https://tintwotin.itch.io/abyssal-ascent
As I got curious, how would a platformer with negative gravity work, I ended up doing this game in html (no game-engine used). 

I've never seen or played a game with such game mechanics before.

So, my question is: how does this negative gravity platformer work for you? 

Is there enough variety over time, to make you keep playing?

> html (no game-engine used).
Love that :) I did the same with my game, it is a lot of fun. TBH, I spend more time reading your code than playing,

Some suggestions/notes:

• The general vibe of graphics and sound feels good for this kind of casual and relaxed game.
• I do not think there is any difference with the 'positive' gravity in terms of gameplay, put the idea of ascent fits well with underwater theme, so I think it is still good.
• The game is tedious to the point of frustration due to the bubbles on the sides which are only reachable by slow-moving platforms. I quit without winning once mostly due to this issue. I suggest to either alter generation code to avoid such bubbles of give player some ability for horizontal dash
• The need to hit 'ascent' platform is non-obvious. I had to restart several times and carefully re-read the rules to understand what is going on. I suggest to visually highlight upper edge as 'dangerous' somehow, and then create a 'safe opening' after the air is collected.
• Using amount of air collected as score seems suboptimal, since it does not depend on player, but only of RNG. Maybe use some other stats like time, number of jumps, etc.
• Game page talks about cloning  but there is no repository link.
• I wonder what editor you use to be comfortable with lines THIS long :)
• Some code seems ripe for refactoring, simplest example being multiple fillText/strokeText pairs.

As I fellow fan of pure-js game programming,  I wish you  good luck with your project!

Thank you, for your in-depth review of the project. Some very good points. 

I realized that it is running at very different speeds across different hardware, some extremely slow other extremely fast, and on my computer, at normal speed. So calibrating the speed to be the same everywhere is a head scratcher. 

Dying floating out of the top screen, is properly the downside of making it upside down, gravity normally kills - negative gravity not so much - but making a closed top and then opening it when all pick-ups are collected, may defeat the abyss ascent feeling. 

Another head scratcher is the minimal traction it is gaining, and I'm not sure, I'll be able to convince myself, that it is worth investing additional time into a game no-one is playing (I did post on Reddit and SoMe - which is where most of the traffic is from, but it's been zero for a week now). 

(Btw. I did work as a professional game dev for a year (3d game engine stuff), and I also code: https://github.com/tin2tin?tab=repositories but this game was mainly vibe-coding test, to see how far I could get that way)   

> calibrating the speed to be the same everywhere is a head scratcher.
To achieve that, calculate time between frames and multiply object speeds by time elapsed (instead of your current fixed gameSpeedFactor).
Note that both Date.now and Performance.now return time in milliseconds, so scale accordingly.

>minimal traction it is gaining
That's normal. Dozens of games are published daily. Mine is not popular either :)

>professional game dev for a year (3d game engine stuff)
You may also look at the engine you develop for more details on time and speed calculations. Which engine it is, BTW?

https://streamable.com/teg4sa

my short review!

@2tgvr3thwryh3t2 Thanks for doing a video for this one too. These games I do are entirely made in HTML, so they're less than 40 kb big. All graphics are drawn or icons used, but no bitmaps are used. Only external file is the music. In your reviews, you miss the work put into how the games are progressing over time. 

They are intended as casual games, mainly for touch devices. Here is the PUSH button visible: