Sorry high workload have prevented me from responding.

Workflow. Creating a Attributeremover (AR) where attribute A is removed after that (at a later stage) attribute A is removed by another AR or mayby by change of input data (illustration).

Firstly - there is no indikation "red star" on the AR - you have to open it to se there is an error.

Secondly - on version ESRI 10.2 these errors will, sometime, mix up the mapping to the source table (the lines will be mixed), On version 10.3 the error will not show at all - how ever data will be wrong.

We noticed on a conversion that a specifik attribute in the model had 6754 elements with values but the resulting database had 7456 elements, when analyzing the problem we noticed that almost all the 7456 values where wrong.

We have implemented a internal workaround where all AR is double checked before running a model.

TAken into account the implication of the error - The big issue is the fact that FME/INterop do not recognize, nor visualize the error..

Best regards

Jesper

I think the reason the cog wheel for the AttributeRemover is not red is because it's debatable whether or not it's an error (its task is to remove an attribute that's not there, so the end result is the same, right?). I do agree it's something that's worth noting though so I'll make an idea for that.

However, I'm still not entirely sure I understand what's going on so if you could elaborate please:

Your process has changed so the AttributeRemover is now set to remove an attribute that isn't there anyway. This should, in my opinion, not impede the process. However, if I read you correctly you are getting values in your output data for that attribute.

Could it be that you're doing an update on an existing table? In that case missing attributes are not updated apparently. Creating it and setting it to empty should do the trick in that case.

If that's not it we should be able to help you out better if you could share your entire workspace.

I can agre that its debatable whether its an error - but the implication to the result is certanly an error.

The problem with the error is that its not consistant. Mening it hard to replicate it.

The history off the error is that we on Interop ESRI 10.2 noticed that the mapping between last operator and the destinaion in some situations where mixed up - usually most off the "attribute mapping lines" where mixed up or removed - the error appeared when saving and reopening a "model".

First move was to upgrade (10.2 -> 10.3) and it seemed to fixe the problem. However we started digging into the error (on 10.2) and found that the mappning mixup - only appeared when there was the above AR errors in the model. Unfortunately the error reaperes on 10.3 but now the "mapping lines" where intakt but the data writen to destination was not correct the. The attributes the error appears on is placed in the bottom off the last operator..

I find it very hard to explain however i have a task in NeXT sprint to make a "video" for internal use maybe that is the best way to illustrate for you to..

We have tried both trunc/update and recreate entire table same result.

best regards

Jesper