Hi, I have experience making whole games with Godot, but I do not use Discord (proprietary spyware).
You can hit me on Matrix though (@kagaku_phreak:hackerspaces.be).
Recent community posts
Hi, I am a FOSS nerd gamedev who hates Discord and that's why ends up doing things solo most of the time.
Engine is Godot, games are weird-ass, graphics is mostly procedural and sounds are mostly recorded body sounds (which works SURPRISINGLY WELL).
Anyway, you can come by and say hi at
#brackeys2020:hackerspaces.be over at Matrix (Yes, matrix, not Discord)
also web link to the same chat https://matrix.to/#/!dAwcNyIysEBzDTgHjS:hackerspaces.be?via=hackerspaces.be/
Have a dayjob so do not expect more than ~4 hours per day though. But should be enough to do sth cool.
Cool upgrade to the Artificial Chemistry!
Though for some reason i can't yet reproduce my favourite multi-layered aggregates (which are stable only because outer shells prevent incoming particles from interacting with the core layers),
Linux'es generally go in different flavors, mainly in terms of default Desktop Environment, like KDE, XFCE and LXQT.
While amog DE's KDE is probably the best in terms of HiDPI scaling and touchpad support (useful for laptops), it is also quite resouce-heavy and can be more easily broken because of its complexity.
The lightest DE's on the ther hand, have varying degrees of touchpad support.
On my personal laptop though i hacked life and just use a DE which DOESN'T REQUIRE MOUSE AT ALL ^_^ (dwm). And it is probably _the_ lightest DE (comprised of around 4 .c files.)
Well, for code experiments generally Arch and Arch-based stuff are the best since they have AUR, which allows to make and share packages very easily, which in turn means that there is a literal sea of packages.
Also Arch and Arch based OS's enjoy literally latest versions of everything as soon as they are available (so called rolling release).
I personally use Arch both as a daily driver and a developer workbench.
However, that incredible flexibility and bleeding edge comes with a cost - even Manjaro in the long run may need a significant maintenance effort (unfortunately again have friends which had troubles with maintaining it).
It is not just sth you can install and forget about it.
And stock Arch is not even meant to be installed by an installer but instead you do it manually (quite easy, takes 10 min when you're used to it, but helps understanding how things work).
If you want almost all the coding stuff and godot (except for maybe some more esoteric and/or youngest unstable tools) without having to go into how Linux actually works and how to maintain it, i think you can easily go for any of Mint (though i haven't checked it for a while) or MXLinux. Maintainance and stability-wise Fedora is also pretty ok though it is a bit more "professional" kind of OS.
Web platform i encountered bugs on is QtWebEngine 5.13, which is essentially Chromium.
When building for web the usual trio for testing is:
1.) Something based on Gecko. For example Firefox.
2.) Something based on Chromium, or Chrome itself
3.) Something based on Webkit (Apple Safari and numerous little opensource browsers).
For Linux in VM better go for sth simpler and lighter than Ubuntu. It is recently going in strange directions, got incredibly fat and not always works nice in VM on Windows (had friends having troubles).
Maybe sth like MXLinux, Mint, (Debian-based) or Manjaro (from another family, Arch based) would be a better choice while still being beginner-friendly (unlike stock Arch, Void or Gentoo [which are great, but are rather for actual Linux fans and not just for quick testing]).
Nice atmosphere, clean crispy graphics and small touches, like colored collars.
Good selection of music theme, wish there was other ingame sound as well.
Balance is rather bad.
There is no real reason for example to have more than one farm at all times, so you just build it, assign someone to it, and forget about this game mechanics altogether because building houses for a ton of soldiers is a far more taxing task then getting food for them.
So this is essentialy a non-mech.
Fog of war in this game is very strange. It is there, but you can see the number of enemy units through it, which kind of ruins the purpose. Fog should at the very least hide enemies to be a useful mech.
Battle is flawed in a sense that enemy units don't behave as a whole but rather as individuals so you actually can beat opponents easily with a much smaller number of your forces.
I believe nobody of my soldiers ever died.
Probably if enemies were at least territorial - i.e get angry at any of your units going on their island, that would be a bit more interesting.
Enemy buildings are not capturable, which makes islands with enemies just islands with enemies - there is no extra reward associated wih them - you can use their farms but again you don't need a lot of farms in the first place.
Unless you go with a small but frequent groups of disposable attackers tactics.
What you REALLY need is housing, but alas you can't capture those "sombody's house"s
(you can force player to use small squads approach by limiting resources to build housing)
Main technical issue i encountered is when building houses, the house ghost is not always there (even when info tab pops up) and when you click on earth with no ghost nothing happens.
So basically you need to try to build a house multiple times which makes it a tiresome process.
Web version ended up unplayable for me. Each time a worker does sth, the game glitches out with resource numbers flickering until the worker is done.
Voxel graphics is ok time-wise.
All the assets (except for te sky) for my NeonHead for example were made in Goxel and it in sum (even with 3d font) it was somewhat about a day of work. For 9-day jam voxels are quite affordable i think.
Have a look (downloadable has better graphics then web cause GLES3). https://houkime.itch.io/neonhead
Also there is openSCAD which let's you literally write 3D models as a constructed solid geometry. If you write a nice set of your own functions it can look very neat and consume quite little time. It can also be wrapped in Python via solidPython for a better syntax.
Now if i wanted to do sth complex which can't be done with Goxel easily i would definitelty go for openSCAD. Especially for models of tech and such. You can look at one of my MIT openscad models here, it has a file and a png preview https://gitlab.com/Houkime/little_flyer
Godot builtin physics (if it is a builtin one) is quirky. I got stuck inside a blue guy and couldn't get out of him with the key.
Also one can't jump when pressing against the wall.
It is actually a third game during this jam which has problems with walls (first two are Crystal Sky ii and Grow Up).
The style is nice though ^_^
Nice deliberately (i believe) trashy style.
Strange physics of the islands. In some cases you can just land and they don't show any induced movement but it others they get pushed around. And spikes too.
I once had a situation when i dropped spikes to the ground and floor became figurative lava.
Dialogues though are wonky since default godot blue selection highlight is too thin and also up key to navigate among response variant in dialogue makes max fly.
Fitting main theme and sounds. ^_^
Nice pixel art and camera motions.
Visually though i think stone, dirt, and forks stick out too much. They are different in style from the character and floaties and stone coloration feels out of place.
Stone coloration weirdness is like when you edit a photo and paste in a human from another photo, and one of the reason a pasted human visually sticks out is because he was shot in another lighting conditions. Unless you manually edit color levels of a pasted human and do a ton of other manipulations he will continue to stick out.
This is one of the reasons i don't really like working with textures, esp coming from different sources.
Gameplay-wise one thing i wish is a reset button or at least a possiblity to kill yourself with forks.
I got stuck on a second level with seeds and couldn't get out of a stone womb below.
Also some minor bugs:
* Inability to lay dirt when pressing one's face against the wall.
* First level with seed introduction i ned to complete 2 times because somehow the exit door just led me to the same level all over again.
I mentioned this both in controls section of description and in game, but probably this should be addressed by making a better first level so that you actually see it for yourself multiple times and learn by doing.
Now the first level is made in a way that you start with one block over your head and it purposefully triggers this situation, but if the roof were bigger or sth so that you should deal with multple columns and maneuver around that would be more useful.
This nuance btw is not just a quirk but a sort of mathematical necessity - otherwise there can be an edge situation when you essentially kill yourself in a very weird way.
...empty... it feels so incredibly empty...
Island is there just to adhere to theme, and everything else conveys little meaning or emotion of its own.
It has such a great parody, creepiness and genre subversion potentials you can almost smell it, but it is not used, and mechanics doesn't really cover for the loss.
If you want to know how to subvert game genres, look up Frog Fractions which is dusguised as a children game. Frog Fractions 2, which is disguised as Glittermitten grove (a management game about fairies) is even more mindblowing.
Game provides no feedback (except for money increasing) when package goes down.
Also no reset and no option to speed up time in more boring periods (like "HOLD SPACE FOR 10x TIME").
Visiblility of black font on a blue background (price list) is bad.
Besides, this price list feels like should not even exist.
Core-wise what is missing is some economical mechanics which gets triggered on physical contact or adjacency.
The approach however feels promising. Keep it up!
Source art is good.
Speed dynamics feels rather frustrating then anything.
Limitations of using cut-out animations were not stretched or artistically accented, so it feels a little cheap.
If you have troubles with making cutout animations look good, consider using OpenToonz for frame-by frame drawn vector animations.
I used it once in a small experiment of mine and man, OpenToonz can totally make a game feel like Hollow Knight.
Nice small touches like voice acting, "Not alive anymore", funny font and the fact that you fall down in a manner that makes you feel that you are REALLY high up.
A bit of annoyance when a player faces the wall and walks into it sometimes player flips his orientation backwards and forwards like crazy.
Also, continuing little touches theme, if you make jumping in this sort of games (which is usually the main action of the game), consider making jumping action feel REAL good and cool. This one is a lacklaster a bit.
E two times when opening a chest and grapping an item feels excessive. One is more fitting.
Ctrl to attack with E to interact is an ergonomics nightmare.
I have not yet found cool people to team up with in such a way that makes resulting collaboration cooler from my perspective than solo work.
People call me in to join projects from time to time, but mostly it ends up being quite dull and not motivating enough and i quit soon.
How one around finding the right people to work with?
Do you have some experience with it?
I guess my games and other work might attract some community eventually and i can collaborate with community people in a typical opensource fashion, but that seems a bit distant. ^_^
I also can try and actively search for creators i personally consider cool and who do FOSS.
(Tested on Arch Linux)
The game has a built-in manual and it is a great plus
It is roughly understandable how to it was meant to be played and what it is all about, however bugs really hinder the possibilty of learning the rest by trial and error.
First, manual button at the start is broken and freezes everything.
Second, and most annoying, is that building screen does not work in any conceivable manner.
When you click a building the FIRST time to select it for construction, it actually builds (although the icon is not changed accordingly when construction is finished. It is changed only when you exit and enter the city)
However THEN strange things start to happen, like when you want to build a second building.
For some reason you can no longer select a place, and once you click, you
1.) erase (at least visually) the first building that you built
2.) build not the second building, but actually the first one.
3.) And it seems like the cycle repeats on and on, building a poor fishery or anything repeatedly till a city dies with tons of food but no water.
The scope was probably a tad bit bigger than 9 days really allowed to produce reliably.
I maybe need to add proper input locks for ENTER (so that you can't accidentially press it too quickly, it is used throughout the whole game but not for dialogues) and have a proper "skip" button for dialogue.
1. No, it is not just you. I was too busy playing with concepts for puzzles and core mechanics. I thought about it for several days straight mostly on paper and in sandbox-y mode, so when the time came to actually load those levels into the game and weave into/around story i was wastly behind the schedule so i made a deliberate choice to include "cool stuff" sacrificing much needed in-between levels of difficulty curve.
It is quite absurd in retrospective - the game even contains a level i can't solve myself (as a bonus), but not the basic stuff :(((
2. Yep, i should have made junk more obvious through first dialogues.
At the same time, it is kinda in line with the fact that it could be really anything instead of junk because it is people who are creative and see things and make them "fun" and not things themselves.
I guess i will have not only have TIME as a factor limiting size and depth of my games, but also my ability to make tutorials for them.
Like, if one can't realistically make a needed shallow tutorial, then reduce the scope even if you can go for cooler things.
I kind of wish for some cool tutorial tech which can introduce mathematically simple but counterintuitive things easily and not as a separate tutorial but like fused with the game itself.
At the same time concepts i usually tackle (even if simple in nature) prevent me from assuming that an average Joe will easily come along without a longish journey. That would make me a mega-accomplished pedagogue if i could do that.
And when you limit yourself to things you can reliably tell and transmit to others, that's a bit different from when you just play yourself with stuff you consider fun (even when it is not really important and can be considered junk).
In other words, i currently have a "wrong" goal for my games - incorporate some cool (for me) stuff and be interesting first of all for me while letting others to play with it as well as a side-effect, instead of considering a game as a communication tool with a "purpose" of communication, i.e. sth _useless_ if not communicating properly.
Or maybe i just need to have two separate types of games, like internal and external APIs - games i want to play myself and games i made to communicate with others.
Ironically a friend of mine when he first saw the drafts (on the first day of the jam) said sth like:
- The main problem with your games are long dialogues and no tutorial. Like i go into a game and have a real hard time figuring what's going on. Make a good tutorial and if you make long dialogues at least voice act them to make them more fun or sth.
I should have listened.
I added basic HOW TO PLAY section in description, but probably the problem here is that i just made the complexity curve WAY TOO STEEP.
I thought that if i couldn't realistically add needed number of levels in time i should just choose the most interesting ones and put them in the order of increasing number of available operations to not pour too much mechanics on the player too quickly.
Obviously that was a mistake ^______________________^
Simple and nice, although maybe too shallow to grab attention for longer.
(One can use less trivial math to add perceived depth without too much code complexity.
Think of it like fractals that are perceived very complexly but are very simple to write down.)
WARNING: DOWNLOADABLES (i.e not web version) WERE ADDED AFTER (~halfday,) THE JAM SUBMISSION TO BE MORE COMPLIANT WITH RULES.
(they are still using unchanged source files that you can fetch yourself from git at https://notabug.org/Houkime/nights_with_junk)
Linux export may not work on Linuxes using old libc.
OSX is NOT SUPPORTED. This is because to compile for OSX (at least in official godot way) i would need proprietary XCode which goes against my principles.
If you're on OSX, go get the sources and play them directly.
Ergonomics and predictability of your next action outcome is really important in hard-metroids and brawler fighters alike i think, but controls (esp when it comes for telekinesis and wall jumping) in FI are not that responsive and/or predictable as a metroid fan would like.
Combined with the relative hardness it is a bit frustrating.
Oddities i encounter:
When a wall is near, a character can enter a sort of walljump mode without the wall.
Rocks sometimes get stuck on spikes or drift randomly by themselves.
The border between bouncing on an arrow and being killed by an arrow is a bit too vague. Sometimes you can just stand in front of arrow dispatcher and only get repeatedly tickled ^_^
I even thought that it is an intended mechanics for arrows but alas vertical ones are still deadly.
You can command a rock to go away while you're directly on it, it won't move (ecpectedly), however when you jump it can slingshot all of a sudden.
Also there is a (probably test level) in the area to the right of the space where you get once you escape the shakey doom. If you go to the right of the test level you will disappear forever.
Not that's it is bad, just leaving an exploratory note here ))
Not really. Sometimes fadeins and fadeouts in audacity to avoid clicks at start and the end, a bit of in-godot pitch randomising and a major pitch downshift for level erasure sound for it to match animation length.
But generally i have a good enough wearable mic (around 30$ worth) so that major interventions are not necessary.
A good thing that you can just make a sound repeatedly in one recording go in Audacity a ton of times and then just cut out a good enough sample.
Procedural generation probably needs to take car physics into account to ensure that there exists at least one gear such that the next platform is reachable.
Number of gears feels a tad bit high. I would settle on 3 or sth.
Also $ usually means that you can buy sth, but you can't. It's a game-player communication problem.
Nice vibes though ^_^
I can't really use synths yet, so for previous NeonHead i went for a mix of simple synth + body sounds, and this one i made entirely with body sounds (except for music) because of how bloody simple they are to make. ^_^
Nice. Enjoyed it a bit though didn't have patience to grind through all repetitions needed. ^_^
I don't seem to receive a piece of floppy after 20 or at least there is no real indication.
Given that levels are not getting harder, 20 is a bit too high number for a mindless grind.
^_^ I knew that given my story-lust i can spend countless hours writing the story, so i intentionally shifted it to the end of production so that one can actually have a game to play and not just hours of dialogues (like it happened to my other game NeonHead https://houkime.itch.io/neonhead)
Nice. Simple, pleasant, short and semi-balanced, though not really that original.
The only problem i had is with ergonomics.
When you hold shift with one hand, your other hand gets busy with the mouse, and thus panning with wasd or arrows is not very convenient. If one uses the mouse to pan, then there is a bug (?) clearing the building target every time when one pans with mouse