Olivier Cailloux

Re-reading, I think I understand that you propose to replace `throws JsonbException` by `throws RuntimeException`. If so, I have no strong opinion either way. It anyway permits the caller to...

Sorry, wanted to think about this then forgot to come back to it. I still think this would be a good addition. If I understand correctly, you argue for declaring...

1. I think that it’s better to keep the Javadoc `@throws JsonbException If any unexpected problem occurs during the close`. Here is why. 1. It does not harm: it is...

I am sorry, I do not understand your first point. In my mind (and in my previous comment), “the implementor” refers to the implementor of the Jsonb spec. That is...

EDIT Forget this, I just realized that your proposal probably is that the implementor should suppress any exception occurring, and try to not throw anything (unless perhaps absolutely required), just...

It springs to my mind that your last point might be about comparing not “my” version to “your” version, but comparing to the existing state of fact. Indeed, constraining the...

Thanks for all the discussion. At this point it appears to me that there are multiple decisions to be taken from a high-level design point of view of the Jsonb...

Time has passed and I believe that we still agree that the status quo is unsatifactory: the spec does not indicate precisely enough the behavior of `close()` wrt exceptions and...

BTW I guess that this proposal is in view of 4.0, as this involves some breaking change for implementors.

> In my view it would be reasonable to assume that `SSLContext.getDefault()` is used by default. I suggest to add this doc to [`ClientBuilder`](https://javadoc.io/static/jakarta.ws.rs/jakarta.ws.rs-api/3.0.0/jakarta/ws/rs/client/ClientBuilder.html). I also think the spec should...