Subject: | |
From: | |
Reply To: | Miller, Keven |
Date: | Wed, 1 Mar 2006 10:58:15 -0700 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
No answer here; just definitions.
As I have learned and accepted
creation date:
The time when a file is created in a directory
through the provided host API (meaning the time
I execute an FOPEN or other file creation statement;
not the time the data is posted to disk).
modify date:
The time when a change to the file's data is done
though the provided host API (meaning the time I
execute an FWRITE or other data changing statement;
not the time the data is posted to disk).
Access date:
The time when a file's data is read or written (same
API note as before).
And with that, I guess thats why its not trivial to just
change a file's filecode, or other structure defining attributes.
These are done with copying it to a newfile, thus giving
new dates.
Some things have been accepted like name change, limit change,
eof change, and with provided special software, filecode change.
Now isn't there something called a status change date?
Maybe that has something to do with file definition changes.
Keven
donna garverick wrote:
> is there a place in our collection of documentation that annotates just
> what actions trigger a change to the modification date for a file?
>
> we're revisiting some sox auditing things here and on the unix side,
> they're reporting (for example) if ownership or rwx for a file changed.
> i was suprised to find that on both hp-ux and mpe changing the owner
> or access to a file does *not* change the modified date. it appears to
> me that *internal* changes to a file will update the modified date but
> not external changes....hence my question. - d
>
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
|
|
|