> Yea, and I didn't make an attempt to report this as a bug or potential > problem. I did email this list a few times and Peter offered several > suggestions which did not provide any resolution. > > Oddly enough, I can pass my 5.8 test suite with 98-99% success. Only > afterwards during run time do the socket problems crop up. I think the test > suite may have been watered down enough for the S/390 platform that one or > more critical tests were bypassed.
Off-hand I can't see that many tests being skipped for $^O eq 'os390'. There's one socket test among those but it's a very specific one, related to the socketpair() call, not vanilla TCPIP client-server stuff. There are more tests skipped because of ASCII-assumptions which EBCDIC does not support, but again these should not affect TCPIP stuff. > If you can offer any suggestions I would be glad to try them out...we would > much rather prefer to deploy 5.8 with our z/OS 1.2 package instead of the > older 5.6.1 build. I'll read the list archives to see what's been already tried. -- Jarkko Hietaniemi <[EMAIL PROTECTED]> http://www.iki.fi/jhi/ "There is this special biologist word we use for 'stable'. It is 'dead'." -- Jack Cohen
