Show HN: Every great read I've come across, compiled into a knowledge graph | Hacker News Reader

Show HN: Every great read I've come across, compiled into a knowledge graph

prashantbarahi.com.np

153 pointsrpac039 comments

As a sort of digital "happening," it's...fun? Interesting? "Neat?"

As something I will ever reference, it's basically useless. Which, I feel bad saying, because that suggests that I'm somehow considered in any way in this graph. But, the reason I mention it is because this and the rest of the "here's all my xyz knowledge/notes laid out for you to peruse" graphs and digital gardens are just bad ways of engaging with someone else's ideas.

The idea is there: start anywhere! Follow the meandering path! But, the experience is more: "well, I guess I need to start at the beginning again" (digital gardens) and "I guess I'll just pick a random node cuz they're all supposed to be equally interesting?" (the OP)

I’ve always felt like the graph view was useful to me, as the creator, but anyone consuming a zettel / knowledge base / digital garden / whatever people are calling it these days - you need a more accessible index formatted specifically for discovery before they are useful.

I think the most useful data visualization is... a timeline. Why? Because life is also a timeline. There can be other organization and visualisations based on that but i think that should be the central way to visualize data.

This takes me back to the importance of "mining the archive" before you start writing about anything. But I not sure if this does it or not. It is always interesting to me how much knowledge we have, and how much effort people have to put in order to acquire good understanding of something. This is still unsolved problem for me.

How does one use this? How is it useful?

One looks in awe at all the nodes and makes a mental note one should also spent more time reading books instead of watching TV or phone mindlessly. After, one closes again because these graphs are not easy to travel in any meaningful way. Especially on one's phone, where likely opened.

That having said. Thanks author for taking the time and trying.

thank you so much for sharing! i think knowledge graphs, knowledge bases and so-called "digital gardens" - compilations of knowledge like these are incredibly powerful in sharing masses of knowledge and information with each other in an async way, and provide a great value to ourselves in solidifying that wisdom in our own brain. bookmarked!

(now my mind is just wandering, but some way to "aggregate" knowledge bases... maybe kind of like RSS but with full-text search and more hierarchical, with tags and multiple layers of navigation... that would be cool!)

I was thinking of making a time series of the bookmarks but yeah hierarchical navigation and FTS sounds infinitely more useful. I'll seek into integrating it into the Algolia index. Appreciate the feedback. Thank you.

Out of curiosity: "great reads" sounds kinda like a proper name / title (it's a plural being used as a singular), are these from lists-of-"great reads" / a layout for some datasource that has "great reads", or just a (large!) collection of things you've found over time?

Or am I just over-thinking that "s"?

Either way, this looks great. Randomly clicking a few entries finds stuff I wish I had found years earlier. Thanks a ton for sharing! I'll have to poke around more :)

I think it is just a typo.

> just a (large!) collection of things you've found over time

This one is to the point. So, the collection was exported from my Pocket account. Whenever I find any interesting blogs, I save it there (with tags so that I can find them easily). The oldest entry dates back to 2017 and over time, I have accumulated about 700-ish articles.

Here's the backstory: https://twitter.com/nottherealpac/status/1545803998854205441

Hope this clarifies :)

I think I've got a couple thousand things in Pocket and Pinboard, though probably less organized and filters than yours seems to be... turning that into a graph for sharing makes quite a lot of sense, I like that idea. Relatively low effort too.

Anyway, thanks again! This looks like a lot of good stuff, it's clearly years of effort and I'm very glad you shared.

I’d say it’s a kind of idiomatic phrase (?) — “great reads” where a “read” is a “thing to read” (like one could use “read” in “comfort read”). In other words:

language

"That book was a really fun read."

"That author is always a good read."

https://www.merriam-webster.com/dictionary/read

> read (noun)

> something (such as a book) that is read

Who doesn't love a graph? I do. As you commented elsewhere in this post, _"Actually, this is more for others than it is for me. I wanted to share my Pocket list with the world..."_ Essentially, this is an _information illustration_ of the cosmos of ideas inside your head. haha. That's cool.

How is it useful?

On the one hand, manually tagged data is the first phase of making data useful in many knowledge systems. On the other hand, algorithmically scraping resources and creating a graph from that data is going to be bananas complicated.

Here's what I'm curious about. Can your curated graph be used to pick out same clusters from a superset of links which include the second graph?

Sure, what I'm describing is a search problem. And it's about words whether they're in a string, vector, or graph structure. And that's all very interesting in it's own right. What's also interesting is a fuzzy search that's fuzzy in a graphy way. Is that a thing? I don't know, but I want to know.

I really want a tool like this to visualize every router, switch, server, VM, and application in my company.

The UI isn't quite as pretty, but icinga / nagios has that feature.

I'm sorry, but I just don't get it. You have a graph of things, but no embedded knowledge. Why are things in the graph, what is the value of the connection between any 2 nodes? There's no annotative trail here at all.

I tried the search, and found none of the following terms:

markup, annotation, micrometer, logarithm, memex, forth, lisp, thread, lathe, precision

I see "good blogs" and there's no hint as to why they are good, etc.

Looks cool and fancy. I would be grateful to see a simple list also, since it’s hard to use it on mobile. “Blogs” tab is small compared to this graph.

Thanks for the feedback. I will work on list and FTS soon :).

And ratings 1-10 and approx time to read,

And ways to sort: best first, and rating divided by time to read?

A graph implies a sort of hierarchy or cycles or at least some kind of order. I doubt, that there is one specific arrangement of the reads in a graph, that will lead to always finding what one is looking for. I think, that freely tagging things might work better.

Nice, thanks.

I like the preview of the individual articles, it would be good if something similar happened when clicking on a tag node. For my taste that would be just a straightforward list of articles tagged in that way but I can imagine there might be other ways.

Im actually more interested in your "Algorithms" tab of your blog.

I had not thought of publicly documenting samples of algorithms I have used in the past. I have an internal library of one-liners and bash scripts but I should publish these

Glad you found it useful :)

> I have an internal library of one-liners and bash scripts but I should publish these

That's a great idea. Bash scripts are sort-of like regex to me (I have to recall them whenever I need them and then I forget them later) and bash scripts samples can really be useful. Looking forward to your Show HN.

Is there a list of tags? I'm not sure I'm seeing it properly: the graph is just a galaxy of tiny dots, to see the text I have to hover the mouse over them.

Thanks for sharing!

Quick question: How did organizing data in this format help you?

Actually, this is more for others than it is for me. I wanted to share my Pocket list with the world, without turning it into a super long <li></li>s and wanted to make it easier to visualize the categories/relatedness of articles (hey this one is "web" and "ui" related; this one looks like its "web" and "performance" related).

Looks really cool. Wish most programs that try to give some crappy image of a graph could use this instead. A lot of potential for a new kind of UI built around this too

What software do you use to visualize the graph like this?

The "i" icon in the bottom left has that info:

So, here is a D3.js network graph that represents all of the wonderful reads I've ever found on the internet. The larger nodes are the tags, while the smaller ones are the links to the article.

Is there a way to download this graph?

Nice graph, have you considered making it a directed graph, and also assigning more explicit semantic meaning to the edges?

So for example, using turtle syntax [1], instead of

<https://engineering.zalando.com/posts/2022/04/functional-tes...> <http://example.com/graph-edge> <https://www.testcontainers.org/>

have

<https://engineering.zalando.com/posts/2022/04/functional-tes...> <http://purl.org/dc/terms/subject> <https://www.testcontainers.org/>

The semantics of http://purl.org/dc/terms/subject is given at the url itself, but in brief:

> A topic of the resource.

> Recommended practice is to refer to the subject with a URI. If this is not possible or feasible, a literal value that identifies the subject may be provided. Both should preferably refer to a subject in a controlled vocabulary.

This would be similar to how wikidata expresses knowledge [2]:

<http://www.wikidata.org/entity/Q28315661> <http://www.wikidata.org/prop/direct/P921> <http://www.wikidata.org/entity/Q750997>

Or in English:

"Go To Statement Considered Harmful"(Q28315661)'s "main subject"(P921) is "goto"(Q750997)

This also makes it easier to query [4], for example, you could get all articles covering a "goto" with the following SPARQL[5] query:

SELECT ?item WHERE { ?item <http://www.wikidata.org/prop/direct/P921> <http://www.wikidata.org/entity/Q750997> }

May help to read the RDF primer [3] also.

[1]: https://www.w3.org/TR/turtle/

[2]: https://www.wikidata.org/wiki/Q28315661

[3]: https://www.w3.org/TR/rdf11-primer/

[4]: https://w.wiki/5RW2

[5]: https://docs.stardog.com/tutorials/learn-sparql

Thank you. Honestly, working with GraphDB and SPARQL inspired me to create this. I did consider if I could create a "real" knowledge graph and even went as far as searching for in-memory graphdb that lives in client browser (maybe one that's built on top of IndexDB. I thought, hey if in-mem RDBMS like H2 exists, is there an in-mem GraphDB available? :D) so that I can query it using SPARQL, but couldn't find anything on it. I wanted to do this without any infrastructure while keeping the bundle sizes low but yes, the way you explained is how it should have actually been done.

I recommend implementing this in a 3D WebXR AR/VR experience for immersive navigation or look at your data through Flow Immersive. Seeing your data in free space arround you is a great way to gain insights.

Looks daunting. Perhaps you could visualize it as a rabbit hole.

But... isn't it just a tags cloud expanded?

it look great but I couldn't find a useful way to use it. A simple list with some tags or tree will probably be better

What does the color of each node mean?

Reply on news.ycombinator.com