Sometimes you realise you’re taking the hard road for no reason.

The board game I’m working on takes place on a 13x13 grid, as mentioned, but it’s rotated at 45° to the board (root prim). My initial viewpoint was that there’d be a 0,0 in one corner and 12,12 in another, which with a simple calculation of offset = X+12Y would reference a list of 169 local coordinate pairs (stored as vectors) to be referred to. Which of course would be great fun to plot and laboriously enter into the script.

So earlier tonight I fired up the spreadsheet and started plugging in some values, and only then did I discover a constant that was hiding in plain sight.

I could have kicked myself.

This constant also acts as a Z-offset to rest the game pieces on top of the board. It spaces both the X and Y axes on the playfield. Moreover, I finally realised, the centre position in the board has local X and Y coordinates of 0,0.

If, instead of 0..12, I thought in terms of -6..6, all of a sudden all I’d need is a multiplier and some simple math. No huge list. Just… one… float.

I think I’ll go back inworld and try it out - a second board and one piece that marches through each cell. We’ll see what happens.

The Second Life board game has taken a slightly retrograde step - but one that to my delight, actually worked right out of the box.

Previously, when I started, I had the sensible idea of having each seat maintain itself, sending link messages to the main script (hiding in the board prim) whenever someone mounted or dismounted. Unfortunately I didn’t understand that whenever someone did so, a changed() event was sent to all the scripts in the object, which caused Interesting Things to happen. So for a while, wanting to get shot of Interesting Things happening, I was contented by having seating and perms control handled by the one main script and one incredibly hairy-looking chunk of code.

Yes, that hairy chunk worked, but I was moving to a state-based system

Today I looked at it again and it occurred to me that, instead of four bulky and unpleasantly similar conditionals, I could have the seats asking: “Hey? That change? Was that me?” and send link messages appropriately.

And my code worked! I think there’s a little adjustment I could make which would be even more elegant, but that can wait for now. After all, for my next trick, I must not only roll the dice, but make it do something very important and cool.

You’ll find out soon.

(This post was originally posted 31 Jan 2011, but Tumblr’s back-dating doesn’t seem to work.)

I just now completed Fallout: New Vegas, nearly a month after purchasing and downloading it.

Initially, the independent ending wasn’t my stated intention; I just wanted to play through and complete the game at least once. So I was playing along with Mister House up until the fifth quest.

Kill an entire bunker of Brotherhood of Steel types? Uh… right. I sided with Yes Man instead.

It wasn’t until that point that I realised what was making me uneasy in my headlong plunge towards completion. It’s the fact that there’s no clearly defined “good” or “bad” endings - technically speaking. There’s three basic paths, but unlike, say, Morrowind or Oblivion, they have more impact than whether you join House Hlaalu or the Fighter’s Guild - they literally close off alternatives.

Despite this, the more I read online about the endgame, the more it begins to feel like Morrowind: inevitably, you find yourself pulled back to Hoover Dam/Red Mountain, then to the final boss fight with Legate Lanius/Dagoth Ur. The main difference is that you have to choose a Great House to join, instead of being able to remain a freelancer. (Which, I suppose, makes sense in a way; I can’t imagine some rogue element wandering through and picking off one of the most feared warriors in the Mojave.)

The fact that completing the game makes me uneasy about how it’s done can only mean one thing. Obsidian’s writing team did one heck of a job.