Sysman, What degree of reality should one attribute to possible "end of file" or "delete data" vulnerabilities in programs?

-- Watchful (seethesea@msn.com), February 05, 1999

Answers

Hi Watchful. First, I want to thank you, et al., for your support and encouragement here! I think 09/09/99 is one area that is being blown out of proportion. First, just about every operating system does have a specific end of file indicator, seperate from the data. In those cases where the programmer did set a "flag", it was, again, usually seperate from the data, WORKING-STORAGE for COBOL people! Finall, if someone did use a date field for something wacky, it would most likely be 99/99/99, not 09/09/99. So, I think possible, but unlikely.

-- Sysman (working@it.edu), February 05, 1999.

also seehere and persuant answer.

-- Mutha Nachu (---@ragingstorm.com), February 05, 1999.

Ok Mutha, most of us have been at GNIABFI. Please don't post on my threads unless you have something constructive, or at least funny to say. Thanks, NEWBE.

-- Sysman (working@it.edu), February 05, 1999.

Sysman, Many thanks for your succint answer.

-- Watchful (seethesea@msn.com), February 05, 1999.

I see...you don't find those posts constructive?!? or is an answer only valid if it doesn't come from 'the other side'? (oohhhhhh)

or did you even bother to read them?

my mistake...thought you were looking for info (those posts happen to AGREE [shock shock] with your first answer.)

-- Mutha Nachu (---@bladeofgrass.com), February 06, 1999.

One reason a 9/9/99 may be more likely to be used than 99/99/99 is that a date field filled with 99/99/99 is likely to be rejected as an "invalid date". Unless of course the field is a simple text field.

In a batch routine connected with a noncompliant backup program where I used to be employed, I chose to use 12/31/99 as "a date that will always be "'later than today'" only because it was as far into the future as the backup program would accept. I hope they are changing their backup software pretty soon. (no grin)

-- D B Spence (dbspence@usa.net), February 06, 1999.

D B Spence, Thanks for the further insight.

-- Watchful (seethesea@msn.com), February 06, 1999.

Mutha, don't get me wrong, I think just about everyone here is looking for constructive input. I also think most of us have been to GNIABFI and know his views, although many newbes here may never have been there. Now, if you pointed to a specific page at ANY site that has info about the topic being discussed, fine. I just don't think it's cool to send us out of a thread to some homepage where we've already been. Just my opinion.

D.B. Your 9/9/99 point is somewhat valid. If a programmer is using a "pre defined function", that function may very well reject 99/99/99. However the point of this type of "exception processing" is just that, an exception. The programmer is very aware of what 99/99/99 means, and would take the needed action. So again, I think possible but unlikely.

Your 12/31/99 point is very valid! That's standard operating procedure on mainframes that aren't Y2K ready. KEEP FOREVER! All files marked as such will "expire" on 12/31. Good news though, they'll only be dead for 1 day, since the next day is 01/01/00, they won't expire AGAIN for another hundred years. Point here, be very careful on 12/31. Would you run your mainframe on that day knowing ALL your files are expired?

-- Sysman (y2kboard@yahoo.com), February 06, 1999.