FHIR Chat · RFC 7386 JSON Merge Patch · implementers

Stream: implementers

Topic: RFC 7386 JSON Merge Patch


view this post on Zulip Igor Sirkovich (Jul 12 2018 at 04:03):

I'm wondering if there was any discussion of RFC 7396 JSON Merge Patch and if there is any plan to implement it in FHIR.

view this post on Zulip Christiaan Knaap (Jul 12 2018 at 13:08):

The PATCH operation for JSON is currently defined as jsonpatch (http://jsonpatch.com/). What would be gained if that would be JSON Merge Patch?

view this post on Zulip Igor Sirkovich (Jul 12 2018 at 14:26):

I might be wrong (I'm just starting looking into this), but it seems JSON Merge Patch is a much simpler solution than normal Patch.

view this post on Zulip nicola (RIO/SS) (Jul 12 2018 at 14:28):

We discussed this here - https://github.com/fhir-fuel/fhir-fuel.github.io/issues/9
Join discussion.

view this post on Zulip Igor Sirkovich (Jul 12 2018 at 17:16):

Thank you @nicola (RIO/SS) . I'm going to present various update options to our development team and if JSON Merge Patch emerges as a favorite approach, I would definitely join this discussion. At this point, I just wanted to gather as much information as possible on Update approaches in FHIR, but couldn't locate anything related to RFC 7386 JSON Merge Patch.

view this post on Zulip nicola (RIO/SS) (Jul 12 2018 at 18:41):

please put your +1 on github ;)

view this post on Zulip Grahame Grieve (Jul 16 2018 at 00:32):

we never discussed rfc 7396 - no one brought it up

view this post on Zulip Grahame Grieve (Jul 16 2018 at 00:32):

how would it be different?

view this post on Zulip Jenni Syed (Jul 16 2018 at 21:56):

Here is a good comparison: https://erosb.github.io/post/json-patch-vs-merge-patch/
I thought we did discuss a bit and the array piece was a killer, as well as invalid json not being able to be failed on the merge patch

view this post on Zulip Igor Sirkovich (Jul 18 2018 at 01:44):

Thank you Jenni, this is a good example. JSON Merge Patch looks simpler, but, as you said, has some serious limitations. I had presented this anyway among other options to our dev team, but the decision was made to avoid any patches and to stick to a regular update using full resource representation.

view this post on Zulip nicola (RIO/SS) (Jul 18 2018 at 08:56):

I think, patch and merge can coexists as xml & json patch do. We can send specific parameter, which defines merge/patch engine - up to piece of js code :)

view this post on Zulip Grahame Grieve (Jul 18 2018 at 09:22):

content-type header would be best

view this post on Zulip Lloyd McKenzie (Jul 18 2018 at 14:51):

The more options we allow, the greater the complexity for both clients and servers. We support XML & JSON because it's very hard to have clients who work exclusively in one syntax or the other submit a PATCH that uses the other syntax. Are there equally strong arguments for supporting two different JSON diff technologies?

view this post on Zulip nicola (RIO/SS) (Jul 18 2018 at 15:52):

They both (json patch & merge) are complimentary standards


Last updated: Apr 12 2022 at 19:14 UTC