insomniac_lemon

joined 1 year ago
[–] insomniac_lemon@kbin.social 9 points 5 months ago

I'm imagining that was intended as a foothold, the guy (still choking) being dragged out of the hole with the other guy's response being a mix of "not again" and "you need a drop for that to work as intended"

[–] insomniac_lemon@kbin.social 1 points 5 months ago* (last edited 5 months ago)

I can't see your reply on my instance, @tal (currently 1 day ago, still not federated) so not sure if you will see this (or if you do, if I'll see your 2nd reply...). EDIT: I forgot the first @ when I first posted, so I don't know if it actually worked as a mention

One was kicked off by someone generating polygonal-style art

It's visually striking (if you don't look too closely), the 1st bonus image is best but I'd still go with a more minimal style. That and aside from hallucination, I would prefer live-rendered polygons. Infinite scalability is the point.

Here's something I made a while ago, animated eye (note: on my end, Imgur links don't work unless viewed in private mode for whatever reason) though a full game with that style is not currently viable for me for multiple reasons (the feature is 4.X only and still an unmerged PR that may not perform well enough for common use, no Nim-lang bindings yet in 4.X).

Carrier Command 2?

I was confused at first, that style of 3D polygonal isn't really uncommon. I don't really buy things (esp not $30 level) and I need a hobby so that's a part of it too*. For 3D art I've done, one of my threads is federated to your instance. Here's the stuff that didn't (these have no textures, only vertex colors):

badgerbadgerbadgerSPACESHIP

banana

office plant

Note the 1st and last show a white background in a new tab but the background is transparent (and show as such on Kbin).

*= I mean I have seen some games that have a nice aesthetic, even better if they're more "real" with it (though that is what's hard to find)

[–] insomniac_lemon@kbin.social 1 points 5 months ago* (last edited 5 months ago)

I was worried the remaster would take some of this away, but it looked just as good.

The PS2 version looks great after upping the internal res (not sure how much other stuff like filtering/other technicals has an effect). I haven't compared it, but like most remasters I'm going to just say the data bloat is probably not worth potential fidelity improvements. That and I'd guess any design issues are still baked in, thus similar experience.

EDIT: In Okami's case I don't know if the data size is due to uncompressed files or just due to higher-res pre-rendered videos, but either seems wasteful to me. What is live-rendered and what is pre-rendered just seems arbitrary to me, I'd get if it weren't viable on older hardware but you'd think a remaster could handle it mostly in-engine.

[–] insomniac_lemon@kbin.social 3 points 5 months ago

For me it's definitely maintenance, largely because my internet is not that fast (~6-8Mbps shared among other people) but also I think package management in general leaves a lot to be desired. Some of it might be better on a new install (or listed in the wiki and I just don't feel like bothering to do the work) though.

[–] insomniac_lemon@kbin.social 4 points 5 months ago

I mean... Teresa certainly wouldn't involve a condom even if aware of having an STD...

[–] insomniac_lemon@kbin.social 2 points 5 months ago

The problem is that they overlap, usually all 3 interlocked. The threads/microblogs I've tried barely get responses (again, federation may be an issue), let alone even answers for even something like Blender. I can use Nim w/o art but I don't have the ideas for it usually (or if I do other issues happen, including just lacking the desire to write for something like a game book).

I've mostly waited for something to improve, but a while ago I started my own simple polygon loader/format and I worked a bit more on that today. I think I made one of my own questions irrelevant (assuming my condition to detect strip vs. fan is correct) and added a couple of other improvements. I don't think it's at a point I'd share it, but I probably could (should) try to make a simple game with it soon.

Though I'd rather have 3D in Raylib (vertex colors not working with Nim bindings, Naylib) or more advanced 2D in Godot 4 (no Nim-lang bindings, and said feature is an unmerged PR that may not be performant enough for full game art).

[–] insomniac_lemon@kbin.social 4 points 5 months ago (4 children)

My niche interests.

  • Nim-lang (this has some activity here, though I often have federation issues)
  • Raylib
  • low-poly+untextured polygonal art and vertex colors (both 2D and 3D)

I don't really consider myself an artist or a programmer (I haven't done much), maybe there's a fediverse instance that could work for me but it's probably too niche even just with those communities.

[–] insomniac_lemon@kbin.social 3 points 6 months ago

Well yeah, but mostly due to lacking any chance and I'm a shut-in (who probably has SzPD). Also time has already made a fool of me.

[–] insomniac_lemon@kbin.social 3 points 6 months ago (1 children)

I want to use Raylib, but mentioning it here on the fediverse doesn't get much of a response (I can't see a raylib community from my instance). My choice of language probably doesn't help, though.

My first issue is wanting vertex colors on 3D models and I am not getting this (this may be a problem with the bindings I'm using, naylib(nim-lang)). The second would be needing guidance for the 2D polygon text loader that I started.

Maybe I could make simple GUI applications with raygui, but I don't currently really have many viable ideas on what I would want to make.


To OP: Another potential option is using Godot w/bindings. Design is pretty fast and flexible, then using signals is super easy.

I've tested some frameworks (specific to my language, so not really helpful to most), the one that I liked more said it was declarative user interface framework based on GTK though I would prefer a similar thing for Qt and there wasn't an ability to automatically scale text size to better fill the available button size (I was testing an adventure-book reader and hoping to use unicode characters).

Frameworks for single page applications (or some other browser-based tech) might be ok for simple stuff. Similarly, I've liked the idea of TUI frameworks (yeah, because htop) but haven't really tried that yet.

[–] insomniac_lemon@kbin.social -4 points 6 months ago

If potential is key, I say keep the context of the MAID process but instead of outright death make it cryonics. Plus other potential relevant volunteer stuff and organ donation stuff lined up. Even if the initial cryonics technique is not even close to viable, other stuff could be transformative. If cryonics has any chance to work, things will get appreciably better in 300-or-so years right?

Hopeful worst is my brain in a jar mostly playing VR and sometimes knitting yarn via robotic arms. Lots of ways it could be better. Also unlike traditional cyborg stuff with all-machine life-support, I would like to still have a complex microbiome if not taking it further with symbiosis.

[–] insomniac_lemon@kbin.social 3 points 6 months ago* (last edited 6 months ago)

Are you challenging me?

For the most part, it's not hard to find them if they're doing the things I said and you pay attention while they do it. Look at how many titles a publisher has on Steam, see if they have a wikipedia page and if so if there's monetary info involved. Recognizing a dev/publisher might also be part of it.

Also with self-publishing never being easier, some of my skepticism starts there. Another is games seeming somewhat shovelware-esque or like they're trying to ride the wave of some other successful game/trend and that's why targeting consoles early-on is likely important to them for the money.


I originally wasn't, but off the top of my head some of the stronger examples:

Just because something is cute pixels that does not mean it's indie. A good introduction to this is the existing discussion of Dave the Diver and its ties to Nexon. EDIT: Also, lootbox controversy with Nexon and Maplestory

One involving unpaid marketing and crowdfunding/early-access: tinyBuild. ~$473m IPO. Publisher of Hello Neighbor, which also has some controversy around it on quality (also mobile games with micro-transactions, because kid audience). While searching on this, I also saw someone angry about them doing testing on Steam and then a post-launch Epic exclusivity. EDIT: Also one of their games not having all content available on GOG.

The game Roots of Pacha had a license dispute (I do not know the cause, but the dev did end up getting the Steam rights) their original publisher had at least 6 different accounts on Imgur (and they also did the crowdfunding/EA thing too, and no it was not like 1 game per account either and some of those accounts are mysteriously gone now). Same publisher was in the news about controversy over boob physics, and I don't doubt it was either suggested by the CEO for the headlines or just marketing clicks if controversy hadn't have happened.


Even if people don't care about stuff like this enough to stop buying the games, I hope they at least try to not enable or reward blatant self-promotion (particularly the more dipping and questionable practices involved) on the fediverse

[–] insomniac_lemon@kbin.social 3 points 6 months ago* (last edited 6 months ago) (2 children)

And don’t confuse high budget indie studios with AAA game developers

On the other hand, there are a lot of publishers out there who really shouldn't have things called indie when they're involved.

The ones who have struck gold (perhaps multiple times) and are already worth multiple millions, publicly traded or even owned largely by investment firms. Some like this still footing everything on the players (crowdfunding and then early access) and on top of all of that going onto places like Imgur and Reddit and doing unpaid marketing there (doesn't seem great for the actual devs, and then there are things like multiple accounts/sockpuppets/deleting+reposting etc).

And even without the unpaid marketing stuff, a publisher has a lot of ways to screw over developers and/or players usually with the goal of money in some form.

 

Made with Blender. If you want to see more (or read details I've written about before) I have other threads but note that most instances I've seen seem to only get one thread or the other (so you may need to look at my threads on Kbin). There's another image (an office plant) in the comments of one too. Nothing really substantial yet, just trying stuff.

Aesthetic note: Some of the look is just a side-effect of a lower resolution with color reduction to reduce the data of the png. For live-rendered content, I prefer high-fidelity (if certain options actually have a benefit maybe I'd use it, but even then something like vertex lighting is more likely).

For potential gamedev I'd like to use Raylib (with bindings for Nim-lang, and I ran into the issue of vertex colors not loading). Godot 3.X could work but I noticed every imported model will have its own material (Godot 4.X would be nice, but no production-ready bindings for Nim yet like 3.X has).

The Blender workflow here could be better:

  • give every new object color data by default
  • ~~render settings eg filmic contrast w/transparency~~ EDIT: this is covered by file>defaults>save...
  • swapping to solid render mode for the vertex paint mode and then back to rendered for object mode
  • an add-on to set vertex colors in a way that allows quickly editing them (sort of like a color index)?
  • do geometry nodes have any use for this style/vertex colors?
    • I made a spiral vase with modifiers easily, but couldn't figure out how to make stripes in the vertex colors without applying the modifiers first and doing it manually...
1
vertex color banana and peel (media.kbin.social)
submitted 7 months ago* (last edited 7 months ago) by insomniac_lemon@kbin.social to c/blender@kbin.social
 

I'm still at the point where I'm not sure what to make, but already wishing there was a way to have a color attribute (and material set up to use it+higher roughness) for newly added objects.

Well, that and it would be nice if I could have vertex paint editor automatically swap the rendering mode to solid (with object mode using whatever is set) considering I can have that set to color attribute with less lighting(/flat shading).

This one doesn't actually have much design when it comes to vertex colors, most of it is in the brown peel. But it was quick to make. Used proportional editing at the start, separated it for the peels. After basic rotation with a moved origin, flattening the peel arms was a bit weird and I didn't see anything better than doing it manually but it's not like it needs to be perfect.

Previous thread on !ps1graphics for more context (and a few other Blender questions, context-specific comments). Kbin link.

 

So the basic idea here is using a low-resolution texture that is designed specifically to be upscaled (anything but the first 3) to a simpler smooth aesthetic. Pixel blob/triangle/hexagon etc in with a 32x-or-lower texture (which is all that is downloaded), crisp shape(s) out. Also probably things like lineart for faces (or other similarly composed pieces), generally less colors hopefully preventing ambiguous details.

Batch conversion / post-processing / basic filter (to preview) are all options, but a live shader in an editor would be more useful. Even better for multiple outputs to see results for different shaders and make designs with their quirks in mind to make specific looks (esp. if multiple shaders can be used), or even just an option to tinker with (say if other pixel-art programs are better).

For an editor, I'd want to use Krita (a G'MIC filter would be fine but the ones I tried weren't great**) or maybe a web editor. GIMP would be fine too.

For a medium I probably want to use Raylib, Godot 3.X (with this aesthetic at least), or maybe even Krita animations? Basic 2D/3D, tile maps (per-layer filtering?), bump maps or other PBR/shader inputs, texture warping, using sprites in 3D, animation etc all seem interesting.

SDF (Signed Distance Field/Function) textures are a similar endpoint (and I would consider using Inkscape for that).


For some backstory, I don't really want the pixel art look (especially because I'm not so good at it) nor do I have the skill for a painted look. I'd rather have resolution-independent* art, but upscaled could allow using raster art's tools/features for better support and easier tools/creation/aesthetic etc. particularly for learning/practice while still allowing for higher resolutions than tested for.

I could probably just go for untextured 3D with vertex colors, but I kinda don't want the extra complexity or needing to use Blender. Though both 2D polygons and 3D meshes could be used alongside upscaled textures.

*= Like THE EYE, but there are issues like it needing not-yet-released-PR which may not perform well if used to create an entire scene, plus AA support not being everywhere even for basic polygons, or language bindings being unavailable for some languages in 4.X (ones that work in 3.X).

**= Upscale [Scale2x] works somewhat (16x with x 8 scale factor to get a 128x texture), though the preview appears broken and you must manually resize your canvas before running it. I suspect this is because it is an older filter.

Meta note: No tags used because the relevant ones I could think of all relate to content, particularly pixel art which this doesn't really relate to aside from specific intention and maybe some techniques.

 

Does anyone have any information about a #polygonal or #vector #tilemap ? Perhaps treating a live-rendered #svg file (or grid-based coordinates in code) as a #tileset for #raylib ?

Polygon tiles on their own wouldn't be so bad, but it'd be nice to have the polygons connect/merge (visuals and performance, basically treating each tile as its exposed edge) plus autotiling (/bitmasking?) and other tilemap features.

An editor would be great, also #hexagon grid. Vertex colors and extra data (shape+color palette+item/decoration) would be nice too.

Or would it be more supported to just go with low-poly 3D? Though even in that case it would be good to have a gridmap.

Note that Godot 4 might allow polygon tiles (using scenes, I haven't tried it, I expect it might be clunky) but I want to use #nim ( #NimLang ) and the bindings aren't where they were for Godot 3 (there are at least 2 people who were working on their own bindings, but I wouldn't know where to begin on testing them).

#gamedev

view more: next ›