jira-importer commented on issue #489: URL: https://github.com/apache/maven-indexer/issues/489#issuecomment-2965144987
**[Milos Kleint](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=mkleint)** commented I'm not so sure that our synchronisation is the fault here. We used to have synchronisation even before 5.0.0 upgrade, all access is a write mutex AFAIK, used to have read mutex for searches for a while I believe but with the previous versions it didn't work out, so we changed it all to write access. so always just one man (thread) accessing the whole thing (currently per context, used to be global) there could some unrelated access to the files from the IDE, like native listeners getting hold of the file changes and the IDE processes them. AFAIK there's no listener watching over the cache folders with maven indexes though. According to the report at [1] there only 2 incidents reported, both probably coming from the same guy on the same same day. so let's call it a user setup error for now. but I believe my initial evaluation still stands, if something unexpected happens, the replace() method doesn't recover well. [1] http://statistics.netbeans.org/exceptions/detail.do?id=193177 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
