Marcin: /* Canonical Takeaway */

2026-01-21T07:03:32Z

Canonical Takeaway

← Older revision		Revision as of 07:03, 21 January 2026
Line 39:		Line 39:
	* Spec → CAD geometry is deterministic		* Spec → CAD geometry is deterministic
	* CAD → BOM and drawings are derivable		* CAD → BOM and drawings are derivable
	* Expertise lives in rules, not in individual drafters		* '''Expertise lives in rules, not in individual drafters'''

Marcin: Created page with "https://chatgpt.com/share/69707955-7ca4-8010-850a-ca6cef5201ee A syntactic schema describes the shape of the data (fields, types, formatting). A semantic schema describes the meaning of the data (roles, intent, constraints, and invariants). Syntactic schemas can validate that an input is well-formed; semantic schemas can be compiled into geometry, BOMs, drawings, and build instructions because they encode what the design is supposed to do, not just how it is written. =..."

2026-01-21T07:02:53Z

Created page with "https://chatgpt.com/share/69707955-7ca4-8010-850a-ca6cef5201ee A syntactic schema describes the shape of the data (fields, types, formatting). A semantic schema describes the meaning of the data (roles, intent, constraints, and invariants). Syntactic schemas can validate that an input is well-formed; semantic schemas can be compiled into geometry, BOMs, drawings, and build instructions because they encode what the design is supposed to do, not just how it is written. =..."

New page

https://chatgpt.com/share/69707955-7ca4-8010-850a-ca6cef5201ee

A syntactic schema describes the shape of the data (fields, types, formatting). A semantic schema describes the meaning of the data (roles, intent, constraints, and invariants). Syntactic schemas can validate that an input is well-formed; semantic schemas can be compiled into geometry, BOMs, drawings, and build instructions because they encode what the design is supposed to do, not just how it is written.

= Takeaway: Semantic vs Syntactic Schemas =

== Definitions ==
* '''Syntactic schema''' = validates ''form'' (field names, data types, required keys, nesting).
** Answers: "Is this spec well-formed?"
** Output: structured data that is ''parsable''.

* '''Semantic schema''' = encodes ''meaning'' (roles, intent, constraints, invariants, relationships).
** Answers: "What is this thing, what rules must it satisfy, and what must change when requirements change?"
** Output: structured data that is ''compilable'' into geometry, BOM, drawings, and build logic.

== Practical Distinction ==
A syntactic schema can say:
* "There is a stud at x=24 inches."

A semantic schema can say:
* "This is a wall module with a 24-inch OC stud system and a window rough opening."

From semantic intent, the compiler can derive:
* where studs must exist
* which studs must be removed for openings
* where kings/jacks/headers/cripples must be inserted
* how sheathing must be panelized based on sourcing rules

== Litmus Test ==
If you change a high-level requirement (ex: add a window, change height, change stud spacing) and the system automatically knows what geometry must change, the schema is '''semantic'''.

If the schema merely records explicit geometry facts and breaks under requirement changes, it is '''syntactic'''.

== Canonical Takeaway ==
'''Syntax''' ensures the spec is well-formed.
'''Semantics''' ensures the spec is meaningful—and therefore compilable.

OSE uses semantic schemas so that:
* Spec → CAD geometry is deterministic
* CAD → BOM and drawings are derivable
* Expertise lives in rules, not in individual drafters

Semantic vs Syntactic Schema - Revision history

Marcin: /* Canonical Takeaway */