[GitHub] [spark] srowen commented on pull request #28661: [SPARK-31849][PYTHON][SQL] Make PySpark exceptions more Pythonic

2020-05-28 Thread GitBox
srowen commented on pull request #28661: URL: https://github.com/apache/spark/pull/28661#issuecomment-635693433 Well, I'd say it differently. A Python person may not know what a JVM stack trace means. Taking it away doesn't itself do much except shorten a big dump of output, which doesn't

[GitHub] [spark] srowen commented on pull request #28661: [SPARK-31849][PYTHON][SQL] Make PySpark exceptions more Pythonic

2020-05-28 Thread GitBox
srowen commented on pull request #28661: URL: https://github.com/apache/spark/pull/28661#issuecomment-635383500 Yeah that stack trace can be the only useful info in some cases, when the real error is on the JVM side. I get that Python users don't understand it as well. But otherwise by def