🦋 Changeset detected

Latest commit: cd2350b

The changes in this PR will be included in the next version bump.

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

trying to understand the use-case better. You want to omit the doctype so you can use Astro pages as partials; does this mean that you also omit the head and body tags? I guess that works since we don't automatically append those.

You want to omit the doctype so you can use Astro pages as partials; does this mean that you also omit the head and body tags?

That is right. I want to be able to generate HTML output that only contains what I made it to contain without any unexpected side effects. More reasons for other can be found in this discussion.

The use case is rendering content for technology such as the nginx SSI module.

We would like to expose this feature, but I think it's unfortunately going to take more than this. Currently our compiler doesn't preserve doctype statements because we expect the rendering logic to add it for us, so enabling this setting would currently break even if a doctype is added manually in your markup.

Thanks for the update @natemoo-re. I looked at @astrojs/compiler to see if I could make the change but I could not.

What would be the next step?

I think I'm going to have to come out against this feature. Astro doesn't at present support partial generation. Any hacks we add are going to make everything more brittle and confusing for users.

It's called the pages/ folder and expects to generate full pages. Partials are something we could support, I have heard some interest from people on the team, so it's not out of the realm of possibility. But it needs to go through the proposal process and be it's own feature if we are going to do it (which probably, imo, means not going through the pages folder). https://github.com/withastro/roadmap/discussions

This is more complicated than a simple change and should go through the roadmap process. https://github.com/withastro/roadmap/discussions

Should it continue in the existing proposal or would creating a new proposal be preferred?

Thanks for being flexible @Sertion! I think we can continue the discussion in withastro/roadmap#266 and hopefully come up with a more fleshed out proposal from there.

This would have been helpful to generate humans.txt file dynamically (inserting the updated date into).
Setting the response header works from humans.txt.astro but it will still emit the doctype header :(