> The docs are more than weak, for something that's supposed to take over the world in 2 years I can't even click half the links in the FAQ?
On my goodness. Thank you. Fixed. I have a script that checks for broken links "npx linkinator -r https://treenotation.org" but I must have screwed something up because those were not getting detected. (Nit: 10 years! not 2 ;) )
> You can make any format "no errors" with little effort, the problem is not creating garbage data in the process.
This is a very valid point. But Tree Notation does give you a lot of meaning -- lines/words/scopes -- without knowing the semantics of the Tree Language. But you do have a good argument here that if people don't come to a consensus on even these things, then Tree Notation could just be garbage. There will need to develop a common language for the universal tree notation semantics.
> And there are infinitely many representations which can take this form.
Good point.
> The question isn't "how can generic encoding work" it's "what does Tree Notations syntax do that others don't already".
Tree Notation is the simplest method we can come up with to do all the things are existing syntaxes and languages can do. By simplest, we mean "tree notation will always have fewer or equal pieces in terms of absolute complexity than a different notation for any given task". And our bet is that simplicity of atomic ingredients triumphs in the long run because less costs less.
But the proof is in the pudding. Tree Notation won't really start taking off (if ever) until there is better tooling and some "killer app" new tree languages.
> But they aren't eliminated at all they are just hidden to produce garbage later on instead.
Clinicians are now using Tree Notation to write their medical records on pen and paper (peer reviewed paper coming soon). There's no such thing as syntax errors. Their "code" looks just like S-expressions, with prettified formatting, except without paranetheses. They can have semantic errors using either one (for instance, spelling "dose as does"). But they cannot have syntax errors with Tree Notation. Imagine instead we were instructing them to use S-Expressions--suddenly their notebooks would be filled with syntax errors for missing paranetheses. They would have 0 benefit and all downside by adding parens.
> If not why blow smoke up everyone's ass instead of just saying "Tree Notation is a S-expression format that uses whitespace instead of parenthesis"?
I'm putting a small fortune and an immense amount of effort into this (and giving everything away), because I believe it will eliminate drastic amounts of unnecessary complexity in so many domains all throughout the world, and will indeed, amongst other things, help us cure cancer, which is certainly near and dear to my heart and I'm sure to so many other people on here. I want to make the biggest and boldest statements I can here, so I can get harsh and sharp critical feedback like yours, that is very helpful and can save us time by pointing out flaws (both in theory and practice like the broken links). If this were something for Lispers, I wouldn't care so much (I love lisp, but it would be an incremental improvement). This is for all the people of the world. We are going to simplify and make better code in all domains. It's a mammoth effort that's going to take billions and thousands of smart people. This is just the beginning.