| 15:34 | <Jack Works> | hello I have a question |
| 15:34 | <Jack Works> | In section 13.15. https://tc39.es/ecma262/multipage/ecmascript-language-expressions.html#sec-assignment-operators-runtime-semantics-evaluation |
| 15:34 | <Jack Works> | it writes: "Let lref be the result of evaluating LeftHandSideExpression." |
| 15:35 | <Jack Works> | but why it doesn't trigger my code when evaluating LHS? |
| 15:36 | <Jack Works> | what does it mean by "evaluating"? |
| 15:57 | <bakkot> | "evaluating" means "invoking the Evaluation abstract operation" |
| 15:58 | <bakkot> | the Evaluation semantics for a.x are here: https://tc39.es/ecma262/multipage/ecmascript-language-expressions.html#sec-property-accessors-runtime-semantics-evaluation |
| 15:59 | <bakkot> | you will note that it returns a Reference Record |
| 15:59 | <bakkot> | which just holds a reference to the property; it does not actually trigger getters/setters |
| 16:00 | <bakkot> | accessors get triggered by GetValue or PutValue, depending on whether you're reading or writing the property |
| 16:01 | <bakkot> | that is, GetValue and PutValue actually unwrap the reference record and call [[Get]] or [[Set]] respectively. but that happens as part of evaluating the thing containing a.x, and depends on whether a.x is on the left-hand side of a = or not. |
| 16:13 | <Jack Works> | Rough outline here: https://gist.github.com/rbuckton/ae46b33f383ba69880c7138c49b5e799 |
| 16:15 | <Jack Works> | that is, GetValue and PutValue actually unwrap the reference record and call [[Get]] or [[Set]] respectively. but that happens as part of evaluating the thing containing |
| 21:22 | <leobalter> | bakkot: do you already have any opened issue for the extending built-ins? I don't have any concerns but the 120min timebox caught my attention. |
| 21:56 | <bakkot> | leobalter: no; we're going to put something together in the next couple weeks |
| 21:57 | <leobalter> | thanks! I'm looking forward to see it! |