Re: YAML-LD: A Missed Opportunity for a "Lightweight JSON-LD"

Hi all, great to see interest in YAML-LD!

Currently, YAML-LD does not change the behavior of JSON-LD algorithms;
round tripping of YAML-LD to JSON-LD and vice versa was one of core
considerations during development.

However, both JSON-LD and YAML-LD support profile parameter. Right now, for
YAML-LD, it is unused, but implementations can define their own profiles.

I expect a reduced subset of JSON-LD features can be such a profile. What
do you think?

Anatoly

On Tue, Feb 24, 2026 at 8:49 PM Christoph <christoph@christophdorn.com>
wrote:

> In my case I do not need anything from YAML other than the authoring
> syntax which can hold JSON so building around JSON is no problem and should
> be no problem as the primitives should work for any tree structured
> document.
>
> The improvements needed are in how properties are mapped in a streaming
> friendly way that is compact in file and allows for co-locating entities. I
> feel that the file boundary must go away so one can compose a whole graph
> in one file or just one node.
>
> Christoph
>
>
> On Tue, Feb 24, 2026, at 8:07 AM, thomas@pellissier-tanon.fr wrote:
>
> Hi Filip,
>
> Big +1 a simpler streaming friendly variant of JSON-LD would be amazing.
>
> I am not sure YAML-LD is the good place for that. YAML is a very
> complicated language with non-streaming friendly features like anchors.
> Building something on top of JSON by removing JSON-LD features seems a
> better approach to me.
>
> Thomas
>
>
>
> Le mardi 24 février 2026 à 03:12, Christoph <christoph@christophdorn.com>
> a écrit :
>
> Hi Filip,
>
> I have struggled with JSON-LD. I would be interested in looking at a
> refreshed proposal for YAML-LD to see how it fits with my use-cases. Happy
> to provide a second implementation in TypeScript to validate.
>
> Christoph
>
>
> On Mon, Feb 23, 2026, at 5:13 PM, Filip Kolarik wrote:
>
> Dear JSON-LD Community and Working Group,
>
> Having some experience implementing JSON-LD processor and encountering
> issues that many users face, I wanted to share some thoughts on potential
> directions for a lightweight alternative.
>
> I see YAML-LD as a missed opportunity, not in syntax, but in design. A
> simpler variant of the JSON-LD context, let's say a subset of JSON-LD
> context expressivity, could enable straightforward, intuitive syntax,
> faster processing, and even better support for streaming. This would make
> it more approachable for developers and easier to use in constrained
> environments.
>
> Isomorphic mapping would also greatly help ensure contexts can be
> processed safely and efficiently.
>
> In practice, most use cases can function without features like scoped or
> inline contexts, reverse properties, nesting, etc. Simplifying in this way
> could drastically reduce implementation complexity and lower the learning
> curve for developers and users.
>
> Please don’t let this opportunity slip by. What is the current value of
> YAML-LD? Is it mainly syntactic sugar? I would be very interested to hear
> the community’s and working group’s perspective.
>
> Best regards,
> Filip
> https://www.linkedin.com/in/filipkolarik/
>
>
>
>