Show HN: SFOR – minimal, no-backtracking, typed data format (experimental)

I have been working on SFOR (Streamable Flat Object Representation) — a structured text format I built after finding that existing formats did not quite fit my needs..

- *Streamable:* Can be parsed as it arrives, without needing the entire payload in memory. - *No-backtracking:* The parser never has to "rewind" to re-interpret earlier data. - *Type-explicit:* Types are clearly defined in the stream, avoiding ambiguity. - *Minimal overhead:* Compact representation without sacrificing readability in its intermediate form.

The goal is to handle cases where formats like JSON or YAML work, but a streamable, predictable, and low-memory alternative is better suited — especially for constrained devices, large datasets, or real-time processing.

I would appreciate your thoughts on: - Potential edge cases I haven not considered. - Comparisons to similar formats you have used. - Whether the "no required backtracking" claim holds up in your experience with other formats.

*Repo:* - [SFOR Repo](https://github.com/brucekaushik/sfor)