next »

[mokesmoe]

Also you can walk up slopes into the ceiling, then jump through the ceiling.

[eigenbom]

(Also PMed directly to some people)

If anyone had intel gfx related issues, I'd greatly appreciate if you could test this latest version on your machines.

Update: moonman 0.11b should be quicker, as it insets a render window inside the main window.

(Older: moonman 0.11)

Some things I'd like to know:
- Does it run now?
- If not, what happens?
- Does it run at 60fps?
- Try resizing the window to smaller than 512x512 and then smaller than 256x256 to resize the internal textures, this may speed up the game or show some bugs.

Thanks a lot!
Ben

P.S. There's now an option in settings.cfg called use_power_of_two_textures which is hopefully a fallback for gfx chips like yours..

[jmcmorris]

Finally took some time to play around with mm. Good stuff Ben! A lot of it felt good. Here are some thoughts I had while playing.

• TIG!
• Building blocks underneath you while jumping up makes it possible to fall through them
• No performance issues but I recently upgraded my computer.
• Holding 'Q' is fun!
• The game crashed when I exited. It did manage to save. Don't have more information than that.

Keep up the good work. I'd definitely be interested to hear what features you are planning to add in the near future. :D

[eigenbom]

Thx for taking a look j-mc!

[eigenbom]

Woo 40000 views of this thread, another numberific milestone! Thanks to all those who tested the 0.11b version and glad to hear that it now runs on low-end machines, albeit a bit jerkily. Also, my moonman facebook page now has 30 likes, woo! More updates later this week.

[Thecoolestnerdguy]

Woo 40000 views of this thread, another numberific milestone! Thanks to all those who tested the 0.11b version and glad to hear that it now runs on low-end machines, albeit a bit jerkily. Also, my moonman facebook page now has 30 likes, woo! More updates later this week.

Congratulations! At the moment that I write that, you're with 40119 views and 31 

[eigenbom]

Yay Update!
Fixed bug that caused the game to crash on exit after spawning 2+ cats.
Reduced mining distance.
No longer can place blocks on self.
Provided power-of-two-texture fix for Intel GMA machines.
Updated website a little.

I decided to start from scratch on the liquid model, some experimental results from a basic ca model follow. They differ in parameters. The brightness of the cell indicates the pressure, so right now it acts more like a heavy gas than a liquid. Mass is still conserved.

[Geeze]

Correct me if I'm wrong but liquid shouldn't stay on slopes/edges/whatever

[eigenbom]

Yeh you're right

[GhostBomb]

I decided to start from scratch on the liquid model, some experimental results from a basic ca model follow. They differ in parameters. The brightness of the cell indicates the pressure, so right now it acts more like a heavy gas than a liquid. Mass is still conserved.

Is there any particular reason why you made it so it could condense like that?  It seems like it would be easier to write code for more realistic liquid that stays the same volume so it seems like a weird thing to implement unless you had a reason.

[eigenbom]

Well for local models of liquid making it the fluid a bit compressible apparently helps with replicating global behaviour. E.g., connecting two water tanks by a pipe and having their levels be equal. But I'm happy to take any advice you've got about this problem?

[rek]

I think the compression/pressure thing is a great idea. It'll obviate some oddness, such as water stacking. It also sells the illusion of depth – the water is flowing in the background, so to speak.

[jmcmorris]

Cool. Good luck with the CA. I'm sure you've done your research but I thought I would share some links I found useful while I was working on my implementation of it.

http://www.drdobbs.com/jvm/an-algorithm-for-compressing-space-and-t/184406478
http://www.gamasutra.com/view/feature/134736/an_intro_to_cellular_automation.php
http://home.comcast.net/~tom_forsyth/papers/cellular_automata_for_physical_modelling.html

I ended up going with a quadtree with lots of recursion. It works great. I do need to refine/optimize the water logic itself though. If you'd like I can share more details with you.

[eigenbom]

I ended up going with a quadtree with lots of recursion. It works great. I do need to refine/optimize the water logic itself though. If you'd like I can share more details with you.

Thx for the links, some videos of your fluid in action would be awesome!

[eigenbom]

Ha, I just read those links and realised we're both building pretty much the same thing.