I'm not a Photoshop user, but I was a GIMP user since my Linux years, and recently I've vibe-coded my own alternative. Why?
Easy, I wanted a middle-ground thing. Gimp, especially in Mac, is riddled with bugs, and it's a very heavy tool. The default Apple preview thing is okay for a single rotation or overlaying text, but not for other simple transformations. I wanted something in the middle.
I'm browsing my photo library, see a crooked image, quick uncrook it with a couple of clicks. I'm def not a "pro", so I don't care about 8k x 8k images or similar. Gimp is too heavy for this workflow, and the default mac app too light. I also prefer it to be offline, so I coded my own alternative.
I only got like 10~20% there few years ago coding by hand, even though I spent something like 1.5~2 weeks on it. I basically built the automated organization and visualization part (datetime extraction, adding tags, removing duplicates, thumbnails, etc), but none of the transformation tools. Now, thanks to LLMs, I've been able to build most of the things I wanted and meaningfully more in a couple of days, _inluding_ processing and editing videos, as well as these for images: perspective cropping, normal cropping, contrast, saturation, resizing, format conversion, etc:
https://x.com/francisco_io/status/2109555827304489202
Easy, I wanted a middle-ground thing. Gimp, especially in Mac, is riddled with bugs, and it's a very heavy tool. The default Apple preview thing is okay for a single rotation or overlaying text, but not for other simple transformations. I wanted something in the middle.
I'm browsing my photo library, see a crooked image, quick uncrook it with a couple of clicks. I'm def not a "pro", so I don't care about 8k x 8k images or similar. Gimp is too heavy for this workflow, and the default mac app too light. I also prefer it to be offline, so I coded my own alternative.
I only got like 10~20% there few years ago coding by hand, even though I spent something like 1.5~2 weeks on it. I basically built the automated organization and visualization part (datetime extraction, adding tags, removing duplicates, thumbnails, etc), but none of the transformation tools. Now, thanks to LLMs, I've been able to build most of the things I wanted and meaningfully more in a couple of days, _inluding_ processing and editing videos, as well as these for images: perspective cropping, normal cropping, contrast, saturation, resizing, format conversion, etc:
https://x.com/francisco_io/status/2109555827304489202