new repo: helix-xml-feed

[trieloff]

Create a new repository with the name helix-feed-service to provide a Helix 3-compatible feed service that can:

• fix the issues with https://github.com/adobe/helix-pages/issues/794
• generate XML feed from a JSON index stored in content bus
• generate sitemap.xml from a JSON index stored in content bus

Supported parameters:

• owner, repo, ref: the usual
• src: name of the JSON source file (will be loaded using helix-data-embed) (we need to discuss if the filtering feature should stay in data-embed) or be removed (https://github.com/adobe/helix-data-embed/issues/364) and implemented in the xml-feed-service instead
• id: name of the table column that has the page URL
• title: name of the table column that has the page title
• updated: name of the table column that has the last modified date
• host: hostname to be used for external URLs, if not specified taken from hlx-forwarded-host, x-forwarded-host (but a Vary will be added)
• href.*_*: name of the table column that has the href for the language *_*

Stuff you can do yourself

• [ ] run npm init @adobe/helix-library or npm init @adobe/helix-service
• [ ] Add topics to the repository, at least helix
• [ ] Add the group "Project Helix Admins" with Admin permissions to the list of collaborators
• [ ] Add the group "Project Helix Developers" with Write permissions to the list of collaborators (Project Helix Guests will be taken care of automatically)
• [ ] Upload a social media image (use this Spark template)
• [ ] Set the repository description
• [ ] Uncomment and adjust the helix-post-deploy/monitoring command in .circleci/config.yaml: decide if your service falls under Development, Publishing or Delivery. (note: failures in the latter two categories will ping on-call engineers on weekends!) See https://status.project-helix.io for reference. (skip for libraries)
• [ ] Set up CircleCI
• [ ] Follow the project in Slack channel #helix-noisy

Open the #helix-noisy Slack channel, then type /github subscribe adobe/helix-…

Stuff you need an Adobe Org Admin for

• [ ] Enable Renovatebot
• [ ] Share the ADOBE_BOT_NPM_TOKEN org secret

[trieloff]

Feedback appreciated:

• would the ability to generate sitemap.xmls be useful or don't we know enough yet?
• we could implement the whole thing (except for filters) using Compute@Edge
• should filters live in this service or remain in helix-data-embed?

[davidnuescheler]

i don't know that i had enough time to think about sitemaps in a hlx3 environment. but if i think of relative priority then sitemaps definitely have a much higher priority than feeds. i am totally fine with helix not supporting feeds and have a manual generation process in a completely project specific fashion, as opposed to having a service that is part of the product.