Remove broken XStream2 test of legacy custom ConcurrentHashMapSerialization #2071
Conversation
…ation This one requires some explanation: there is a bug As custom serialization of ConcurrentHashMaps was removed circa 2010 (see eb51d46) We can assume that testing conversion is no longer a high priority. Root cause appears to be a subtle change in the XStream attribute access behavior when run under JDK8 only. This appears to be the only case that triggers it.
|
@reviewbybees specifically @jglick and @daniel-beck who had requested this be included. |
| @@ -296,11 +296,6 @@ public void concurrentHashMapSerialization() throws Exception { | |||
| } finally { | |||
| v.delete(); | |||
| } | |||
|
|
|||
| // should be able to read in old data just fine | |||
There was a problem hiding this comment.
The Javadoc for the test is now incorrect—it is no longer testing that we
still can deserialize to [from?] older, verbose format
Since it is no longer testing Jenkins code at all, I would rather suggest just deleting the whole test case—the part you left behind was only a kind of control case for the real test which is now deleted.
|
I would propose just deleting |
|
This pull request originates from a CloudBees employee. At CloudBees, we require that all pull requests be reviewed by other CloudBees employees before we seek to have the change accepted. If you want to learn more about our process please see this explanation. |
… conversion dating back to 2010 and does not work with Java 8
|
@jglick Removed them both, as separate commits in case any issues arise from the ConcurrentHashMapConverter Edit: specifically Java 7 users upgrading from ancient installations. |
|
🐝 (maybe @kohsuke can explain what this was about to begin with) |
|
🐝 |
|
🐝 I was hit be this today in junit test suite. If the test suite has |
|
@reviewbybees done |
|
This pull request has completed our internal processes and we now respectfully request the maintainers of this repository to consider our proposal contained within this pull request for merging. |
|
PR build does not seem to have worked. |
|
Oh for goodness sakes. Interrupted errors and bizarre socket quirks. Once this one finishes it might need yet another build sigh. |
|
No word from @kohsuke so I am going ahead with a merge. |
Remove broken XStream2 test of legacy custom ConcurrentHashMapSerialization
…s it prevents use of JDK 8 to run core tests.
This quirk causes failure of the the Jenkins test suite when run under JDK8.
This one requires some explanation: there is a bug with testing deserialization of the custom ConcurrentHashMap serialization that existed before 2010 (see eb51d46). Given the age, we can assume that testing conversion is no longer a high priority.
Root cause appears to be a subtle change in the XStream attribute access behavior when run under JDK8 only. This appears to be the only case that triggers it. I've left in as much as possible of the test for coverage, while removing the offending piece.