Feature Requests

I thought of this after I responded to the beta testing email for the knowledge graphs. I was wondering if it would be feasible to create new connections between cards directly from the graph view. Since a common use case for using the graph view is to find connections between cards that you may not notice otherwise, I thought this might be useful. What I was thinking was if you right click on a card in graph view the system would allow you to drag a line between that card and another card that is not connected. Then the cards themselves would be updated with the connections/backlinks.

I would be sold if we have an option to have Mindmap along with summary. A good example is Mapify (formerly chatmind) - https://mapify.so/app/new

Editabile agents crawling all the cards for extracting, connect and relate logically specific topics on user input. A new card is then created with the reasoning within it. Reasonings card can be related to each other and edited/fed with new info.

In the mid-90s Apple Computer had a technology called Hot Sauce / Project X that allowed users to surf the web by "flying through" search results. I think it was only ever released in beta but at the time was a fresh way of navigating the web. In Recall's graph view, although I love the existing functionality, it would be great to fly through it visualising the links, largest concentrations, contexts, etc. There is not much material left about Project X but these give a small glimpse: https://en.wikipedia.org/wiki/HotSauce https://www.facebook.com/groups/AppleMacintoshEnthusiasts/posts/2729057797451785/ https://macintoshgarden.org/apps/apples-hotsauce-technology-project-x

I read this awesome Co-founder guide on the recall docs page: https://docs.getrecall.ai/use-cases/the-co-founder-guide-to-recall I went to try organise my connections but found that I needed to manually edit each connection. Let me bulk edit the knowledge graph!! And bulk edit connections on each card.

MOTIVATION: I'm a recovering metadata nerd. I studied it in grad school. In many ways, connections between nodes are more important than the nodes themselves. Right now, the knowledge graph does an amazing job at node management, but connection management can be improved. If this is done correctly, it will accelerate insight generation and improve user experience. REQUEST: Add an automated connection summary between two or more nodes. This can be placed directly in a connection card or elsewhere. Rather than limiting summaries to just nodes, treat connections as important by giving them summaries as well. As a user, I can see the connections between nodes, but they I am still forced to investigate those connections manually. I propose that the knowledge graph adds treats relationships as first class entities. SOLUTION: Users should be able to interact with connections directly. Just like how a user can click a node, they should be able to click a connection. This will reveal a panel just like the panel for nodes. The difference is that this panel will provide detail on the nature of the connection itself. Users should be able to save and manage connections. This is particularly useful when using the "pathfinder" feature. The summary will compile each connection summary and rapidly overview the relationship between the starting and ending nodes. IMPLEMENTATION: It feels like the knowledge graph is currently designed around RDF triples, which deprioritizes relationship fidelity. I recommend leaning further into the property graph approach. The graph already utilizes this to describe [NODE] <-out/inbound-> [NODE] connections. However, by giving the connections priorities too, users will gain more insight. This can be done by shifting to a [SUBJECT] <-predicate-> [OBJECT] connection view to provide semantic context. Here's the difference spelled out in an example: (Current State) Node Relationship: [Richard Hamming] references-> [Bell Labs] [Bell Labs] <-is referred to by [Richard Hamming] (Future State) Semantic Triple: [Richard Hamming] was employed by-> [Bell Labs] [Bell Labs] employed-> [Richard Hamming]

Zettelkasten notetaking matter is really a hot topic for every notetakers. I think many people more or less want to try it, but eventually fail. Most of the reason are 1 .they feel hard to see the big picture when card increase. They either need to use a different tool (most likely mindmap) other than their notetaking tool to display the map, which is troublesome and hard to persist, or they use the plugin in the notetaking tool to kind of mimic the mindmapping experience, which is not the same, unuseful and often lack important features like dragging/rearranging/making connection they are hard to generate summary/MOC for the cards/section of cards they generate. Which help them to recall their old knowledge and link to the new problem they face. I think for the reason 1, recall if going to generate mindmap function, can seriously consider overcoming this. This will give recall a really competitive advantage in the market when there are not really good enough tool in the market providing this feature. For the reason 2, I think if we can automatically generate the map's MOC, which will be very helpful and also will be a handy job for you as recall has been access to AI (though not yet being able to access to internet) I think this is important because you can see majority of the youtuber talking about this notetaking style are not actually presenting their method pretty clearly, due to the tools they are using are not suitable for the process.

The app needs a separate tab to show all of the entities created (people, places, things, etc) from the Connections feature. I'd really like to be able to see them all in one place and click through them to explore the different connections they have throughout my saved content.

Certain documents within a folder have unique words and phrases which the AI is easy to pick out. It's not good at picking out common phrases between documents. If the user notices this, the user should be able to update connections for the AI to learn off of.