Eric Deandrea
> This is great work. but if there is anyway it jeoperdizes the 1.0 release (this is for @dliubarskyi and @mariofusco to decide), then we should hold off on it...
> Totally understandable, but 1.0 is the absolute priority for the team (for many reasons) Agreed. Thats why once a week or so I will merge back with `main` so...
@geoand I'd really like you to take a look through this too to ensure its generic enough and there are enough extension points where we don't have to rewrite things...
> I assume you have already attempted to build Quarkus guardrails onto this? I have https://github.com/quarkiverse/quarkus-langchain4j/issues/1284 but I have not started working on it because it'll involve a lot of...
@cescoffier / @geoand Should we deprecate at least the existing guardrail annotations & interfaces, as they will most likely be replaced by similar things upstream? That way people know that...
Also tracking https://github.com/langchain4j/langchain4j/pull/2571
@geoand / @mariofusco Now that guardrails are upstream I will start working on this in the next couple of weeks. Question - should we leave the existing Quarkus implementation there...
Yes we will have to do something using upstream on `Multi`. I think we can make it work.
@cescoffier Do you have an opinion on this? > Question - should we leave the existing Quarkus implementation there and just deprecate for a few versions leaving both implementations? That...
As a bare minimum we can swap the upstream annotation for the custom Quarkus one but keep the custom implementation we have in quarkus