Back to the original problem, as reported in post #3 and #7. The fact is that in many cases, we can successfully test a task by clicking "Run selected task", but then, in real life it won't run.
Two main technical reasons:
1) The test is PATH-aware, the real task (using cron) is not
2) The test is insensitive to the output mode ("Default" or "X application"), the real task is not
I think this is indeed a bug worth to be fixed.
At least, the message that appears when testing a task should be improved ("Note about working directory of executed tasks", it is confusing.) Additionally, the "Run selected task" function should be improved.
Back to the original problem, as reported in post #3 and #7. The fact is that in many cases, we can successfully test a task by clicking "Run selected task", but then, in real life it won't run.
Two main technical reasons:
1) The test is PATH-aware, the real task (using cron) is not
2) The test is insensitive to the output mode ("Default" or "X application"), the real task is not
I think this is indeed a bug worth to be fixed.
At least, the message that appears when testing a task should be improved ("Note about working directory of executed tasks", it is confusing.) Additionally, the "Run selected task" function should be improved.