4 ms·

>In fact, the whole programmer's fetish with writing everything with a plain terminal text editor needs to die. Markdown, YAML, JSON, INI files. It's the year 2

by vegai_ 5y ago

>In fact, the whole programmer's fetish with writing everything with a plain terminal text editor needs to die. Markdown, YAML, JSON, INI files. It's the year 2021 for christ's sake, and we're composing documents with tools designed for green CRTs! Alternative take for this same sentiment: It's the year 2021, why has nobody figured out better ways to present knowledge than simple text files?

Tagbert 5y ago
Define “better”. There are a plethora of alternate ways to represent knowledge. Each has its own set of trade-offs. Which do you prioritize? Ease of editing Interoperability/readability in multiple systems Expressive styling Optimized for configurations Optimized for readable text Etcetera Fill out that list, stack rank it and tell us which format is better.
- vegai_ 5y ago
  Better technology works like this: some people pioneer in it and most people are skeptical initially. Eventually, almost everyone transfer to using the better technology, and will wonder how they ever coped with the older technology. I know, we don't get many of these in this field anymore. I think this means something.
throwaway984393 5y ago
There are better ways [1] [2], there's just no paying customers. You can express knowledge using a Semantic Web Stack[3], encode data in N-Triple/JSON-LD/RDF/XML and query it with SPARQL or Turtle. Wikidata[4] is an example of this. But (afaik) nobody's selling a product that builds and queries a large semantic knowledge base for consumers. Some industries have Semantic Web-based solutions (i'm not aware of commercial ones but I assume they exist), but nobody has yet explained to the consumer why or how they should build ontologies or keep data in triples. The terminology itself is probably a barrier to market, and all the interfaces are (imo) clunky. If I were to build a consumer product, I would hide all the technical jibberish and make a slick UI that walks the user through linking semantic data, leaning heavily on Wikidata to fill in gaps. Gameify it so that people can compete for points to link more data, add a subscription fee, add social components: poof! Billion dollar personal-knowledge-base app to replace Evernote. Value add/tagline: "Know yourself better." (inb4 selling the free version's users' anonymized data to advertisers...) [1] https://www.wikidata.org/wiki/Lexeme:L302468 https://www.wikidata.org/wiki/Lexeme:L302468 [2] https://www.wikidata.org/wiki/Q90889591 https://www.wikidata.org/wiki/Q90889591 [3] https://en.wikipedia.org/wiki/Semantic_Web_Stack https://en.wikipedia.org/wiki/Semantic_Web_Stack [4] https://medium.com/@swaapniil/a-guide-to-using-sparql-queries-playing-with-wikidata-cc5e2f6a5ffc https://medium.com/@swaapniil/a-guide-to-using-sparql-querie...
- vegai_ 5y ago
  I claim that technology that nobody uses is not better. Also see my other comment: https://news.ycombinator.com/item?id=28902738 https://news.ycombinator.com/item?id=28902738

by vegai_ 5y ago

Tagbert 5y ago

Define “better”. There are a plethora of alternate ways to represent knowledge. Each has its own set of trade-offs. Which do you prioritize? Ease of editing Interoperability/readability in multiple systems Expressive styling Optimized for configurations Optimized for readable text Etcetera Fill out that list, stack rank it and tell us which format is better.

vegai_ 5y ago
Better technology works like this: some people pioneer in it and most people are skeptical initially. Eventually, almost everyone transfer to using the better technology, and will wonder how they ever coped with the older technology. I know, we don't get many of these in this field anymore. I think this means something.

throwaway984393 5y ago

There are better ways [1] [2], there's just no paying customers. You can express knowledge using a Semantic Web Stack[3], encode data in N-Triple/JSON-LD/RDF/XML and query it with SPARQL or Turtle. Wikidata[4] is an example of this. But (afaik) nobody's selling a product that builds and queries a large semantic knowledge base for consumers. Some industries have Semantic Web-based solutions (i'm not aware of commercial ones but I assume they exist), but nobody has yet explained to the consumer why or how they should build ontologies or keep data in triples. The terminology itself is probably a barrier to market, and all the interfaces are (imo) clunky. If I were to build a consumer product, I would hide all the technical jibberish and make a slick UI that walks the user through linking semantic data, leaning heavily on Wikidata to fill in gaps. Gameify it so that people can compete for points to link more data, add a subscription fee, add social components: poof! Billion dollar personal-knowledge-base app to replace Evernote. Value add/tagline: "Know yourself better." (inb4 selling the free version's users' anonymized data to advertisers...) [1] https://www.wikidata.org/wiki/Lexeme:L302468 https://www.wikidata.org/wiki/Lexeme:L302468 [2] https://www.wikidata.org/wiki/Q90889591 https://www.wikidata.org/wiki/Q90889591 [3] https://en.wikipedia.org/wiki/Semantic_Web_Stack https://en.wikipedia.org/wiki/Semantic_Web_Stack [4] https://medium.com/@swaapniil/a-guide-to-using-sparql-queries-playing-with-wikidata-cc5e2f6a5ffc https://medium.com/@swaapniil/a-guide-to-using-sparql-querie...

vegai_ 5y ago
I claim that technology that nobody uses is not better. Also see my other comment: https://news.ycombinator.com/item?id=28902738 https://news.ycombinator.com/item?id=28902738