Bug
Auto commit off: dobject state during trigger execution
Issue description
Numerous regression tests failed due to the fix for the causing issue. There seems to be an issue of what is considered a new topic in FastWriter. This is important for various reasons and seems to fail now, so that old topics are considered new.
Developer comments
The actual reason here was that the set_name change on a tobject was not recognized as an update and there for the AFTER INSERT triggers fired instead.
The triggers need an accurate and 100% uptodate representation of the topic map data in all tobjects. This was done too late, after a root package (together will all child packages caused by triggers were saved. Now it is done after each individual package ensuring that the executing trigger performs just like when the data would have been persisted in a previous request. Basically triggers that operated on a new topic were reading from MySQL, since the tobject had a null dobject.
|
Work sessions3
Start |
2024-02-10T12:35:43
|
End |
2024-02-10T14:10:46
|
Participant |
Robert Cerny
|
Start |
2024-02-10T17:00:00
|
End |
2024-02-10T19:54:16
|
Participant |
Robert Cerny
|
Start |
2024-02-11T09:06:26
|
End |
2024-02-11T13:15:43
|
Participant |
Robert Cerny
|
|
We are sorry
This page cannot be displayed in your browser. Use Firefox, Opera, Safari, or Chrome instead.