Re: existence of directory

On Dec 7, 2:19 pm, glen herrmannsfeldt <g...@xxxxxxxxxxxxxxxx> wrote:
GaryScott <garylsc...@xxxxxxxxxxxxx> wrote:

(Richard wrote)

Directories are much more common today. As in I could make the opposite
statement - I can't name a single system that I have used in recent
years that doesn't support them. But I think some exist, and are even of
significant importance, just not on systems I happen to use today. Some
IBM mainframe operating systems come to mind.
VM (including CMS) has supported directory tree structure for about a
decade, but you have to do special things to enable and use them.
They don't really fit in well for the way some things are done.

Well, even without tree structure VM/CMS gave you 26 mini-disks,
which sort of work like 26 directories (no tree, though).

Yes, but in actual implementation, most users only got an "A" disk,
everything else was read-only, part of the system, or provided
"libraries" and "tools". And space allocation was tiny by today's
standard (as was the user memory space to run applications, often
about 1500k). I'm still perfectly comfortable with the way CMS works,
but expanded support for tree structured directories allows easier
interface to the rest of the world. And most of the old extreme
memory/space limitations are gone now. Heck, you can even multitask
within a single VM now (woo hoo)!

MVS and predecessors give 44 character filenames which were normally
divided up such that part was a user identifier.  Physically a flat
directory structure (across the whole disk) but often used as
a logical tree structure.  That system catalog was tree structured.

Then there is the PDS which does have a directory, but again no
tree structure.  (Partitioned Data Set, similar to what are known
as libraries on other systems, such as unix ar files.)

Yup, I did use MVS more than I was one of the least user
friendly systems ever created (thus the need for VM).

-- glen