What I'd like to have happen is for imported images, instead of being embedded as base64, to be queued to be committed as new files in the same repository. This would result in ideal storage of the images (i.e. no bloating of the tiddlywiki HTML file), and trivial loading from the corresponding gitlab pages URL.
Has anyone implemented this? If not, I may try to find time to do so.
Thanks,
- Chas