Then let's kick XMLBeans to POI 5.0 and drop Java 6 ASAP if it's preventing Java 9 support.
Java 9 is just around the corner (July 27). We should aim to support Java 9 the day it's released so we're not a reason for other software projects to delay Java 9 adoption. Have we addressed the open issues with POI 3.17 beta 1 that we'd be comfortable with a 3.17 final release? We could make 3.17 final and 4.0 nearly identical, released the same day or a few days apart, with the only difference being dropping Java 6 compatibility and adding Java 9 compatibility. On Jul 9, 2017 5:47 PM, "Andreas Beeker" <kiwiwi...@apache.org> wrote: > +1 for switching now or after POI 3.17 ... if we will do a 3.17 final > soonish. > > I'm actually bringing this up, as I was facing again problems with > generics bugs in Java 6 and independently had a devops discussion with > fluxo yesterday, questioning me/us why we are still on Java 6 level. > > > If we're talking about POI 4.0, we should also think about replacing > > XMLBeans, since that can't be done gradually like refactoring the code to > > use Java 7 language features. > > Switching to Java 7 is probably just a declarative issue in the beginning > and we gradually upgrade the source in the releases afterwards - I don't > see a XmlBeans replacement coming so soon and we shouldn't pack to many > changes in version 4.0. Furthermore I don't think replacing XMLBeans can be > done gradually - at least this should happen module-wise in my point of > view. > > I also wouldn't spend any time to think about Java 6 mitigations - as the > long term support is running out for most customers - at least that's what > happened to my last project in a quite update-reluctant company with a IBM > java environment. > > Andi > > >
