mdgunn

Submission for the "Cirque des Jeux" Game Tournament. Got some crashes going on as I do my usual thing of leaving it till the last minute to get a complex level together and ditching pretty much all the interesting work done up till then. Splines are roughly assembled so path is a bit wacky right now. Single level on rails shooting gallery thing.

When I've seem this it is usually because the model has not been triangulated before import. In Blender stick triangulate in your stack and it will get applied on export - unless you modify default export options. Whilst I think you don't HAVE to do this it can help as otherwise the engine has to decide which way to triangulate a quad which I think can lead to problems sometimes if it decides differently to your 3D authoring program. Not really an expert but if you're having problems this may help. If still problems maybe start a new thread on it.

Try creating new project, make sure you select the VR one. You should be seeing it in the list if you are on the newest version of leadwerks (4.5).The template name has VR in the name so if you don't see one you may be on old version of Leadwerks. You can then copy any files that are complained about in your actual existing project, or move them the other way. Make sure the VR template runs first though with zero modifications so you can see it working with your setup. I have used the VR template and at the time I think the sky seemed to be looking for a skymap that was not present (may be since fixed) but apart from that worked fine. I could teleport around fine and seeing the teleport model in use in game.

Glad to see the tournaments still in place. Really hope to have a decent finished article next time.

I think you are coming across the bug I mentioned here: where the tutorial is not fully in line with the code. The actor class is there and should work out OK but it's unfortunate that the tutorial and code don't fully align. For a tutorial this is an unfortunate (perhaps even unacceptable?) oversight.

OK maybe I'll check that out. Thanks.

The map is very nice and there is a lot of nice content but I'm getting FPS score a bit lower than I might have hoped to see. I also expected that the map might not be as highly optimised as a final game level so it was partly expected. I'm running the map on an RX480 and getting a bit over 30fps typically. I would hope to get around 60fps at 1080p on a such a scene if well optimised (I mean in a typical published game that might be on a different engine with different ways to optimise). Don't think there are any LODS in effect yet many largish building remain visible at quite long range. Don't think LUA should be much of a factor given there is (AFAIK) almost no scripts set up (given it starts with a spectator camera). I expect trees probably eat a lot and close vegetation shadows but for such as scene they probably can't be changed much to gain much speed and still carry the quality of the scene. Anyone got any opinions on what they would do to get this map to run at nearer 60fps on mid-range card such as a RX480? I expect it might be a lot of work so just curious as to opinions rather than something I plan to attempt. As an official release it provides a good reference for examining performance and ways of tackling it.

Any chance of the MAX files being replaced with a more generic format like FBX or OBJ? This seems an unusual choice. Maybe time was a factor?

I'm using Trinus for PSVR with PSMoveService. I think it pretends to be Vive. You see the controllers as Move controllers, as in screenshot above, which is nice. Haven't tested if all the buttons map out correctly but the teleport certainly works OK.

Small note. Skymap missing in from the VR template in my setup (possible I'm on pre release beta maybe). Got VR working with PS VR and move controllers. My PSVR setup terrible, not properly configured, but can confirm teleporter working OK with move controllers.

Another milestone for v4. Good work Josh!

Turns out the code I found have for Temples and Towers was massively behind where I got up to with it and didn't include any of the map I did above. It's probably backed up onto some hard drive somewhere but god knows where. Would be starting pretty much from scratch. Shame.

The VR stuff is shaping up nicely. I have a game idea I worked on some time ago I might return to for checking out VR.... Obviously the template has a clear VR focus and needs to stay on track, but I'd like to mention that I think it would be really good if we had a VR template (or engine support) to show a mechanism of downgrading to run in a NON-VR state as well (wasd and mouse, or controller). Dishonoured uses a teleport feature just fine. and you pick stuff up. Should we really make our games ONLY accessible via VR? Are they really un-workable or immersion braking on a WASD + mouse or controller setup? Maybe the answers for some people is yes and yes. But if some think no why exclude them? A nice example of degrading from VR, or extending to VR (whichever way you want to look at it) is a useful feature for a template (or engine) to handle. Yeah we could all code our own solutions but I'm pretty sure a good few users would appreciate a solid template or detailed example. Keep up the good work. Looking forward to how this develops.

Never got round to mini maps yet so out of interest is the mechanism presented here the most obvious choice to go with or what are 1 or 2 other ways that might be considered? I imagine it depends on your game of course (as always) so I suppose I mean in most cases what methods might have been used?

Thanks for the article and demo code. Much appreciated.

Good work. Slightly confused by the ending. Does this mean the framework will remain an internal project only for your own games? Whichever way you want to go, OK with me. Thanks for the time you took to share some of your work. Have a good Christmas.

Had a quick google and there seems to be a degree of confusion, among Blender users themselves, and in the thread I read some users cite Blenders manual and others informing that the manual is/was incorrect. https://blender.stackexchange.com/questions/55975/what-is-the-role-of-auto-smooth My conclusion from the thread was that auto-smooth is pretty much doing the same thing as split mesh modifier (including respecting sharp marked edges). Someone said it doesn't add (extra/duplicate?) vertices, though some said game engines/gpus in the end usually do. So why 2 similar things, auto-smooth, and split mesh modifier Usually this is because modifiers sometimes add more control, or by being a modifier you can make their operation occur at a particular place in the modifier stack. Open to further corrections.

I still get a bit confused about the different options here but I think you may want to look at auto-smooth in blender. I think having this on means that resulting models have edges SPLIT (at a threshold - so avoiding individually marking this) so that they must be represented as truly flat by other programs. Auto-smooth values of 30 degrees usually work for most things but depending on the model you may need higher or lower. You can apply the same thing locally to any edge. I think this is 'Sharpen Edge' command. I always get this confused with creasing too but I think making stuff to be considered sharp is what will work when exported outside blender. I could be totally wrong so anyone feel free to correct me.

I'm trying to practice modelling now and then so I had a go at a similar couch. Your's is more detailed. Mine turned out at 417 faces 852 tris. Wouldn't call myself an expert but image attached if any use. I'm using a number of simple unconnected intersecting meshes (same single blender edit object) with hidden faces removed. Not trying to model out a single box modelled block. If you know what I mean? The cushions of course take the most geometry. I like the idea of trying to re-use many simple 'materials' in a larger shared texture which I think would boost performance significantly and maybe offset to a degree medium-poly modelling. I wouldn't say I've settled on a particular method myself though. Currently I've found it quite satisfying and quick to take things to a low or mid level of geometry with multiple simple materials (which you can then drag and drop to switch in leadwerks), rather than regenerating maps all the time. Thanks for the vid @ScrotieFlapWack . I follow that person on YouTube already but I'm always forgetting to go back through his stuff.

I've never done that. I'd need to look into it. Maybe if I get round to doing a bit more I'll take a look. Thanks for the interest.

Oh..forgot to say...glad to see the update and hear you're still moving forward

There are some free or good assets and tutorials on Fusion 360 GumRoad E.g. this guy https://gumroad.com/mrdavids1# Think he and others are also on CubeBrush.

Thanks for the update. I've tried to do a driving game a couple of times and found it a bit frustrating, so this sounds like a significant step forward. Driving game or VR game for xmas competition....hmmmm......driving AND VR maybe?

I'd be interested. Can't promise I'll have the time to provide much useful feedback but I can try. Personally I use PlayStation VR. You might want some non oculus/vive feebback? Can be a bit janky as I don't have everything set up as well as you can (in principle you can do passable/decent sitting or even roomscale VR with head tracking and controller tracking). I have previously had LeadWerks working with the PSVR and it worked fine (at least to get a simple static scene going). Cool to see your own world in VR. Am certainly interested in getting it a bit more functional and simple to use. I'd probably be considering a simple VR (or VR supported) game for xmas competition (depending on some further success and a idea to engage me).

Thanks for the round up. I didn't know some of these had been submitted ( I must be looking in the wrong place!).