Ayo (Node.js fork)

[ksmithut]

Feel free to close this out, but since it hasn't been said yet, there's a major fork of the node project: https://sourcecontribute.com/2017/08/22/node-js-has-forked-into-ayo/. I'm not sure if it will stick, or if they even have all of the processes in place to be able to support the types of releases required for nvm to consume it, but here it is: https://github.com/ayojs/ayo. Given they it was started by some of the node.js leadership, I think it could potentially stick. Given that it's similar to iojs, I was wondering if nvm would ever plan to add support for ayo.

I'm in no way affiliated with either group, I'm just a bystander wanting to build software.

[ljharb]

Thanks for filing the issue.

It's not similar to io.js at all. That was an effort by major contributors to node who hadn't seen a release in years, to get technical progress moving, and to add non-corporate governance.

This seems like an opportunistic PR play after 4 people quit a committee, 3 of whom are still publicly involved with the project, after a controversial decision, while there are many changes in the works that would allow that decision to be soon revisited.

I'm not going to add support for a fork lightly; I don't think this will stick; and I don't think a fork is an appropriate or well-timed response to what's been happening. Either way, I wouldn't add support for a fork that adds nothing but a partial documentation commit - that has value for nobody.

[ljharb]

Thanks for filing the issue.

It's not similar to io.js at all. That was an effort by major contributors to node who hadn't seen a release in years, to get technical progress moving, and to add non-corporate governance.

This seems like an opportunistic PR play after 4 people quit a committee, 3 of whom are still publicly involved with the project, after a controversial decision, while there are many changes in the works that would allow that decision to be soon revisited.

I'm not going to add support for a fork lightly; I don't think this will stick; and I don't think a fork is an appropriate or well-timed response to what's been happening. Either way, I wouldn't add support for a fork that adds nothing but a partial documentation commit - that has value for nobody.

[ksmithut]

Thanks :)

[ksmithut]

Thanks :)

[ljharb]

Now that a few days have passed, I should clarify some things.

At the time I wrote this comment, there was one commit and virtually no discussion on the ayo.js repo. Since then, there's much more activity, and the goals of the project are much more clear - had this been asked today, I'd not have responded so strongly.

I still don't think a fork is the right solution to the issues that precipitated it, but I do wholeheartedly agree with the aims of the fork, and with the morals and values of those who created it - and I fully support their goals.

In general terms, for a fork to be added to nvm, at a minimum the following would need to be true:

1. there exists at least one official release
2. the project follows semver
3. the project has a way to list versions, preferably identical in format to http://nodejs.org/dist/index.tab and https://iojs.org/dist/index.tab
4. The project ideally provides binaries for at least all the platforms that node supports, and if not, provides a concrete set of rules I can implement to help decide when a user can use a binary or not
5. there exists "enough" interest that the added complexity is warranted (this is a fuzzy requirement, yes, but hopefully it still makes sense to people)

If anyone wants to make a case for nvm supporting ayo.js, please do, but please note that any anti-CoC or similar trolling will be dealt with promptly.

[ljharb]

Now that a few days have passed, I should clarify some things.

At the time I wrote this comment, there was one commit and virtually no discussion on the ayo.js repo. Since then, there's much more activity, and the goals of the project are much more clear - had this been asked today, I'd not have responded so strongly.

I still don't think a fork is the right solution to the issues that precipitated it, but I do wholeheartedly agree with the aims of the fork, and with the morals and values of those who created it - and I fully support their goals.

In general terms, for a fork to be added to nvm, at a minimum the following would need to be true:

1. there exists at least one official release
2. the project follows semver
3. the project has a way to list versions, preferably identical in format to http://nodejs.org/dist/index.tab and https://iojs.org/dist/index.tab
4. The project ideally provides binaries for at least all the platforms that node supports, and if not, provides a concrete set of rules I can implement to help decide when a user can use a binary or not
5. there exists "enough" interest that the added complexity is warranted (this is a fuzzy requirement, yes, but hopefully it still makes sense to people)

If anyone wants to make a case for nvm supporting ayo.js, please do, but please note that any anti-CoC or similar trolling will be dealt with promptly.