[Progress]

Terms such as those used to described patterns are just handy for communicating. That's all they are, descriptions, not implementations. It's a surprisingly common misconception among developers (where I work anyway).

[Progress]

Command pattern, for undo functionality. Needs a project to be designed that way from the outset. But if anyone's creating anything with multi-level undo, look up command patterns for inspiration.

[Leadwerks 3 begins closed beta test]

Depends what kind of testing is being done. My day job is development/testing of medical software, we pretty much try to ensure every release goes out with zero defects. I can write you a test plan and some test specs for Leadwerks3D that you can expand on, then you'll have some consistent release testing. Gawwwd knows Leadwerks 2.x could have done with that.

 

Wouldn't have time to test anything myself but Mac is a capable guy as are others but Mac

[Terrain Test]

It's more about the technique than the quality of art though. More tools for those that can make best use of 'em.

[Level Construction Set Part 3: An Instancing Issue]

We also have the opposite problem.

 

Models which have many similar child objects but not instanced. Cockpit interiors have 30+ toggle switches, 30+ rotary controls and cockpits can be duplicated.

[Terrain Test]

Just not added or clear in the scene. But models do cast shadows. In addition you have the advantage of using baked light-maps and shadow maps if you need them. Baked normal maps can also be used for extra detail.

[Level Construction Set Part 3: An Instancing Issue]

I'm sat here in my lunch hour reading this nodding, "yes, yes, yes. Totally ugly yes, spot on."

 

I like your workaround so much I'm going to use it in my next bit of code and owe you a virtual pizza. I'll send you a real one when I get the chance.

 

Too bad there isn't an easy way to instance the materials. One of the problems we have for squads of armor is how to deal with vehicle damage and animating tank tracks.

 

Tank tracks we managed to do by having a material for the tracks that scrolls the UV offset of the track based on the object colour (Maclebee did the leg-work on that). Colour can be different for instanced entities.

 

But since players can fly (up to 4) hero ships in different skins, duplicating materials is expensive. Since a hero ship has multiple 4096x4096 textures having duplicates should be avoided.

 

One of the solutions I'm consider is a damage texture that is blended in splats according to RGB model colour. (Haven't the faintest idea how to do it yet.)

 

This is consistent with how we already apply damage to other objects where intensity relates to damage%. (Destroyed vehicles are left with a dark smokey/burned appearance).

[Terrain Test]

Just as a matter of idle interest is that exported as a regular triangle mesh or irregular mesh?

 

That's pretty much how you do it for mobile apps. For a while there was no native terrain object for a popular iPhone engine. Exporting as an irregular triangle mesh worked really well, the downside was the very limited shader support.

 

This is the same process but additionally makes use of the coverage maps (masks used for splats). I have a UDK material for using exported terrain meshes in a similar way, so you get the worst of both worlds

 

There's a lot to be said for using the same tool to export terrain as a mesh, a split-mesh or height-map for multiple engines. I've tested the DAE export in Leadwerks before. Only issue one has is making sure patches fit into 64k buffers.

 

It's more than a workable means to create terrain, it's reliably used in games for other engines already. Leadwerks is easier than UDK as it's not so constrained with regard to format (these terrain layer materials can be awkward to set-up).

 

When it comes to 'uber mega streaming' (which I utterly like the sound of), creating massive terrains in the order of 250,000 sqr km just as easily as 5km sqr. with the tools we're now using. Exporting them as a single data files for continuous streaming or slicing them up into nice chunks (for paged streaming or non-streaming) is just a different mouse-click away.

 

I would have included a chapter on Leadwerks in my upcoming book on this subject however there simply was not enough space

[Visual Studio]

Non of the devs I speak to are interested in VS2012 for a variety of technical reasons. XP support is vital for us though.

[Final touches on the script editor]

All looking quite interesting. I've been working a lot the past two weeks with Unity and UDK editors doing the same things. When using Leadwerks you appreciate the simpler and more open layouts without having tons of stuff in your face but miss basic features like this. Looking forward to having the best of both worlds.

 

I have a question about the scripting. Are app level events going to be easily available? So we can script map loading...well pretty much everything from start to finish?

[Level Management: A Bucket 'O' Walls]

Ah, domestic violence game would go down great with my kids here. They have to settle for blowing each other up in Minecraft. Oh the howls of laughter and joy.

 

@Ken, as for CombatHelo, I return to that in September when I've got my book out of the way. As for realism, I can take it or leave it, depends entirely on the game and what it's trying to do.There'a Air Traffic control sims and there's those little air-plane lander games that you swipe with your finger.

 

The occlusion issue I totally forgot about, "layers" in other engines you're not supposed to name get around this. Your method should work fine in this respect.

 

Macklebee had a kind of x-ray shader thing going on.

[Level Management: A Bucket 'O' Walls]

Thanks Ken that was a nice walk-through.

 

I remember first time I saw fading walls in Die Hard, a shooting game on the old Playstation which had 3 game types bludgeoned together. One of them was 3rd person.

 

It's possible without too much effort to do fading walls at the shader level but I like your approach as it has character and fits with the lovely retro music you've got going on. You could even combine the two.

 

But if anyone wanted to play with a shader to do it, off the top of my head...

 

The rotation angle of the model subtracted from the camera angle in the vertex shader will give you the parameter needed to determine how much the wall can be faded. This is passed to the frag shader which blends the rgba output accordingly.

 

Alternatively use the vert colour to set the alpha (and set the colour of the model in a LUA script based on it's rotation). Kind of messy but affords a level of control (since you can set keys to start/stop the fade).

 

Just some thoughts. The level manager looks great too. It will reap rewards in terms of control over performance. That's something you can't do with large levels containing big models. I've heard people say GPUs are more efficient than culling routines but in the real world it doesn't work out like that.

 

Good work, I wish I had half the energy you have

[Particle Update: Video Blog Style!]

You are spoiling us with these new toys.

[Hang Out with Leadwerks]

I recorded it but it had no audio which sucks.

 

Topics covered were:

 

Cool stuff being done by the Leadwerks community recently; Shadmars Ocean shader, Josh was clearly stoked (and rightly so) about Hookwink on Steam, some thoughts on the LCP and focus on gameplay.

 

Then a question on Leadwerks3D and Josh talked about focus on gameplay features (it will have tutorials on that).

 

And I can't remember much else except a dog barking and trying to mute it but not being able to find the right one.

 

I hope we can do it again sometime, I think we can share desktops and running apps so a show and tell should be possible.

[One of the Happiest Days of My Life]

And amazingly this debug out looks slightly more elegant than iTunes.

[H.A.A.R.P An Extremists Approach To Data Driven Design]

I say something in Unity for like $100 bucks a pop but I think it's really smart. Even more so as it's in Leadwerks.

 

Why is it called HAARP out of curiosity?

[Let's Make a Game - Procedual Content Creation (part 03)]

Nice of you to say so Roland. Of course I only use the word cell as reference to an area, a virtual space of n size.

 

If you make the patches small and room height to something like 8 patches then and add vertex noise you get some very lumpy cave like walls, there's plenty of room to experiment.

[Leadwerk Community Project 1. Setting Up]

Long boring article? What's wrong with those? I'm very good at them.

[Bars, Girls, and...... Videogames?]

I see you have to plug-in to recharge. Is there something you're not telling us?

 

He's a CYLON!

[Let's Make a Game - Procedual Content Creation (part 01)]

It would be going up this weekend, got some booky stuff to get out of the way before tomorrow.

 

I'm having some code issues with getting the corridor geometry joining with room walls. There's an interesting TSurface.Slice(...) command that's not exposed in LUA which I'm guessing is a helper function in 3D World Studio. That might have been interesting to play with.

 

Put some AI robots in the rooms to shoot at you, add some laser guns and power ups then you're good to go.

[Totally Tubular]

I feel a 6DOF tunnel shooter coming on. Anyone else?

[Freaky Friday]

My week not so productive, between debugging helicopter physics and implementing mob movement I've been browsing the gaming news and press. The Utah teapot video was interesting, I spent nearly 5 mins trying to come up with a cleaver sounding word for a 16x16x16 grid and best I could do was "sextusdecicube".

 

Rumor mill has it that Playstation 4 and XBox Next will both feature what I dubb 'FU technology', a feature to prevent you from playing old/pre-used games. Consoles are already less relevant.

 

Patents being filed for even more exotic controllers and methods of interaction. It appears there's a huge room of monkeys coming up with ideas and they are being flung at the patent office in the hope that some of them stick. Is it a Kinect? A tablet? A console? Do I stick it up my nose? Never mind creating content, that's what people want.

 

The UK has the biggest internet based economy as a percentage than anywhere else.

 

At the weekend the UKs National Computing Museum hosted the 30th birthday of the BBC Micro, brining back together the players who created it and then designed the ARM processor now used in nearly every smartphone. The Raspberry Pi was demoed running a BBC emulator adding to the list of working emulators for classic computers. Not bad for a such tiny device. They have been delayed yet again due to radiation emission testing (the HDMI port was being over-driven and causing excessive RFI)

 

And now I've taken to writing long comments to kill 15 mins instead of getting on with some group steering code.

 

The next comment will be far more interesting with less waffle I'm sure

[Another Design Puzzle]

We spent a lot of time refining simple compound structures which I think could have been built VERY quickly in this manner, they are simple walls and buildings joined together in various combinations. What's the term for that, greeble or something? We sacrificed the ability to blow down walls and inflict local structural damage for performance. I have no regrets about that.

 

I totally get it now and see a use, indeed would have had a use for it but just didn't know it.

 

As for editing of objects, I think (taking some experience from watching my kids play LittleBigPlanet) you can get away with an awful lot by drawing an outline with various shaped brushes, then extruding it along an axis. And doing the same thing with a different shape and joining the two parts together. Repeat.

 

It's easy enough to do with a joypad and child's play.

[Another Design Puzzle]

Ahhh, I see, that makes sense. I remember seeing early QUAKE level editors which it reminded me of. I didn't realise there was still a demand for it. So it's good for creating rapid prototypes of level geometry.

 

OK I can appreciate that. Basic canyons decorated with Pure3D rocks come to mind as well as implementing colliders which are a pain. I learn something everyday Thanks for all the replies.

 

< gets it now

[Another Design Puzzle]

Can I ask a silly question, who is this for? I might be missing something (quite probable in my deteriorating age).

 

I can't imagine if you were making a game that you would be creating assets in anything other than some dedicated package to get decent results. This is fine for throw away programmer art I guess.

 

But looking at it from a practical "I want to do this in my game..." I want to punch a hole in some object (a logical operation) or morph it in some way.

 

Beyond that I find it hard to see an artist wanting to do much with it. Gluing multiple objects together (combining meshs) is something an artist and programmer might want to do at run-time. That has a ton of uses. Blowing a hole in the side of a mesh is another (breakable walls with specified extent).

 

I just come at it from the point of view of "how does this enable me to create a game." And the artist I work with use Max to create content. If I want houses there's a ton of nice ones from Dexsoft or Arteria etc. that look professional for minimal cost.

 

Again apologies if I've missed the point.