irrlamb 1.0.1

irrlamb 1.0.1

Postby jazztickets » 05 May 2019, 16:12

The object of the game is to touch all of the glowing orbs by skillfully maneuvering an energy ball through a variety of challenging levels.

Releases:
https://github.com/jazztickets/irrlamb/releases/

Trailer:
https://www.youtube.com/watch?v=6iex-aMhUjM

Image

Image

Image

Image

Image

Image

1.0.1
- Added secret levels to menu when unlocked
- Added campaign progress to menu
- Added player/orb sounds to replays
- Added option to enable/disable gamepad
- Added short messages for config changes
- Fixed various bugs in house level
- Fixed some compile warnings
- Minor bug fixes

1.0.0
- Added 16 new levels including a secret level
- Replaced Bullet physics engine with ODE
- Physics now runs at 500hz
- Updated UI which now scales with resolution
- Added replay sorting
- Added freecamera mode to replays
- Added gameplay options
- Added mouse sensitivity to options
- Added option for disabling multiple lights
- Added replay validation
- Added rolling friction
- Updated some tutorial levels
- Windows build now uses MinGW-w64
- Added irrlicht source to project and fixed many irrlicht bugs
- Increased quality of in-game screenshots
Last edited by jazztickets on 11 May 2019, 15:53, edited 2 times in total.
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby dulsi » 05 May 2019, 18:32

You really need a description with this. The readme on github doesn't tell you much. I watched almost the entire video. It took me a while to notice you were trying to get the other balls. Initially it looked to me like a physics demo.
dulsi
 
Posts: 570
Joined: 18 Feb 2016, 15:24

Re: irrlamb 1.0.0

Postby fluffrabbit » 05 May 2019, 19:08

I remember something called Irrlamb (or Irr-something else?) from years ago before I was doing much C/C++. I thought it was an Irrlicht extension library at the time; turns out it's a game. Neat!

As someone who needs to implement physics in his game rather soon, I am curious about what motivated the switch from Bullet to ODE.
fluffrabbit
 
Posts: 557
Joined: 11 Apr 2019, 11:17

Re: irrlamb 1.0.0

Postby fluffrabbit » 05 May 2019, 20:54

Toilets aside (what is it with this site and toilets?) I was so mesmerized by the trailer that I had to play the game.

It's weird that you're building Irrlicht in-tree. Though not unheard of, distribution maintainers prefer dependencies to be external. It's hard to tell which dependencies are internal and which are external; there are a number of them and it's a mix. When I built it, the binaries ended up in "bin" when they should have ended up in "working" for things to run. This is by no means a poorly-structured project, but it could be improved a little.

Also, you should check out ReText or a similar Markdown editor. Everybody likes a fancy README.

As for the game, I have been a little apprehensive about Irrlicht after some experiences with it not playing nice with my window manager. However, this game is immersive enoough that I never felt the need to alt-tab or alt-F4. The tutorials are pretty boring, but there's a nice atmosphere to being an advanced energy being in such a physically precise world.
fluffrabbit
 
Posts: 557
Joined: 11 Apr 2019, 11:17

Re: irrlamb 1.0.0

Postby drummyfish » 05 May 2019, 21:58

To a big fan of Neverball, which I am, this looks hugely amazing. But -- I've built the game, installed, and when I run it, my mouse (cursor that is, not my physical mouse, that would be scary) keeps going top left constantly by itself, i.e. I can't play. Any idea what this could be?
socialist anarcho-pacifist
Abolish all IP laws. Use CC0. Let's write less retarded software.
http://www.tastyfish.cz
User avatar
drummyfish
 
Posts: 448
Joined: 29 Jul 2018, 20:30
Location: Moravia

Re: irrlamb 1.0.0

Postby jazztickets » 06 May 2019, 00:56

I have a write up on Bullet->ODE that i'll probably post to the ODE mailing list at some point. Basically Bullet has terrible internal edge handling in its collision response: https://code.google.com/archive/p/bullet/issues/706

There are way too many bugs in Irrlicht to not include and fix it in source. Global glTranslate(0.375, 0.375) for all 2D graphics calls (ugh), no alt-tab support in fullscreen, terrible refresh rate detection... the list goes on. The README file tells you exactly which external dependencies are required and how to build and run the game. A lot of projects separate binaries from media for some reason (/usr/bin and /usr/share anyone?). Perhaps it's just for cleanliness or consistency...

The game is probably detecting a game controller (or thinking one of your peripherals like a fancy gaming keyboard is a controller) and is stuck at -32768, -32768 for its input. The same thing happens with my ps3 controller until you actually turn it on. Open up ~/.local/share/irrlamb/config.xml and set joystick_enabled="false".
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby Julius » 06 May 2019, 08:09

No idea how complex that might end up, but have you considered switching to the SuperTuxKart engine? It is also Irrlicht derived so it might be feasible.
User avatar
Julius
Community Moderator
 
Posts: 3297
Joined: 06 Dec 2009, 14:02

Re: irrlamb 1.0.0

Postby fluffrabbit » 06 May 2019, 08:47

Technically STK uses Antarctica as an alternative to Irrlicht apparently, but same difference. Other rendering engines miss a lot of Irrlicht's game features, so I don't know if I can strongly recommend anything as distant as OGRE, Panda3D, a raw graphics API, or one of those single-header things.

Referencing a Google Code issue for Bullet doesn't exactly shine a positive light on it. It's been a long time since that site was open. ODE is pretty old too, so I suppose there isn't really an age argument. I have to wonder what people prefer about Bullet over ODE, Newton, etc. The most game physics experience I have is raycast collisions in DarkBASIC and various things in Godot's in-house physics engine a few years after DarkBASIC stopped working and before Godot switched to Bullet. When you post that write-up, Google Groups is kinda nasty so maybe you could post it here as well, as its own thread on how the two engines compare. :)
fluffrabbit
 
Posts: 557
Joined: 11 Apr 2019, 11:17

Re: irrlamb 1.0.0

Postby drummyfish » 06 May 2019, 18:26

jazztickets {l Wrote}:set joystick_enabled="false".


Works! Thank you very much, am gonna play now :)
socialist anarcho-pacifist
Abolish all IP laws. Use CC0. Let's write less retarded software.
http://www.tastyfish.cz
User avatar
drummyfish
 
Posts: 448
Joined: 29 Jul 2018, 20:30
Location: Moravia

Re: irrlamb 1.0.0

Postby jazztickets » 06 May 2019, 19:13

drummyfish {l Wrote}:
jazztickets {l Wrote}:set joystick_enabled="false".


Works! Thank you very much, am gonna play now :)


So was a controller plugged in after all or something else? What's the console or irrlamb.log say?
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby drummyfish » 06 May 2019, 23:40

I had no controller plugged in AFAIK. I'm on a different computer now so can't see the logs, but the joystick was set to true in the config, which was causing it. Maybe it should be off by default? I'll try on my current PC as well.

BTW I love the game, great job.

EDIT: current computer is okay (Debian, the previous was Mint).
socialist anarcho-pacifist
Abolish all IP laws. Use CC0. Let's write less retarded software.
http://www.tastyfish.cz
User avatar
drummyfish
 
Posts: 448
Joined: 29 Jul 2018, 20:30
Location: Moravia

Re: irrlamb 1.0.0

Postby jazztickets » 07 May 2019, 17:37

jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby fluffrabbit » 08 May 2019, 07:29

jazztickets {l Wrote}:Here's the bullet/ode write up:

https://groups.google.com/d/msg/ode-use ... I92DW4BAAJ

Thank you for that. I will do my best to interpret the information.

Problems with Bullet:

The main reason I switched is that Bullet has poor internal edge handling with its collision response. Sometime after 2010 Bullet updated its collision response and I noticed bumpy movement over flat trimeshes.

Video of internal edge problem: https://youtu.be/71jd0g6PsdE
Original bug report: https://code.google.com/archive/p/bullet/issues/706

In the video, when the sphere hits the plane it sometimes dips into the plane a bit and bounces back, which is a phenomenon I have experienced in some game engines. If I am interpreting correctly, this is due to the seam between triangles.

Project from the video: https://github.com/jazztickets/btest

Someone referenced the issue in the libGDX issue tracker and linked a stackoverflow question. Following the trail of breadcrumbs, it looks like the issue finally found its way to Godot.

Bullet's Erwin did his best to resolve the issue, but it appears to be more fundamental to how Bullet works, so Godot's solution was a workaround. This is where my understanding of the problem ends. Apparently setting a certain parameter to 0 lessens trimesh collision issues, and of course there is the "just use more physics steps" solution, but I don't know anything beyond that.

Problems with ODE:

-- Damping --

One minor thing I noticed in ODE was that the timestep affects certain parameters like damping. For example if you change your timestep from 100hz to 1000hz, you need to multiply your damping by 0.1. Same with ERP.


-- Update Rate --

There was one issue in ODE I was unable to solve that only happens at sufficiently high physics rates (e.g. 1000hz).
Certain levels in the game require you to ride on top of a large object to move it around the map, and I found that I was unable to roll a large sphere with
relatively small mass (1.5 vs player's 1.0) and also a medium sized cylinder with a bigger mass (6.0).

Video example of sphere rolling:
https://youtu.be/UB8vnEBsJCM

Switching from dSINGLE to dDOUBLE in ODE helped a little bit. Increasing the ERP helped a tiny bit but that introduced too much bounce.
In the end, I had to settle for 500hz, even though the problem still occurs in one level (though it's not a big deal).
I imagine the problem occurs because some small value somewhere in the solver is being rounded down to 0. This was not a problem at all with Bullet, even at 10000hz.

Conclusion:

ODE is better for game physics.
fluffrabbit
 
Posts: 557
Joined: 11 Apr 2019, 11:17

Re: irrlamb 1.0.0

Postby Wuzzy » 08 May 2019, 10:42

Sadly, I can't build this one. I have gotten all the dependencies including libvorbis, followed the instructions in the README, but CMake still fails:

CMake Error at CMakeLists.txt:63 (find_package):
By not providing "FindVorbis.cmake" in CMAKE_MODULE_PATH this project has
asked CMake to find a package configuration file provided by "Vorbis", but
CMake did not find one.

Could not find a package configuration file provided by "Vorbis" with any
of the following names:

VorbisConfig.cmake
vorbis-config.cmake

Add the installation prefix of "Vorbis" to CMAKE_PREFIX_PATH or set
"Vorbis_DIR" to a directory containing one of the above files. If "Vorbis"
provides a separate development package or SDK, be sure it has been
installed.


I'm on Arch Linux with libvorbis 1.3.6.

EDIT: Nevermind. After a full re-clone and re-build, it worked.
User avatar
Wuzzy
 
Posts: 989
Joined: 28 May 2012, 23:13

Re: irrlamb 1.0.0

Postby jazztickets » 08 May 2019, 15:15

@Wuzzy You were missing the cmake submodule. For anyone else, you'd want to do 'git clone https://github.com/jazztickets/irrlamb.git --recurse-submodules', OR clone the repo then run 'git submodule update --init'.

EDIT: There's also the AUR package: https://aur.archlinux.org/packages/?SeB=m&K=jazztickets
Last edited by jazztickets on 08 May 2019, 16:47, edited 1 time in total.
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby fluffrabbit » 08 May 2019, 15:56

IMHO submodules is another problem to resolve as a developer before the big mainstream release.
fluffrabbit
 
Posts: 557
Joined: 11 Apr 2019, 11:17

Re: irrlamb 1.0.0

Postby jazztickets » 08 May 2019, 16:57

Yeah, i had found those links a while back and sadly the gist link from here results in 404: https://github.com/bulletphysics/bullet ... t-63558700. I didn't find a patch for anything, but if anyone does feel free to submit a pull request to either https://github.com/jazztickets/irrlamb/tree/bullet or https://github.com/jazztickets/btest. Although at this point i don't care too much anymore =)
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby fluffrabbit » 08 May 2019, 19:51

Oh... I was only hopelessly posting links I found, but if you believe the problem with Bullet can be fixed, that's great. Sorry, I have zero experience with physics APIs beyond a simple raycast. :)
fluffrabbit
 
Posts: 557
Joined: 11 Apr 2019, 11:17

irrlamb 1.0.0—REVIEW!

Postby Wuzzy » 08 May 2019, 23:26

OK, thanks!
You made me spend basically half the day with this game playing … :D
Guess I'm hooked … for now.
So I have a lot to say now.

First of all, congrats on the 1.0.0 release! I remember test versions of this game from years ago.

Boy, this game is hard. :D
After you complete the tutorial, the game almost instantly jumps into hard mode.
Completing any other levels takes me much, much longer. I guess I will be busy with this game for a while … :D

So, the idea of the game is interesting. Controlling the ball is really tricky and hard, it takes a long time to fully grasp how the ball truly works, even beyond the tutorial IMO. And to actually master the ball movement takes even longer. The control scheme alone makes the game quite unique, I guess. I'm not sure if I ever saw a ball game with such a tricky control …

The gameplay is as simple as it can get: Just touch all the orbs. The tricky part is with figuring out how to control your ball and how to interact with your environment. There are no extras, no collectibles but the game doesn't need them. It's a strong focus on physics.

About the levels … The difficulty raises very sharply. It goes from easy-peasy tutorials to EVIL MODE in an instant. The only easy non-tutorial levels are Hills, Skate Park I and Skate Park II. I consider everything else to be hard to super hard. Heck, I even found the Caves level to be tricky as the orbs are placed in some annoying positions.

Don't get me wrong, I don't think the hard levels are bad (in fact, you should absolutely keep them!), but for future versions I'd suggest to add some easy to intermediate levels so you can get used to the weird ball controls. It's not good when the difficulty raises all of a sudden.

It seems in the Caves level there is a nasty point of no return. At the end, there is this giant cave and it seems there is no way to return … it goes downhill very sharply. So if you have entered it, but missed a few orbs behind you, you must restart the level. It took me a long time until I realized I had to restart. Or I have completely overlooked a secret path … Idk. I think the hardest part in Caves is with me being confused where up and down is … Haha. Everything is dark and the walls are basically a big tube, and sometimes I really had trouble in figuring out where gravity will pull me to. But an interesting level nonetheless. Maybe disorientation was also totally intentional in this level … If that's the case: Good job! :D

The current level in the Caves category I'm on is Boulders (the 3rd) one. I mean, I totally get what I must do there, but I am too stupid to do it. But I am getting better … slowly. I will beat it eventually. My current record is 2 orbs.

I really liked the Sinkhole level, rolling in circles is fun! Also really makes you think in which order you hit the orbs because you don't want to get them near the hole … I believe this might even be my favourite level.
I haven't managed to beat Trees yet. It's way too easy to fall off. :D

The skate park levels I, II and III are cool, and the third one demands some crazy tricks. But with some thinking, it's doable. It took me like 30min or more. Spake Park IV I don't understand (yet). I'll think about it later. There's also a graphical glitch when you fall outside this level (just go left or right). Everything looks so flat and unreal and some surfaces just disappear when viewed from the other side …

The Tutorial is well made. Its good it teaches various tricks, they are really important for the levels. The only level that confused me for a while was the lava level (“Long Jump”). After only 75 deaths, I made it. XD The trick you have to do here is neither easy nor obvious. It's not enough to follow the on-screen instructions, you apparently also have to hit the keys at the perfect moment. Maybe this level should be completely redone. Maybe it's easier to understand if it goes in the forwards direction.

I found the secret level! And believe it or not, it “only” took me >150min to complete it! I swear, if my computer or the game would have crashed, I would have gone berserk. ;)
This makes me wish the game would save the current level state when I quit so I can resume later. Anyway, cool level design and I totally understand why it's a secret. :D
And boy, this level is huuuuuge! No wonder it took me so long. But the most time I spent figuring out what I am supposed to do.

The most annoying part was the toilet. First, it took me ages to figure out what I am supposed to do here. Then I figured I could use the toilet as ramp, but even after a billion attempts I fail here like 19 out of 20 times, but I am getting better here. I have not found a reliable way to climb up. I just try and try again, hoping to succeed. Eventually I did.
In my 2nd playthrough, I looked for secrets and I found one secret pizza piece at the window!
But something odd happened when I jumped on the television. The game teleported me to a Drug level?! WTF? And it seems there is no way to return. I speeded on this for like 10 minutes but nothing happened. Also no orbs anywhere. Not fair as it forced me to restart the level and lose all progress. :( Thankfully, this didn't happen in 1st playthrough. Maybe something happens when I touch the surface of those pillars, but I am too dumb for that. Is this supposed to happen?

Finally, I suggest to make the secret level accessible through level category menus as soon you found it. Having to wait ca. 20 seconds every time gets annoying fast. Plus, this would expose the level records.


The controls with gamepad are not good. I figured out the camera is much easier to control with mouse. The controls with keyboard and mouse is good. It works.
Btw, I noticed in the README there's a shortcut for flipping the mouse axis and some other tweaks. It would be good if the game would write some short message if one of the function keys is pressed. It's possible to hit those keys by accident and it is probably confusing when the mouse axis suddenly flips for no apparent reason.

I would recommend to players to stick to keyboard and mouse for now. The problem with the gamepad is that the camera control is too hard to control and also too slow. A precise and fast camera control is, however, very important for many levels. The gamepad moves the mouse (which is nice), but it's too slow on large resolutions (like 1920×1080). It takes ages for the mouse to move between buttons. Maybe increase the gamepad mouse speed on larger resolutions?

Plus, the camera movement with gamepad is on the wrong axes for my gamepad. My gamepad has 2 thumbsticks and the X axis of the first thumbstick moves the camera vertically and the Y axis of the 2nd thumbstick moves it horizontally … That's sub-optimal.
Flipping the axes does not help either, it won't fix the weird mapping.

Note that each gamepad uses its axes differently, you cannot predict which axis is X and which axis is Y, or if it is even a thumbstick at all. Hardcoding the axes into the game will only work with your own gamepad and a few lucky players who happen to have a similar gamepad. Therefore, I wish that the gamepad controls were configurable.

The sound effects are rather boring. Given the idea of irrlichts and orbs, the game would be much better if there is more atmospheric noise or a nice atmospheric background music, but there almost none. Some levels really are made for some background atmosphere: Volcanic atmosphere, cave atmosphere, space atmosphere, etc. The whole game is basically made out 3 sound effects: Your ball moves/rotates, and the orb hums or gets hit. Yes, there are other sounds, but they are less common. Needless to say, this is getting very repetitive.

What's worse, replays don't have any sound at all. This makes them boring to watch. I don't understand why you disabled sound in replays. It doesn't make sense. I have never seen a game which does this! Until today.

The graphics are rather poor. Everything is very simplistic and the textures are just meh and very repetitive. For dark levels the simplicity is acceptable but not for the bright levels like Skate Park III. The lava texture is especially bad, it's not even animated and looks very blurry and low-res.

The GUI is minimalist and does its job fine. It's like how game GUIs should be. The buttons and texts are large enough to be readable. I am mostly happy with it and wouldn't change much. The only minor thing I'd suggest is to add some check mark or another icon for completed levels/levelsets.

Also, I agree with the previous posters that the README should say what this game is about. That's the point of a README. ;)
So, PLEASE describe the game properly in the README. Also in the first forum post. I doubt many people actually care about a changelog when you never posted a previous version here. Maybe also add a screenshot or two. :) This increases your chances of your game being taken serious. <3

Anyway, overall it's a nice little game with unique gameplay. The game would benefit a lot from improved graphics and sounds, about the general “atmosphere” of the game. To make the game appeal to a larger audience, more easy-medium non-tutorial levels would be an improvement. But the current amount of levels is healthy, considering it takes a serious amount of time to proceed in the game. I am not going to run out of content anytime soon. :)
User avatar
Wuzzy
 
Posts: 989
Joined: 28 May 2012, 23:13

Re: irrlamb 1.0.0

Postby jazztickets » 09 May 2019, 02:49

Great feedback!

The game is hard for sure. I grew up mostly on NES and old PC games, where games were brutally difficult back then, mostly because you had start over from the beginning if you ran out of lives or from the lack of feedback in what you're supposed to do.

I don't know if you've played X-Moto or not but it's in a similar camp. Some of those levels are ridiculous. Imagine being a top player and still spending 35 hours just to complete one level. X-moto is an inspiration for sure. Fun fact: I created the skate park mod for that game back in the day (http://wiki.xmoto.tuxfamily.org/index.p ... e_Park_Mod)

irrlamb is certainly a puzzle physics game, where you learn the full extent of what your little ball can do, and what interesting interactions can occur with other objects through experimentation.

Turns out creating effective tutorials that communicate the idea you want to convey is hard. When I watch people play, they usually want to just run in guns ablazing, ignoring any tutorial message and spam the jump or forward key until they win. For example in the jumping tutorials, the idea is to watch how close you are to the stairs and jump slightly before. Most people just spam jump until they get up. Once they do, they're like "Yay I did it!", but they didn't learn any useful skills. Then when they get to a harder level, they're stuck cause they didn't go through the proper "training".

More ideas for tutorial levels:
- Jumping and landing on a small pillar below you, with full control afterwards.
- Advanced sphere rolling
- Gain speed by jumping and spinning in air (similar to long jump, but for making longer jumps)
- Tightrope control (like on the rails in the skate park levels)

I've thought about rearranging those orbs in 'Caves', but sometimes I hit them first try.. other times it takes like 10 tries. :shrug: I think there is some trick to it though.
It definitely requires some camera work for sure cause sometimes you'll be free falling but don't know it. It also probably doesn't help that I develop on a VG248QE monitor that's super bright on 144hz mode. "Looks good on my machine. Count it!"

Are you on 60hz? I tried playing on a few of my 60hz monitors and I'm like "holy shit this is way harder". And yes I agree, controllers suck for this game. You need WAY more control than a controller can give you. Although I have beaten House and a few other challenges with my Steam controller.

What controller do you have? The default inputs should work with xbox or PS3. The inputs are configurable in the xml file generated in the save location, but I'd just stick to keyboard/mouse.

The hardest way to play would be on a TV (20ms input lag for most tvs w/ game mode enabled), plus a controller, plus vsync enabled (up to an additional 16ms lag). That's the true hard mode.

The long jump level I made sideways since if it were back and forth, the camera would be right up next to the player and it would be hard to tell what the hell's going on. Plus I think it helps show just how much further you can jump. Although I think adding that tutorial level I mentioned earlier would be a good fit before that one, since it should teach you about spinning in air first.

Kudos on the House completion. That last jump to the window is kind of bullshit, eh? We had a party last Friday and it took my friends about almost two hours to beat! Ha. The toilet was part tricky at first, but they eventually figured it out. Although we discovered you can get up just using the books, but you have to be *very* lucky to get a good bounce.
The key to the toilet is to get good speed and jump right before the bottom lip. https://streamable.com/xiims

I'm actually surprised you found that 2nd secret level legit. I guess in the old physics engine the jump was way harder, so I didn't think anyone would try it. Did you really quit at 10 minutes in? It takes about 10m15s to beat if you're holding forward the whole time. Hah!


Sure, I agree about the art and sounds. But after 500 attempts and 2 hours in (when you're ready to smash your monitor in), are you *really* gonna care about having more than one texture in a level? =)

Hell even xmoto has "ugly mode", where you can disable all the textures and play in wireframe mode. Though i don't think that would work too well in 3d...

My protip for the harder levels: Take a break and come back to it the next day. You will play better after sleeping.

To add sounds to replays I would have to record sound events, the pitch of the player's sound, etc, since it's not actually running the simulation, it's just recording camera and object orientations every frame. The experimental Validation mode has sound but with limited playback options (since it's literally playing the game with the inputs it recorded).

That brings me to a point about the irrlicht engine. It hasn't had a meaningful update in like, 10 years. The blender exporter 'irrb' doesn't work in 2.8 and the engine itself has tons of issues. The stubborn problem I have is if I switched graphics engines I couldn't call the game irrlamb anymore... right? Hah.

I'd love to explore the genre more, and actually put true heart and soul into it (like I'm doing with choria)... but it has to be with my own engine, otherwise it's hard to stay motivated. Though it's funny how I still put in like 50-60 hours a week for 2 months in this last update. I think it's more about getting closure at this point than anything. I was just really ready to be done with the game in the last few weeks, and the lack of polish shows =)

It definitely needs better marketing, but it's also a very niche game as you've discovered. I don't think many people would stick to it like you have. I'd love to hear your thoughts (and curse words) later in the game when you get to the real bullshit levels. Thanks for playing.
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Re: irrlamb 1.0.0

Postby Wuzzy » 09 May 2019, 21:47

It was definitely the correct decision to actually finalize the game, even if not everything is perfect. There are just too many games in the free software sphere that are not quite finished and stay dormant forever. So any addition to the free software game sphere is good! :)


I now managed to also find all 3 secrets in the House level! It was fun to search for them. :)

Quote: "I'm actually surprised you found that 2nd secret level legit. I guess in the old physics engine the jump was way harder, so I didn't think anyone would try it. Did you really quit at 10 minutes in? It takes about 10m15s to beat if you're holding forward the whole time. Hah!"
Well, this is bullshit … There are not even any orbs. It's a really weird level. It looks more like a test level than anything else. What I don't like is that if you trigger the Drug secret, you lose all progress on your House level. Maybe make it so that to trigger the teleportation, you have to fall into something from which it is clear you cannot escape. Like a big whisky glass or mug into which you can fall in, but never go out, with a ruler or something leaning into it. But that still doesn't make Drug a good level …

I have another small suggestion for the House level: Currently, when you drop in the ventilation, you teleport to the monitor. But when I look up, there is nothing. I think it would make more sense if an air ventilation shaft is placed above the monitor. So it looks like you have been ventilated through the house. ;) Maybe spawn the ball inside the shaft, so you reall drop out of the ventilation shaft. But of course in such a way that the ball doesn't touch anything before hitting the monitor.
Currently, this feels more like a kind of random teleportation.


I have something to say about the Corkscrew tutorial: Initially I managed beat the level without understanding what I was supposed to do. Maybe I just lucked out. I rolled close to the left brick wall and won at first attempt. Therefore, I somehow completely missed the lesson that you have to roll the opposite way. I figured it out later, in a Skate Park level, eventually.

I just beat the Boulders level! After 136 min and >300 plays! The level after that, Descent, was much, much easier, compared to that. It only took me 5 min and 12 plays. Looks like the level order is a bit out of place here. I think Boulders is so difficult because every tiny mistake means instant game over. Descent is more forgiving. Here you just need to be strategic. I think Boulders should come much later.
User avatar
Wuzzy
 
Posts: 989
Joined: 28 May 2012, 23:13

Re: irrlamb 1.0.1

Postby jazztickets » 11 May 2019, 15:59

Good job! I dunno, for me I didn't think that level was going to be that hard. Of course, i'm the best irrlamb player in the world so what the hell do I know :p. Although it's nice to have an "easy" level following it, right? I think it's important to have plateaus in the difficulty curve to keep you motivated.

You should upload some of your replays so i can test the validation function. Also, new version is out.
jazztickets
 
Posts: 19
Joined: 07 Jan 2010, 05:51

Who is online

Users browsing this forum: No registered users and 1 guest