For those who played it, what did you think?

For me, Nifflas's games are like the Nintendo of indies in terms of polish. Everything seems extremely well thought out and nearly flawless. Even if you don't like the type of game he makes, I think everyone can agree that he always delivers a level o polish that few other indie developers achieve.

Must have taken me something like 3 or 4 hours to get to one of the endings (although I was also working towards the other ending at the same time).

The game gives you a lot of freedom to explore and many different ways to achieve your goal. There were a bunch of times where it was clear that I captured certain creatures in ways that weren't intended. The stores let you buy some of the stuff instead of having to explore to get them, and farming cash isn't exactly difficult, so you can get some items that may make you a bit too powerful.

It was definitely a very enjoyable experience, although there were a couple times where I had to stop for a little bit from the frustration of losing progress.

Took me a while to set it up. My PC isn't the greatest, so I had to keep lowering the anti-aliasing settings, only to discover that the game still doesn't run smoothly, forcing me to restart it yet again. In the end, I had to turn it off entirely in order to get it to perform well, which doesn't look too great, but is still tolerable. Wonder what kind of anti-aliasing technique the game uses. Seems to be a pretty resource-intensive one. Not sure if a 2D game really needs that.

As for the game itself, it's really fun so far. It has a very playful feel to it that encourages creativity. It's an interesting change of pace that this game doesn't kill you off nearly as often as FiNCK did, yet gives individual deaths far greater consequences. I was also pleased to discover how non-linear the game is. Furthermore, the physics are a considerable improvement over FiNCK, which feels a bit rough around the edges in comparison.

I also find it impressive that Nifflas made a few last-minute improvements to the game. I watched the three preview videos it received prior to release, and it seems he added new explanations about the difference between large and small doors as well as how to use birds in response to one of the Let's Players not figuring these things out as quickly as intended.

Regarding the soundtrack, I'm not consciously noticing most of the individual ways in which it adapts to the gameplay, but it feels a lot livelier than in Nifflas' past games as a result of its adaptiveness. I'm curious to see what he'll do in this regard in the future.

I've unlocked 3 or 4 doors so far. I'm really enjoying the gameplay and the puzzles, and the apparent open nature of the world. Knowing Nifflas, I'm sure there's an entire world of secrets I've already missed.

But while the gameplay is excellent, the controls/mechanics are intuitive, and the puzzles are challenging but fair (and fresh) there is absolutely no story or worldbuilding to speak of (so far.) This is a major departure from almost every other game he's made (that I've played) and I am missing its absence.

Wonder what kind of anti-aliasing technique the game uses. Seems to be a pretty resource-intensive one. Not sure if a 2D game really needs that.

I assume the game doesn't actually use bitmaps for the sprites, relying instead on polygons or vectorial images. It makes sense with the aesthetics of the game and means you can play in any given resolution without hindering visual quality, at the expense of performance.

there is absolutely no story or worldbuilding to speak of (so far.) This is a major departure from almost every other game he's made (that I've played) and I am missing its absence.

Have to agree there. It's kinda funny how there's all these annoying NPCs roaming around with no clear goal, though.

Uurnog was a pleasant surprise that I almost completely overlooked. I try to play all of Nifflas' games, but didn't even realize what it was in Humble Monthly until someone pointed it out. I have just finished the second ending.

It's weird. I was really hoping it would be another exploration-oriented game like Knytt, but it was more puzzle platformer like FiNCK. Still, I enjoyed it. Uurnog is probably the only puzzle platformer I've enjoyed in the last two years. I'd like to congratulate Nifflas for making something novel and good in a genre that has basically been done to death.

One big point of difference to previous Nifflas games is that Uurnog doesn't have a map or hundreds of unique collectable items or anything else that lets you know whether there's anything you've missed. I'm pretty sure I missed a few rooms somewhere. I don't like not having an indication that I've 100%'d the game in a game like this. This is probably my only criticism, and it's based on expectations built from playing previous Nifflas games.

Everything felt really polished. The music really helped express the atmosphere. The graphics were great. The puzzle design was interesting. The AI of the other human characters was weird, but that just added to the charm.

Overall, I really enjoyed it, and like all of Nifflas' games it has left me wanting more.

I really enjoyed it. It felt more organic that any of nifflas's other games to me, due to the interactions between different blocks, and the music, so it wasn't disappointing that it wasn't a huge game like knytt underground and the original knytt stories levels. After I finished the game i remembered that i still haven't finished lyle in cube sector, so i'll probably try to finish that soon.

I love the visual aesthetic (especially the trees) and the procedural music, as well as the sheer number of approaches possible for each puzzle. Managing the items in the save room was an entertaining sub-game as well.

The penalties for dying felt pretty steep at first, but once you memorize how to obtain certain valuable items it's not too painful. I was kicking myself more than once for pulling something out in a cursed zone without thinking about it though.

I'm in agreement with several of you about the NPCs. Although it was frequently humorous to see them pop out of doors with active bombs, their random behavior (more so than their muteness) made them seem less alive than the humanoids found in Nifflas' other games, which often had distinct personalities.

Overall, a very fun and pretty game. I was personally a fan of FiNCK so not disappointed at all to see a return to that style.

It's a tightly designed experience that takes a bit too long to click.  By the time I'd really settled into making goals for finding creatures, the game was just about over, and the second ending is easy enough to achieve that I felt myself really wanting more out of that puzzle style.  The lengthier stretches of puzzle are quite nice, though.

Something I was also hurting for until near the end of my playtime was a better labeling system for each door.  Other than the town and opening area, the letters didn't really cut it for me, and I found myself wanting proper names - "Waterfall," for example.  I was also kind of annoyed at the fact that the keyboard controls were not at all to my taste and I couldn't rebind them.  Mercifully, the gamepad controls are comfortable and intuitive, but the lack of rebinding is still a bit of a glaring issue.

Overall, it's a nice, polished experience, and I'm happy to have played it.  My desire for more might have me give a shot at designing my own puzzles, provided I can understand the confusing developer interface.

Hi! I'm glad to hear you seem to enjoy the game generally! To answer a few things:

I assume the game doesn't actually use bitmaps for the sprites, relying instead on polygons or vectorial images. It makes sense with the aesthetics of the game and means you can play in any given resolution without hindering visual quality, at the expense of performance.

This is actually completely correct. The story is that I wanted big antialiased pixels to make the game resolution independent and work with rotating objects. Originally I wrote a shader that could accomplish it, but there was a corner case where it'd look bad. I mean that as literately a corner case, because a transparent pixel in a corner between two non-transparent ones can only have a single color. The GPU interpolates towards the transparent pixel's channels independently (r, g, b, alpha), and that would make the color bleed over where the AA happens.

So, I wrote an algorithm that can convert a pixel image into a 3D mesh (though I only use two dimensions). The triangulation is a bit too advanced for my skills, so for that I used LibTessDotNet. At some point I realized it was very easy to use groups of four pixels as units instead of one, allowing a "block" to be split into two triangles. This also means that graphics eat more resources the more detailed the meshes are, not necessarily how big they are, so I had to stick to big solid color areas for memory reasons. That's basically how the game's style came into existence, by me not being able to fix a shader issue.

Have to agree there. It's kinda funny how there's all these annoying NPCs roaming around with no clear goal, though.

That part is basically inspired by any Zelda game. I love how Link runs into someone's house, start throwing pots all over the place (you know you did this). I wanted all NPCs in Uurnog to be like that.

I was also kind of annoyed at the fact that the keyboard controls were not at all to my taste and I couldn't rebind them.

Yeah, I agree. If I expand the game in the future, I definitely want to improve that.

Hey Nifflas, if improving Uurnog rendering performance is ever an issue, a solution you might attempt (although it could be difficult and possibly not worthwhile in the end after running performance tests to make sure) is to break up the static turf tiles into segments that are 1024x1024 pixels (or similar power of two sizes) large relative to the display resolution during gameplay, then render those tiles to a static texture and render the static texture as just one big wall (two triangles) on the appropriate layer between all of the animated features. Since the map tiles don't move (unless they're tree leaves or grass and the like) you'll be able to save thousands of polygons per frame off of the otherwise static meshes with no degradation of visual quality, having to only render those polygons once when the player enters the zone or gets close enough to see the associated tilemap. As a micro-optimization, this may even work for the player object, since the player doesn't rotate. The key factor is that it has to be done at run time, rather than during editing, because it's dependent on the end-user's display size, otherwise you'd be better off just fixing your shader.

The reason I think this can be a good idea, despite the potentially large use of graphics card memory, is because I believe it's how true type text and other static vector content is generally rendered in a hardware rendering space. Besides, since all the polygons are untextured monocolor, you're really not using that video memory for anything else.

I don't know, my graphics card seems to be beefy enough to handle any number of polygons like this just fine, so I didn't even notice it. It's actually really neat having vector graphics in the editor, where you can zoom in and see all the perfect detail of any little object.