Editorial: quick fixes

[jmdyck]

... for some recent and not-so-recent commits.

Fixes: #1971

[michaelficarra]

The grammar parameters commit goes against our editorial resolution to have grammar parameters always appear in SDO definitions (in the production name and in each alternative, but not any alternative "guards"). Please drop that commit.

[michaelficarra]

LGTM other than the grammar flags commit.

[jmdyck]

The grammar parameters commit goes against our editorial resolution to have grammar parameters always appear in SDO definitions

Hm. Was that resolution stated somewhere? Note that 5.2.2 Syntax-Directed Operations still says:

When an algorithm is associated with a production alternative, the alternative is typically shown without any “[ ]” grammar annotations.

[michaelficarra]

@jmdyck It was in our editor's update slides and we discussed it again 15 days ago but I seem to have forgotten to post the resolution in that thread. Anyway, I guess the above comment will suffice.

[jmdyck]

@jmdyck It was in our editor's update slides [..] but I seem to have forgotten to post the resolution in that thread. Anyway, I guess the above comment will suffice.

On that slide, you're presumably referring to the bullet that says:

ecma262 #1885: include flags and subscripts in SDO grammar snippets

• (I guess "flags" means grammatical parameters.)
• It's interesting that you reference Editorial: Add subscripts to BindingIdentifier static semantics. #1885, since that isn't about an SDO, but rather two Early Error rules.
• Moreover, in Editorial: Add subscripts to BindingIdentifier static semantics. #1885, the justification was that the accompanying prose specifically references the grammatical parameters.

So all I got from #1885 and that bullet was that productions should include grammatical parameters when the accompanying prose references them. Now you're saying that every production in an SDO or Early Error rule should include its grammatical parameters? I count roughly 2000 productions where the parameters would have to be reinserted. Why do you want that? (And do you think anyone will be willing to do the work?)

[bakkot]

Yeah, I also saw that it was going to be thousands of places and reconsidered. I'll bring it up again sometime.

And do you think anyone will be willing to do the work?

I could do it automatically with ecmarkup at this point, I think.

[michaelficarra]

@jmdyck At least one of the mentioned motivations during the editor call was the fact that readers have many times been confused about why the parameters were not present in the SDO definition and asked about this. I can look for examples if you like, but it might be hard to search for them.

[jmdyck]

(force-pushed after rebasing to master and adding a commit.)

[michaelficarra]

Please drop the commit that removes the grammar parameters in SDOs.

[jmdyck]

Please drop the commit that removes the grammar parameters in SDOs.

I disagree with the rationale, but I'll drop it if that's the consensus of the editors. (So far, I'm getting mixed signals.)

[jmdyck]

(force-pushed to rebase to master and add a commit to do another step-reference fix)

[bakkot]

I'll drop it if that's the consensus of the editors. (So far, I'm getting mixed signals.)

The editor group will talk about it again at some point, but until then let's avoid churning it - that is, I would also prefer you drop the commit in question, for now; we can always add it again later if we decide we are in agreement to omit all parameters.

[jmdyck]

Fair enough.

[jmdyck]

(force-pushed to remove the "Drop grammatical parameters" commit)