Mark 'the Landin man' writes:

>That doesn't make it *right*. And, right or wrong (however such things
>are judged), it's a good example of what kind of lobotomy MPE
>would have to go through before it could exist in a UNIX world.

Whether it is *right* or *wrong* is a matter of option. BTW, this is NOT a
good
example since MPE already supports this behavior.

>You dig your hole deeper. :) So now we can't even trust the directory
>... it may or may not list files which are actively being used by
processes.

Yes we can trust the directory. Because the object in question has been
deleted, no future access to this object is allowed. To have the directory
show the file would mislead you into believing the file existed when in
fact it has been deleted. Whether the file is listed (flagged as deleted)
or not is a matter of implementation.

>I still fail to see why this kind of behavior is considered a "good thing".

I will try to answer this by reversing the scenario. Please help me
understand how the original MPE behavior is a "good thing":

* You are the creator/owner of the file, it can be read by others
* You want to remove the file so it can't be used because it has a problem
* Someone is currently reading the file
* You 'purge' the file, but the o/s says: sorry the file is in use - can't
purge
* Until you either abort the accessor of your file or get them to end the
  application that is reading the file you cannot purge that file - and
others
  can go ahead and access that file. And you must wait to purge *YOUR* file.

IOW - it seems totally senseless to me that a file that I have ownership
over cannot be purged when I want it purged.

And while you are pondering this why don't you explain to everyone why
you think that the original MPE limitation that only the creator of the file
can
rename it is also a good thing...

I suppose you also think that 8-char limits on file names is also superior
to
the unix style where you could accidentally give a file a name that might
actually mean something...

duane percox

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *