Skip to content
This repository has been archived by the owner on Jan 25, 2022. It is now read-only.

The semantics of (null)?.b is now fixed #121

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

claudepache
Copy link
Collaborator

As there is no canonical semantics, links to the relevant discussion threads for the various points of views. Closes #69. Closes #65.

As there is no canonical semantics, links to the relevant discussion threads for the various points of views. Closes #69. Closes #65.
@rkirsling
Copy link
Member

I don't mind adding links here, but we shouldn't remove the existing text...? It precisely describes the reasoning behind the semantics that we went with.

@claudepache
Copy link
Collaborator Author

I don't mind adding links here, but we shouldn't remove the existing text...? It precisely describes the reasoning behind the semantics that we went with.

In fact, the existing text presented my original reasoning, which I retracted later (and I have not counter-retracted). So, while it was indeed the reasoning at the origin of the current semantics, its preeminence is primarily for historical reasons.

Since I am not motivated to spend time in trying to present a balanced summary of the opinions of the various people involved in that proposal (in case I would be able to), I have just linked to the appropriate discussion threads. And the historical argument is not lost, as it is clearly quoted in the top comment of the first thread linked.

@rkirsling
Copy link
Member

But the question isn't asking for an origin story or someone's reasoning, it's asking for the rationale behind the spec, and that's exactly what the current text provides. I agree that the links are helpful, but only to exemplify the amount of debate that occurred on the subject—someone looking for a simple answer should not need to waste their time.

@claudepache
Copy link
Collaborator Author

claudepache commented Aug 21, 2019

@rkirsling However, we’re disagreeing on what the rationale or the simple answer should be. You think that the old text provides it, I think that the new text provides it.

I am open to reinstate the original argument, but in a way that makes it sufficiently clear that it is more the historical than the canonical reason.

@rkirsling
Copy link
Member

By "rationale", I mean "what does this approach buy you?". There was also fierce debate surrounding the topic of #93, but we have not written the story behind that decision, we've described the benefits of the approach. The existing text here is not "historical" or "uncanonical" in any way I can see; it explains how one is meant to conceive of the feature. The current state of this PR makes it sound like it simply won a popularity contest, like the committee's decision was basically arbitrary in nature.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

(null)?.b should evaluate to null, not undefined Changing null -> undefined potentially dangerous
3 participants