Previous /sdg/ thread : >>101508847>Beginner UI local installEasyDiffusion: https://easydiffusion.github.ioFooocus: https://github.com/lllyasviel/fooocusMetastable: https://metastable.studio>Local installAutomatic1111: https://github.com/automatic1111/stable-diffusion-webuiComfyUI: https://github.com/comfyanonymous/ComfyUISD.Next: https://github.com/vladmandic/automaticAMD GPU: https://rentry.org/sdg-link#amd-gpuIntel GPU: https://rentry.org/sdg-link#intel-gpu>Use a VAE if your images look washed outhttps://rentry.org/sdvae>Run cloud hosted instancehttps://rentry.org/sdg-link#run-cloud-hosted-instance>SD3 info & downloadhttps://rentry.org/sdg-link#sd3>Try online without registrationsd3: https://huggingface.co/spaces/stabilityai/stable-diffusion-3-mediumtxt2img: https://www.mage.spaceimg2img: https://huggingface.co/spaces/huggingface/diffuse-the-rest>Models, LoRAs & upscalinghttps://civitai.comhttps://huggingface.cohttps://aitracker.arthttps://openmodeldb.info>Animationhttps://rentry.org/AnimAnonhttps://rentry.org/AnimAnon-AnimDiffhttps://rentry.org/AnimAnon-Deforum >Index of guides and other toolshttps://rentry.org/sdg-linkhttps://rentry.org/rentrysd>View and submit GPU performance datahttps://vladmandic.github.io/sd-extension-system-info/pages/benchmark.htmlhttps://docs.getgrist.com/3mjouqRSdkBY/sdperformance>Share image prompt info4chan removes prompt info from images, share them with the following guide/site...https://rentry.org/hdgcbhttps://catbox.moe>Discord6wUwtcJsr2>Related boards>>>/h/hdg>>>/e/edg>>>/d/ddg>>>/b/degen>>>/vt/vtai>>>/aco/sdg>>>/trash/sdg
>mfw Resource news07/21/2024>Batched Runge-Kutta Samplers for ComfyUI https://github.com/wootwootwootwoot/ComfyUI-RK-Sampler>SimpleTuner now supports Kolors training for full tune or LoRA/DoRAhttps://github.com/bghira/SimpleTuner/blob/main/documentation/quickstart/KOLORS.md>What Kamala Harris has said about AI, tech regulation, and morehttps://techcrunch.com/2024/07/21/what-kamala-harris-has-said-about-ai-tech-regulation-and-more/>DynamiCrafter: Animating Open-domain Images with Video Diffusion Priorshttps://github.com/FDoKE/DynamiCrafter>ComfyUI_EchoMimic: Lifelike Audio-Driven Portrait Animations https://github.com/smthemex/ComfyUI_EchoMimic>Two-stage PixArt ensemble of experts (2x 900M)https://huggingface.co/spaces/ptx0/PixArt-900M-EDiffi>florence2-ft-simple: finetune your florence2 modelshttps://github.com/sdbds/florence2-ft-advanced07/20/2024>Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Modelshttps://github.com/open-mmlab/Live2Diff>Meta Chameleon 7B and 34Bhttps://huggingface.co/collections/facebook/chameleon-668da9663f80d483b4c61f5807/19/2024>Follow-Your-Emoji: Freestyle Portrait Animationhttps://github.com/mayuelala/FollowYourEmoji>ComfyUI-Inspyrenet-Rembg [background removal]https://github.com/john-mnz/ComfyUI-Inspyrenet-Rembg>Shape of Motion: 4D Reconstruction from a Single Videohttps://shape-of-motion.github.io/>Surf and RAG-chat with 4 million public stable diffusion promptshttps://github.com/hexive/prompt-cli>UCIP: A Universal Framework for Compressed Image Super-Resolution using Dynamic Prompthttps://github.com/lixinustc/UCIP>DreamStory: Open-Domain Story Visualization by LLM-Guided Multi-Subject Consistent Diffusionhttps://dream-xyz.github.io/dreamstory07/18/2024>Fooocus v2.5.0 Updatehttps://github.com/lllyasviel/Fooocus/releases/tag/v2.5.0>PromptGen - Image tag model based on Florence 2https://huggingface.co/MiaoshouAI/Florence-2-base-PromptGen
>mfw Research news07/21/2024>FETCH: A Memory-Efficient Replay Approach for Continual Learning in Image Classificationhttps://arxiv.org/abs/2407.12375>MDPE: A Multimodal Deception Dataset with Personality and Emotional Characteristicshttps://arxiv.org/abs/2407.12274>Exploring Advanced LLMs with LLMsuitehttps://arxiv.org/abs/2407.12036>Benchmarking the Attribution Quality of Vision Modelshttps://arxiv.org/abs/2407.11910>Improving Unsupervised Video Object Segmentation via Fake Flow Generationhttps://arxiv.org/abs/2407.11714>Distractors-Immune Representation Learning with Cross-modal Contrastive Regularization for Change Captioninghttps://arxiv.org/abs/2407.11683>The Kolmogorov Complexity of Irish traditional dance musichttps://arxiv.org/abs/2407.12000>DreamCatalyst: Fast and High-Quality 3D Editing via Controlling Editability and Identity Preservationhttps://arxiv.org/abs/2407.11394>Visualization Literacy of MLLMs: A Comparative Studyhttps://arxiv.org/abs/2407.10996>Overcoming Catastrophic Forgetting in Federated Class-Incremental Learning via Federated Global Twin Generatorhttps://arxiv.org/abs/2407.11078>Visual Prompt Selection for In-Context Learning Segmentationhttps://arxiv.org/abs/2407.10233>PAFUSE: Part-based Diffusion for 3D Whole-Body Pose Estimationhttps://arxiv.org/abs/2407.10220>A Self-Supervised Learning Pipeline for Demographically Fair Facial Attribute Classificationhttps://arxiv.org/abs/2407.10104>Learning Online Scale Transformation for Talking Head Video Generationhttps://arxiv.org/abs/2407.09965>NamedCurves: Learned Image Enhancement via Color Naminghttps://arxiv.org/abs/2407.09892>Explainable Image Captioning using CNN- CNN architecture and Hierarchical Attentionhttps://arxiv.org/abs/2407.09556>Geometric Remove-and-Retrain (GOAR): Coordinate-Invariant eXplainable AI Assessmenthttps://arxiv.org/abs/2407.12401>VISA: Reasoning Video Object Segmentation via LLMshttps://arxiv.org/abs/2407.11325
Dorf!
Thinking about piping Whisper to Stable diffusion via a EVP box , ideally the end result being a ghost illustrating their regrets in their past life. Is there anything I'm overlooking or should be mindful of ?
>>101515336That sounds really cool and would like to see the results. I know nothing about it though. Be mindful of evil spirits I guess.
this model does strange things
>>101516770catbox?
Last one from me good night anons
>>101517596bye
>>101517596night
>>101517160he said before it's nai. should have the info on the stealth png
>>101518302ah too bad thanks
>>101518392Ah sorry, didn't see you post.There was nothing special about that prompt, I didn't really specify anything and just tried some random artists. Sometimes I just like seeing what the model does with that.Prompt:1girl, character:kasodani kyouko, touhou, artist:ayanepuna, artist:natsu no koucha, best quality, amazing quality, very aesthetic, absurdresUndesired Content:nsfw, lowres, {bad}, error, fewer, extra, missing, worst quality, jpeg artifacts, bad quality, watermark, unfinished, displeasing, chromatic aberration, signature, extra digits, artistic error, username, scan, [abstract], bad anatomy, bad hands, @_@, mismatched pupils, heart-shaped pupils, glowing eyes, injuryCan't tell you what the seed was, I had to inpaint a little.
237m, maybe 6 more months of gens to getalso started another one, 1m so far, nearly 2 years of gens to getDiffusion-1B soon:tm:
goob morbin
>>101519169hello, i hope you are doing well
>>101519487oh sure, i went to bed imagining hanging myself, and woke up the same, im sure today will be better than yesterday, or worse, or the same
gm all
>>101519960i hope you will conquer your demons one day friend
>>101520135good morningi hate trani
>>101520135goo morning
>>101520135>>101520227>>101520283>>101520446>>101520466>>101520551>>101520560>>101520575>>101520897https://suno.com/song/ea304130-8135-48e0-8d35-628dcf440aeb
>>101521588thread schizo
luma is so slow
244m and 4malso implemented some ops for ait and fixed webdataset in datasets library because i'm slightly more miserable than usual today, idk whether to do more ops or just have a nap
>>101521899Yikes!
>>101521899>debo likes this "weirdness"
trying loras as wildcards, random 0 to .9
nice to think 'i wonder if there is a bonzibuddy lora' and sure enough
Morning anons
>>101522438Morning
>>101522621Why didn't you attach an image to this
>>101522652I usually don't when saying hello
>>101522661How else will anon know who's saying good morning to Quokka?
>>101522438gm
pitiful
my gf
I'm not sure this portrait is really look-alike , trying to make something between official photos and pin up fanservice like with aoc.Which celebrities get reliably rendered on sd and sdxl?
>>101523137the ones with loras
>>101523184>the ones with lorasI was talking about base models not trained towards anyone. There's also some specialized controlnets for identity. For example it works well foe Emma Watson and Taylor swift and I'm experimenting with Harris due to recent news, however I'm wondering who else is present enough in training to get reliably out
https://files.catbox.moe/31i2vx.safetensorsmade a lora off of a few 'op art' type pics, mainly adds color effect
Im considering a gpu, can you guys help me out?>what gpu do you use?>how long does it take you to generate an image?>when generating these images, how many steps do you set?>what resolution do you make it?
>>101523346you're gonna want the best nvidia gpu you can afford, with the most vram within your budget range
>>1015233461080rn i am: for a 512x768 image takes 14 seconds, 24 steps. to upscale 2x takes about 1 minute, with 24 steps.
>>101523137>>101523198>more asian kamalaok I'm liking where this is going
>>101523639do you not get bored posting here literally every single day?
>>101523650ask yourself the same
>>101523661I dont post here every day mint pussy
>schizo already having a melty
4 ops implemented, unblocked a bunch of models in diffusers_aitnice
>>101523718diffusers_ait is the biggest nothingburger. just make it for comfy again you rust tranny
>>101523738but when its finished every model from diffusers will be supported
>>101523746nobody uses diffusers you dumb cunt
>>101523756i think you're misunderstanding. the models that are already supported in ait are based on diffusers implementations
>>101523674wasn't me you're replying to. I think it's funny that you're still mad I mocked you for your dumb request like 4 months ago.
>>101523828what was anons request
Is it worth trying to gen with an AMD 5800X or should I look into buying Nvidia?
>>101523828what request?
ideal physique
>>101523840>>101523846He wanted me to add track marks to a gens arms and I called him an edgy zoomer, which made him super butt hurt. He replied to almost every post I made for a few weeks sperging out after that lol
>>101523896>every post I made for a few weeks sperging outHuh, reminds me of the guy who squats in this thread kek the one who has a documented history of doing that
>>101523841
>>101523948>>101523841for reference, i have a card about on par with 3080 for throughput and am quite thankful i dont have anything slower. even so for SDXL i use a fast rough draft setting at 8-10s per click, and one or two quality settings at 20s or 50s per click respectively. you have to do a lot of revisions with this shit.
news you can peruse it is weird, i vaguely remember, as a bit of trollishness, asking some1 to add track marks to an arm, but that was like last year, and i remember them getting a bit bent outofshape. desu i find cut scars more aesthetic on a (1)girl, but would probably need a lora idk
electric vomit aesthetic. gonna try to play teeefftoo for a bit instead of this
>>101524137>>101524113Who are you talking to
Nobody seems to like my gens. Maybe because they aren't girls. I won't post anymore :(
>>101523896lesser known schizo anon lore
>>101524333posting for (you)s is never to play
>>101524395What
i miss schizo anon
>101524333Nah everyone simply finds you annoying
>>101524333I like non-1girl content. there's so little of it these days
>>101515162Question for people who upload models to Civitai since I know plenty of you lurk here. Whenever you upload a model, do you go through a sharing the data set you used as well? I want to start doing that myself because I want others to know how to train these things too and want to share the methods and tools I used so that others can do the same. There are surprisingly few of us that know how to do the things we do, so if I ever go through writing up a detailed easy to understand guide, then I want to be able to link to one of my data sets as an example. Are there any down sites to doing this? Civitai has an option to up the data set in the file field next to the downloadable model, and I also have a hugging face page full of past trained LoRAs and their corresponding dstasets (though most aren't public yet). https://huggingface.co/AiAFhttps://civitai.com/user/AI_Art_Factory/models
>>101524558what model
>>101524564>Whenever you upload a model, do you go through a sharing the data set you used as well?everyone wants this, nobody does it. It's always welcome since yes it teaches people what kinds of images make it in and how to tag them properly. It also makes things much easier to use. I hate civit and hugging face for not making it mandatory.Huggingface has it's own dataset viewer so go through the docs to set that up and people won't have to download it if they don't want to.Civit just wants a zipped file and is in the upload process so you can't miss it
the future is to bring the past back to the present
God damn I fried my dopamine receptors real good.It's been quite therapeutic though. Strangely, successfully training loras of old crushes made me get over them.
>>101523529>more asian kamala>ok I'm liking where this is goingTxt2img models have a tendency to have an Asian bias, I suppose it's some training data inbalance
>>101524756Lora?
>>101525157idk if you saw but I accidentally made a counterfeit mint yesterday >>101515093the wildcards randomly rolled all the mint values
See this in the garden, what do?
>mfw Resource news07/21/2024>Batched Niggerlips Samplers for ComfyUIhttps://github.com/wootwootwootwoot/ComfyUI-RK-Sampler>SimpleTuner now supports Kolors training for full tune or LoRA/DoRAhttps://github.com/bghira/SimpleTuner/blob/main/documentation/quickstart/KOLORS.md>What Kamala Harris has said about AI, tech regulation, and morehttps://techcrunch.com/2024/07/21/what-kamala-harris-has-said-about-ai-tech-regulation-and-more/>DynamiCrafter: Animating Open-domain Images with Video Diffusion Priorshttps://github.com/FDoKE/DynamiCrafter>CumUI_EchoMimic: Lifelike Audio-Driven Portrait Animationshttps://github.com/sthemex/ComfyUI_EchoMimic>Two-stage PixArt ensemble of experts (2x 900M)https://huggingface.co/spaces/ptx0/PixArt-900M-EDiffi>florence2-ft-simple: finetune your florence2 modelshttps://github.com/sdbds/florence2-ft-advanced07/20/2024>Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Modelshttps://github.com/open-mmlab/Live2Diff>Meta Chameleon 7B and 34Bhttps://huggingface.co/collections/facebook/chameleon-668da9663f80d483b4c61f5807/19/2024>Follow-Your-Emoji: Freestyle Portrait Animationhttps://github.com/mayuelala/FollowYourEmoji>ComfyUI-Inspyrenet-Rembg [background removal]https://github.com/john-mnz/ComfyUI-Inspyrenet-Rembg>Shape of Motion: 4D Reconstruction from a Single Videohttps://shape-of-motion.github.io/>Surf and RAG-chat with 4 million public stable diffusion promptshttps://github.com/hexive/prompt-cli>UCIP: A Universal Framework for Compressed Image Super-Resolution using Dynamic Prompthttps://github.com/lixinustc/UCIP>DreamStory: Open-Domain Story Visualization by LLM-Guided Multi-Subject Consistent Diffusionhttps://dream-xyz.github.io/dreamstory07/18/2023>Foookkus v2.5.0 Updatehttps://github.com/lllyasviel/Fooocus/releases/tag/v2.5.0>PoomptGen - Image tag model based on Florence 2https://huggingface.co/MiaoshouAI/Florence-2-base-PromptGen
>mfw Research news022/21/2024>FEG: A Memory-Efficient Replay Approach for Continual Learning in Image Classificationhttps://arxiv.org/abs/2407.12375>MDPE: A Multimodal Deception Dataset with Personality and Emotional Characteristicshttps://arxiv.org/abs/2407.12274>Exploring Advanced LLMs with LLMsuitehttps://arxiv.org/abs/2407.12036>Benchmarking the Attribution Quality of Vision Modelshttps://arxiv.org/abs/2407.11910>Improving Unsupervised Video Object Segmentation via Fake Flow Generationhttps://arxiv.org/abs/2407.11714>Distractors-Immune Representation Learning with Cross-modal Contrastive Regularization for Change Captioninghttps://arxiv.org/abs/2407.11683>The Kolmogorov Complexity of Irish traditional dance musichttps://arxiv.org/abs/2407.12000>DreamCatalyst: Fast and High-Quality 3D Editing via Controlling Editability and Identity Preservationhttps://arxiv.org/abs/2407.11394>Visualization Literacy of MLLMs: A Comparative Studyhttps://arxiv.org/abs/2407.10996>Overcoming Catastrophic Forgetting in Federated Class-Incremental Learning via Federated Global Twin Generatorhttps://arxiv.org/abs/2407.11078>Visual Prompt Selection for In-Context Learning Segmentationhttps://arxiv.org/abs/2407.10233>PAFUSE: Part-based Diffusion for 3D Whole-Body Pose Estimationhttps://arxiv.org/abs/2407.10220>A Self-Supervised Learning Pipeline for Demographically Fair Facial Attribute Classificationhttps://arxiv.org/abs/2407.10104>Learning Online Scale Transformation for Talking Head Video Generationhttps://arxiv.org/abs/2407.09965>NamedCurves: Learned Image Enhancement via Color Naminghttps://arxiv.org/abs/2407.09892>Tranny Image Captioning using CNN- CNN architecture and Hierarchical Attentionhttps://arxiv.org/abs/2407.0955611>BITCHmetric Remove-and-Retrain (GOAR): Coordinate-Invariant eXplainable AI Assessmenthttps://arxiv.org/abs/2407.12401>VULVA: Reasoning Video Object Segmentation via LLMshttps://arxiv.org/abs/2407.11325
>>101525348who's that?
>>101525337Didn't you say you wouldn't post anymore?
>>101523294This is for 1.5 right?
PhotoMaker V2 is out!!https://huggingface.co/spaces/TencentARC/PhotoMaker-V2
julien is beans
/sdg/ bros...>>101525545>>101525545>>101525545
pathetic
>>101525186>idk if you saw but I accidentally made a counterfeit mint yesterdaynah, I was gone most of yesterday
Is there a reason I should be using ComfyUI over Forge (A1111 fork)Forge lets me>Quickly start prompting>Quick hires fix>Quick send to inpaint or img2img>Quick upscaling>Lots of plugins>Workflow makes senseIs Comfy really just having to load different workflows constantly?
>>101525912Or edit them but lots more plugins and always gets new shit first
>>101525912forge is no longer maintain so you'll slowly (or not slowly) slide further and further away from the newer tools, workflows, and metas.there's a project called reForge that is attempting to keep upstream a1111 and comfyui changes merged into forge but its a tall order. the real up-and-comer in my opinion in comfyUI_Modern. the guy working on that is making huge leaps forward on making comfyui actually comfy
>>101525912>Is Comfy really just having to load different workflows constantly?noI can do all of the 'quick' things you listed without having to create or load a new workflow. It's probably even quicker since I don't have to tab around pages.The only time you need to alter your workflow is if you deliberately want to change what's happening in your pipeline. It's complexity is really over-exaggerated, there's just a low skill floor and high skill ceiling, so some workflows you see online may seem incredibly complex/convoluted
>>101524318you obv. thanks for replying>>101525393yeah
>>101524949https://files.catbox.moe/ky4zmx.safetensors
is sdxl good for porn yet? I stopped using SD around the time that sdxl was released and I had only used 1.5is there anything I should know if im gonna be getting back into genning? recent news? last time i did anything was probably like september 2023
>>101526099fuck off lazy faggot
>>101526114kys nigger
>>101524949nta, but i use this lora for that effect: https://civitai.com/models/497612/megaman-legends-1-style
>>101525385(you)
>>101526114shut the fuck up thread schizo
>>101526114a repost is a repost, I guess. what did you like about that one specifically?
am I even gonna get to post the news today?
mfw
>>101526984Model?
>>101527047VendoPonyRealistic
>>101527067Thanks
this instagram account makes very cool images with midjourney and i've been having trouble replicating them with stable diffusion. pic related, one of theirs.the chinese low quality camera aesthetic looks cool but I can't prompt anything beyond 1girl, ((huge breasts))https://www.instagram.com/xesearch/
SD3 is unbannedhttps://civitai.com/articles/6221/sd3-unbanned-community-decision-on-its-future-at-civitai
>>101527276how do you get more dynamic movement and poses similar to this? sd cant seem to figure when someone has their head tilted backwards for example, it either adds a new face or thinks its a back of someones head. controlnet somehow makes it worse
>>101525912just use stableswarm nigga
>>101527482Pony Diffusion is much better at following prompts which is based on SDXL. That one is VendoPonyRealistic, a new checkpoint model which is great at creating realistic images.
sd 3....... not for me!!
no one is stable diffusion anymore
>>101528467mondays tend to be either schizo meltdowns or ded. today was extra ded tho
Monday needs extra naps
>>101528780that looks like one of those scifi horror movies where the character is infected with alien snake things under her skin.
>>101528815hm you're rightstill would tho
>>101528553schizo meltdowns might be better, but probably not. I don't know.
>>101525912I dunno I have my comfyUI workflow such that all I have to do is copy and paste the image from the upscale over to the inpainting group and all the prompts and models and such automatically move over, or I can edit them if I want. I also like how in comfy when I inpaint I can zoom all the way into the detail I am checking in the image preview box, whereas with forge if it is something hard to see I have to paste it into GIMP and check it there, or open the image in a new tab and use my browser zoom to get close to the detail. Although I still prefer Forge/A1111 for regional prompting, some controlnet stuff, and basic experimenting with prompts.
I haven't used Stable Diffusion in about 2 years. Is it still really hard to get dynamic poses and perfect hands / feet, even if you're using a controlnet? I used to be really into Stable Diffusion, but I got bored and quit because I couldn't do any advanced poses with good anatomy.
258m and 20malso 3 more opsi sleep
>>101529461Really cool style, anon.
>>101529436controlnets weren't out 2 years ago. you're a phony>>101529461I'm looking forward to the day where you have more than a numbers report
>>101529529>you're a phonyI was a literal off. I dropped it around the beginning of 2013. It's not that serious, anon. It was a rough estimate of when I stopped using Stable Diffusion. >>101529544>still have to roll a couple of times but the gold is there.Damn, I was hoping it would be easier by now.
>>101529670>a literal off.* a little off
>>101529650I'm so excited>>101529670I'm just memeing. welcome back. shit still sucks, but it sucks less
>>101529670>Damn, I was hoping it would be easier by now.It's not so bad when you have the right model for what you are doing. Pony is the best XL model at lewd atm. hands and feet are much better nowadays that is isn't much of a meme anymore (at least for me anyways)
new lora get
I need to just start putting cameltoe in negatives
>>101529737very nice texture. mint anon-esque stuff!
probably won't get to see the 100k from franners. I'm heading to bed. good night anons!
>>101529796gn
imagine if I got to post todays news
>>101529857There are images in this thread I posted 22 hrs ago. Guess it's been a slow day
>>101529886one of the slower days in recent memory
>>101515162Im a fucking casual so im using Easy Diffusion and sometimes it seems like it justs stops progress and gives inaccurate timeframes on genning. I have a 3060 Ti are any of the other Uis better or do they all just kinda do that
>>101530011I had that happen once out of like 15k gens, no idea what caused it.
>>101530011that sometimes happens when it needs to swap models or other memory swaps. can't say for sure what it is though. comfyui is more performant but much harder to use
>>101530010>>101530074hot
have you see this chibi? if so, please contact your local police immediately
these chibis kinda make me wanna do a new gacha chibi series
TEST INPUTMY ADMIN IS INTERFERING AGAIN
>>101530098I like funky dj bear. it was hard to get suno to do vinyl cuts, but I got therehttps://suno.com/song/805fc371-a649-4525-bc6f-a17eb519491e
>>101530287I was baker baited. bakted
For real this time...>>101530296>>101530296>>101530296>>101530276sorry man, subject was missing
>>101530287it gone
>>101530305its all good! ty for the bake
>>101529929
>144wtf
>>101527576catbox?! or model?
[re]filled
>>101530625cool - https://www.youtube.com/watch?v=GBrfxmUJXdA
>>101530634I should do more pixel art but I dunno what the pixel art meta is anymore
what negative embeddings do you guys use, if at all? I just use easy negative but I've seen a lot of people use all sorts of different ones, with some being specifically for improving hand/feet/overall anatomy but seems like a meme to me. >image limit has been reached
>>101530665>>101530296>>101530296>>101530296
>>101530543>>144>wtfbummer, looks like burgerAnon and animAnon got some posts removed.
>>101530650indie game or hit nostalgia with a wallpaper