- Andy Sylvester recently wrote a post on organizing information for use based on his experience developing a web app that presents a "list" of outlines in OPML, the OPML files could be stored on other servers. I am going to riff a bit on this topic here. If one were to look back at past tools used to collect and store information and produce knowledge like CommonPlace books and Zettlekasten I see a common denominator that both had an index that provides a structure to the information contained so that it could be found. Recall also that before Google there was Yahoo, which aspired to be the Index for the Web. #
- Of course, the Web grew so fast that Yahoo's original manually curated Index did not scale, thus Google search won and Yahoo is now really a distant memory. However, while one does not see it, there is an index to the web that Google automatically produces, it's just that how we use the index is by searching for words or phrases rather than browsing the index. #
- Search is fast, but what is lost? I think what is lost is the ability to visually see the relationship between entries in the index. As I understand it the Zettlekasten methodology is basically using the index one builds to discover related items that are not obvious. Recent applications like Roam, Logseq, and Obsidian enable Zettlekasten-like ways to "discover" relationships by making it easier for one to link information with each other. #
- I personally use Roam, but if I am honest, I have not "discovered" anything new with it, probably because I don't really "work" within it but instead use it for storage and retrieval. For me the problem with Roam is that while it does outlining it is not an outliner, it is a text editor that does outlining and wiki links. For me, an outline is more akin to an index rather than a document. For example, look at this outline of the technology that I use, it's a list of top level items. If you want to "discover" what tech I use every day, click Every Day Tech to expand the list of tech items I most use. When you click the iPad Mini 6 you see some details and links to more information about the iPad Mini 6 and you know that the iPad Mini 6 is one of the tech items I use every day.#
- The problem with this page is that the top level list makes most sense to me, but it might not make sense to you. What if you could create a top level index that made most sense to you? Perhaps you want a list of Apple Products so you make that the top level node of our index, but what if you could then simply include my iPad Mini 6 node into your index? Imagine right clicking the node and selecting "Copy outline node", and then imagine going back to your outline and selecting "Paste outline node" that automatically created an include of that node of the OPML file I manage. #
- The use case I describe above requires something deeper than OPML includes, it requires OPML node includes, which I don't think currently exist. The benefit though is that whenever I updated that piece of information in my file you would see that update with no real effort on your part. What if you could atomize other snippets of information (a paragraph or two) that transformed into an outline node and updated live...could this be done by "copying" the RSS behind that snippet, and "pasting" that RSS snippet into an OPML node? (Of course, that requires each paragraph on a page to be a separate RSS item.) #
- I would love it if something like this could be made, I think in fact that tools like Hypothesis come close, but coming back to my main point, I think there is value in indexes that one actually sees and interacts with. A few years ago I built a wiki and I found a plugin that automatically creates a page that is an index of all the pages in my wiki in alphabetical order. My wiki site index is like the index you see in the back of some books, like text books. While I can search for pages in my wiki, I often just go to the site index page and quickly scroll down the page to find the title of topic I want to open. The site index page would not exist if it weren't automatically produced because it would take too my effort to manually maintain it. I think Andy's app shows a way to publish an index in outline form that one not only sees as an index, and it's relationship, but can also interact with as an outline. The "magic" would be in the automation and simplification of using something like it for one to make what ever index they prefer. #
- Andy Sylvester recently wrote a post on organizing information for use based on his experience developing a web app that presents a "list" of outlines in OPML, the OPML files could be stored on other servers. I am going to riff a bit on this topic here. If one were to look back at past tools used to collect and store information and produce knowledge like CommonPlace books and Zettlekasten I see a common denominator that both had an index that provides a structure to the information contained so that it could be found. Recall also that before Google there was Yahoo, which aspired to be the Index for the Web. #
- Of course, the Web grew so fast that Yahoo's original manually curated Index did not scale, thus Google search won and Yahoo is now really a distant memory. However, while one does not see it, there is an index to the web that Google automatically produces, it's just that how we use the index is by searching for words or phrases rather than browsing the index. #
- Search is fast, but what is lost? I think what is lost is the ability to visually see the relationship between entries in the index. As I understand it the Zettlekasten methodology is basically using the index one builds to discover related items that are not obvious. Recent applications like Roam, Logseq, and Obsidian enable Zettlekasten-like ways to "discover" relationships by making it easier for one to link information with each other. #
- I personally use Roam, but if I am honest, I have not "discovered" anything new with it, probably because I don't really "work" within it but instead use it for storage and retrieval. For me the problem with Roam is that while it does outlining it is not an outliner, it is a text editor that does outlining and wiki links. For me, an outline is more akin to an index rather than a document. For example, look at this outline of the technology that I use, it's a list of top level items. If you want to "discover" what tech I use every day, click Every Day Tech to expand the list of tech items I most use. When you click the iPad Mini 6 you see some details and links to more information about the iPad Mini 6 and you know that the iPad Mini 6 is one of the tech items I use every day.#
- The problem with this page is that the top level list makes most sense to me, but it might not make sense to you. What if you could create a top level index that made most sense to you? Perhaps you want a list of Apple Products so you make that the top level node of our index, but what if you could then simply include my iPad Mini 6 node into your index? Imagine right clicking the node and selecting "Copy outline node", and then imagine going back to your outline and selecting "Paste outline node" that automatically created an include of that node of the OPML file I manage. #
- The use case I describe above requires something deeper than OPML includes, it requires OPML node includes, which I don't think currently exist. The benefit though is that whenever I updated that piece of information in my file you would see that update with no real effort on your part. What if you could atomize other snippets of information (a paragraph or two) that transformed into an outline node and updated live...could this be done by "copying" the RSS behind that snippet, and "pasting" that RSS snippet into an OPML node? (Of course, that requires each paragraph on a page to be a separate RSS item.) #
- I would love it if something like this could be made, I think in fact that tools like Hypothesis come close, but coming back to my main point, I think there is value in indexes that one actually sees and interacts with. A few years ago I built a wiki and I found a plugin that automatically creates a page that is an index of all the pages in my wiki in alphabetical order. My wiki site index is like the index you see in the back of some books, like text books. While I can search for pages in my wiki, I often just go to the site index page and quickly scroll down the page to find the title of topic I want to open. The site index page would not exist if it weren't automatically produced because it would take too my effort to manually maintain it. I think Andy's app shows a way to publish an index in outline form that one not only sees as an index, and it's relationship, but can also interact with as an outline. The "magic" would be in the automation and simplification of using something like it for one to make what ever index they prefer. #