'Jumper'. Game inspired by Portal. Windows and Linux Demo Now Available!

Original Post

Hey guys.

So to take a little break from my main game, I’ve decided to start on a simple tech game.
My idea and goal for this game comes from the first Portal game. Simply to test 1 single mechanic, nothing else and to do that one mechanic perfectly. Then create puzzles using that mechanic. In very clean crisp levels like you find in Portal I. I love that!

The mechanic: Jump pads that give the player a jump boost

Can you give me feedback about how the player jumps on the pads? The aim is to get it to feel good quality and as a player would expect.
Also, I want to work on the level design. I think the modular approach I’ve taken with it so far will help keep a clean, tidy and uniform look which I’m going for. I want to work on the current modular pieces and add some more pieces to add a bit of variety. Again, I want to go for same kind of look and feel of Portal 1 maps. I love that clean look.

One of my motives for this, is that I’ve always felt that the quality of a game is so important, more so than graphics etc, so I’m doing this as a quality exercise to improve my future games.

Also make sure only layer 1 and 4 are visible to play. Start from the main menu scene.

For those viewing the thread for the first time, I’d like to point out that although this started off as a Blender game engine project, I have since migrated it to UE4. So although the game is no longer really a Blender game engine project, I’m keeping the thread active for continuation purposes. However, I’m still using Blender and your other typical open source game creation tools for models/art/textures etc.



Nearing end of production. Making a final push to create a decent amount of stages, finish creating remaining needed assets, and polishing over everything. It will then be a case of doing some final optimizations and then rebuilding all the lighting etc to production level before finally packaging. So around a month or two I hope!


Linux now supported! Demo available to download along side windows version!


Windows only at this moment. Should run on both 32/64 bit.

If you are going to play the demo, please can you leave me some technical info afterwards:

  • Exact Windows version
  • RAM
  • CPU - Type and Power in GHZ
  • Graphics card
  • Other standard PC specs
  • Monitor size

The game has been package only at preview quality, so the lighting isn’t perfect at the moment and so this demo is not indicative of the final level of quality. There are some stages that still have elements blocked out with simple cubes awaiting to be replaced by new assets.

You may also notice some debug messages when starting a new stage. I’ve done this intentionally to make sure the auto save feature is working correctly. While on the subject, could you also make sure that the continue button on the main menu is working for you as you would expect it to. If not, how would you like it to behave differently?

If you have a controller, please try it out. I’ve tested XBox 360 controller which works fine. A = Jump. Start button will bring up menu. All other controls are your typical FPS movement controls. Menus not supporting controllers yet. Only the actual game itself.

Demo download and product page: http://cypher2012.itch.io/jumper

Preview Video and Images:


Update demo

Extra level
More modular pieces
Mechanics tweaked



Updated demo
Improve camera focal settings.
Remodelled jump pads
Animated opening doors

Could somebody tell me if I have set up the occlude boxes correctly?

Added lighting to the jump pads

Hi, I tried it out, and it was quite a nice demo.

If you want people to download and give feedback I’d recommend keeping the file size quite small. Also you could try packing the textures rather than having them in the same folder, most of them were not applied properly.

Things I liked:
Nice smooth movement, simple and clean design, nice gameplay concept.

Things that could do with work:
When i jump my head (camera) goes through the ceiling, so I can see outside the level just for an instant.
Moving was a little bit too quick with the default walking speed, it made lining up jumps a little frustrating.

I think this could be the basis for a fun game, but it needs something a little more going on in terms of gameplay. Maybe some puzzles, or a shooting mechanic where you have to shoot targets while jumping to open doors etc…

Try running your game in wire-frame mode, to test Occlusion.
You will be able to see through the walls, and see if objects are being clipped out from occlusion.

I gave your download the college try, and gave up after the 4th failed download attempt.

Thanks for the feedback. I agree with the points made. I might make the jump pads a little bigger too, as well as slowing down the player movement to make that a bit easier.

I need to address the issue of being able to see through the map. I think its because the players ‘head’ doesn’t have any collision. When I’m home I’ll take a look at that. Ill play around with maybe moving the head in front of the body instead of above so you cant see through the ceiling. If not, I’ll make the ceilings higher were they need to be.

I’ll also try to make subsequent uploads of the demo a bit more efficient to make it a bit easier for people to download and try, I was just being lazy :slight_smile:

Also, it is my intention to make some puzzles out of it. The sections that are there at the moment, are intended to be your kind of introductory stages and the stages will progressively get harder. I may go and add another mechanic, but the whole point of me doing this demo was to get some practice implementing one single mechanic but as best quality as I can and then creating some puzzles with it. Never the less, it’s coming along nicely. I’m trying to be a perfectionist and making lots of tiny tweaks to iron even the slightest things out. I’m hoping to end up with a demo that is very clean and smooth to play.

@Nines, I took this method, but I didn’t really see anything happen. So I added a whole load high res spheres to see what happens and still nothing. I could still see them.

Thanks again for your feedback guys

That tells me that your occlusion is not working.

I was able to download your game and look at it. You simply have your occlusion boxes set up incorrectly. It’s an easy fix though!

An occlusion object only occludes things which are on the opposite side of the viewport. If the viewport is inside the occluder, it won’t occlude anything.
Any objects that are inside the occluder also wont get occluded.

Rather than encasing your rooms with occluder boxes, what you want to do is create a sort of ‘cage’ of occluders around the outside of your world.

Here is a side-view of the first room of your level. The red boxes are proposed placement of occluder objects. I drew a little stick-man with arrows pointing in various angles to show how the occluders block your view. Green check marks show objects that should be drawn, while red Xs show objects which will most likely be occluded.

Also you will need to split the level in to sections. I seem to remember it being one big chunk… occlusion only works with making individual objects invisible, not parts of a big object.

Anyway, keep rockin’ it’s looking good!^^

@nines and @smoking_mirror thanks for the info guys. Looks like I have some further reading to do. Although, as the levels aren’t going to be that highly detailed anyway, do you think it is worth the effort?

@smoking_mirror thats right about it being one chunk. Although I’m taking a modular approach to designing the sections, I’ve merged them together once done.


Help prompts + pop up help menu (H key to bring up help menu)
Player can boost forward
Pad lights can be toggled on and off if your computer can’t handle it with L key
Map extended
Played about with the main lighting. Im not sure if I’m happy with it just yet
Adjusted player move speed
Added walk sounds
Improved player collision related to bug seeing through ceilings

NOTE: I’ve reduced the size/quality of textures and packed them this time so they should all show properly now :slight_smile:


JumpingGame.zip (5.42 MB)

Your textures are not packed with the blend file. Use File->External Data->Automatically Pack into .blend
Afterward, the only file you should have to give us is the .blend file itself!

You’ve packed the .blend1 file as well as some screenshots. Those don’t need to be there.
There are missing textures. It is trying to look in a folder on my desktop which does not exist. Packing your files will fix this.

Looking at the meshes for your level. You should have each ‘panel’ of wall seperated out into its own object. If you do this and set up your occluder objects how I suggested in the post above, your game should run much smoother, and allow you to add many more detail objects to your world.
The only face each panel which will be visible to the player is the one facing inward, so don’t worry about modeling the back-side of the walls. All that is doing is doubling your poly count.

Hey nines,

You must have tried one of the old uploads. I’ve deleted the old ones. Try the only one that is available at the moment.

So having lots of individual pieces, in this instance, will actually be better? What I’ve been doing is combining all the objects together once I’ve finished a section, and then joining them all together, and then removing all doubles. This really reduces the vertex count. Youre right about the outside facing walls, I should delete those. Just laziness :slight_smile:

Thanks for the feedback :slight_smile:

My bad! I failed to see the update in the original post.

As for breaking up your level into smaller objects: Think of it this way, you don’t want to worry about the poly/vert count of your entire scene so much as you want to worry about the poly/vert count of what is currently being drawn on the screen.
If the player can’t see a particular part of the scene (it’s behind a wall, or out of the camera’s field of view) there is no need for your computer to put resources toward drawing it on the screen.
This is the whole idea behind occlusion; to only draw what you can see, and ‘forget’ about everything else.

Occluder objects only work per-object. That means, the only time an object is going to be occluded is if the whole object is hidden behind the occluder.
If your whole room is made of one object, there is no way any occluder is going to entirely hide that object, unless it’s between you and another room.
If your occluder box is surrounding the room, it will never be ‘behind’ the occluder box. If the player’s view is inside an occluder, nothing will ever be ‘behind’ it.

If your room is broken up into several smaller objects, than there’s a good chance that many of those objects will be out of the player’s current view, and thus candidate for getting occluded. This could mean more verts in more ‘visible’ areas of your level, but over-all it should be more efficient. An extra hundred or two verts on your screen should not have much impact on framerate. Not nearly as much as having an extra thousand or two verts being drawn in places that aren’t visible.

Ah, I see. Thank you for the information :slight_smile:

but does any of this happen automatically? Whether that be on the code level, or the users GPU. So, if you don’t use any explicit occlusion whatsoever, will your computer do some occlusion anyway?

What do you think of the boost feature? I was thinking of how to make some challenging steps. So I made that little hall section, where you have to combine the jump pad and boost to get across.

I’m pretty happy with the mechanics of the game. I need to make some nice puzzles/levels now with the mechanics. I want to work on the sounds and environment. Although the game is basic, I still want it to feel nice and polished by time I’m done. I’ll probably bake the lighting in the levels. Or at least the shadows. My GPU can handle all those lights, but I presume they drag a slow computer right down, but they look so nice I think.

There is a lot of stuff out there about puzzle or mechanic tutorials. Extra Credits I think has something about it. And Half-Life episode 1 has developer commentary about it.

I can’t really remember, but I believe it went something like this in the Half-Life:

  • Show the player in safe environment how the new mechanic works (Half-Life had the car that covered the ant lion burrows)
  • Reinforce the mechanic in a similar way that isn’t super dangerous
  • Now place the mechanic in the regular gameplay (like where people are now shooting at you)
  • Ramp it up

Portal test chambers forces you to know the new mechanic before you can move on. I noticed you have a ramp by the first jump pad.
I’d say remove it.

Found it:

Hey cam.dudes,

Thank you for the input! You seem to get my inspirations! :slight_smile:

I enjoyed the video! I agree with the ramp, it’s a bit pointless it being there. Better to force them to learn :slight_smile:

I downloaded and played some Team Fortress for some more inspiration! I also love their design!

What do you think of level design so far?