- 2D reset canvas button under the left side toolbar. on click prompts a
confirmation to clear the canvas.
- all prompt boxes persist prompts, aspect ratio, system prompt and
reference images for that session until refreshed.
**fixes:**
- moved canvas fit to screen button to inside prompt box
- canvas fit to screen logic works properly and always centered
- canvas doesnt overflow screen and annoyingly scroll anymore
- brush now has an active state and border and shows brush icon when not
active
- hidden eraser button for now since it doesn't work
TODO:
- missing features
- bug fix for expanded modal conflicting with resized size (not going full screen)
- gallery modal now resizable with set minimum size
- set a minimum size for lightbox resize
Model router priority tab in settings:
- drag and drop to update
- function calls on order change
Login refactor:
- fixed login (opens modal on log out)
- fixes bug where login modal opens even before login state is checked
(annoying)
- uses zustand states now
- pricing modal with tiers: free, plus, pro and max
- set up and ready to integrate with stripe
- small bug fix with the library modal buttons being squished
- upgrade button in TopBar is commented out for now
Moved Login to top level.
Minor bug fix later., logout after login in and exiting, log out and the button
won't trigger the modal
---------
Co-authored-by: CrossPr0duct <alicextensor@gmail.com>
- Create 3d model from image works now (accessed via wand button next to
the + asset button).
- Modal pops up, or if you dont have a FAL api set in settings it'll pop
up an action reminder to set it first.
- You can choose image from library (opens gallery select mode) or
upload an image from computer.
- Pressing create enqueues the image to 3d and closes the modal.
Spinning loading toast on the top right says Generating... until it
finishes.
- When done, currently it just pops up in the library and you can drag
it onto the scene yourself.
**Current bugs/missing features:**
1. No 3D object image thumbnail in gallery
2. If you open up the gallery and try add the newly generated 3d object
to the scene, it will be stuck at "loading" text because of a CORS error
(not sure why). Pressing the button to refresh the gallery before
dragging it in seems to fix that though.
- reusable model selector component
- uses zustand with separated states for each tab
Image model list:
- GPT Image 1 (GPT-4o)
- Flux Pro Ultra
- Recraft 3
- Flux.1 Kontext
Video model list:
- Kling 2.0
- Runway
bflatastic
Brandon Thomas
ArEnScandCrossPr0duct
