0.0.24

Fixes to annotation indexin

0.0.23

• Fixes and improvements to indexing and initial setup

v0.0.10 - Improve Onboarding Flow with download screen

• new download screen for the time it downloads the embedding model

v0.0.9 - More robust indexing logic

• Improvements to the indexing robustness
• Potential fix for windows bug

v0.0.8 - Greatly improve onboarding process

• Apple-signed application
• Show notification when app is connected and installed

v0.0.7 - Mac code signing & fixes to detection of substack feeds in imports

add code signing for mac

v0.0.6 - Better queuing of feeds-to-index

Add better queuing of different feeds where indexing started at the same time

v0.0.5

• Adding local embedding and vector database for new Rabbit Hole feature

A few notes:

• Downloads a 130mb text classification model after install
• Uses somewhere between 300-500mb ram (will be optimised soon)

v0.0.4 Better error tracking and (pot. fix of server shutdown)

We had a couple of reports of the server not being responsive after a while (if that happens to you still then try to restart it)
Rest assured that pages/annotations you made are buffered in the extension to be sent off when the extension can find the server again.

v0.0.3 First publicly usable version

Just start up the file, and a small "m" icon will appear in your menu tray.
It won't have a dock icon. It just runs in the background.