[rules-users] Upgrade options and considerations

[Stephen Masters]

I'm not sure there's a huge benefit. The main reason I was looking at upgrades from 5.3 recently was to sort out the problem in Guvnor where LHS and RHS rows are misaligned in Guvnor web decision tables. That was fixed in 5.4.

Having just attempted the migration, I would agree with avoiding 5.5.0.Final. Too many Guvnor features were broken, so I had to just give up on that project. The breakage of rule templates wasn't a serious issue for me, but the loss of data enumerations was a killer. For now, we're keeping 5.3 decision tables short, and hoping that 5.5.1 or 5.6 might resolve some of the critical regressions.

So that would be my tip … unless you're experiencing problems on 5.3 that were resolved in 5.4, just hang fire on the effort of upgrading until a patch or two have gone out on 5.5.

Steve


On 11 Apr 2013, at 17:06, Lance Leverich <lance.leverich at gmail.com> wrote:

> My company started out using Drools/Guvnor v5.1.1, for a single product/project. Later, we decided to start building an enterprise oriented service, which was partially completed and is based upon Drools/Guvnor v5.3.0.
> 
> Recently, the decision has been made to get all of our Drools/Guvnor instances running the same version. I ran into issues (mostly related to dependencies and the use of rule-flows) when I initially attempted to just get everything running on v5.3.0. But those issues seem to have all gone away when I replaced my rule-flows with BPMN's and moved everything to a v5.3.5 instance.
> 
> My question, for this group, is whether there are good reasons to move even further forward and upgrade everything to v5.4.x? If so, what might those reasons be? Are there any known, significant, issues in one or more of the 5.4 versions? Is there any consensus on which patch version of the 5.4 tree would be best? I have already been told by upper management that they would not support moving to v5.5.x, so that's not even on the table.
> 
> Any feedback would be appreciated.
> 
> Lance Leverich
> _______________________________________________
> rules-users mailing list
> rules-users at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-users