quoting IRC conversation: <sebersole> sannegrinovero: ah i remember why... <sebersole> Configuration <sebersole> we have tests that explicitly test serializing Configuration <sebersole> and services not being Serializable can cause those to fail <sebersole> of course thats a transitory concern <sebersole> like i said in the email prior to emmanuel's semi-harsh reply... i dont think it is really necessary <sannegrinovero> ah, if it's transitory would you accept a patch to remove the comment and the "extends Serializable" ? <sannegrinovero> I can replace the comment with a "soft requirement" for some services explicitly needed in tests <sebersole> no, i would not accept such a patch... <sebersole> but you could make that change if you want <sebersole> :) <sebersole> just run the tests first <sannegrinovero> you mean I should push it directly? <sebersole> yep
On 26 September 2012 13:38, Emmanuel Bernard <emman...@hibernate.org> wrote: > On Wed 2012-09-26 13:24, Sanne Grinovero wrote: >> I just found this comment on org.hibernate.service.Service: >> >> /** >> * Marker interface for services. >> * <p/> >> * NOTE : All services must be {@link Serializable}! >> * >> * @author Steve Ebersole >> */ >> public interface Service extends Serializable { >> >> >> What kind of bad things will happen to us if violating this recommendation? >> >> In OGM we have for example connection pools managed in a Service, we >> can hardly serialize that. > > Well with a bit of work we could make such services serializable but it > seems the equivalent services for ORM are not necessarily making this > effort. _______________________________________________ hibernate-dev mailing list hibernate-dev@lists.jboss.org https://lists.jboss.org/mailman/listinfo/hibernate-dev