Conversation
|
Thanks @sztosz. Could you add unit tests that document this functionality? |
|
Sure, will add them tomorrow. |
|
@markusfeyh I added unit tests, and they work with Ecto 3.5, but they are not backward compatible with earlier versions because embeded_dump and embeded_load were added only in Ecto 3.4.1. Should I just drop the tests with embed_as, which de facto tests not the Ecto.ULID library but if Ecto does it job correctly, or rather try to rewrite those tests to be backward compatible? |
|
Would it be possible to add a check if the ecto version is greater than 3.5 so those tests are only run in that case? |
|
Probably the best solution. But I also am not sure any more If it should be |
|
What about doing like I did in my PR ? |
|
I've got other things on my mind right now, so changed it to draft. I may find some time later in the week. Changing behaviour to just use may work, but it does not enforce that public API stays same in proper way. |
|
Agree with @xward, I sent the same fix #5 over a year ago. 🙂 @dcuddeback If you are no longer interested in this project, may I take over? I rely on this package heavily and evidently many do. |
|
@jayjun If you fork it, update, and fix it, I'll be glad to make the project I work on switch to it. We're doing daily a lot of builds in our pipelines, so it may give new library some stats in hex.pm that would make deprecation of this package easier, that's of course with assumption that @dcuddeback will not respond in timely manner. I totally understand that @dcuddeback does not maintain it any more for whatever reasons, and I'm OK with that. I can only thank him, and thank any person that will carry on with making ULID's compatible with current and further versions of ecto. |
|
@sztosz I managed to contact dcuddeback and long story short, TheRealReal has assigned people to look into this. They seem like an active company so I won’t consider forking yet. |
|
Any updates on this? Is there a better fork? |
Adds required callbacks implementations
embed_as/1andequal?/2for Ecto.Type behaviour