Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Pathname delimiter ch

1 view
Skip to first unread message

Paul Rogers

unread,
Feb 3, 1995, 9:06:00 PM2/3/95
to
[v]>I know that MSDOS uses \, UNIX uses /, and MacOS uses :, but what do
[v]>other OS's use for pathname delimiter characters (i.e. separating the
[v]>directory or folder names in a file's pathname)? And what were the
[v]>motivations behind a particular character being used?

Well, MVS uses .
---
* SLMR 2.1a * Sects! Sects! Sects! Is that all Monks think about?

Paul Rogers

unread,
Feb 3, 1995, 9:29:00 PM2/3/95
to
PT[v]MVS/SP with TSO and SPF didn't have subdirectories. But they did allow yo

[v]use long dataset names with dots in them. And they had partitioned datase

[v]where each dataset had a bunch of members. Unfortunately, editing a membe


I'm sorry, but MVS does indeed have what we call subdirectories. Each
of those periods delimits a level of the directory tree. No directory
name between periods may be longer than eight characters, the sum must
be no more than 44 characters.
---
* SLMR 2.1a * It's as bad as you think and they are out to get you.

Paul Tomblin

unread,
Feb 4, 1995, 11:24:09 AM2/4/95
to
In a previous article, paul....@nwcs.org (Paul Rogers) said:

>PT[v]MVS/SP with TSO and SPF didn't have subdirectories. But they did allow yo
>
> [v]use long dataset names with dots in them. And they had partitioned datase
>
> [v]where each dataset had a bunch of members. Unfortunately, editing a membe
>
>
>I'm sorry, but MVS does indeed have what we call subdirectories. Each
>of those periods delimits a level of the directory tree. No directory
>name between periods may be longer than eight characters, the sum must
>be no more than 44 characters.

I wouldn't call them subdirectories, because in order to list all the files
in say, TCEC.PTSD, the listc or listcat utilities had to sequentially search
the VTOC of every single disk pack in the entire installation. If they were
real subdirectories, then all the files for that level would be on one
location in the VTOC.


--
__ _ Paul Tomblin
/ / (_)__ __ ____ __ This is my computer, and I speak for me.
/ /__/ / _ \/ // /\ \/ / (http://watt.oedison.com:8080/~tomblinp/)
/____/_/_//_/\___/ /_/\_\ ... It's not just for breakfast anymore. (mdw)

Geoff Lane

unread,
Feb 7, 1995, 7:50:40 AM2/7/95
to
Paul Rogers (paul....@nwcs.org) wrote:

: Well, MVS uses .

And RiscOS

--
Geoff. Lane. | Internet: zza...@cs6400.mcc.ac.uk | Janet: zza...@uk.ac.mcc
CS6400 Sys Admin, Manchester Computing Centre, Oxford Rd, Manchester, M13 9PL

Greg Limes

unread,
Feb 20, 1995, 4:47:49 PM2/20/95
to
In article <792463...@warpfive.demon.co.uk> a...@warpfive.demon.co.uk (Andrew Clegg) writes:

| In article <3hcj47$g...@osfa.aber.ac.uk>
| a...@aber.ac.uk "Alun "Da Penguin" Jones" writes:
|
| > Not a problem. The convention under RISC OS is to use h.stdio, so
| > all your headers go in one directory, C files in another, objects
| > in another etc. The compiler magically does the swap as it works,
| > so #include "fred.h" works fine to include "h.fred". It also translates
| > Unix style filenames - #include <sys/time.h> will translate to
| > sys.h.time.
|
| Isn't this a wart or a misfeature or something? After all, what if you
| really did have a file called h in the directory sys/time that you wanted
| to include?

Then you would say "#include <sys/time/h>" and RISC OS would
helpfully grab the file "sys.time.h" ...
--
-- Greg Limes li...@3do.com, li...@netcom.com
"Your reality check is in the E-mail"
Not speaking for my employer, of course
PGP key available on request

0 new messages