Both the PR build and the commit build should include a javadoc build once we fix those. It's quite easy to break javadoc refs. You just add javadocs, refactor stuff and boom.. javadoc gets broken.
On Tue, Jun 9, 2015 at 3:46 PM, Daniel Kulp <[email protected]> wrote: > >> On Jun 8, 2015, at 6:47 PM, Clebert Suconic <[email protected]> >> wrote: >> >> True, >> >> although I think it makes sense to fix the javadoc stuff anyways to >> not require that setting > > I started looking at this. Definitely a lot more involved than I was > expecting. It was definitely the right call to add the setting to get the > release out. > > At this point a “mvn -Pdev install” will now work without that setting. > That’s a good start. The javadoc run in artemis-website now will run > without an error. (plenty of warnings though) However, javadoc in the > individual modules still fails. Since the javadoc in the modules is part of > the deploy (and release) process, we still need the switch. > > One step at a time….. > > Dan > > >> >> On Mon, Jun 8, 2015 at 6:41 PM, Hiram Chirino <[email protected]> wrote: >>> I would not recommend exerting too much effort in maintaining Java 7 >>> support since Java 7 is EOL. The only type of lib I would say should >>> keep old Java support for is client libs. There are some platforms >>> out there that don't rev that quickly (stuff like GWT, Android, etc). >>> Would it makes sense to keep clients libs in builds outside of >>> Artemis? >>> >>> On Mon, Jun 8, 2015 at 5:57 PM, Clebert Suconic >>> <[email protected]> wrote: >>>> The first try was using profiles but it was kind of messy. This was >>>> done very closely to the releases and I didn't have time to evaluate >>>> any other options back then. >>>> >>>> If all we need is to fix javadoc, I would say we fix javadoc and >>>> remove the java8 dependency (at least for now). I'm not sure yet how >>>> difficult that would be though >>>> >>>> >>>> >>>> On Mon, Jun 8, 2015 at 5:45 PM, Robbie Gemmell <[email protected]> >>>> wrote: >>>>> Do you mean the -Xdoclint:none option? Javadoc is stricter with Java8 >>>>> and will refuse to process things that 'worked' with Java7. You can >>>>> make it lenient again using the -Xdoclint config option, but that only >>>>> works when using Java8 and so setting it then makes javadoc processing >>>>> fail when building on Java7 since it doesnt understand the new config >>>>> option. >>>>> >>>>> When I first hit this elsewhere I just updated all the javadoc to >>>>> remove the errors seen using Java8, allowing things to work on 7 or 8 >>>>> without disabling doclint. You could possibly use profiles to apply >>>>> the config selectively. >>>>> >>>>> Robbie >>>>> >>>>> On 8 June 2015 at 22:23, Clebert Suconic <[email protected]> >>>>> wrote: >>>>>> The only thing I remember we had to add JDK 1.8 to the equation was >>>>>> some option we needed for building javadocs. Perhaps there is a better >>>>>> way to solve that. >>>>>> >>>>>> Right now the codebase is not using anything specific to JDK 1.8. (I >>>>>> mean.. at least that's the idea. we could have slipped something.. but >>>>>> I don't recall anything specific to java8 in the codebase) >>>>>> >>>>>> On Mon, Jun 8, 2015 at 2:35 PM, Daniel Kulp <[email protected]> wrote: >>>>>>> I’ve updated a bunch of things so that Artemis now loads fairly easily >>>>>>> into Eclipse without any errors for all the non-example things. I >>>>>>> haven’t attempted the examples yet. >>>>>>> >>>>>>> Just have a couple of questions: >>>>>>> >>>>>>> 1) In the poms, we specify that Java8 is required to build, but java7 >>>>>>> is used for the source/target. Thus, Eclipse will pick up the Java7 >>>>>>> runtime. It seems to work OK so I’m kind of wondering why we require >>>>>>> Java8 to build. Maybe in the examples someplace? >>>>>>> >>>>>>> 2) artemis-dto has a profile for jdk-1.5. I assume that is not needed >>>>>>> at all as there is no way it would ever be triggered. I think the >>>>>>> ibmjdk profile in there is irrelevant as well? (seems to reference >>>>>>> things about differences between 1.5 and 1.6) >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Daniel Kulp >>>>>>> [email protected] - http://dankulp.com/blog >>>>>>> Talend Community Coder - http://coders.talend.com >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> Clebert Suconic >>>>>> http://community.jboss.org/people/[email protected] >>>>>> http://clebertsuconic.blogspot.com >>>> >>>> >>>> >>>> -- >>>> Clebert Suconic >>>> http://community.jboss.org/people/[email protected] >>>> http://clebertsuconic.blogspot.com >>> >>> >>> >>> -- >>> Hiram Chirino >>> Engineering | Red Hat, Inc. >>> [email protected] | fusesource.com | redhat.com >>> skype: hiramchirino | twitter: @hiramchirino >> >> >> >> -- >> Clebert Suconic >> http://community.jboss.org/people/[email protected] >> http://clebertsuconic.blogspot.com > > -- > Daniel Kulp > [email protected] - http://dankulp.com/blog > Talend Community Coder - http://coders.talend.com > -- Clebert Suconic http://community.jboss.org/people/[email protected] http://clebertsuconic.blogspot.com
