Assembler is certainly powerful and flexible, but that should be used to make
things easier to follow and more productive. As an example, my structured
programming object-oriented macro language which I used to write the IBM CICS
coupling facility servers allowed me to code things like the following:
USING MYOBJECT,R10
ALLOCATE MYOBJECT
That would allocate heap storage for an instance of the object MYOBJECT, copy
in the initial value (addressed via an A-type constant) and set R10 to its
address (which was an interesting technique in itself). If the MYOBJECT object
definition was copied into the module which owned it, it generated an
initialised copy with the structure name available as an entry point. The
initial value typically included method definitions which contained addresses
of entry points in the owning module. If it was copied into some other module,
it generated a DSECT instead, and any fields which were not in the public
section were automatically made anonymous so they could not be referenced by
name. Provided all public fields preceded all private fields, it wasn't even
necessary to recompile any modules apart from the owner if the private fields
were changed or extended, because the length used for acquiring storage was
obtained from a length prefix in the initial value.
Jonathan Scott
-----Original Message-----
From: IBM Mainframe Assembler List <[email protected]> On Behalf
Of Martin Ward
Sent: 17 February 2026 19:53
To: [email protected]
Subject: Re: Clever code to celebrate
On 17/02/2026 19:11, Seymour J Metz wrote:
> I would call it obscure. Coding instruction as DC, and not even
> honoring instruction boundaries, serves no purpose exceptmaking it
> harder to read.
Isn't the purpose a celebration of the power, flexibility and diversity of
assembler?
--
Martin
Dr Martin Ward | Email: [email protected] | http://www.gkc.org.uk
G.K.Chesterton site: http://www.gkc.org.uk/gkc | Erdos number: 4
