Split implementor interest and implementation bugs from PR template

[saschanaz]

Closes #207

I did not completely replace the existing template with the WHATWG one, although I would be okay with that.

Once we do this, we should make sure other Editing WG repositories get the same templates.

[snianu]

(and none opposed)

Any reason why this is an "and"?

[snianu]

Also tagging few more people to comment on this. @dandclark @sanketj @johanneswilm

[johanneswilm]

This is fine for specs that have largely been implemented and where only minor things are to be changed/added over time (like this one).
But the charter of our working groups has quite a different focus, opening for experimental ideas and new suggestions on how things could work that are being worked on collaboratively over time without there being a consensus among browser makers that this is actually the approach one ends up choosing.
In conclusion: this template could work for this repository and some others if the maintainers of those documents agree with it, but it should not automatically be applied to all our repos.

[sanketj]

I agree that a template like this would be helpful, but I'm not sure if we should use "At least two implementers are interested (and none opposed)" as the condition for being able to land normative spec changes. This is a space where browsers already do not align with one another on several parts of the spec. Examples include: clipboard permission policy, web custom formats, sanitization. In the absence of interop, it has worked well in the past to at least document the behaviors of each engine in the spec. We should absolutely strive for interop with every change we make, but perhaps we need a better condition that allows for cases where browsers may have by-design differences.

Thoughts?

[annevk]

If engines cannot come to agreement, that is what implementation-defined is for. Which if used to the extent you suggest is more indicative of failure of the standardization process than anything else.

A specification shouldn't discuss concrete implementations. That's not what it's for.

[annevk]

We don't do that anywhere in the web platform. I'm not sure why we'd start here.

[johanneswilm]

This has been done over the past 12-15 years, starting with changes observed in execCommand ( https://w3c.github.io/editing/docs/execCommand/ ) that were found that couldn't easily be removed.

[annevk]

There's a number of notes there that I don't think make sense to publish as part of a standard, but I don't think there's anything normative there that does this.

[johanneswilm]

@annevk That is the work of the previous editor and I agree that the wording doesn't follow what I would expect for this kind of document. Maybe the current editor, @zcorpan, will want to change that eventually. Nevertheless, that was the beginning of documenting differences between browsers.

I don't think we would want to document differences in a "normative" way. I guess that would mean that one specific browser HAS to work in a specific way that is different from another browser.

[saschanaz]

I don't think adding non-normative editorial thing would require implementor interest even after this PR.

[sanketj]

I agree that a template like this would be helpful, but I'm not sure if we should use "At least two implementers are interested (and none opposed)" as the condition for being able to land normative spec changes. This is a space where browsers already do not align with one another on several parts of the spec. Examples include: clipboard permission policy, web custom formats, sanitization. In the absence of interop, it has worked well in the past to at least document the behaviors of each engine in the spec. We should absolutely strive for interop with every change we make, but perhaps we need a better condition that allows for cases where browsers may have by-design differences.

Thoughts?

[johanneswilm]

Having a template like this or a similar one seems like a good idea to the extent that it ensures that browser bugs are filed correctly. But it should not change what we are working or how we are working on it in the WG.