Why are dynamically created functions named "anonymous"?

[woess]

Could somebody please clarify why functions created by the Function constructor are assigned a name property with value "anonymous"? i.e., 19.2.1.1.1 Runtime Semantics: CreateDynamicFunction:
39. Perform SetFunctionName(F, "anonymous").
This strikes me as odd and inconsistent. This is the only place in the spec where an anonymous function gets to have a name property, and usually Get(F, "name") would return "" (inherited from Function.prototype) for anonymous functions.

I suspect this might be another case of a web compatibility issue or web reality turning into spec but I couldn't find any information in that regard.

Similarly, I find it a bit strange that in the Function.prototype.toString revision proposal, the source text is synthesized as function anonymous(... when in fact the function is to be parsed without a BindingIdentifier.

[ljharb]

My assumption is that it's to match legacy browser behavior.

[ljharb]

My assumption is that it's to match legacy browser behavior.

[claudepache]

Both Function.prototype.toString output and Function.name were implemented in browsers long before they were standardised, so yes, it’s just web reality with no compelling reason to change.

[claudepache]

Both Function.prototype.toString output and Function.name were implemented in browsers long before they were standardised, so yes, it’s just web reality with no compelling reason to change.

[bmeck]

@woess does this satisfy your question / can we close this?

[bmeck]

@woess does this satisfy your question / can we close this?

[woess]

Thanks for the clarification, @ljharb and @claudepache. That answers my question.
Not very happy about this special case, but I agree there's no compelling reason to change.

[woess]

Thanks for the clarification, @ljharb and @claudepache. That answers my question.
Not very happy about this special case, but I agree there's no compelling reason to change.