See, that’s what the app is perfect for.

This is just going to be one of those trivial notes on working with artists.  Similar things have been said many times before me, but, as usual, before you step into it yourself, you just won’t get it.
We have a beautiful game interface, but its support is turning into a bigger and bigger pain by the day.  Time and time again, here and there we come across non-extensible menus, with running out of space for buttons and information on the “soldiers” icon, with the duplication of the same information in various interface spots.
I really wish I could say that all of this is our concept artist’s fault, but it it’s not.  It’s my fault for delegating this to a concept artist.  Thousands of my colleagues’ reports stressed that this is a task for a technical specialist, but we really wanted to make it beautiful.
Ok, it’s beautiful. But whether this beauty will be dragged into the release, or will have to be abandoned for something more technical and functional – that I don’t know.
But our concept art is really amazing – not every AAA project has works that are comparable in quality.
Their high-resolution versions are here.

29.09.15, game designer Smaka

Here is where we finally get to the actual journal entries.

Having extended the deadlines for a year or two (today StSq is, without exaggeration, our life.  Skills and experience that we’ve acquired while working on it already exceed all of the cumulative game development experience we’ve gained before we started).  We were heading into the homestretch. We’ve seriously reconsidered the game mechanics’ deficiencies.  We’ve planned the changes to the gameplay, so that it would work the way we wanted and it’d be fun to play.  We made a checklist. And started going at it.

To make going at it more fun we do it sprinter-style, in weekly stretches.  For the last month now we’ve released a new Early Access build every week.  It keeps us awake =)

But this makes the funniest things sometimes.  Like when we sketched the destroyable obstacles, and everything was working fine.  I embedded it into the level, assigned destroyability to the boxes. A stack of boxes. When I threw a grenade, the top box was just left hanging in the air.  And what are you going to do about it? You can’t embed physicality into the environmental elements at this stage of development.  And it’s not really required by our genre.  I’m sitting there, giggling, looking at our 3D modeler, and he’s smiling, too.  But in my mind (and his too, probably) a wave of panic is rising – this should be in the release 2 or 3 iterations later, may be we should forget the destroyability, but that’s a shame, we spent so much time writing it, but what can we do…

Well, it’s clear enough – you have to go through each level and fill it in rationally.  Which means that if the boxes are stacked, they should have a common damage collider.  But all we could do at that first moment was giggle.

A bit later, in the evening, when I played this thing, I figured out that this feature doesn’t fit well into the current level design, and that it’s something that needs thought.  And I don’t even know how to start.  It’s not very clear what can be destroyed, and what would wreck irreparable damage to a script of any level.  X-COM had destroyable obstacles and hideouts, I should play that…

28.09.15, game designer Smaka.

So, the huge X-COM terrain tiles, as if constructed from 1.5 meter cubes annoyed the hell out of us, just as the complete lack of the technical equipment (armored vehicles and stuff like that).

We made two excellent decisions:
- To create an uninterrupted terrain without any reference to the grid, that would have the opportunity to rotate any of the environment’s elements at any angle.
- To create equipment, and not just any equipment, but one with a path search capability that would take into account the wheeled transport’s turn radius. (As far as we can tell, no one before us has thought of the same useless and extremely expensive feature)

We created the non-gridded terrain pretty quickly, and it really does look amazing.  But various difficulties come up often enough.  For example, if someone is hiding behind a stack of bags, he sometimes may shoot through it, and sometimes shouldn’t even try, since he has a vertical play of about 30 cm due to local bumps.  In conjunction with the completely physical shooting it is sometimes very counter-intuitive, and therefore learning to play StSq is fairly difficult (considering there’s still no tutorial).  Anyway, we wanted it to be played approximately like Warhammer 40к, and that’s approximately how it came out, which is not exactly fast and with a serious effort.   Plus the rippled terrain isn’t always visible on levels with bright sunlight.  

Having implemented automatic squatting (all of the soldiers now shoot standing up, and thus shoot over their hideout) and having rearranged the locations, we’ve solved the issues mentioned above.  Now it’s all nice and smooth.  But the amount of time and nerve cells spent is incredible.  And this approach to terrain really doesn’t add much, if anything, to game play.

As for the equipment with turn radius, it’s simpler and scarier.  Considering the lack of adequate analogs or any ready-made solutions, our lead programmer was writing the code for it from scratch.  Considering the research done, that kept him busy for over half a year (the time we originally thought the completion of StSq would take).  What resulted was an awesome solution for searching in 3D space, where x and y are the coordinates, and the third variable is the angle of rotation.  Under these conditions (due to the above-mentioned uninterrupted terrain) the 150x150m terrain takes up 451*451 units on the (x,y) plane, so much of the research was dedicated to the optimal way for the path to be found in this 3D space in tens of milliseconds, not tens of seconds.  But the lead programmer, who previously had nothing to do with game programming, gained some serious experience in convex geometry. I’m particularly happy to see this as my educational background is in mathematics.
 Of course, we can now have tiny cars going around, and in the future games we’ll have authentic-looking horses, and all of this is really incredible.  But wow, it was far from easy (psychologically) to extend deadlines again and again, while this was being implemented.

But when I look at our locations now, it sometimes seems to me that it was worth it.

28.09.15, game designer Smaka.