Does GNU Parallel leak memory? about jobflow

[rofl0r]

maybe it was just the specific perl version i used, however in my case parallel's memory usage with big jobs increased slowly but steady over time.
here's a maillist report of somebody that seemed to have a similar problem:
http://lists.gnu.org/archive/html/parallel/2012-01/msg00020.html

[ole-tange]

As you can see from the message you linked to, the code has been fixed, so there is now an optimization that forgets arguments already seen, so if there is only one input source the memory requirement will remain steady.

The testsuite has since Aug 14 16:54:27 2016 included a memleak test.

So if you cannot come up with an example that shows GNU Parallel has a memory leak, could you please remove the memory leak statement in README.md?