I have been working with different kinds of computers for around 28 years- my first exposure was in high school where we had a "computer club". Students would teach other students about Fortran and Assembler language for a (even at that time) obsolete mainframe system. You would use a keypunch machine (like a very unforgiving electric typewriter that putholes in cards) to type up your program, give it to someone and get backyour cards and whatever got printed out. Usually something about a syntax error. I got to work with a "terminal" connected to a larger computer when I went to a different high school. At least you got nearly immediate feedback when you had typed something wrong. This was in 1973 and all of the city of Chicago schools were connected to a single computer. Some friends of mine figured out interesting ways of becoming a "privileged user" on the system and looking to see what other people were doing. In some ways we learned quite a bit about the system, but the guys running the computer center got rather upset with us - they thought we were going to crash the whole system or delete something vital. One thing that we were a lot more interested in doing was turning off some of the limitations that they had imposed so that only "small" programs could be run. Then, we could write bigger programs that would do more than add two numbers and the like. I thought I wanted to work with computer hardware design, especially the stuff that I thought was really interesting - processors. Well, from a 1968 perspective, I guess that was where things seemed like they would be interesting. Today, I am really glad I didn't really have the grades to get into the college electrical engineering program that I thought I wanted. Most of the people that I know that did go through that aren't doing anything really creative. However, I think there would have been other things that I could have done if I had a college degree. After high school I ended up working for a number of companies doing various kinds of programming on mainframe systems, eventually ending up working mostly on a system called VM/370. This is conceptually a lot like having a PC today - the system created a virtual environment for each user so it was like having your own machine. One of the nicest things was that if whatever you were doing crashed, it didn't affect any other users. This was significantly different from other environments at that time where usually one problem caused everything to come apart. Needless to say, if you didn't feel like working at night or on weekends there were a lot of things you just didn't do in these other environments. Yup, I worked a lot of weekends. For a while I worked on writing video games for home game console machines, like the Atari 2600 and Colecovision. I got to work from home, with some pretty interesting hardware. A lot of it I had to assemble myself. CP/M machines typically came in even more pieces than a PC does today - you had separate boards for the CPU, the disk controller, memory, and so on. When you got everything together, you still had to then write BIOS routines so the operating system could communicate with the disk controller and other devices. This was a challenge that was beyond most people and even most computer dealers, which is pretty much why we have the IBM PC hardware standard today rather than CP/M. The game console computers had to be programmed in Assembler - there was no memory for anything else. You also had to count machine cycles in a lot of cases to make sure that your program didn't run over hardware limitations - such as how many machine cycles there were per scan line on a television set. Pretty primitive machines. Sadly, video game programming today isn't too much better - everyone always wants to do more and more with less and less hardware support. Today a video game console like the Sony Playstation has a lot of memory and a graphics coprocessor, but to get really responsive 3D graphics from it you need to write some stuff in Assembler - it turns out that it isn't quite powerful enough for what people want from it. I got back into mainframe computers around 1984 working for another software company. This was all mainframe stuff and we had a rather large amount of contempt for PC's - they mostly crashed all the time and there were lots of compatibility problems. Sounds sort of like where we are today, 14 years later. One of the biggest problems then was producing large manuals for the products we were working on. Again, where I am working now we have the same problem: producing large manuals for the products. Seems that Microsoft Word isn't up to the task and keeps blowing up, so we had to use something different. Same problem, 14 years later. Around 1990 I started working on PC stuff and Windows and doing a mix of mainframe and Windows programming. For a while I had a software company in Ohio and we owned around five PC's and a mainframe system. Unfortunately, the product we were trying to sell was something that a lot of people wanted but couldn't justify the budget for and we ended up closing down. That was the last time I have worked with a mainframe, around four years ago. I would say that the biggest issues in the 1970's when I started working with computers was bugs, bug fixes and making computer programs maintainable. Today, we have the same problems on PC's, although the approach is a little different - on mainframe systems the software is actually maintained by the companies that sell it and they have to manage the distribution to customers of the fixes for their bugs. Today, PC software companies don't bother with fixes, just new releases. Still, the biggest challenge with computer programming everywhere is making things that can be adjusted, fixed and modified in the future - usually by someone that didn't write it originally. That is what all of the stuff about "object-oriented programming" and languages like Java and C++ are about. Maintability is also one of the big Y2K problems that companies are facing.