Progress update

I had parked development of AxleTramp for a while because it was getting to the point where I'm starting to finalise some of the code but I didn't want to go too far down a route where I'd have to rework it again to add multiplayer. This was before the new Unity multi-user called UNet had been released, actually over a year after it had been announced but a release wasn't looking likely on the roadmap any time soon.

When developing projects I like to hit as many 'quick wins' or get unintended bonus code/products as possible by spinning out key waypoints and sub-systems that can add extra value. This seems to be at odds to the usual interpretation of Agile I see in business. In my experience, if development of a product was a journey from A to Z then many will want you to go directly to Z, not passing any other points and just deliver tomorrow. This is often accopmanied with a hard-line 'agile' blinkering- just do the bare minimum and nothing else, leading to conflict over attempts to break down the product into tasks small enough to be done in a sprint, as identifying an order of precedence or dependencies is all a bit too Waterfall somehow.

In an earlier update I mentioned that spinning out some of the wing aero code helped me find a logic error in my implementation of the downforce aero code. Cars, even how I drive them are largely 2D in their motion so side-projects to the rescue again! Another of my man-child like interests is flying drones, however I'm not really any good at it unless you count turning expensive bits of electronics into burnt and broken pieces! Multi-rotor drones have the sort of erratic 3D movement which is great for tuning multiplayer networking! If you watch some of the best drone pilots' videos the races or freestyle flights are often done around derelict buildings or multistorey car parks, so modelling some of the assets could be shared between the games too as I always loved the dirty post industrial style of the earlier Wipeout games, so traffic cones & signs, street-lights, shipping palettes and industrial items like work lights and oil drums might have some cross-over. If you fancy playing an earlier alpha version of that then head over to DroneKill on the Play store.

Axle Tramp demo level: Loops of DEATH!

WHAAAT? A playable demo rather than dodgy renders and rants about Javascript??? Yes, enjoy, send feedback, spam, hate mail, pro-Javascript propoganda ... no please don't send info about TotallyAwesomeWhatever.js just let me know what you think of the demo.

This is still an extremely early prototype, so may have poor performance, crashes, explosions, implosions, damp fizzles or any possible negative outcome. Given the modern day war on evil plugins allowing you to do anything other than make kitten websites you may need to try this in a different browser to your usual one. You may also have to install the Unity plugin or un-disable it if you are using Chrome or stop it re-disabling it.

This is just a short test track to show off the loops. However to make it super easy to do two loops in a row I've raised the ride height, stiffened and also damped the suspension more as the centrifugal force is quite high so by the time you get to the second loop probably now having been full throttle and burning NOS for a good few seconds it was as likely to end up on its roof as it sounds being full throttle and burning NOS. This does make it grip less, either understeering on flat turns or quite often lift off or kerb grinding related oversteer.

Browser support-

• Internet Explorer (works)
• Firefox (works)
• Chrome (needs you to set a flag to un-disable the plugin and will nag about it wanting to re-disable it)
• Internet-explorer-on-windows-ten-wait-it-isnt-internet-explorer-it-is-called-edge-now (doesn't work on mine but I still have a beta of Win10!)
• Safari ... who knows, I assume it doesn't work.

Control methods

Welcome back, I knew you were just going to jump straight into the game! Early prototype remember ... no UI to tell you anything in the game. No it isn't somehow telepathically controlled. No worries if you were trying that, you can play with either a game pad (highly recommended) or on the keyboard like an 80s track & field superstar, but for best results use a game pad! Plug an XBox controller into your compouter if you have one. You may need drivers but who would have thought that would work?! I don't have a steering wheel so that almost certainly won't work.

It has manual gears with auto clutch. I have written an automatic gearbox but it still has some bugs and ties itself in knots when switching from forwards to backwards or vice-versa so it needs to switch from the inputs from brakes to accelerator or going through the gears like cars do in GTA.

Keyboard controls

• Steering: Left & Right arrow keys.
• Accelerator: Up arrow (even if in reverse)
• Brakes: Down arrow (even if going backwards or even sideways ... it won't help in freefall from the top of the loop though!)
• Handbrake: Space
• Gear up: A
• Gear down: Z
• Recover: R

... and that is all the keys. Oh, Nitrous? Press 'Q'!

XBox controller (PC)

Apparently the inputs can be different on an iBin or via the command line, it 'should' work if you dont have 'doze, but I can't say.

• Steering: Left stick horizontal.
• Accelerator: Right stick forwards (even if in reverse)
• Brakes: Right stick backwards (even if going backwards or even sideways ... it won't help in freefall from the top of the loop though!)
• Handbrake: A
• Gear up: B
• Gear down: X
• Recover: Y

... and that is all the keys. Oh, Nitrous? Press 'Left shoulder button'! And, yes I used the same stupid jokes for both.

"Amateurs don't use nitrous oxide. I've seen how you drive. You've a heavy foot!" - Harry

Tips: How to drive / do a loop.

This is a 300BHP car that weighs practically nothing, has a very short wheelbase and very hard suspension ... so carefully is a good place to start. It does have a lot of grip but also a lot of power so will spin the tyres therefore counter-intuitively despite what you see on TV, sometimes less throttle will go faster. Short-shifting will also help the extra 200BHP that comes from the Nitrous system. It will slide at low speed as well as flip over if cornering too fast, so, get a feel for how it drives before going crazy.

It can do the loops without using Nitrous, but it makes it easy. You need to start off at the top of the hill in first or second slowly enough for it to just roll over the edge. When it is going down that slope it is time to floor the accelerator and be ready to shift up a few gears quickly as it will accelerate even faster down hill, of course. If you are going to spray your way through the loop then you only need to hit the NOS at the foot of the first loop, making sure to short-shift it otherwise the rev limiter will spoil your fun. Short-shifting it is also useful if trying without NOS as it will make the wheels less likely to spin if they break contact momentarily. Stop the NOS at the top of the first loop, otherwise it will be bottoming out on the short straight connecting the two loops, probably making it slide sideways. It is a good idea to also lift off while coming down the tail of the loop, but don't use the brakes as this just unsettles the car and won't scrub enough speed to get you back in control if you were going too fast to begin with. The second loop should be roughly the same procedure, except that there is a hump after the loop to launch your car if you are going too fast. Nitrous into it or try to stay in control ;p

General troubleshooting of car handling

• If it isnt going anywhere, press 'Gear up' and try the accelerator. It starts in neutral!
• If it's going slow and screaming like an angry electronic bee. Change up a gear or try less throttle, it doesn't need to be flat to the floor.
• If it's going even slower and sounds like a tractor, try a lower gear.
• If it's rolling over forwards going down a hill, go slower, it has quite a lot of 'wing' on it so it will pitch it into a roll way before the rear tyres are clear of the road. If you want to jump the other cars are better for that.
• If it rolls after going up a ramp, go slower, it will also probably backfire and scream a little as you approach the speed where it will start to roll forwards, if it is starting to do this then lift off a little before the ramp.
• If the nitrous is gone, post it to a friend, ask if they can get around without crashing and hit refresh in your browser to show them who is boss as you have obviously mastered it by now!

Axle Tramp demo level: Loops of DEATH!

A short test track to give a quick demo of how things are going. I've got a variety of corners, ramps, loops and spiral ramps. It takes a bit of a run up, a bit of nitrous and a special tune on the suspension to get it to do two loops at once but it will do it. I used a little of the old Daytona super oval technique while recording this but every time I do it I can do both loops without touching the sides or crashing over the jump after the second loop ... honest!

Please help by purchasing a t-shirt from my Zazzle shop.

Finished modelling the body of the Muscle car, there are still some mechanical parts to do before it is available in-game.

So ... I have been putting off this particular post for a couple of weeks, which turned into a few months. This was because there is nothing really visual to show, and it's pretty hardcore nerdery that will only really make sense to some developers, it's been hard enough trying to get any 'buzz' going without rambling on about something only used by a small proportion of developers and understood by far less. Consider it developer karma for having posted videos of Forza in a previous post, or mostly of 3D rendering eye candy. Ok, so most of the renders have been matt grey radiosity style renders so the eye candy in question here is probably a dull beige near tasteless sweet for oldies.

I first started using Unity about five years ago... it seemed like it was what Macromedia Director should have been like ten to fifteen years ago had 'Macrodobia' not had what seems like a love of self cannibalisation hailing its death and long live the new king- Flash. Up until they decided that too was getting too successful. [No bias and resentment here]. I'd still say Unity is very much like Director, as with a few clicks and dragging & dropping a few 'behaviours' you can put together something half decent in no time. One of the first things I built in Unity was an approximation of what I spent months building in Director for my undergraduate final project, except in unity it took maybe an hour to do and looked way better- like a console game equivalent of my project in a few clicks. Simultaneously very impressive and slightly depressing!

"You almost had me? You never had me - you never had your car... Granny shiftin' not double clutchin' like you should. You're lucky that hundred shot of NOS didn't blow the welds on the intake! You almost had me?" - Dominic Toretto

One of the main goals I had for making a game in Unity was that it would help force me to engage with c# a bit more. I never like being the cut-and-paste-king, found-code-expert or behaviour-guru ... of course there is not much point re-inventing the wheel but I believe that if you couldn't (eventually) write the code you shouldn't be using it. I don't want to be that guy who when the client says that is perfect, but can you just make this blue has to say no because he doesn't know how.

I've been looking at putting the game into Steam Greenlight but it seems they now don't take things without a working demo. I think this is largely due to a rise in 'asset flipping', if you don't know what this is then see this amusing / odd video: The Asset Flip. I'll try not to go off on an anti-Javascript rant about how 'modern developers' are only capable of plugging This.js into That.js but this is kind of the motivation of the work in this post. In more layman terms, it's like being the sort of person that takes their car to the garage for every single mundane thing. So, as this is what I do for a living, in the same way as in the Dom Torreto quote- you need to get your shit together before you go racing, its no good just bolting all the best bits together!

It's all Facebooks fault really ...

I was getting the Facebook-Unity SDK working ... it doesn't take much, it's all basically plug and play. The real trouble is syntactic sugar vs syntactic salt and code smells. Facebook obviously strongly backs Javascript, which I felt added a slight Javascript smell in that it was all in one 'behaviour' in their demos at least, used too many 'callbacks' for my liking and using it seemed to spread the Facebook code wherever it was used.

OK, the real reason- I've been doing Flex for Enterprise and large corporations for a while now and things are done a certain way. Flex is already completely asynchronous, even though Flash is still largely single threaded you don't normally know or particularly care when something is going to give you a result or fault, you just give it a request and as long as you are set up to handle whatever comes back then it's all good. Now that is fine for a single view or very small application but will quickly go wrong once you ramp up the complexity. Here is where we like to increase the complexity by some orders of magnitude in order to gain an odd sense of simplicity. If its not so far abstracted that it still seems to actually do something then it is too concrete, things should be completely decoupled from absolutely everything else. Clean!

Micro-architectures like Parsley and Robot Legs give you dependency injection, commands, decoupled bindings, messaging and probably more that I've neglected to mention, for BBC-like impartiality I should say that there are others but these are the ones I've already mentioned and most importantly can't be concatenated to form a funny sounding name. In the same way that Flex being async gives you an abstracted freedom and confidence in cause & effect; these frameworks free you from the more mundane logistics of exactly how things are wired together. With what can seem like a bit of magic if you want something then you can have that dependency injected, or if you need something to happen then firing off a message to trigger a command ... Command patterns It's a hard habit to crack too.

So, that is 'Monkey Brain', basically a clone of Parsley in C# proof (to me) that I have levelled up my C# to a point where reflection is no problem whatsoever. I think that is it for this post then. An incredibly abstract concept that I didn't really go into in any detail and mostly ranted about plug & play development, but this at least reassures me that I'm not that guy, I'm a different type of guy which is its own problem :D

This isn't the only thing i've been doing since the last post...

So, it has been a while, I feel i should give a summary / table of contents for imminent updates.

• Re-written more correct physics based Engine, Transmission, Nitrous and Suspension.
• Added ABS, TC, Gearbox and Anti-roll bars.
• Broken Gear selection in trying to add full-manual, semi-auto (manual with auto-clutch), full-auto & 'arcade style' automatic.
• Decided that continuing to effectively brute force my knowledge of other ECMA / C like languages into C# would eventually get me into trouble so firgured reflection is probably the most involved thing you can usually do with a language, and creating an inversion-of-control / dependency-injection framework is usually the best way to give yourself that requirement.
• Attempted to port some Apache Flex-ness to Unity UI systems. Read this as tried to port (re-write) the entire Flex framework into Unity.
• Crazy track parts! Curved ramp. Loop, Spiral ramp (think fancy car-park).
• Particle based oil leak / no-grip weapon.
• Started a contract.
• Finished a contract.
• Fixed the wing aero issues when downforce not parallel to gravity when moving perpendicular to the ground. IE- schoolboy error I made assuming the transform's up would automatically be translated into world space but hadn't noticed due to a lack of excessive landscape gradients & crazy track parts.
• Made a simple flight sim to prove to myself that I have got the forces aligned correctly. Got carried away & modelled some parts for a plane and sent them off to be 3D printed.

... and finally ...

Do you remember that picture at the start of the post? Yes the one you have now forgotten about if you read all that blurb? Well previously while trying to roughly block out some models for the next few types of cars I came up against a design problem: the driver! One of the original concepts for the game was to keep the graphics fairly simple funky-retro-looking and to make the cars look like stylised toys. So, a decision needed to be made- how does the driver look in a normal car? Is the driver still a minimalistic pill shape or should it have a body and more modelling to the helmet?

If you are a regular visitor to this dev blog you will know I did have a go at modelling a more realistic racing helmet, and a far less realistic weird monkey-man abomination with a manic grin. I'm still undecided on whether it looked OK, while in the F1 style car it didn't fit as well, naturally seeing as the cockpit (entire car) was modelled around the pill shaped driver. Another far less complicated way to go is to just have opaque windows...

This is still a work in progress ...

This is intended to be an american v8 muscle car and probably demonstrates the intended stylised design of the cars a little better than the formula style car. In the second of the renders (this was actually the first render) the box on the hood is a boxed out intake sitting on top of a supercharged engine. For now the v8 is a simple mod of the v6, changing the angle to 90° and adding the extra pair of cylinders, intake & exhaust ports. As part of modelling this car, I am also modelling the chassis and mechanical parts so the current engine model is for size / fitting only at this point.

Some thoughts for other vehicles-

• Hatch back
• Supercar
• Rally car
• Drift car
• Van
• Police car
• 'Big rig' truck
• Ambulance
• Bus
• Shopping trolley

WAIT! Shopping trolley?!

Yes, why not ...

Hauppauge HD PVR Rocket Test: Forza 5, McLaren MP4-12C

First test, of the Hauppauge HD PVR Rocket. The default settings seem to be terrible ... saturation, brightness & contrast are all way off.

I love the way this McLaren drifts. Ok so it's a lazy 4WD drift but even with my 'subtle' drift technique it drifts very nicely. This wasn't even a particularly good one either.

Hauppauge HD PVR Rocket Test: Forza 5, Ford Coyote

Second test, of the Hauppauge HD PVR Rocket. Tweaked the settings a little from the defaults ... still completely terrible!

I think I've done a reasonable tune on this car- when fully upgraded but with the default tuning it really likes spinning. I've just about tuned that out now so it can be thrown into corners a lot more. It's still a coyote and wants to bite your face off so its like a pet coyote. This was a pretty slow run- the fastest lap was something like 3 secs off my best with this car as I haven't used it in a while.

Hauppauge HD PVR Rocket Test: Forza 5, Datsun 510

Third test, of the Hauppauge HD PVR Rocket. Tweaked the settings quite a lot, looks pretty close to me.

This Datsun 510 is one of my 'best' tuned cars in this version of Forza- basically no suspension or damping with the shocks set to exaggerate body roll and all the wheel alignment to make sure it doesn't go in a straight line ... it's quite a rodeo tune when its engine swapped, stuffed full of turbos and has drag tyres!

Well a face anyway...

I've been trying to model some other cars, except the spaceman caspsule head i put in the F1 car makes it a bit difficult to do anything with a roof. So should I model the insides of the car or not ... if I do then I need a driver model ... they wont all look like weird little monkeys with manic grins ... unless thats the avatar you choose.

The helmet model and proportions of the driver fit with the car. I was going for a Toy Story-esque look to the face ... a little bit Buzz a little bit Woody maybe? The helmet and HANS device look pretty good in the F1 car at least.

I do like the minimal toy look of the existing capsule head though but I can't find a good way to do a driver podium. I've been considering putting the car on one as it it was a scene from Cars.

Having proven that the method of building the track from sections is a good way to go as it allows me to fairly quickly build a track I decided to give a bit more flexibility in the pieces. In the previous iteration there was only one type of collision geometry per visual track section type.

Ground

To allow the track to work in multiple places the ability to use the track on any number of settings the 'ground' collision geometry has 3 options. (Moving front to back in the image above.)

• Track Only: If the track is flat / on the ground plane then it needs no additional colliders.
• Bridge: A Smaller supporting structure to avoid the track looking like it floats in the air. Note supports will be added as additional detail pieces.
• Embankment: Where the underlying terrain is uneaven or to have the track surrounded by <spoilers>water / lava</spoilers>.

Kerbs

• Soft verge: With the addition of different ground types meaning you are no longer so constrained to the track the first problem is getting back on track once you go off.
• Rumble strip: Mostly I wanted a bit more modelling detail in the rumble strips which were just painted kerbs. This doesnt show up too well in the above render but can be seen in the serrated edge of the kerb in the headers between posts.
• Pavement: A nice hard edged kerb to really get in the way. Its hard to do city races without them.
• Bank: Not shown in the render, but a more gradual blending of the kerb to wall for a rally section.

Walls

• None: Life is too short to be constrained to the track ... see earlier spoiler about water / lava ;p
• Short: Basic constraint of a tyre wall or Armco barrier.
• Mid: Short barrier but with a wire fence? This is the collision geometry, the actual 'skin' could be anything.
• Full (+ ceiling): City streets, tunnels, ramps and/or crazy stunt sections like loops may need full collision.

An early render of a complete track built from sections eg- straight, fast left, sharp right etc. From initial game testing on this level it was immediately obvious that the straight ramp sections were a problem on a number of grounds, but mostly highlighting why real cars don't have such short wheel bases and very rarely almost completely square ones!

On trying to get onto the ramp the front of the car would bottom out, especially if approaching at an angle. If it didn't spin you around completely just the fact its now broken traction meant you had a hard time getting up the slope with it either wheel spinning down to the bottom or flipping over backwards with softer suspension. This version also didn't have anti-roll bars or frics. Yes that may have been deemed illegal in F1 now but having a control on front-to-rear movement helps stop the car rolling all the time.

Getting up the ramp was in comparison fairly trivial to getting down it! Again most notable was approaching the ramp at an angle where the wheel would leave the track and spin the car. If you did go straight down though, the front would catch at the bottom and roll you over forwards. These are of course largely problems with the set up of the car, especially given that I hadn't done any work on the stability but having enough ground clearance and soft enough suspension to get down the slope at the time meant you would roll over backwards trying to get up the ramp in the first place. So, new ramps and stability control!

As previously mentioned I want the graphic style of the game to be somewhat stylised and fairly toy like while (hopefully) not looking entirely childish by remaining suitably detailed. I wanted to keep it quite low poly too so that the cars look quite blocky- eg the exhausts are octangonal but do have some modelling along their length. The brake ducts are probably more detailed than they should be but I wanted of offset how retro the engine looks.

There will be various fings available eventually but this is an attempt at a modern-ish looking F1 front wing & brake ducts.

A 120° NA v6 with DOHC. I'm actually a little unsure about the fuel system- I think that would be a pair of fuel rails with plate injectors rather than carbs. To try to get the racing as close as possible the engine will be largely cosmetic anyway acting as black box crate engines. For more details on v6 engines see this video by Engineering Explained.

Spot the gear box ;p I think I've just invented a new power train layout ... not quite Mid-Rear or Rear-Rear ... Top-Rear? Obviously with the proportions of the car I can't get the gear box in the correct place. I Should be able to fit a v8 in the car though. It's almost as amazing drive system as the Koenigsegg Direct Drive in the Regera as that doesn't have a gear box either!

"Axle Tramp: A form of wheel hop which is usually found in live rear axle cars. It occurs when sudden torque loads on the suspension cause the driven wheels to shake violently by slightly rotating the wheels and then springing back." - Automotive Dictionary

"Tire shake: A term in drag racing, when the engine is putting out more horsepower than the drive axle can handle, causing the rear tires to shake violently. This results in a loss of speed, and can also result in loss of steering, and occasionally, lead to on track accidents. Refer to in other disciplines sometimes as axle tramp." - Wikipedia

First off to clarify / explain the name. Although it sounds like this could be an entirely different type of site- one of the NSFW variety (as if games are either), 'axle tramp' is an automotive / motorsport term for a type of vehicle behaviour resulting from a combination of excess power and/or poor automotive design. Not to be confused with the intentional wheel hopping in cars with hydraulic suspension, axle tramp -also called 'bouncing' or 'tire shake' is an increasingly violent vibration in the suspension and drive train that can be very dangerous.

It should be fairly obvious that that the tyres are the only part of a vehicle in contact with the ground and are responsible for all acceleration, braking and steering so if the tyre is not in contact with the ground then it isn't going to be doing what they are meant to and things are probably going to go very badly very quickly!

So why use this name? Why call a game after dangerous handling? Is that the aim of the game? No. The point is that any vehicle at some point, under some circumstances will display some unwanted or unexpected behaviour whether that is oversteer, understeer, a lurching crashy ride etc. I should probably have but some sort of disclaimer everywhere by now to obey traffic laws and drive safely as there is no reset button in real life. It is when a vehicle is pushed to the limit of it's performance envelope on a test track, not even a racing circuit that the worst of these behaviours will be found. Axle tramp sounds better to me at least than 'push the envelope' ... that sounds like a bad ass bureaucratic desk simulator ... wheres the copyright key?

A shiny new website to start us off!

Having said that, I'm just doing the absolute minimum on it so I can spend more time making the game :D Seeing as I've put off making a website for so long (my own company website is a great example of that) there will probably be a load of posts from today. Might look weird, thats all.

I didn't want to use an off the shelf blog or something like that. I've built plenty of my own content management systems in the past but to avoid me spending ages faffing about with getting one of those just as I want it for this i'm going to try to keep it simple- eg if I'm going to write a few hundred thousand lines of code then its best if its in the game! ;p

If it isn't obvious, this should be / will be a dev blog type thing. Most recent at the top etc ... I don't know about this stuff, I usually make more software orientated things.

- edit: I've rebuilt most of the HTML and also taken the dates off seeing as I didn't upload the site when I first made it. Problem solved :D