# \#iso

**URL:** https://swi-prolog.discourse.group/tag/iso/203.md

[Latest](https://swi-prolog.discourse.group/latest.md) · [Categories](https://swi-prolog.discourse.group/categories.md) · [Tags](https://swi-prolog.discourse.group/tags.md)

---

## [What is the real purpose of Annex (formal semantics) in the ISO Prolog standard (1995)?](https://swi-prolog.discourse.group/t/what-is-the-real-purpose-of-annex-formal-semantics-in-the-iso-prolog-standard-1995/9566)

<div class="topic-metadata">

**Author:** [@iaroslav\_baranov](https://swi-prolog.discourse.group/u/iaroslav_baranov)\
**Replies:** 0\
**Last updated:** [April 10, 2026, 8:31am UTC](https://swi-prolog.discourse.group/t/what-is-the-real-purpose-of-annex-formal-semantics-in-the-iso-prolog-standard-1995/9566 "2026-04-10T08:31:30Z")

</div>

The standard already gave the execution model as a stack of states with goal/substition/backtracking info for each state. It looks like they try to redefine the execution model (and also some build-in predicates) using t…