Hello everybody, hello Bubble team!
While trying to track user data, I noticed that the Bubble Rich Text Editor sometimes fails to update the base64 encoded image with the one that has been stored in the s3 amazon service. It looks like it fails client-side when the user interrupts the whole auto-save process (while using the auto-bind feature, I haven’t tried it without the auto-bind). Larger image are more prone to this since it takes more time to upload them and then if the user would exit or navigate elsewhere then the image source isn’t updated with the s3.amazon link, thus the base64 encoded string stays. However, the image transfer is successful but it remains unlinked to anything really and just takes up space.
Editor example (it is in tooltip mode and thus you can’t see the icons):
I copied and pasted some sample images as shown below.
Here you can see from the console, that the larger image didn’t sync it’s source because of client-sided interruption.
Here is a view from the File Manager. The image was uploaded.
This is a minor bug of the Bubble Rich Text Editor. If the base64 image is replaced server side, this would probably solve the issue.
It is a problem because it creates untracked images. I can regex and find those base64 images but I don’t have their respective s3 link. So I can’t replace the base64 images them manually, nor can I delete the s3 file, since I also can’t “Do a search for>image” to find unused images with a backend workflow.
A solution or bug-fix would be awesome!