On 2015-02-02 17:59, John McKown wrote: > Long ago, 1980s, we used VM/CMS for program development. On identical > hardware, it beat the dickens out of TSO/ISPF. I still have fond memories > of XEDIT. Of course, ee ran MVS on a separate machine. We ran MVS under > VM/SP1 for testing, and with shared SPOOL to submit jobs to run on the > production machine. z/OS might run well if it would use the VM "helper" > interfaces like VSE. > We tried using VM/CMS for program development, circa 3 decades ago. Rapidly gave up because shared (LINK ... MR) access to MACLIBs with concurrent updates was a nightmare. We had a service machine that serialized updates to the MACLIBs, so they were never actually corrupted, but we got I/O errors because the clients' in-storage caches of the directories were stale and appeared to be corrupted. We moved our MACLIBs to MVS. Did you confront that problem?
Nowadays, the logical representation of a MACLIB would be an SFS directory. I doubt that IBM has implemented that. CMS ISPF makes things even worse. > On Feb 2, 2015 6:46 PM, "Ed Gould" wrote: > >> So, it was IBM saying if you don't run VM, FY? >> Only if you want to have informally developed tools, and see source. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
