Skip to main content

HI,

I use a FME Server 2016 to run different workbenches. For the update on FME Server 2018 I have to upgrade these workbenches, too.

Therefore I upgrade all transformers, which are upgradable. With the dissolver I get some problemes:

old workbench:

new workbench

In the new workbench I get less objects then in the old workbench, but the result must be the same.

The new dissolver has the addional parameters: Connect Z Mode, Tolerance and Aggregate Handeling.

Which are the correct configurations the get the same output like in the old workbench?

 

Thanks and best regards

 

 

What you're observing is the main reason why FME doesn't automatically upgrade the transformers: to avoid any surprises in transformer behavior due to changes in the underlying transformer algorithm.

Concerning the mentioned settings :

  • Connect Z mode: this is only relevant if you have 3D data (Z values). It is also available in FME 2016 and the default value "First Wins" is identical with FME 2018.
  • Tolerance: This is a new setting in FME 2018. The default value of "None" is identical to the behavior in FME 2016. You can also read more about this setting here.
  • Aggregate handling: in FME 2016 all aggregates were automatically deaggregated. In FME 2018 you can choose to send them to the <Rejected> port instead. If you leave the setting at the default "Deaggregate" it should behave as in FME 2016.

If you still get differences, there are two things I'd recommend:

  1. Create a small sample workspace with data that reproduces the issue and send to Safe support so that they can determine if it's a regression or as expected. You can have both versions of the Dissolver transformer in the same workspace.
  2. Do not upgrade the old Dissolver transformer, leave it as-is in your workspace to avoid any changes in the output.

I think the new Dessolver should generate the same result as the previous one, by default.

Make sure that the Tolerance parameter is set to 'None' (default).


Reply