Operating System - OpenVMS
1748255 Members
4028 Online
108760 Solutions
New Discussion юеВ

Re: File Revision Number XAB$W_RVN (Fortran 90/77)

 
SOLVED
Go to solution
HDS
Frequent Advisor

Re: File Revision Number XAB$W_RVN (Fortran 90/77)

Thank you John.

When one designs and codes an "off-the-wall" approach to something like this, it is refreshing to see that someone else out there does something similar. Even though this is really off-subject from what I was originally asking, it is greatly appreciated to see that a similar approach is used and that it is found useful elsewhere. For that alone, it was worth the point assignment.

Your choice of attributes to compare are of interest. I neglected to mention that I too am using the EOF (End of Data), but intentionally chose not to use the ALQ because sometimes the comparisons are across devices that have different cluster sizes. However, others that you mention may be very applicable...I like the idea of the FFB (First Free Block), but had not previously thought of that one. I would need to confirm that the FFB is a valid test when tested across the devices with different cluster sizes...but, off the top of my head, I think that it is a good one.

Very nice ideas. And, I must add, reassuring to see that someone else is doing something similar to that "fingerprint" approach.

Very appreciative.

-Howard-
John Gillings
Honored Contributor

Re: File Revision Number XAB$W_RVN (Fortran 90/77)

Howard,

I'm using the lexical function F$FILE_ATTRIBUTES names for attributes.

> FFB (First Free Block),

FFB is First Free BYTE not BLOCK. For sequential files, EOF tells you the block containing the end of file. FFB gives you the location of the next byte beyond the end of file. If two files are identical, both EOF and FFB must be the same, regardless of cluster size.

In RMS terms, FFB is XAB$W_FFB from XABFHC.

Note that FFB only makes sense for sequential files. My code depends on it being returned as 0 for all other file types.
A crucible of informative mistakes
HDS
Frequent Advisor

Re: File Revision Number XAB$W_RVN (Fortran 90/77)

Good morning.

Thank you for that clarification. It now makes sense that FFB would be a valid test for seq files...regardless of disk cluster sizes.

I thank you all for all of your advise and suggestions. This has been a most rewarding and informative forum thread. It touched on topics that I had not expected...but am very pleased that such occurred.

Much obliged to all who participated.

-Howard-
HDS
Frequent Advisor

Re: File Revision Number XAB$W_RVN (Fortran 90/77)

I have received much useful information; so much so, that I have enough to not only address my original issue, but also enough to fuel new enhancements and to provide a basis for new utilities/tools.

Thank you.

-Howard-