3 ms·

Disclosure: I work on Tree Notation. It’s the future of file formats, IMO. The idea is to have 2 levels: a simple, minimal syntax/notation (think binary) calle

by breck 7y ago

Disclosure: I work on Tree Notation. It’s the future of file formats, IMO. The idea is to have 2 levels: a simple, minimal syntax/notation (think binary) called Tree Notation, and then have higher level grammars on top of that, called tree languages. It works for encoding data and also for programming languages, regardless of paradigm. https://github.com/treenotation/jtree https://github.com/treenotation/jtree

philwelch 7y ago
This sounds a lot like s-expressions.
- breck 7y ago
  Thinking of it as s-expressions without parens is a very good analogy. That ends up making a huge difference.
atombender 7y ago
Your project looks interesting, but I looked through the Github project and site, but I couldn't find a language specification, reference manual, BNF-like grammar, or anything to indicate what the syntax is, beyond a very trivial example in the Github. To be blunt, I think you need to start with a spec to get any traction. That allows people to understand your data model and particular text encoding of it. If they like it, they might use your tool and perhaps port the system to other languages.
- breck 7y ago
  Good feedback, thank you. You may have seen the spec, it’s just quite minimal (here’s a more elaborate one: https://github.com/treenotation/jtree/blob/master/spec.txt https://github.com/treenotation/jtree/blob/master/spec.txt) Here’s a BNf: https://github.com/treenotation/jtree/issues/1 https://github.com/treenotation/jtree/issues/1 There’s a FAQ as well. Docs needs work, in particular I’m hoping people will create their own explanations of the ideas in external places, as that might be a better way to understand it. Happy to provide help to anyone that is interested in that.
svnpenn 7y ago
wow that is awful
- breck 7y ago
  It’s terrible! A waste of time! No one should use it! It’s definitely not the future!
- dang 7y ago
  "Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something." https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html Edit: we've had to ask you multiple times already not to be a jerk on HN. Would you please review the guidelines and take the spirit of this site more to heart?
majewsky 7y ago
From a quick glance, my issue with Tree Notation would be that it's not enough syntax, i.e. it does not provide enough structure for me to grasp the overall structure with a cursory glance. Maybe it would work better if GitHub had a syntax highlighting for it. (But requiring syntax highlighting to be readable is a large red flag on its own.) Or that's just a feeling that would be mitigated if I saw larger files.
- breck 7y ago
  GitHub syntax highlighting is coming. If anyone wants to help with that that would be awesome. One of the top outstanding issues. Syntax highlighting is there for sublime and codemirror but one bug left in those implementations and would love help getting syntax highlighting generation going for Monaco and linguist so we’d have it everywhere on GitHub. Your impression is correct though, without highlighting it’s awful. Try the Tree Language designer app to see it with highlighting, type checking, autocomplete etc...there are interesting ways to accomplish everything without syntax, often not obvious. But tooling is essential to make it better than existing options. Help wanted!
h1d 7y ago
Reminds me of Stylus over CSS.

by breck 7y ago

philwelch 7y ago

This sounds a lot like s-expressions.

breck 7y ago
Thinking of it as s-expressions without parens is a very good analogy. That ends up making a huge difference.

atombender 7y ago

Your project looks interesting, but I looked through the Github project and site, but I couldn't find a language specification, reference manual, BNF-like grammar, or anything to indicate what the syntax is, beyond a very trivial example in the Github. To be blunt, I think you need to start with a spec to get any traction. That allows people to understand your data model and particular text encoding of it. If they like it, they might use your tool and perhaps port the system to other languages.

breck 7y ago
Good feedback, thank you. You may have seen the spec, it’s just quite minimal (here’s a more elaborate one: https://github.com/treenotation/jtree/blob/master/spec.txt https://github.com/treenotation/jtree/blob/master/spec.txt) Here’s a BNf: https://github.com/treenotation/jtree/issues/1 https://github.com/treenotation/jtree/issues/1 There’s a FAQ as well. Docs needs work, in particular I’m hoping people will create their own explanations of the ideas in external places, as that might be a better way to understand it. Happy to provide help to anyone that is interested in that.

svnpenn 7y ago

wow that is awful

breck 7y ago
It’s terrible! A waste of time! No one should use it! It’s definitely not the future!
dang 7y ago
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something." https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html Edit: we've had to ask you multiple times already not to be a jerk on HN. Would you please review the guidelines and take the spirit of this site more to heart?

majewsky 7y ago

From a quick glance, my issue with Tree Notation would be that it's not enough syntax, i.e. it does not provide enough structure for me to grasp the overall structure with a cursory glance. Maybe it would work better if GitHub had a syntax highlighting for it. (But requiring syntax highlighting to be readable is a large red flag on its own.) Or that's just a feeling that would be mitigated if I saw larger files.

breck 7y ago
GitHub syntax highlighting is coming. If anyone wants to help with that that would be awesome. One of the top outstanding issues. Syntax highlighting is there for sublime and codemirror but one bug left in those implementations and would love help getting syntax highlighting generation going for Monaco and linguist so we’d have it everywhere on GitHub. Your impression is correct though, without highlighting it’s awful. Try the Tree Language designer app to see it with highlighting, type checking, autocomplete etc...there are interesting ways to accomplish everything without syntax, often not obvious. But tooling is essential to make it better than existing options. Help wanted!

h1d 7y ago

Reminds me of Stylus over CSS.