On Fri, 31 Aug 2018 06:52:57 -0500, Jerry Callen <[email protected]> wrote:
>Andrew Rowley wrote: > >> When using source control you STILL need to make sure that 2 people >> are not updating the same file at the same time - it is just the >> window that is smaller. > >On z/OS you could solve that with DISP=OLD (though that's not practical >in all situations). Not if you have some other process, such as with Zowe, that does not use the ENQ that the system uses to prevent simultaneous updates. >Your git server becomes the canonical >reference, and you update the PDS (via an automated process) when >changes are merged into the "master" branch by that server. I remain skeptical that an automated process can in all cases resolve the updates that were made by more than one person to a single element. There have been cases where I have made changes to a source module and had to consult with someone else who was also making changes to that same element in order to resolve the differences. That after having determined exactly what changes the other person had made. And sometimes two people can solve the same problem in a module in different ways >> We are where we are - it is important that existing functions continue >> to work as expected. So, please, make Zowe edit compatible with ISPF >> edit serialization. > >I think the proposals by Matt Hogstrom and Kirk Wolf solve the same >problem in a better way. I don't think Zowe should perpetuate a >practice that, IMO, is actually broken. The problem was solved by the use of ENQ over 50 years ago. If Zowe does not always use the enqueues, it breaks that solution, and those proposals cannot solve the problem at all, let alone in a better way unless either: 1. Everyone who uses ISPF or batch processes to update data sets is compelled to immediately change to use Zowe, and there is some mechanism to enforce that, or 2. ISPF and batch processes are updated to use the same procedures that are used by Zowe, and those updates are installed on all systems that have the potential to update these data sets. Has anyone even submitted an RFE to try to get such changes made? Not that I would guess that there is much likelihood that such an RFE would be accepted. Note that for case 1, you have painted yourself into a corner where a set of changes that cause an IPL to fail can result in the inability to recover. The insistence that we should do things a "better way" without maintaining compatibility is short-sighted at best. -- Tom Marchant ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
