>From my perspective, the mainframe and the open systems (including Mobile Apps) have different agendas and different concerns. On the mainframe we have many things working concurrently (IMS, DB2, CICS, MQ, etc). It seems in the open systems world the applications are limited to one server (or server farm). I am not aware that multiple functions can co-exist on a server.
This is not to say the open systems and mobile apps that do not share these concerns - however The mainframe has massive applications that if they go down the amount of time to recover the application and/or fix it could take many hours/days that costs the business income. Or cause an LPAR wide outage that affects many more working applications (MQ, DB2, IMS, CICS). That is bad for the bottom line. Any "freeware", "Shareware", etc... brought in to a mainframe environment will eventually become a critical piece of production applications - no matter how much you say " THIS IS NOT FOR PRODUCTION" Yes, there is a need for tools. However, in the past, system programmers used assembler programs to create those processes and then they mentored juniors to support that code. In today's environment, rarely do shops have assembler programmers, there are not enough young people going into the Mainframe arena, and management is reluctant to provide the training they did in the past to be able to support these tools. On the mainframe we are also very mindful of resource utilization and performance. I cannot place a REXX tool into production because we do not have the REXX Compiler, and REXX is a resource consumer. Therefore the edict is no REXX. I can have REXX for my personal use. But I cannot share the REXX outside of my group. If REXX goes into production - warning lights and bells go off and the REXX Police come out and haul you away. There is also a concern that when you have an operating system or hardware upgrade, the "tools" will break. I had that happen when IBM moved the UCB Above the line and a 24bit program expected the UCB and not the Address of the UCB. It was something we missed in our analysis because we did not know someone had done that or that it was in production. Fortunately it was localized to a few production batch jobs, but it did take a while to determine why the S0C4 was occurring. I would suspect that once the old guard is gone - the young pups supporting the mainframe will start to do exactly what you are proposing and therefor the mainframe production environment will be more "open" So, say I bring in Python for $99 to read logs and I can do that because my manager has approval for software purchase under $1000.00 . And it is restricted to my use. But someone sees me use it, asks to have the same thing for their logs, and BOOM, now it is rampant in production and a "critical" app. And it all started out innocently enough. I believe that mainframers have learned over the decades there is no such thing as a one-time use or limited to just me ported tool. Things have a way of spreading out quickly and invariably winds up in a critical production path. >From observation, it seems the open systems and mobile apps have a higher tolerance for ported tools and outages - even though it might impact production, and it does not seem to be as high as a concern. I have seen open system apps down for several days because someone used a ported tool that no one knew was there. We can see what happened when IBMLINK moved from a green screen product to a web based application. And how quickly its priority for uptime has seemed to lapse. Just my two cents worth Lizette -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of David Griffiths Sent: Thursday, October 03, 2013 8:15 AM To: [email protected] Subject: Re: Interested in up to date open source software or low cost utilities? On 3 October 2013 15:59, Lizette Koehler <[email protected]> wrote: > I have several downloads I use from the CBT Tape. But I do not incorporate them into a production - if this dies the system dies - process. If the tool I have from the CBT TAPE dies, it does not impact anything but my statistics or analysis functions. Do you think that - along the lines of the CBT Tape - that there is much of a potential market for non-production-critical tools? For instance a company buying a $99 copy of Python for their log processing needs? Cheers, Dave ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
