From 7d9f364c9f7659135a4d56044143d6d5085a41c1 Mon Sep 17 00:00:00 2001 From: Ms2ger Date: Wed, 13 Nov 2019 12:40:08 +0100 Subject: [PATCH] Correct non-normative summary of the overload resolution algorithm. --- index.bs | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/index.bs b/index.bs index 8895f370..fe1a34d3 100644 --- a/index.bs +++ b/index.bs @@ -11088,10 +11088,14 @@ Note: The HTML Standard defines how a security check is performed. [[!HTML]] and there is an overload with an optional argument at this position, then we will choose that overload. If there is no valid overload for the type of value passed in here, then we throw a {{ECMAScript/TypeError}}. - The inspection of the value at the distinguishing argument index does not have any side effects; - the only side effects that come from running the overload resolution - algorithm are those that come from converting the ECMAScript values - to IDL values. + Generally, the inspection of the value at the distinguishing argument index does not have any + side effects, and the only side effects in the overload resolution algorithm are the result of + converting the ECMAScript values to IDL values. + (An exception exists when one of the overloads has a [=sequence type=] or [=frozen array type=] + at the distinguishing argument index. + In this case, we attempt to get the {{@@iterator}} property to determine the appropriate + overload, and perform the conversion of the distinguishing argument separately before continuing + with the next step.) At this point, we have determined which overload to use. We now convert the remaining arguments, from the distinguishing argument onwards,