On Sun, 17 Jul 2011 23:35:30 +1000, Clement Clarke wrote: >All Operating Systems run Programs, use Data Sets or Files, and have >some sort of Command Language. > >If Z/OS Users didn't have to be concerned about DCB, SPACE and so on, >then they could write JCL that was very similar to Unix or Linux. > >If all that information is stored in a Data Base of Data Sets, then it >would be quite possible to allocate new data sets when necessary, and >hide all the normal JCL information from the User. > Isn't much of this what SMS is designed to do nowadays?
>All we would have to do is remember which order to put the file names >in, and have a program get all the DCB etc information for output data >sets, and either dynamically allocate them, or generate JCL for them. > Do I understand two separate modes of operation, dynamic and generative? Does it provide en masse enqueues as JCL does, to avoid deadlocks? I suppose this would be intrinsic in a generative mode. How does it interface with JES3 setup processing? In a dynamic mode, are there any restrictions on running APF-authorized programs? Must the interpreter be APF- authorized? Ia it portable? In at least a generative mode, can it run on a non-z platform? I keep much of my JCL as here- documents (in-stream files) in self-tailoring shell scripts on a Solaris platform (could be z/OS Unix except for performance and flexibility). And does it address more of the recurrent JCL complaints: o PARM>100 characters? o In-stream data sets in procedures? (Ah! that's coming in z/OS 1.13.)? o Substitution of dynamic system symbols? o Symbol substitution in in-stream data sets? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
