| Description | Test #1:
After duplicating a survey, we tried to remove the conditions through the relevance equation (didn't know it was a condition, thought it was relevace). We were able to do it and it worked good.
That lead to false expectations as you may set the relevance equation to 1 but the condition are still on the DB. When exporting and importing conditions took precedence and things were not working as expected. (see steps below).
Test #2: While truobleshooting, ...
When entering a condition in the designer, the relevance equation is not updated.
Also, the relevance equation is not set to readonly nor the "there is a condition" tip appears. |
|---|
| Steps To Reproduce | This is what happened to us:
1- On LS4, we duplicated an existing testing survey which was using conditions.
Conditions are shown as equations on the "relevance" attribute (the one use for showing and hidding), so they look kind of the same.
On v3 you can tell if there is a condition or a relevance equation because the relevance equation is readonly when there is a condition, as well as being a tip letting you know about the situation.
Still, on v4, the readonly mechanism and the tip seemed to have gone away. So you can't tell at first sight if there is a condition or a relevance equation.
2- We removed the relevance equation (set it to "1" - alwasys shown) and things looked good on our environment. Still, seems that on the database, the conditions were still there.
3 - When exporting the LSS, the conditions were exported as well.
4 - When importing the LSS, it seems the conditions took precendece over the relevance equation, so some questions were shown depending on the answers of others.
The initial reaction and interpreation of what happened was "questions are shown randomly".
Still, it wasn't that but the conditions doing their work.
Later, when we found out, we removed the conditions from the condition editor, exported and then everything worked out good. |
|---|
