You still need z/OS domain knowledge? I'll step up to the plate. Dan Phipps Infrastructure Management Consultant ITO Global Service Delivery ACS, A Xerox Company 10309 Wilson Blvd Blythewood, SC 29016 p 803-753-6547 f 803-753-6794 m 803-223-3770
-----Original Message----- From: Nicholas Clark [mailto:[email protected]] On Behalf Of Nicholas Clark Sent: Tuesday, February 07, 2012 9:05 AM To: [email protected]; [email protected] Subject: Re: porters wanted: building, testing, patching on z/OS On Fri, Jan 27, 2012 at 05:08:41PM -0600, Craig A. Berry wrote: > And I'm adding a cc to perl-mvs (just like I did back in September) > since one thing that earlier thread demonstrated is that while no one > on perl5-porters is admitting to knowledge of or access to z/OS, there > are people on perl-mvs who are currently running Perl on that platform > and would like to see that capability continue even if they can't > currently contribute to making it happen themselves. They are the > biggest stakeholders in this, after all, though perhaps some of them > were satisfied with IBM's answer that if IBM decides a current port is > needed, IBM will do it. I don't know if IBM are different from Microsoft, Apple or Google, but but generally when it comes to "bigcorp", one doesn't have a good negotiating position if "bigcorp" decides something that doesn't fit with your own needs (or survival requirements) > I'll just note in addition that in the early stages of a renewed > porting effort, domain expertise may well be more important than Perl > expertise. Getting a build environment set up, flinging the Perl > sources at it, examining the broken pieces for clues about what to do > next, and so forth, will likely go for several rounds before one gets > very deep into Perl's innards. Of course both Perl and z/OS expertise > would ultimately be required. But, really, there's no *requirement* for the initial individual(s) to do the entire hit. Right now, someone with just z/OS domain knowledge of "how to build things", how to navigate around the file system, what the local equivalent of "cd", "less" and "diff" are would get a long way. Just being able to report back with build failure output would let everyone else solve the initial problems, and get things to a much better state than they are now. This would be a heck of a lot better than what we have now, which is a lot of effectively dead code, some of which is actively getting in the way of progress. And there's an easy option on that, which no-one on the perl-mvs list would like, but if nothing else changes, it's going to be the only solution left on the table. Please don't all step backwards at once* Nicholas Clark * Can't find an image, but it's the 4th bullet point on this list http://tvtropes.org/pmwiki/pmwiki.php/Main/GotVolunteered
