Cobol has to be the shittiest programming language EVER. That is all.

LOL

I hate it too, but that's what keeps us old guys employed. One of the batch systems I maintain is written in COBOL with embedded PL/SQL.

Just remember, it was designed to be coded on card decks. I never worked in a batch mainframe shop, but I remember a consulting gig from the early 80s where most of the client's programmers worked in COBOL on the mainframe. (I was working on a realtime system.) When they arrived in the morning, they picked up their card decks and printouts that were generated from the overnight batch run from the operations window and checked for errors. They repunched program cards that needed changing, desk-checked the code (you didn't want to waste a batch run on a stupid syntax error the compiler would reject), and resubmitted the card deck at the operations window for the noon batch run. Most then read the paper and went to lunch early. After lunch they picked up the results from the noon batch run, made additional changes, desk-checked it, and submitted the card deck for the overnight batch run. Those folks only got in two runs a day. Larger shops had development systems so programmers could submit batch build and runs all day (well, as fast as they could be submitted and run by operators), but when even a minimal mainframe cost at least $500K (in 1980 dollars), many companies only had "the" computer that had to be shared by production and development.

Thanks for the memories C5. I remember my first CS class. We watched a video of a guy in a white lab coat in front of a huge computer espousing the power of computing and claiming that "you and I will probably never see one in our homes". To coin another ancient phrase "you've come a long way baby".

LOL

I hate it too, but that's what keeps us old guys employed. One of the batch systems I maintain is written in COBOL with embedded PL/SQL.

Just remember, it was designed to be coded on card decks. I never worked in a batch mainframe shop, but I remember a consulting gig from the early 80s where most of the client's programmers worked in COBOL on the mainframe. (I was working on a realtime system.) When they arrived in the morning, they picked up their card decks and printouts that were generated from the overnight batch run from the operations window and checked for errors. They repunched program cards that needed changing, desk-checked the code (you didn't want to waste a batch run on a stupid syntax error the compiler would reject), and resubmitted the card deck at the operations window for the noon batch run. Most then read the paper and went to lunch early. After lunch they picked up the results from the noon batch run, made additional changes, desk-checked it, and submitted the card deck for the overnight batch run. Those folks only got in two runs a day. Larger shops had development systems so programmers could submit batch build and runs all day (well, as fast as they could be submitted and run by operators), but when even a minimal mainframe cost at least $500K (in 1980 dollars), many companies only had "the" computer that had to be shared by production and development.

Eeesh. I dunno, maybe it's just the fact I came from a PHP background.

I'm certainly not an expert in PHP, but I find myself able to work through problems 10 million times more quickly than with Cobol.

Of course it's annoying as all hell when you have to define variables before you do *anything*. Not just that, you have to specifically describe them.
DATA DIVISION be damned.

Saw this gem and thought it was kind of humorous since I've been constantly just rewriting my old assignments for class for new ones.
It has been said of languages like C, C++, and Java that the only way to modify legacy code is to rewrite it - write once and write once again; or write once and throw away. On the other hand, it has been said of COBOL that there actually is one original COBOL program, and it has only been copied and modified millions of times.

If you think COBOL sucks, try doing all your programming in machine language. It took another guy and I two days to write an echo program that would display the press of a key on the keyboard on the printer of a Honeywell DDP-116. In the land of the blind, the one-eyed man is king.

Obviously you've never programmed in RPG...

In the land of the blind, the one-eyed man is king.

Indeed. However, today that language is *still* under development.

Obviously you've never programmed in RPG...


Ahh, IBM System 3 fill-in-the-blank programming with the little 96 column cards.

Programming a DEC PDP 8 through the tab keys sucks. Programming a PDP 8 with a tab loaded bootstrap and an ASR-33 is a little bit better. We thought we'd died and gone to heaven when the paper tape reader showed up.

I did two years of COBOL, a Semester of RPG-II and dabbled in BASIC.
My instructor's favorite saying was "BASIC causes brain damage".

I loved COBOL but it is really wordy and can be a mess depending on who coded it.
We did our first year of COBOL and the 2nd was "Structured" COBOL.

RPG was pretty easy... as long as you had coding sheets.
I did all my stuff on a Wang VS-65 and an IBM S/34.
Got out of computers though, too much compiling and debugging. Drove me crazy. Plus the whole sitting at a desk thing wasn't for me. I need to move around and use my hands for something besides pecking on a keyboard.
I have wanted to learn some of this fancy new stuff, but I don't have time to immerse myself in more syntax.
I have an 8 inch floppy at home with a system I wrote to maintain inventory in a video store. :histeric:

Oh, in my electronics training I did some programming of a basic processor in hex, one instruction at a time with a keyboard.
I had to build a circuit with the "display" on a breadboard below the programmer and see if I got the right results.
It was pretty cool but man it would suck to still be in that era.
It was an old motorola processor... and I mean old.

I did some work on old military gun fire contol computers that used gears, cams, linkages, synchro XMTRs and RCVRs, and black magic to compute firing solutions. It worked pretty good as bad as it sounds. At a range of 5000 yards the first round would land within 25 yards of the target even if both you and the target were moving.
http://dcoward.best.vwh.net/analog/ford.htm

dkozloski: That's an interesting piece right there. Interestingly enough, I'm now in a FA unit... yay! I'm a cannon cocker for the time being

Btw, for those who do some php devel, you might have heard of this, but if not, it's pretty nifty
http://en.wikipedia.org/wiki/Xampp