Has anyone tried JASONLines?

General
1

This looks like a superior way to use JASON:–> JASONL

2

Am I understanding this correctly that this is all valid JSON, just with a convention to use newlines for human readability?

3

Agreed. It is nicely readable, but that isn’t primary purpose of JSON.

I do like the idea of some sort of script that could convert JSON into that format just to preview (or edit I suppose).

4

No - it is faster and doesn’t load the entire code before running it - it does it line by line and skips errors in layout. It looks like a much better way to use JSON syntax

5

why they not show a example with decimal, date time values or binary data.

json is horror if everything can be there or not if you use a rest api.

repeating the columns as key every row is for bigger data exchange impractical.

6

But it will zip efficiently.

As an XML user, I see this as much more approachable than ‘standard’ JSON

7

It depends on the use case.
For further processing, unzipping adds another step.

OT
if possible i like to use Object Serialization and Deserialization.
xojo should provide this essential feature and look how ms c# do it.
the good thing at Serialization/Deserialization is you can save and load(restore) complex hierarchy of objects.
in asp .net you can have a web service input with json or xml and it create/map it to a handy object.

8

Sorry, I had to ask ChatGPT for the differences:

The main difference is structure:

JSON

A single valid JSON document containing one value—often an object or array.

{
  "name": "Alice",
  "age": 30
}

Or multiple records in an array:

[
  {"name": "Alice", "age": 30},
  {"name": "Bob", "age": 25}
]

JSONL / JSON Lines

A text file where each line is a separate, valid JSON value, usually an object.

{"name": "Alice", "age": 30}
{"name": "Bob", "age": 25}
{"name": "Charlie", "age": 35}

Key difference

JSON JSONL
Structure One JSON document One JSON value per line
Multiple records Usually an array Separate lines
Streaming Less convenient Excellent
Common use APIs/config files Logs, datasets, ML pipelines

Rule of thumb: Use JSON for structured documents; use JSONL when you have lots of independent records that you want to process one at a time.

9

I have used JSONL once already (at Codex’s suggestion). I can’t remember the exact case for it (probably serialization of lots of records), but I do remember that it fit the bill for what I needed.

10

Not a fan. I usually pack my data in pure JSON near to TOML.

Regular JSON: (99 bytes)

[
  {"name": "Alice", "age": 30},
  {"name": "Bob", "age": 25},
  {"name": "Charlie", "age": 35}
]

JSONL: (87 bytes)

{"name": "Alice", "age": 30}
{"name": "Bob", "age": 25}
{"name": "Charlie", "age": 35}

Structured JSON: (72 bytes)
I designed the structure, so I know that the first line carries the columns’ titles and the rest are values

[
  ["name", "age"],
  ["Alice", 30],
  ["Bob", 25],
  ["Chalie", 35]
]

Same as above, just one liner packed for computers, not humans: (55 bytes)

[["name","age"],["Alice",30],["Bob",25],["Chalie",35]]