....TWO DAYS OF TESTING TO CHANGE ONE LINE OF CODE....
Is that an average for any other shop in this country.
Mike once told me about three programmers from India that discussed changing one line of code in a system he worked on. He thought that was excessive. Now, I come here, and find its normal to talk small changes to death. And its normal to test things to death. And its normal not to have test data. Its normal not to have decent documentation. And its normal not to make obvious changes. And, if you try to improve anything, you are wasting time.
I tried to create a testing procedure for one program that took both Jeff and Mike each almost two days to test.  If my math isn't wrong, that's over 30 man hours of work for two lines of code. If I had finished that program (given a few more hours of work), it would never take anyone longer than a few hours to test changes to that program, because my program would have generated good test data.  So I'm dumb for doing that. I was wasting time. I was stupid for not editing 3x290 byte records in Edit, TEdit, Notepad or something simular. Thats smart...yea..sure. And, I created the only decent test documentation that exists. But, its smarter, it seems, to have to bug someone, and repeatedly ask them what you have to do to test one program, over and over and over again. Wasted time and wasted money at Uncle Sam's expense. If I were the one who was fiscally responsible for those systems, I'd keep someone like me, and get rid of the people that condone two days of testing for one line changes. And I'd get the coders good test data.

 

Speaking of test data, JP Morgan must have a decent test system.
They go over our changes. They have a QA staff. Why isn't their test system duplicated to our system. They surely have documentation. Why is all of ours written only by us?
Also, for a single line of code change, there is a minimal of three lines documenting the change.
To show what got changed. In 1987, I was the administrator of Control at Valley Bank. That's what it did. Anybody from the 1990's on up would say, why don't we let Source Safe do that? Source Save cost $200.Why don't we have version control? What century is this anyway again? To read the code, you almost need to delete all of the comments, just to see the current code.
Like I said, these systems are administered by people who don't want to keep current, don't want to simplify their jobs, and don't want to listen to suggestions. That door is closed.
I saw code errors.

If, I had pointed out the code errors I had seen, I would have been reprimanded.
And that is just not right.