cc @annevk @domenic would you mind reviewing this?

It might also be worth to serialize aggregateError.erros and suppressedError.suppressed.

@legendecas The spec already mentions them indirectly:

<li><p>Any interesting accompanying data attached to <var>serialized</var> should be
     deserialized and attached to <var>value</var>.</p></li>

.cause also should be serialized.

#5749

@zloirock No need to mention them one by one due to:

<li><p>Any interesting accompanying data attached to <var>serialized</var> should be
     deserialized and attached to <var>value</var>.</p></li>

@anonrig this is a rather vague formulation. For example, Safari still does not serialize it, and in the mentioned above PR it's handled directly.

@domenic thanks for the review. This is my first time contributing to html spec so any help/guidance is extremely helpful. what are the next steps for moving this forward? I can update the pull-request without any problem.

Thanks for your contribution!

Per my last message, there are two major blockers:

• Creating a fully-detailed specification for serializing and deserializing that extra data. (And then updating the tests to test the edge cases of that. E.g., what happens if someone re-defines suppressed as enumerable, or deletes it from the instance, or deletes it from the instance but re-defines it on Object.prototype?)

• Getting implementer interest from 2+ browser engines.

FWIW, you can see an earlier failed attempt at this in #5749. We were unable to get implementer interest, sadly, as only Gecko seems to be devoting any engineering resources to structured cloning.

I'll try adding the agenda+ label to attract interest from others.

WebKit is supportive of doing this.

SuppressedError is part of the Explicit Resource Management proposal and currently at stage 3.

I personally don't think there is much point in adding these without also specifying that their properties like errors of AggregateError are serialized.

I personally don't think there is much point in adding these without also specifying that their properties like errors of AggregateError are serialized.

I'll update the PR and add those fields.

@evilpie in the previous TC39 meeting, this proposal received stage 4.