Open drinehart1 opened 2 months ago
That is a good idea. Maybe we can put all the JSON in firestore, or maybe that postgresql JSON database you mentioned. I had to increase the MB of form data to 10MB when a user hits 'Save' in Neuroglancer. And that is just a hack.
According ref (https://github.com/google/neuroglancer/issues/526), the native json can handle around 10K annotations. After that there may be instability. As we craft a better approach to annotation CRUD, we may wish to focus on segregating the Neuroglancer core data from the data.
This would allow better tracking of events as well