Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Drop .host-meta discovery if no known implementations #40

Closed
aaronpk opened this issue Nov 1, 2016 · 10 comments
Closed

Drop .host-meta discovery if no known implementations #40

aaronpk opened this issue Nov 1, 2016 · 10 comments

Comments

@aaronpk
Copy link
Member

aaronpk commented Nov 1, 2016

Are there any known implementations of advertising the hub using the .host-meta method? If not, it's probably worth dropping that from the list of supported discovery methods.

@tantek
Copy link
Member

tantek commented Nov 1, 2016

FWIW we also made a similar decision for Webmention and other SWWG specs, based on that and group resolution to prefer "follow your nose" methods of discovery, settling on link rels being sufficient.

@julien51
Copy link
Collaborator

julien51 commented Nov 8, 2016

I'm ok with that.

@julien51
Copy link
Collaborator

julien51 commented Nov 8, 2016

Actually... let me change my mind :)
For these content types which do not have a clear "linking" mechanism and which are hosted on platforms which do not allow for Link headers (Jekyll + github), this is a convenient way to set up hub discovery. So I'm in favor of keeping it!

@aaronpk
Copy link
Member Author

aaronpk commented Nov 8, 2016

Do you know of any publishers or subscribers that ever actually implemented .well-known discovery?

@julien51
Copy link
Collaborator

julien51 commented Nov 8, 2016

Good point. I don't.
But it's cheap to keep it.

@aaronpk
Copy link
Member Author

aaronpk commented Nov 8, 2016

It's actually not cheap to keep it. At the very least we should mark it as an "at risk" feature. In order to get to Recommendation, each feature of the spec must have at least two implementations. Part of the testing tool I'm building is keeping track of how many implementations implement each feature of the spec. You can see what this will look like for Micropub, and the less pretty version for Webmention.

Keeping this feature also means that:

  1. Every subscriber will have to check for the .well-known value because they can't assume the publisher will send a link header
  2. I will have to build support for it into pubsub.rocks

@julien51
Copy link
Collaborator

Since it's already in the spec, we cannot remove it. BUT, we should put it "at risk" so we can later on deprecate if we see no usage of it.
Also, the current wording around that is not clear enough and uses the wrong link. I believe @aaronpk will clarify!

@pfefferle
Copy link
Contributor

the correct link to the host-meta spec is https://tools.ietf.org/html/rfc6415 (instead of the well-known) and the correct URL path is /.well-known/host-meta (instead of .host-meta)

@akuckartz
Copy link

👍 for "at risk"

@aaronpk
Copy link
Member Author

aaronpk commented Nov 22, 2016

This has been done in #61

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

No branches or pull requests

5 participants