23, but can Mirror Image .
The next build is also Str based. I am not a big fan of the unmodified Magus. While I have yet to play one, the Magus' schtick seems to be casting casting spells through his weapon greatly increasing his crit chance with them. They are very spikey damage dealers. The problem is that as a medium BAB class, the Magus has problems hitting high AC mobs. That's why I prefer some of his archetypes.
The Myrmidarch is basically a fighter with some spell casting ability. He gets armor training, weapon training (allowing the awesome gloves of Duelling), and Fighter feats. He loses most of the Magus's casting flexibility of some of his spells. I still think the trade is worth it. The Myrmidarch has a + 23 to hit, 2d4+18 for damage (+20, 2d4 +27 Power Attacking). His AC is a self buffed 27 but also can Mirror Image. His Saves are slightly better at 14, 8, 12.
The last two builds are both Dex based. First up is another Magus build. I always liked the idea of the Kensai but it took a while to find a trick to make it work. Since the Agile weapon feature is from a 3.5 Adventure Path it isn't technically Pathfinder material and so requires GM approval. The only other way to get Dex to damage to to use a scimitar and take the Dervish Dance feat. The Kensai Archetype is focused around a singular weapon and becomes very effective with it. Any time the Kensai scores a critical hit he can spend a point from his Arcane Pool to increase his critical modifier x1. It doesn't increase spell critical damage but it is still pretty sweet. The Kensai comes in at +20 to hit, 1d6 +15 to damage. Too bad there is no way to get either Power Attack or Piranha Strike with this build. AC is 26ish which is pretty good for no armor. Mirror Image is also an option here. The Kensai will end up with one of the highest ACs in the game. Saves are quite strong at 12,14,12
The last build is a much better team player than the others. This Gish is built from a Bard. The Arcane Warrior archetype makes a decent Gish at low levels but falls behind the To Hit curve by level 12 and starts having a tough time hitting things. I have recently spotted a better archetype for my purposes. The Dawnflower Dervish Archetype is also Dex based but uses Cha as a casting stat. Like most bards, this Gish has great skills selection and high skill values. He has more spells than the EK, but his spells tend to be more utility and buffs than offensive damage spells like the Magus. What makes the Dervish a good Gish is his Battle Dance class feature. Instead of Bardic performances effecting everyone, the battle dances only effect the bard but they are TWICE as effective. He can also cast Cure spells on himself 3 times a day as a move action. The Dervish's to Hit is +24, 1d6+18 damage (assuming 31 rounds of battle dancing is enough to cover ever battle). His AC is 28, he also has access to Mirror Image, and has a high UMD score to use Shield, and Barkskin wands for more AC goodness. His saves are the best of the bunch at 11,18,13.
I can post exact builds but I didn't want to fill up too much space with walls of stats. Ask if you want more details. I think all of them are quite playable and each has a very unique flavor. By the way if you guys have not read it, pick up Inner Sea Magic. I think I enjoyed it more than Ultimate Combat.