I agree that the type of use of the grid in 4e (and this was also true in 3.x, though distances were still measured in feet or inches or something) is materially different. In AD&D (1e particularly) it is fairly unclear exactly what the expected process is. The game is all built on the classic Chainmail style 'just measure on the table' concept, with what [MENTION=42582]pemerton[/MENTION] quoted showing that minis physical size basically dictated what could move where, who you could attack, etc. Often we would mount our minis on fixed-size bases, which is also a technique borrowed from minis wargaming (both Battlesystem and Swords & Spells perpetuated this for D&D mass combat).
However, then there are hexes and squares. My understanding at the time is that these were really just a convenience to obviate the need for a ruler, you could count off squares and get a pretty good idea of the range, and hexes were even better (but don't work so well in built areas). These were also common in minis wargame play, and the wargame club I belonged to in the late 70's had terrain cut from MDF and styrofoam which conformed to the hex grids ruled on all their (12 or so) 4' x 8' gaming tables.
AD&D's DMG seemed to imply that you could actually use the squares/hexes as a positioning mechanism where PCs 'occupy a square', but beyond what Pemerton quoted (and even there it doesn't actually state outright that things occupy squares) the idea doesn't get any support. Still, as time went on and our wargamer beginnings faded into myth and legend we did tend to that kind of technique (say by the late 90s).
Still, you can argue a lot of things from the 1e DMG. Page 70 has a paragraph titled "Who Attacks Whom:" in which it is stated that it is "generally not possible to select a specific opponent in mass melee." This is rather at odds with any notion of a system in which creatures have an exact position, as that would tend to make who attacks whom fairly obvious. You could however go with the randomization technique next proposed only within the bounds of what you are positioned near, but the text doesn't seem to assume that!
Page 71 has "Example of Melee:". This is rife with contradictions to the above rules WRT how exactly you might use positioning. It starts by describing the positions of the two parties in moderate detail. Almost immediately one of the fighters in party A rushes forward to specifically target a given member of party B, which seems potentially to be in contradiction to the 'generally you can't target a specific opponent in melee', but MAYBE this is an exception because he's closing? We don't know, there isn't any stated rules on how that works! Nor does the text describe the GM or players moving miniatures around or anything similar.
The example then continues in the same vein with the various opponents selecting targets and attacking, even though by now a general melee has certainly broken out. Eventually one of the casters tosses a Web and some of the PCs from each group are caught, but it is never really explained HOW this determination is made! It appears there was some kind of calculation involved, but the example either leaves out all the information related to positioning (IE where the minis were moved or whatever) or else the DM truly winged it. If he did and I was part of party B I'd be pretty miffed about the Web, as who was caught in it seemed fairly arbitrary.
I'd also note that the entire encounter doesn't mention any use of terrain at all, even though there is a corner which could have factored in. Overall the 1e DMG is pretty hazy on what exactly the process for position in combat is. It doesn't really ever explain how movement is adjudicated either. The party with initiative gets to react first, but it doesn't say that they execute moves first! How movement is handled vis-a-vis initiative is also a mystery left unsolved by the rules!
