Add section about metadata

[ann0see]

Short description of changes This adds a section about adding metadata to the server. I think we should get this in soon (for 3.10.0) in a minimal way and make it updatable to have quick adoption. The standard format proposed by @pljones was "easy".

This is JUST A START and discussion about the standard should continue in https://github.com/orgs/jamulussoftware/discussions/3086 - thus it is a KB entry.

Context: Fixes an issue? Related issues

Related to: https://github.com/orgs/jamulussoftware/discussions/3086

Status of this Pull Request Ready for review

What is missing until this pull request can be merged? Short ! review. The standard is to be discussed elsewhere.

Does this need translation?

YES

Checklist

• [x] I've verified that this Pull Request follows the general code principles
• [x] I waited some time after this Pull Request was opened and all GitHub checks completed without errors.
• [x] I'm sure that this Pull Request goes to the correct branch

[ann0see]

@mcfnord I think you should review this PR as you seem to be interested in this topic. The standard should NOT be discussed here.

[ann0see]

(needs a local check - especially if the HTML comments are correctly parsed)

[ann0see]

Fixed. @mcfnord Also the link works locally.

[mcfnord]

My conclusions come up a lot for me when considering this stuff:

a) We are removing the CLI example from headless guidance, b) --welcomemessage used in systemctl introduces inherently tricky location file requirements that my recent documentation improvement only hints about. I had hoped to deliver much more helpful guidance on this. c) A default /etc/jamulus/welc.html would circumvent this by providing a file already configured for use where we could ship with default metadata, or just an empty file, which might behave the same as no welcome file, the current default.

I want to give people a reasonable way to use this interface, before giving them the interface. Without some additional steps, we are making welcome messages harder to use, while increasing the features hidden there. Formula for pain. I want to score a major victory for welcome message usage on systemd headless Linux by delivering /etc/jamulus/welc.html as an empty file in v1.

[ann0see]

We are removing the CLI example from headless guidance,

I don't see where it got removed unfortunately.

--welcomemessage used in systemctl introduces inherently tricky location file requirements that my recent documentation improvement only hints about. I had hoped to deliver much more helpful guidance on this.

Yes, but out of scope of this PR.

I want to give people a reasonable way to use this interface, before giving them the interface.

I'd like to have a draft out for adoption - just as KB entry, which means it's a living document and can be updated at any time. The documentation however (Running a server) should already have a link as this needs translation and isn't easily updatable if no release is happening.

[mcfnord]

For these reasons, I think the KB is limited in scope to bot policy. I don't want to advocate for describing a generalized metadata infrastructure to users (even if we have one) because I don't want to invite anyone to think metadata is where they communicate to users.

[ann0see]

I still think a generalised standard should be documented. But maybe not that prominently? How else should admins know about it?

[mcfnord]

I don't see where it got removed unfortunately.

In the next release, we remove:

The following minimum setup is required to run a Registered Server:

jamulus --nogui --server \ --directoryserver genreServer:port \ --serverinfo "yourServerName;yourCity;[country ID]"

This means we will mandate usage of systemctl, unless you have the Secret Knowledge of CLI usage. And that's why I go so hard against the serious complexity involved using a welcome file. Previously I advocated for a two-part Running a Server page: a Quickstart that concluded with the server running at the CLI on a public directory, followed by the deeper dive into usage based on systemctl. I still recommend that design because I believe it would lead to more operator successes.

[ann0see]

Ok. I wasn't aware of this @gilgongo can we add this back?

[gilgongo]

Ok. I wasn't aware of this @gilgongo can we add this back?

I seem to recall I wanted it in (it was me who put it in originally), but there was some discussion about default configurations and it being misleading in some way? Can't recall.

[ann0see]

@gilgongo could you please re-add it (with a PR)?

[ann0see]

Can we agree to merge and discuss the content in the mentioned discussion? Maybe we add a disclaimer to the KB entry that it's a WIP.