[Red Hat JIRA] (DROOLS-5913) [DMN Designer] Improving UX when Decision Table is relying on not updated data.
by Donato Marrazzo (Jira)
[ https://issues.redhat.com/browse/DROOLS-5913?page=com.atlassian.jira.plug... ]
Donato Marrazzo commented on DROOLS-5913:
-----------------------------------------
I believe that this use case would be almost address when this issue will be addressed:
https://issues.redhat.com/browse/BAPL-1562
In fact, the key aspect is that the decision table headers (input and output expressions) are FEEL expressions.
When users add a new column, they will get intellisense suggestions to set correctly the expression to variable field (e.g. "Vehicle.field").
Regarding the second issue, I believe that it can be addressed by a proper FEEL expression validation that underlines the wrong expression.
Maybe the final resolution would be a complex refactoring feature that is able to keep references between the Data Models and their usage in the FEEL expressions.
> [DMN Designer] Improving UX when Decision Table is relying on not updated data.
> --------------------------------------------------------------------------------
>
> Key: DROOLS-5913
> URL: https://issues.redhat.com/browse/DROOLS-5913
> Project: Drools
> Issue Type: Enhancement
> Reporter: Yeser Amer
> Assignee: Guilherme Gomes
> Priority: Major
>
> We should find a way to improve User Experience in the case a Decision Table is relying on Data Type which was modified by the user.
> Original Message:
> {noformat}
> TL;DR a modification of an Object (aka Structure) in the Data Model does not get reflected in the DMN Model.
> Maybe someone can point out the error in the following scenario.
> (1) Using the DMN editor in BC RHPAM.7.9 I defined the "Ve" Object (aka Structure) and the "Ve_data1" simple variable in the Data Model as custom data types
> (2) "Ve", named as "Vehicle", was used as data input to a DMN model and a decision was defined backed by a decision table
> (3) when the decision table was first created the columns for the "Vehicle" attributes were created as individual columns.
> (4) Modified the "Ve" structure by removing some attributes
> (5) The deleted attributes are still available in the decision table.
> I have two issues in this scenario.
> (I1) Step (3). By creating the columns as individual columns of the "Vehicle" variable in the decision table any sense between them belonging to the "Ve" structure is lost. Furthermore the decision table allows for changing the data type of the columns despite the fact that these belong(?) to the "Ve" object.
> (I2) Step (5). I was expecting that by modifying the columns in the data structure that change would be visible in the decision table as well. Yes, a warning is generated, but it is cryptic, lost in a table full of equally cryptic messages and no visual indication of the actual problem in the decision table. I can check the messages and find out what is going on, but explaining this to my customer proved to be quite a challenge.
> We are still targeting DMN to business analysts I presume.
> {noformat}
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
3 years, 4 months
[Red Hat JIRA] (DROOLS-5913) [DMN Designer] Improving UX when Decision Table is relying on not updated data.
by Yeser Amer (Jira)
Yeser Amer created DROOLS-5913:
----------------------------------
Summary: [DMN Designer] Improving UX when Decision Table is relying on not updated data.
Key: DROOLS-5913
URL: https://issues.redhat.com/browse/DROOLS-5913
Project: Drools
Issue Type: Enhancement
Reporter: Yeser Amer
Assignee: Guilherme Gomes
We should find a way to improve User Experience in the case a Decision Table is relying on Data Type which was modified by the user.
Original Message:
{noformat}
TL;DR a modification of an Object (aka Structure) in the Data Model does not get reflected in the DMN Model.
Maybe someone can point out the error in the following scenario.
(1) Using the DMN editor in BC RHPAM.7.9 I defined the "Ve" Object (aka Structure) and the "Ve_data1" simple variable in the Data Model as custom data types
(2) "Ve", named as "Vehicle", was used as data input to a DMN model and a decision was defined backed by a decision table
(3) when the decision table was first created the columns for the "Vehicle" attributes were created as individual columns.
(4) Modified the "Ve" structure by removing some attributes
(5) The deleted attributes are still available in the decision table.
I have two issues in this scenario.
(I1) Step (3). By creating the columns as individual columns of the "Vehicle" variable in the decision table any sense between them belonging to the "Ve" structure is lost. Furthermore the decision table allows for changing the data type of the columns despite the fact that these belong(?) to the "Ve" object.
(I2) Step (5). I was expecting that by modifying the columns in the data structure that change would be visible in the decision table as well. Yes, a warning is generated, but it is cryptic, lost in a table full of equally cryptic messages and no visual indication of the actual problem in the decision table. I can check the messages and find out what is going on, but explaining this to my customer proved to be quite a challenge.
We are still targeting DMN to business analysts I presume.
{noformat}
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
3 years, 4 months