Cobol - Memory management (was: Perfrom Thru

Author

Memory management (was: Perfrom Thru

William M. Klein

2004-04-13, 11:30 pm

Interestingly enough, even when the Standard *did* include a "segmentation"
module, it made not implicit or explicit statements about "loading" or not
loading specific segments. It certainly ALLOWED for code not to be "loaded into
memory" - but it made no such requirement.

This is similar to a (misconception) about the CANCEL statement. There is
nothing in the Standard that says that a CANCEL statement must "free" from
memory the storage of a subprogram.

The Standard does talk about the SEMANTICS of CANCEL (and previously
segmentation) but these can be fully met with or without ever doing "memory
management".

--
Bill Klein
wmklein <at> ix.netcom.com
"Paul Knudsen" <HughG@dodgeit.com> wrote in message
news:mr6p70hvpusjv3t9mg7fck68b3c1vfvfj7@
4ax.com...
> On Tue, 6 Apr 2004 16:43:19 -0700, "Chuck Stevens"
> <charles.stevens@unisys.com> wrote:
>
>
> To only load rarely run code when it is needed.
> --
> Get on the NRA Blacklist: http://www.NRAblacklist.com

Paul Knudsen

2004-04-14, 11:30 pm

On Wed, 14 Apr 2004 02:05:54 GMT, "William M. Klein"
<wmklein@nospam.netcom.com> wrote:

>Interestingly enough, even when the Standard *did* include a "segmentation"
>module, it made not implicit or explicit statements about "loading" or not
>loading specific segments. It certainly ALLOWED for code not to be "loaded into
>memory" - but it made no such requirement.
>
>This is similar to a (misconception) about the CANCEL statement. There is
>nothing in the Standard that says that a CANCEL statement must "free" from
>memory the storage of a subprogram.
>
>The Standard does talk about the SEMANTICS of CANCEL (and previously
>segmentation) but these can be fully met with or without ever doing "memory
>management".

I guess with modern storage management it doesn't much matter.
--
Get on the NRA Blacklist: http://www.NRAblacklist.com

Sponsored Links