Fix to the way the 23 handles the object duration in type plus image fixes.#1310
Fix to the way the 23 handles the object duration in type plus image fixes.#1310
Conversation
…ture/remove_old_builds
…ture/remove_old_builds
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## develop #1310 +/- ##
========================================
Coverage 95.04% 95.04%
========================================
Files 111 111
Lines 5368 5369 +1
Branches 33 33
========================================
+ Hits 5102 5103 +1
Misses 266 266
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
lwasylow
changed the title
|
@jgebal Can you review and approve? |
| end; | ||
|
|
||
| overriding member procedure on_event( self in out nocopy ut_reporter_base, a_event_name varchar2, a_event_item ut_event_item) is | ||
| l_event_item ut_event_item := a_event_item; -- ADD THIS LINE |
There was a problem hiding this comment.
In the 23 apparently there is an issue with a duration of the object when casting/treat objects. I think it's oracle bug and is patched in some versions that object is no longer in memory when not using in out.
It's probably something that oracle should fix but easier to do it here then rely on their patch
There was a problem hiding this comment.
I have worked around that problem in our pipeline by using version 23.26.0 instead of 23.26.1.
It is strange to see this exception especially as a_event_item is actually in not in out.
I am not a big fan of making changes in code to workaround Oracle bugs but if it solves the problem permanently than it's OK.
Since it is a solution for a problem, please create a separate issue for that specific error we see, so the error is documented in repository and it should be referenced in code comment.
That way there is clarity why we use this redundant and unnecessary assignment.
With this we minimize the risk of introducing a regression at some point in future, when we are no longer using Oracle version 23.26.1 for builds and testing. Some users may still use it and be prone to the issue.
There was a problem hiding this comment.
I read your comment on #1310. This is resolving this problem then.
I will add a comment in code just to make it super clear.
2 participants
Updating builds and removing old versions of Oracle