I used to make a point of buying a new set of dice every con I went to. Each character I play has their own die that are colour coded based on the class.

What. So I'm a little OCD, get over it.

I have a d30 kicking around in my backup dice bag that's a hang over from playing Tales of the Floating Vagabond years ago.

Obviously they were playing 2e and using "Complete Guide to Unlawful Carnal Knowledge".

I've been using minis for D&D since 3e and have found it speeds up combat hugely and reduces the amount of time people bicker about whether they can hit or be hit in a given combat.

The amount you ROLEplay will depend on your group and your GM more than your rules set. I've run systemless games than where all combat and system heavy games that were all characterisation. It's entirely up to the group how it will play out.

In the Australian convention RPG gaming scene there were rumours that SMOGs[1] existed in the background pulling strings. It sounds like you have also glimpsed their presence.

[1] Secret Masters Of Gaming.

Having played every edition since AD&D (1st ed), I actually quite enjoy the way 4e works. It's generally reasonably smooth, you get a sense of progression in the amount of damage you deal and receive, without the sill number of attacks you used to roll in 2e (for example).

I don't miss THAC0, adding up 1d4+14+2d8 is enough math for me on the average afternoon (yes that is a rogue with a dagger and combat advantage - add a couple of d12s if I crit).

I know it uses the same launcher and agent platform as the WoW:MoP beta based on blue posts in the tech forums.

If you think about the size of the game world, comparing D3 to WoW (for example), even with random generation of some terrains, D3 is probably smaller than a single continent in WoW with less mob types. It operates in isometric view. WoW currently clocks in around 25Gb I think for the Beta folder (there was a 17Gb download on top of my previous 17Gb directory, but it overwrote a heap of stuff - say a max of 34Gb if it was all new content).

For your last point, that's why I made the clarifying comment about using education for entry level. I strongly disagree with past experience being hard to verify. Sure, the exact details may be hard to verify, what you can verify is whether or not they developed the expected level of skill in the given amount of time. If I ask for 5 years of java experience, that's really a proxy for having developed core competency with the language (along with some of the technologies in the ecosystem), and that's what I'll test for in the actual interview.

People who get fired for incompetence or social problems are both pretty easy to pick out in an interview.

I've always found past job experience + #2 for confirmation to be best. Better, even, than #1 + #2, for anything other than entry level.