My uncle was right.

[Cap5750]

A few years ago, I was visiting with my uncle.

 

We both are in similar professions (he's a programmer for a phone company) and I'm an unemployed database administrator/analyst.

 

Anyway, one afternoon he and I were sitting around talking about programming.

 

"You 'young guns' have it all too easy and don't know real programming." he told me.

 

"What? You're crazy!" I exclaimed. "I've written tons of code and done just as much complexity as you always have."

 

"Sure you have." he said. "I'm not downing your skills. But when I was learning programming, we had to write everthing. Today, you guys have a lot of tools that does a lot of the hard stuff for you.".

 

Now that conversation initially bothered me at first.

 

But as I'm sitting here going over tutorials on programming the Atari 2600 and dabbling in small changes, I'm now understanding what he means.

 

I mean, just to play a sound, the things I had to keep in mind are amazing.

 

I've always had a respect for those that programmed in the 70s and early 80s due to the limited space they had along with the limited power and huge time and costs it took to develop even the simplest programs.

 

But as I sit and go through these, I'm beginning to have even more of an appreciation.

 

Cap

[Bryan]

Not only did programmers have to manage all the hardware directly, but they had to do it on lousy development systems, often with little or no debugging information when things went wrong.

 

Of course, before that they had to set 8 switches and press a load key for each byte.

 

-Bry

[Cap5750]

I remember a college prof. telling me once about a program that was being worked on. He said he was pushing the cart with all those cards across to load them when it accidently tipped over. He had to sit there and resort them!

 

Cap

[Vigo]

It may be that today, we have more helpful tools creating programs. But on the other hand, todays programs have achieved such a level of complexity which could have never been achieved 20 years ago. So i guess it´s a tie.

 

Of course doing old stuff with todays possibilities is easier. If you wanna code your atari really the old fashioned way, you would have to build a primitive eprom burner, with switches to enter the address and data in binary by hand, and assembling the 6502 code by hand, translating your program into hex-codes, then into bit patterns. Wouldn´t this be fun?

[Cap5750]

Uhm....no?

 

But I do agree with you. I mean, come on...I can open up my Delphi IDE and with three clicks of the mouse, write over 1,000 lines of code producing a fully working program.

 

What I was really getting at though is that by going through these Atari demos and testing a few things on my own, I have a greater appreciation of what these guys had to go through in order to made Pong 2600 do some of the things it did.

 

Cap

[Bryan]

The biggest difference is that you don't really need to know exactly what your resources are today. If you need a 16K or a 1MB buffer, you just alloc it. You write code that performs the functions needed without much regard to how many cycles it executes in (or really even how it gets implemented by the compiler).

 

Compare that to the 2600 where you can execute maybe 4000 instructions per refresh, and you have 128 bytes for your program and your stack, Period. Not only must you manage your memory wisely, but often you need to manage your bits within those bytes wisely. When it comes to code, you'll find yourself writing the same routine over and over trying to shave a few cycles and bytes. You literally must think like the machine.

 

I think people just like the challenge. It's sort of a game in itself.

 

-Bry

[eric_ruck]

There is another big differences, in that you don't have to manage your virtual resources anymore. Just remove the last reference to a Java object and "poof" (eventually).

 

Also you don't need to understand the low level resources, such as stacks and interrupts.

 

Problem is, once in a blue moon something along these lines grinds a major development effort to a halt, and if nobody knows this stuff the whole project tanks (or they have to hire me as a consultant to fix it).

 

I think every Comp Sci should have a class where they have to write something for a limited resource system. I think EE's do this all the time. It's also why I tend to avoid hiring CompSci's and instead look for people with liberal arts degrees who learned to program because they like it.

 

Eric

[Vigo]

Which of course means that absolutely no CompSci likes coding this way...

 

Thanks to you, now i have the answer why my JavaSDK supports no Atari binaries!!! I´m so grateful! You saved me!