Hello!
Posted: Fri Jul 18, 2014 8:25 am
Hello -
My name is Bruce Yelen and I've been involved with mainframes since 1975 (yeah, I'm an "old-timer"). I "stumbled" across the forum while searching for a new position, and decided that participating will be a great way to determine what others are doing.
For many years my focus has been CICS, both as an applications and a systems programmer. For the last 15 years I've been writing middleware that allows client applications using Sybase database servers on Windows and Unix/Linux systems to make RPC calls and language requests to the mainframe (via CICS) that looks to the client as just another database server. I also support an api that allows CICS applications written in COBOL/PLI/HLASM to return data in response to these RPC calls. This involves some very heavy "down and dirty" use of socket interfaces, exits, and so forth to accomplish. I'd say 70% of our code is assembler, with the other 30% being written in "C". Our only use of COBOL is for example code that we provide our customers.
In the "Ancient History" department, I started with the very earliest versions of the operating system we now know as z/OS. If I haven't left anything out, the progression went something like this: OS360/MFT -> OS360/MVT -> OS/VS2 -> MVS/XA -> MVS/ESA -> OS390 -> z/OS. MFT and MVT stood for "Multi Processing with a Finite number of Partitions" and "Multi Processing with a Variable number of Partitions". With MFT, you could only run as many jobs as partitions, and each partition was hard-coded into the sysgen with a specific amount of storage available. MVT did not use partitions, but you could not exceed your total amount of memory. OS/VS2 was the first in the series that supported the new virtual memory features of the old IBM System/370 line.
I've also worked with ACP (Airline Control Program) which was a forerunner of today's TPF. It's been an interesting ride and just gets better!
- Bruce
My name is Bruce Yelen and I've been involved with mainframes since 1975 (yeah, I'm an "old-timer"). I "stumbled" across the forum while searching for a new position, and decided that participating will be a great way to determine what others are doing.
For many years my focus has been CICS, both as an applications and a systems programmer. For the last 15 years I've been writing middleware that allows client applications using Sybase database servers on Windows and Unix/Linux systems to make RPC calls and language requests to the mainframe (via CICS) that looks to the client as just another database server. I also support an api that allows CICS applications written in COBOL/PLI/HLASM to return data in response to these RPC calls. This involves some very heavy "down and dirty" use of socket interfaces, exits, and so forth to accomplish. I'd say 70% of our code is assembler, with the other 30% being written in "C". Our only use of COBOL is for example code that we provide our customers.
In the "Ancient History" department, I started with the very earliest versions of the operating system we now know as z/OS. If I haven't left anything out, the progression went something like this: OS360/MFT -> OS360/MVT -> OS/VS2 -> MVS/XA -> MVS/ESA -> OS390 -> z/OS. MFT and MVT stood for "Multi Processing with a Finite number of Partitions" and "Multi Processing with a Variable number of Partitions". With MFT, you could only run as many jobs as partitions, and each partition was hard-coded into the sysgen with a specific amount of storage available. MVT did not use partitions, but you could not exceed your total amount of memory. OS/VS2 was the first in the series that supported the new virtual memory features of the old IBM System/370 line.
I've also worked with ACP (Airline Control Program) which was a forerunner of today's TPF. It's been an interesting ride and just gets better!
- Bruce