Post by Kenneth Dodrill in S.M.I.L.E jam comments

Indie Horror Month » Entries » S.M.I.L.E

Viewing post in S.M.I.L.E jam comments

Kenneth Dodrill291 days ago (1 edit) (+1)

The level design reminds me a lot of the intro scene to Metal Gear Solid. I like the visuals and the music sounded good (could be turned up by default some more).

I ran across some bugs/issues though. My framerate wasn’t very good throughout, I haven’t used Unreal Engine before so I’m not sure how you could try and fix that. Some of the boxes in the first level rotated with my character, so jumping on them sent me flying. Sometimes it felt like I couldn’t shoot as fast as I could click.

If you are trying to make a DOOM-like game, maybe try to make an IWAD for DOOM / GZDOOM. There are a lot of tools out there for it and they come with support for basic item placement and stuff like that. Might be quicker to prototype out that way. Trenchbroom is a map editor I have used before.

Like Reply

Secresteyn291 days ago (1 edit) (+1)

Hi Kenneth,

Thanks very much for taking the time to try out the game!

I'm happy you recognized the callbacks to MGS in the level design and the retroShooter gamefeel.

I started this project working from a doomWad project and a pitch to mix elements of doom and the MGS1 demo disc!

Brief:

Depending on progress, pick one or more clone or baseDoom wad pack.

Make a short, microgame to enter into horror jam and publish on itch.io.

Use doomClone and create an eerie loop that subverts expectations.

Pitch One:

MGS1 Demo Disc elevator loop.

Each lift exit triggers a new horror mutator.

Make each loop more abstract and horrorfied.

I will be looking into a full optimization pass in a future update. My testing platforms are a moderate-tier PC and SteamDeck.

I found that adjusting settings from high to Medium improved performance from 12fps to 60fps capped on SteamDeck using a power limit of 8 watts. I attempted to force settings to medium when loading the first level - but it's a very hacky solution and there's lots of room for optimizing the hell out of this moving forward.

I will try my best to address optimization and performance in upcoming builds!

Gameplay Feedback:

Thanks for pointing out the issue with boxes - I'll be sure to work on a solution to this as soon as possible.

Your point about firing and input delay is a valid concern and it's something I will rework at the earliest opportunity. I am not satisfied with how the weapon system works currently and will be overhauling it significantly. I hope I can make it feel a bit more responsive.

With regards to Doom/GzDoom, it was my first port of call when planning out the project.

Project Smile started as a prototype in UlimateDoomBuilder before switching to Unreal. It may sound a little backwards - but I thought I'd build a baseline in Unreal and then match it back to Doom from there.

My hope is that building out this project will better prepare me for delivering on some lofty ideas I have for some custom iWad campaigns!

I would like the flexibility of a modern engine that allows me to work with features from other classics. I am particularly fond of the increased interactivity from BuildEngine games as an example.

I also have a few more elaborate design ideas I would like to explore to add a bit of a twist on that classic shooter feel that may be more feasible in Unreal than GzDoom.

I like the depth and breadth of content and creative minds found in the doom Modding scene and I will be releasing some iWads in future to help me refine the level design and gameFeel for my final solutions.

I've been playing a lot of Doom, Duke Nukem3D and ShadowWarriorClassic to help me solidify that core retro feel.

I do think the core mechanics and game feel of doom is lighting in a bottle and it still feels so smooth and "juicy" all these years later. I hope I can work towards something equally satisfying over time.

Thanks for pointing me in the direction of Trenchbroom I've come across it a couple of times when doing research and will look at it again.

More than anything, thanks very much for taking the time to offer some insightful feedback and checking out the project.

Playtesting and fresh eyes are exactly what I need to help elevate this project to the next level.

Happy Gaming!

Like Reply

Kenneth Dodrill291 days ago(+1)

Very interesting that you started with UltimateDoomBuilder! I suspect Unreal has some kind of DOOM template through an addon or something, did you use that or did you make all this on your own?

I like the idea of the different levels mutating the difficulty and/or enemies.

That’s interesting about the steam deck fps, I’m curious: are you running the game natively on there using the Linux version, or are you using Proton with the windows version? I’d be curious to know how each performs. I played this game on Linux by the way.

That makes sense about the features of Unreal vs GZDOOM. Unreal certainly has a large toolkit.

The last time I used Trenchbroom I was using a plugin for Godot that imported maps from it into Godot. I enjoyed using it, making maps was pretty easy. It may be more targeted at Quake though, not sure.

Oh, I forgot to mention last time that I finished a few levels and then I accidentally walked off the platform as it was going up. So I couldn’t get to the next level. Maybe you can block the player in when they get on the elevator.

Anyways, no problem for the feedback. Glad to help.

Like Reply

Secresteyn291 days ago(+2)

Hi Kenneth,

I will add some more collision volumes around the map boundaries and tidy up collision meshes in the next build.

Thanks for the tip!

The code is based on a Course by Uisco / Unreal Dev:

Available Here: https://www.unreal-university.com/courses/retrofpscourse?affiliateId=Zlel6rp15Qy...

I took this course at the start of the jam and got my prototype up and running and then started to add and tweak logic and features to match my goals.

I also deliberately chose to start with Unreal because I plan to rewrite each of the core systems from scratch for Unity / Godot in the future.

An advantage of Unreal5 is that I can really quickly prototype ideas using code blueprints and then polish, refactor and port the core logic to any target platform.

Unreal5 has a great userbase and community generating easy-to-digest tutorials and documentation.

I find the visual blueprint system a bit unwieldy at times but it's a good way to "mentally map out" basic logic.

I am also looking to learn a bit more C++ and highlight gaps in my general programming knowledge.

Again, it's a slightly odd way to learn but I find working across each of the major engines helps me get a better grasp of core programming concepts as I often have to adapt what I am learning to fit a new platform.

I tend to seek out programming content in different languages depending on what topic I am most interested in learning about.

Unreal seemed like a good place to start whilst leaving me scope to eventually port and rewrite my own completely custom implementations for Unity / Godot in my next projects.

I am also taking the time to study SourceCode for Doom, GzDoom, Quake and some other popular open-source projects to get an insight into how old-school devs approach issues and how modern engines handle things differently.

In all honesty, the Linux build was a bit rushed and I am beginning to suspect it may be slightly out of sync with the Windows build so I'll be sure to update them and sync my bug fixes on the next iteration.

When testing during development on SteamDeck I stuck to using Windows with the latest version of Proton.

Other than some minor hiccups installing prerequisite packages and adding the .exe as a non-steam game this was the fastest and easiest way for me to playtest on deck.

The native Linux build cooks and packages the content differently and I have yet to fully profile performance gain/loss between native Linux and Proton builds on the handheld.

It's something I'm very curious about moving forward.

One of the main hurdles I encountered when prototyping my maps in UDB is that building multi-storey maps is a bit of a nuisance.

Some of the ideas I have for level design studies require and utilize multiple z-axis levels.

I also decided to pursue Unreal5 to build as my previous (unpublished) games have been made using Unity (C#) and I have been further developing my Python skills to help augment my blender workflow.

I am working on an addOn for Blender that allows me to build levels procedurally using procGen algorithms, geometry nodes and traditional modelling techniques.

Long term, I will be building different versions of the "retroShooter" in Unreal 5, Unity & Godot to hammer out import/export and editor compatibility issues between my Blender addOn and the major game engines.

With a focus on adding more advanced AI, Behaviour trees & procedural generation systems.

I am very interested in how viable Godot will be for a project of this nature moving forward. I will be sure to test out Trenchbroom when I move onto my Godot build.

I am a big fan of open-source projects and would prefer to open-source as much of my work as possible moving forward.

I have been waiting for Quake to go on Sale to replay it and gather some thoughts and research notes.

I've also been playing PRODEUS recently, and that has a nifty little level editor built right in and is a good idea of where I hope to move my project moving forward (graphically, at least)

If you're a fan of Trenchbroom and Doom and haven't already tried out Prodeus I'd say your missing out!

Thanks again for the feedback - I'll be sure to let you know when I squash those pesky bugs.

Happy Gaming

Like Reply