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

Composite posts - how to recognise? #6

Open
kevinmarks opened this issue Dec 14, 2015 · 9 comments
Open

Composite posts - how to recognise? #6

kevinmarks opened this issue Dec 14, 2015 · 9 comments

Comments

@kevinmarks
Copy link

Longer form posts often have multiple media types included. For example, http://www.kevinmarks.com/ampreaction.html has an audio citation and a reference to a video. I marked these up with u-audio and u-video so they could be seen as part of the post when parsing as microformats. However, this would be classified as a video post by the current algorithm.

Newer post authoring tools on sites like medium and buzzfeed encourage composite posts with many images and/or videos - how doe we ensure these show up as article, which is presumably the type desired?

Or is there a new type needed as an a collection post?

@tantek
Copy link
Owner

tantek commented Dec 15, 2015

In the example you gave of http://www.kevinmarks.com/ampreaction.html - your uses there of u-video and u-audio seem to be on others's content that you are citing, not your own, nor the primary content of the post, thus it would be more appropriate to mark those up as h-cite citations, with the u-video / u-audio respectively as properties of h-cite, not your post.

There is a collection post defined here: http://indiewebcamp.com/collection with a few real world examples. You may want to look at that and see if it fits what you consider a "Composite post", and if so, use markup accordingly in your post.

If you want recognition of "collection" posts added to the algorithm, go ahead and use collection markup in your post, and we can analyze the real world examples to see if we have interoperable convergence in markup among 3+ real world examples.

Will optimistically leave this issue open in the expectation that we'll achieve that goal.

@rhiaro
Copy link

rhiaro commented Jun 7, 2016

Objects with multiple types are possible in AS2, so maybe we need wording that suggests implementations assign multiple types if necessary or if the input is ambiguous.

@tantek
Copy link
Owner

tantek commented Jul 28, 2016

Post type discovery is based on empirical publishing practice, not what is "possible", thus it is preferable to avoid any wording that suggests theory ("if necessary").

If implied multiple type posts are found in the wild, please cite them in this issue. I'll leave it open for a bit in case there was one that just didn't get mentioned, but when I revisit (in at least a few days) if this is still theoretical, I'll close it.

@rhiaro
Copy link

rhiaro commented Jul 28, 2016

I have posts that are simultaneously likes and bookmarks or likes and reposts. My templates are bad and just pick one to show though. I'll try to fix that up and link..

@rhiaro
Copy link

rhiaro commented Jul 28, 2016

Do you consider quote tweets a reply, retweet, citation or all? (Or none?)

@tantek
Copy link
Owner

tantek commented Sep 22, 2016

A quoted tweet is a quotation: https://indieweb.org/quotation#Twitter

And upon further thought, the use-case for Post Type Discovery are clients that need to assert a post is a specific post type or is a / has a primary post type.

Clients with an intrinsic or explicit multi-type design (e.g. a personal site with flexible multi-type templates) do not need to use Post Type Discovery because they can conditionalize all their multi-type facets based directly upon each property's presence/absence or specific values orthogonally.

So to answer:

  • @kevinmarks your post does actually have a primary type, as determined in a previous comment in this htread.
  • @rhiaro your multi-type post handling may not need post type discovery at all, since it appears your templates do not need a singular or primary post type in order to function.

Will await satisfied comments from both of you before we should close this issue (no changes needed, except perhaps a non-normative addition regarding the use-case and explicit alternative guidance for multi-type aware clients).

@tantek
Copy link
Owner

tantek commented Nov 22, 2016

Update: based on implementation feedback (what implementations are doing), it makes sense to expand post type discovery to return a list of post types, with a primary post type, so that implementations that already handle multiple post types on a post can continue doing so, while providing the "primary" for implementations that still want just one post type for any given post.

@tantek
Copy link
Owner

tantek commented Dec 19, 2017

Updating again. As much as I'd like to generalize to returning a list of post types, existing implementations (that consume posts and do post type discovery, e.g. for displaying a specific UI) just don't do this yet so I don't think it's appropriate to document something that aspirational at this point.

I do think the idea does have merit though, and that expanding the algorithm to return a list of post types (the first of which is perhaps what the current algorithm returns) is worth exploring and I think the right place to do that brainstorming and iteration is on a wiki page, not on this note.

I'd like to close this issue without changing the algorithms to return a list of post types.

@tantek
Copy link
Owner

tantek commented Dec 19, 2017

Additionally, on the publishing side, we don't have documented evidence (permalink citations) of 3+ "composite posts" (only one specific example from @kevinmarks above) that are using similar markup that is the existing documented threshold for adding existing "types" to the algorithm (per https://tantek.github.io/post-type-discovery/index-src.html#scope ), so that side is also not quite ready for documenting in the spec as well.

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

3 participants