Where does the money go?

goaliedad

Parent
10-Year Member
Joined
Apr 7, 2009
Messages
1,594
http://news.yahoo.com/special-repor...s-conceal-epic-waste-144950858--business.html

Yet another story by journalists that don't understand the basics of managing a large enterprise. Don't get me wrong - there are lots of things wrong with how the Pentagon conducts business. I've heard plenty of stories from a sibling who is a fairly high level manager in an oversight agency mentioned in the article.

However, the journalists lost me with this paragraph:

Reuters has found that the Pentagon probably won't meet its deadlines. (See related article [ID:nL2N0J00PX].) The main reason is rooted in the Pentagon's continuing reliance on a tangle of thousands of disparate, obsolete, largely incompatible accounting and business-management systems. Many of these systems were built in the 1970s and use outmoded computer languages such as COBOL on old mainframes. They use antiquated file systems that make it difficult or impossible to search for data. Much of their data is corrupted and erroneous.

This is blaming the computer for the lack of controls.

This guy who spent plenty of years working in a mainframe environment knows for a fact:
1) A large percentage of our nations financial and banking systems run on those "outmoded computer languages such as COBOL on old mainframes". 2) The relational databases and flat files on these old dinosaurs are the simplest places to search for data ever invented.
3) Computers do not create erroneous data. Users input it. Data corruption is a function of media management. Using appropriate redundancy and maintaining your media is required no matter if you are using a mainframe or an Intel box to manipulate and present your results. In fact, both types of systems use the same disk and tape systems for their storage.

That being said, COBOL is one of the easiest languages to audit. It reads much like English and is comprised of fairly basic computing logic to follow. This makes the review of controls in the code fairly easy to follow. Try to find qualified auditors for some of these object oriented languages and you will discover how difficult dissecting modern information systems can be.

My understanding of the basic problem here is that basically the computer automation of controls for operational systems has not kept up with the changes in procedures and practices over the past couple of decades. Blaming the software that was not written for the practices in place for the failures of the practices in place shows little understanding of the problem at best.

The sloppy accounting in DoD is a cumulation of the implementation of business practices without adequate development of the underlying systems to control those practices. Place the blame for this where you may, but don't blame the hardware.
 
^^^^^
Did you actually build these machines? You seem so protective! :biggrin:
Actually, I worked in the industry for 38 years and completely agree with you. But journalists and the general public don't understand... and it's human nature to "blame the computer". I must say though, besides some courses in college, I'm certainly glad I never had to program in COBOL. Way too tedious for me.
 
COBOL wasn't that bad. Did a lot of Medical on NCR in COBAL. Could copy a lot of code from other programs. GIGO still the reason. When I started it was with punch cards in college. Now that was a pain.
 
COBOL wasn't that bad. Did a lot of Medical on NCR in COBAL. Could copy a lot of code from other programs. GIGO still the reason. When I started it was with punch cards in college. Now that was a pain.

Well, it helped to know how to program that drum on the keypunch machine. Really speeded things up for me. :biggrin:
 
COBOL wasn't that bad. Did a lot of Medical on NCR in COBAL. Could copy a lot of code from other programs. GIGO still the reason. When I started it was with punch cards in college. Now that was a pain.

Well, it helped to know how to program that drum on the keypunch machine. Really speeded things up for me. :biggrin:

Fortunately, most of my mainframe coding days were spent in PL/1 which allowed many more types of processing techniques, although I had my share of COBOL. Strangely enough, I actually enjoyed Assembler the most. It reminded me that data structures don't grow on trees. :yllol: And anything I wrote in that ran soooo much faster. Moved on to database administration in my later days on big iron.

I refuse to go back to coding again having looked at the JAVA that these kids slap together today. Runs like a dog and encourages bad coding practices (like failing to declare variables before using them). COBOL almost forced you to organize your thoughts and Assembler punished you severely if you didn't. And I didn't need to patch COBOL, PL/1 or Assembler every week!

I didn't get to do punch cards growing up, as my first shop was in the process of converting what was left of their code to electronic media after a halon event occurred in the computer room during night time batch processing with all the decks staged for execution... :yikes:
 
COBOL, SNOBOL, ALGOL, PL/1, APL, FORTRAN, Pascal...

Am I dating myself???

Steve
USAFA ALO
USAFA '83
 
COBOL, SNOBOL, ALGOL, PL/1, APL, FORTRAN, Pascal...

Am I dating myself???

Now if you had told us that you had to test the vacuum tubes before replacing them....THAT would be dating yourself. :wink:
 
Now if you had told us that you had to test the vacuum tubes before replacing them....THAT would be dating yourself. :wink:

What's a vacuum tube! :shake:
Actually, I remember those puppies in TV sets. An entire industry or two died when they disappeared... TV repairmen if nothing else.
 
Back
Top