Update 55: #NoCrappyBuild

Today is the day we predicted Build 11 would be coming out.  And it almost did (on the beta branch), but further reflection revealed no one is served by a crappy build.  It's better to test it, polish it, and make it as good as we can before subjecting people to it.

Why's This Keep Happening?

Because we're really not as bright as we pretend to be. :-)

No Really, Why?

There is a steam forum thread about Build 11 that I update every so often about what is happening.  If I could summarize the last 8 months in one sentence it would be:

We keep thinking the game engine/simulator is more done than it really is.

So to complete Task X necessitates going and doing some work on the game world/engine/simulator before returning and actually doing Task X.  Usually the estimate to Task X is accurate, but the refinement or additions to the simulator are completely unknown.

I like the line in the Sully when Tom Hanks says "Everything is unprecedented until it happens.

In game dev everything seems to be working, until it isn’t.

For illustration purposes I'll relay this week's issue:

My task was to make an NPC soldier job who would help fight back against monster incursions to town worked hexes.  Based on having everything I needed: NPC jobs already worked, NPC interactions with monsters and lairs already worked, so I estimated 2 days.  Well it turned out:

  1. NPCs couldn't equip weapons.  It turns out the whole concept of equipping anything was built for the player only, so it needed to be moved up (base class) to any character (NPCs, Monsters) in the game.  Now any NPC can have a full equipment loadout, just like the player.  This means enemies can have real equipment loadouts that drive their combat stats.  Someone attacking you with a powerful sword?  Kill 'em and you can have it.  Guards in a rich city will have better weapons and armor and do more damage than in crappy towns.  This is much closer to Pen & Paper RPG.
  2. Soldier had to realize he didn't have a weapon, so go and buy one.  Based on his weapon skill and money available, and what the weaponsmith is able to create.
  3. The way Lairs "owned" territory was creating issues, so we brainstormed a new better way that, ironically, is closer to Pen & Paper rpgs: the patrol.  Monsters would be recruited at a lair, then launched as patrols/raiders to travel within a certain distance.  While better, this is a pretty big change to how lairs function, and low impact for Build 11 so we decided to fix what we needed for the soldier by rewriting part of the lair to pave the way to monster patrols, which you will be able to see on the map, in the near future.  This took better part of a week.
  4. With all the above done, create a new way of triggering an interaction between NPCs and monsters.  That then results in combat (if it is a soldier) which we had to then create a quick simulating version.

#1, 2, 4 took about 2 days to do.  It's the extra week of #3 that caused the big delay.  It left no time for testing.  The build gets pushed back.

Every time the simulator doesn't do what we need, we revisit it, improve it, and it becomes one step better in being the most advanced role playing engine ever devised.  And while we still encounter "holes" in it, each "patch" makes it more robust enabling it to do wildly more than we ever imagined.

So when is Build 11?

Sometime near the end of Jan.  I am away for almost 2 weeks, so we have to see where we are at when I return.  We're close.  Real close.