Replies: 3 comments 1 reply
-
Welcome back @gnidan , My motivation for contributing is like many open-source projects : to have new contributions / contributors and users. That said, in the past I'd started a plan (not perfect and still open to discussion) : 1. Reflexions about the resolving library & usages hereMany libraries are out :
First thing would be to compare the libraries : what they offer, their maintenance, strenghs and weakness, ... 2. Design the proposal (how to handle it, display it, ...) extensivelyThis is the point with the most functional, technical and visual questions. 3. Getting users feedbacks about the proposalMany actors (like ajv ) acknowledge that draft-07 is most widely used version of JSON Schema specification. Remember on the roadmap page, I mentioned which version of the specification a keyword was linked to ? 2019-09 and 2020-12. I saw that many other projects (like monacco I also used) started considering these versions thanks to user requests like in microsoft/monaco-editor#3160 or microsoft/monaco-editor#4114 If we manage to find feedback of any kind on it ( for example, https://json-schema.org/#community have a Slack but not only : other teams / companies could be involved from my opinion ), it would make the product / solution better Side NoteIn the future, the UnsolvedRefs component will somehow likely still remain useful in case users ask "Why my references weren't resolved" (as people can make mistake like everyone). If the project had more contributors, we'd be able to make faster progress on the subject. |
Beta Was this translation helpful? Give feedback.
-
Update: I was able to find a way to avoid the use of dynamic references (with only a small loss of expressiveness). For now, I won't be digging into this too much more, although we'll see... other parts of my schema might end up needing those later. |
Beta Was this translation helpful? Give feedback.
-
Update for the first point : new libraries are out like |
Beta Was this translation helpful? Give feedback.
-
Months ago, I told myself I wouldn't fall into an old habit of making everything generic... yes,
$dynamicRef
and$dynamicAnchor
are quite interesting, esp. how the JSON Schema blog suggests a very nice way to specify aList<T>
schema, but it's madness, right? Certainly it's no problem that the excellent docusaurus-json-schema-plugin doesn't support it... no problem at all that @stoplight/json-ref-resolver (used by this plugin) is now unmaintained, so even if I wanted to use$dynamicRef
, it'd be quite unfeasible to get support...Anyway, hi, nice to see you again. Turns out that dynamic references are exactly the tool I need for one of these schemas I'm writing.
I might be opening this discussion a bit early, since I've only just started playing around with the new UnsolvedRefs customization (which I think I'll need in conjunction with swizzling), so I don't yet have a lot of findings. I plan to add to this discussion as I see what the current capabilities are (and how much of a pain in the neck they are ;).
In the meantime, do you have any near-term goals for addressing this (beyond the unsolved refs stuff)? I apologize if I am forgetting some discussion in this repo somewhere, but I'm wondering what your thoughts are regarding the stoplight library, since I imagine you're probably not too happy about its end of life... will this plugin need to find another alternative, or maybe add custom support itself for all these latest JSON Schema features? I ask in case there's a way for me to implement my customizations that'd be appropriate for eventual merging upstream
Beta Was this translation helpful? Give feedback.
All reactions