I made a simple note with a title. The only thing in the body was a URL. I expected myReach to follow that link and summarize the website but it didn’t. Is this a bug or a feature request?
Thanks for your feedback. It depends how you save the link. Ideally, if you’re wanting to save a website, the best is to directly save it as a website (rather than a URL in a note). You can then access the Summary that is automatically created in the Properties Tab of the Node.
However, if you’re wanting to add a URL within a Note, there are two ways:
- If the website is already saved in myReach, add it as a @mention. This will connect the two, and you can go to the website node to see the summary.
- If it’s not in myReach, click “Add” and then “Web URL” from the options in the note. This will add the link within the text of the note. When you click on it, you will have the option to open the URL or create it as a website in myReach.
Hope this helps!
Thanks for that, Sofia but isn’t that a bit counterintuitive? I’ll certainly use that but it seems that the note should be parsed anyway so it could pick up the URL and incorporate that into the summary
It essentially comes down to what you’re wanting to save.
In essence, if it’s a website it makes more sense to save it directly as a website, so all the metadata and content can be extracted and learnt. This way, you’re able to get relevant information and knowledge from the source, ie. by querying the AI Chat.
If it were saved as text within a note, it wouldn’t be able to extract and show metadata from the source. We will look into adding automatic detection of URLs in notes in the Development Plan for future development.
This approach myReach takes also means that imported items have to be edited on order for their embedded URLs to be recognized as web links.
I have over 500 items to import and over 350 of them have URLs embedded in them. They’re all markdown so they’re recognized as URLs in that they are clickable but the link isn’t followed when the note is learned so the actual information isn’t recorded. Further, when I add the URL as an embedded URL, of still isn’t followed during the learning process.
I really wish you’d reconsider your collective position on this.
Thanks for your feedback
Let me add a bit more context to the feed – a node represents a knowledge item, these are split into different node types based on the type of data that knowledge item is made off. For each node type then we apply the most appropriate strategy to extract all relevant information. For a document we parse the content, a website we extract the HTML and clean it up, audio file we transcribe, etc…
Once the content of the node has been extracted we then identify the most relevant relationships and offer them as recommendations. In addition to that for all node types except NOTES we identify from the parsed text the URLs and offer these as relationship recommendations.
If you then click on the relationship recommendations you will see the complete list of URLs that have been identified in the text of the node. You can then chose which ones you want to create as seperate nodes – these would then be automatically connected and have all their data processed and analyzed.
As mentioned earlier this is currently available for all Node types except NOTES – we are however working on adding the functionality.
Will keep you posted when we have it ready for notes
Thank you for the update. Anxiously waiting for the update so I can import my notes.