The Sudden Urge To Blend

The other day I had a sudden urge to figure out the new Blender. A few years ago I spent some time with Blender, and learned how to use it at a basic level. I first learned to use Python by writing a file exporter to export from Blender into my engine. But that was version 2.49. When version 2.5 came out they changed the interface. I tried to find some tutorials on how to use the new interface but it was hard to separate between tutorials on the new and old versions, and I was too busy to put the time in.

But for some reason I decided yesterday to try again. It took me a while to get Blender's camera controls to play nice with Ubuntu, but eventually I was able to get it running smoothly. I found a nice simple tutorial on youtube, and the result was this:

I think that's not a bad start. The new interface seems cleaner than before; I think it will be easier to use. I might have to dig out my old file exporting script and see if it works with the new Blender, but since my engine isn't running at the moment I think that can wait. I'm going to continue following tutorials for a while and try to reach a higher level of skill with Blender than I used to have (well, I'm going to try; I'm way too busy and have way too many hobbies these days). It would be nice to be able to turn out some simple animations or something.

In my last post I talked about game ideas. I didn't discuss some of the smaller features that I was thinking about trying to get my engine to support. One thing is to try to create a generic character "layout", so that it's easy to mod by adding characters that follow the file format and description.

A much more ambitious idea is to take that a step farther and create a generic human mesh. That mesh could be modified to be any male or female character as long as the vertices were only moved around a bit and none were added or deleted. I think there would need to be some redundancy in the mesh to make it easy to modify, but the idea is then that clothes etc can be modelled separately, based on the default mesh - the idea is that the clothes are "anchored" on vertices that do not change (only move around a bit). Then the engine can easily add any clothes to any model and they should fit fine, by simply fitting to the standard vertices. The idea of course is to have a character customisations system. This idea is not particularly well thought out I suppose - I haven't done any research or anything - so it probably won't be all that simple in practice. Well, it's something to keep in mind for the future.

Stealing Ideas

I was thinking about my old game engine the other day. I've barely touched that thing in over four years. Yet still I kinda feel like someday I want to do something with it. I don't know why, I just can't let it go. Some time late last year I put a few weeks into cleaning up the lowest-level module and converting it over to build with a makefile system (I'm running Ubuntu now, no more MS Visual Studio). I was having fun for a while trying to design a powerful multiplatform makefile, and I got the first module to build with it. I guess I ran out of time and/or steam shortly after starting to look at the second module, and of course then the momentum was gone.

I'm thinking about maybe going back to that second module, trying to get that to build under the new system. If I could get the engine running on Ubuntu that would be nice. I just have so little time these days, and to be honest I have less motivation than I used to; I don't really play very much these days, I have multiple time-consuming hobbies, I'm very busy at work, and with the Indie market huge on Steam these days there's seems to be less reason to try to produce a game for it's own sake, not when I don't exactly have any big ideas that I feel the need to see made.

Speaking of ideas, maybe I'll just put down what ideas I have had in the last year or so. I might have mentioned some of these before so bear with me.

First of all, back when I started this blog I had an idea for a game which I tentatively titled "Ghostwalker". The general idea was that as you got closer to death you would start to see into the spirit world; this could be an advantage since you would be able to see enemy souls through walls etc. A later idea involved multiplayer games where different players saw the world in different ways.

Well, recently I was thinking about that again and came up with a different take. This is heavily inspired by a short comic I read online a while back (I believe it was called Milk Run, although I remember something similar in a Spiderman comic once), where a man was stuck in a powered suit that fed him altered images of reality, making him think he was a knight battling demons. Remembering this, I thought, why not do that in a video game?

It would work something like this: the game would never break from first person. There would be very little narration; as much as possible the "story" -which would probably be quite minimalist - would just be communicated through events in-game, and as much as possible without trying to take control away from the player. There would be no HUD (except I suppose the pause menu). You would play a character who remembers nothing. This might not need to be explained outside of their body language in the first few seconds of the game, as they look at their own hands with unfamiliarity; I think perhaps it's enough that no backstory is given; he is as confused as you are (well, more-so since you know it's a game), which I think helps to put the player on the same page as the character they are playing, even if it can be clichéd. But hey, what isn't a cliché these days?

So however the game starts, the main character is quickly fighting for their life against demons or monsters or something, in old-fashioned or possibly unearthly environments (this would take some thought, but one element would be that all text that you come across would be in unreadable runes, as part of making the world seem unfamiliar). Early on you pass a mirror - and see your reflection: you are fully armoured (which you might have guessed from the sight of your gauntleted hands earlier). I'm thinking that at some point - probably the very first few seconds of the game, before the player is handed control - the character tries to take the armour off, but can't. It might be a good idea to scatter mirrors through the game so seeing yourself is a normal occurrence. You discover that you are capable of firing energy blasts from your hands (this might not even be scripted or explained; when the player naturally starts pressing buttons on the controller he discovers the "attack" buttons?). Perhaps there were glowing runes inscribed in the gauntlets' palms.

So the game goes on for a while. Some kind of narrative might start to take shape - I haven't put much thought into this, other than the idea that (in keeping with the comic) you probably learn about someone who you have reason to kill. Perhaps you are presented with a "vision" of the persons face, and he is made to look evil / shown at the head of the army of monsters/demons who you are fighting. Perhaps you occasionally come across text leading you to him, or perhaps you are just guided by visions.

After you've been playing the game for a while, there starts to be some bugs. Graphical glitches and so-on; animations that look off, etc. Nothing big, but it starts to get noticeable. Occasionally the whole screen glitches a little, but only for an instant, then it's back. This should be fairly subtle, always in-game (never trying to draw your attention to it in a cutscene or anything heavy-handed like that - heavy-handedness is the mortal enemy of plot twists), and it should go on for several hours of gameplay. Then it starts to get a bit more noticeable - occasionally you see models that look like they are from a completely different game - humans in modern clothing and so on. The model will probably be replaced after a moment with something more fitting in-game, like a demon. Then it starts to get even worse. As you pass mirrors, for just a moment (so at first the player will almost certainly miss it, or at best only catch it from the corner of their eye; eventually it might happen for long enough that if they are looking for it they will see it) the mirror doesn't reflect the knight in armour, but a man in a robotic powered suit. Finally - perhaps when taking damage from certain types of enemies - the whole screen starts to flicker and show you a completely different view, one of a slightly futuristic world populated by humans and robots rather than monsters and demons. During these glitches text suddenly becomes readable, the incomprehensible sounds your enemies are making are replaced by people yelling to each other in English, and so on.

I'm not exactly sure how exactly the full truth should be revealed (although I think it should be mostly obvious by now), but eventually you discover that you were locked in the suit and manipulated into killing a target (whether you realise this in time to not kill the target could be left open to player actions?). Perhaps there needs to be some motivation provided, like the suit is very powerful but people won't wear it willingly because it fries your brain or something, so in order to reach a well-protected target you were kidnapped, brainwashed and strapped in, and it's been feeding you an altered view of the world, but as it sustained damage it started to fail. What I do know is that surprise and journey of discovery is everything; it must be subtle, slow, and it MUST NOT BE REVEALED IN THE BLOODY TRAILER!

Ah-hem. Well, to be honest that's a pretty similar idea to a great game idea Yahtzee wrote about once, though the actual execution is obviously different.


Anyway, some other ideas I've had include:

-A split-screen multiplayer first-person arena deathmatch shooter that takes place on the inside of a miniature Dyson-sphere like structure. The general idea being that a) there aren't enough split-screen multiplayer games around, and b) a small arena where you can pretty much always see everyone else (as all you need to do is look up to find them) should help solve some of the problems associated with a "deathmatch" style game with only a few players playing.

-As above, but with fancy gravity mechanics, so there would be floating structures in the middle of the sphere, and if you get close enough (by jumping or using a jet pack or something) their gravity starts to affect you and pull you down onto the them, so you can go leaping across this space by jumping between the gravitational pulls of these floating asteroids and things.

-A strategic shooter (X-Com style) which obeys strict line-of-sight, so you can only see on-screen what is in the field-of-view of your characters, everything else is just blackness (except things that they have seen, which is then a muted grey to represent the fact that it's their last view of the place, but it may no longer be accurate).

-As above, but multiplayer with timed turns, so eg 30 seconds to decide on a character's action, then the character does it, now the next player has 30 seconds to decide what to do with their next character etc. Commands would involve things like "move then go into overwatch" or "move then scan for enemies" or whatever; you would have to que up all commands before triggering the execution, after which you have no control until your next turn. Not sure how well this would work, but I think the idea has potential - it would probably need lots of balancing though to give players a reason to advance without them very slowly stepping forwards while just hugging cover. I know that there's a game that works like this (can't remember the name), only everyone decides their moves at the same time then they are all executed at the same time, and I don't think there's any line-of-sight limitations? Dunno if there's anything closer to what I'm thinking than that, but I wouldn't be surprised if there was.

-A cartoony Street Fighter-like game that ONLY has special moves and ultra-over-the-top-super-moves, no regular jabs and kicks. Silly and probably relying much more on luck than skill, but could be funny for a few minutes.


As I've made clear, none of these ideas are very original, mostly they are slightly different takes (at least to the best of my knowledge) on what's already been done. Some of them I think have potential, but I suspect that even with modern game dev tools they are mostly too ambitious for just one man even if he had the time and skill, which I certainly do not. Which is why I'm writing them down; I'll obviously never be able to actually create them.

An idea for weapon selection in console games

In an old post I talked about a small idea for console shooters: basically instead of the current system of reloading, where you can drop an almost-full magazine without losing any ammunition, you would tap the reload button to quickly drop the magazine and lose the ammunition (which could be picked up again manually later) and load a fresh one, or hold the button to remove the magazine and stow it before loading the new one - this would be much slower but you would keep the ammo. The idea is for use in games where ammo isn't plentiful and resource management is a bigger deal, obviously.

Well, I just finished Army of Two: Devil's Cartel, and one thing I really liked about that game was the weapon selection system. I had an idea for a slight tweak to the formula. The basic idea is that the player can carry a number of main weapons and a handgun. It would probably be two main weapons which would typically be long-guns that can be slung behind the player's back, while the handgun fits in a holster; that seems reasonably possible in real life (not necessarily practical, but possible). In a similar way to Rage, a single tap would quickly switch to the handgun and back while holding the button would open a weapon-wheel.

The twist is that when performing this quick-draw the player doesn't put away his primary weapon, instead he keeps carrying it in his off-hand while shooting the handgun one-handed. Because he doesn't need to hang up the primary weapon the draw can be extremely fast, but because he's firing the handgun one-handed he won't have as much control, meaning accuracy and rate-of-fire suffers. Of course he wouldn't be able to reload the pistol with both hands full, so hitting the reload button would holster the pistol and reload the primary. unless the character has some kind of rig that holds magazine up so that the handgun can be brought down onto them to reload, like Lara Croft sometimes uses - that could be an upgrade item you pick up later. When selecting the pistol using the weapon wheel, he holds in in both hands so he can fire and reload it normally of course.

If the primary weapon could reasonably be used one-handed - another pistol or perhaps even a small submachine gun - then quick-drawing the pistol would allow the user to dual-wield the two weapons. I think it would be reasonable to be able to reload while doing so (the animation could involve tucking one pistol under the arm while reloading the other, for example).


This probably all sounds quite complicated on paper, but I don't think it would be that complicated in use, and it would complement my older reloading idea. So why bother with all these complicated micro-management systems? Well, because they add depth to the basic shooter gameplay. They force you to make rapid tactical decisions on the fly: do you reload and risk getting shot or killed before you can return fire, or quick-draw a pistol to fight your way out of trouble but delay getting your primary weapon back up? Hopefully it would make the right game more engaging. At any rate I expect it would appeal to some gamers; unfortunately those are probably PC gamers.

Idea for in-line kill-cam

I was playing Black-Ops at a friends house, the fact that you can play split-screen online makes it fun again - I might have to buy it just for that. Anyway, I died many times without any warning, which is a bit annoying. Of course it's fun when you insta-kill another player, which is why it's possible, but when you just fall over dead and you're not sure why (until you see the replay of how you died, if it's enabled), it can be annoying.

Then I had an idea. In some one-player games, when you kill an AI enemy in a dramatic way (e.g. a headshot with a sniper rifle), you get a slow-motion view of the kill from a more dramatic angle. As a repeatable example, in Stranglehold there is a special ability (I forget the name) where you take a single shot and, if it hits, you get to see the hit in slow motion from a better angle. So why not do that in an online game like Black Ops?

Let me clarify. In Modern Warfare and Black Ops, after you get killed you see a replay from your killer's point of view. My suggestion is; instead of waiting until after the player dies and falls to the ground, as soon as the player is officially "dead": stop, zoom the camera out a little to show the player, and in slow motion show them dying. If the get hit by a sniper bullet you will see the blood fly (and perhaps the bullet hit), if they trigger a claymore you can jump to an angle from above the player that shows the claymore, then show it explode and the player fall.

It won't work well for everything (so it can be applied selectively), but for some instant deaths (like the claymore or long-range sniper kill previously mentioned, which are not immediately comprehendable at the moment of death), it can avoid the momentary "how did I just die?" frustration while looking pretty cool. And there's room for a little cheating too. For example, if you do get killed by a sniper, the game can actually go back in time a little and follow the track of the bullet (camera exits the player's body, flies down to focus on the sniper, shows the shot, follows the bullet from behind on it's flight until it hits the player and we see them fall in slow motion) - basically the player himself is shown only in the frame he died in while the sniper can be shown only in the frame he took the shot, even if the two are actually a few frames apart, it won't look wrong because the bullet is supposed to be that fast. Whether or not that would look wrong for, say, a grenade-launcher kill I'm not sure, but it's probably worth a try right?

It won't impact the gameplay since the player is already dead and has lost control for a moment when it happens anyway, if the spawn time is shorter than the view time the player may be given the option to just skip it and get right back to the killing. Since these games save enough information to show the replay anyway, it shouldn't be technically hard to do, and of course it can be disabled by the player just like the current kill-cam. The player can view either the regular after-kill view or this in-line kill-cam view depending on the type of death.

Well, it's an idea. I wouldn't be surprised to see it done soon, the way slo-motion is being used these days - or perhaps it's already been used and I just haven't seen it?