A few observations:<p><pre><code> - Lots of reference to military applications
- Nuclear reactors and shielding
- FORTRAN!
</code></pre>
An interesting note about the applications on the nuclear side: I worked in a semi-high security clearance environment on certain nuclear operations. One item we dealt with in particular was a very old algorithm implemented in FORTRAN. We were attempting to scale the system involved, and the implementation of the algorithm was a major bottleneck.<p>The algorithm was phenomenally complex (it's nuclear science, after all). And, we had a challenge in documentation that was impossible to clear up with the original algorithm implementors: most of the team involved had passed away more than a decade earlier.<p>It was one of the neatest programming challenges I've ever encountered. Those old-school engineers were cool, and I wish our industry could keep more of those people around to pass along what they learned and help teach our industry going forward. The technologies may change, but logic never goes out-of-date.
The saddest thing for me is just what an institution IBM has been for America over the years. Having gone to UC Santa Cruz, the Computer Engineering department was started by IBM alums (my advisor was Glen Langdon of Langdon and Rissanen / Arithmetic Coding fame. Other friends/faculty spent time working at IBM Almaden Research where several advances in storage technologies have been produced.<p>These days, IBM is in decline. Friends/tech blogs/even sites like Cringely -- <a href="http://www.cringely.com/" rel="nofollow">http://www.cringely.com/</a> -- note how IBM is quickly trying to shed any semblance to it's old self in names of meeting investor expectations.<p>IBM isn't the only legacy company in the same situation, we have seen this happen with HP as well as AT&T over the years. Microsoft Research, Google to a point, and Xerox Parc (after a period of decline) are stepping up for some long term/basic research. But, I wonder, will we every see the hey day of IBM Research, PARC, Bell Labs, etc. ever again?
The funny thing about the problems stated at the top is I could imagine seeing a job posting today (maybe even from IBM) with the same problems and thinking, "Wow, I have no idea how I'd go about solving those problems."<p>It makes me wonder what they were really doing back then vs what a job in those fields would look like today.<p>Related, I find it somewhat annoying when people abstract the job to such a degree that you can't see the tangible things you'd be working on in that field. I'm all for a "change the world" vision -- I really am, not just qualifying --, but sometimes I'd like to hear up front how they plan to solve that problem.
Compare this to a more recent job posting, which more frequently reads like "Looking for a code ninja to make our photo-sharing app beautiful. You need to be awesome and make shit happen." Funny and sad at the same time.
I'm noticing that a lot of the expectations / requirements are a lot lower. Most are "up to 2 years" experience.<p>Last time I was job hunting, everything was asking for 5+ years experience in one software stack and multiple frameworks. Sometimes 5+ years in multiple fields. What changed?
I love the fact that the problem domains remain so static. Well, that's not strictly true: we have a lot <i></i>more<i></i> problem domains with which to contend now, but even our tremendously improved knowledge in operations research, military science, and meteorology haven't led us to consider these "solved problems".<p>Though I personally wish we focused much more on (1) and (3) than (2) as a society.
Most of these problems, and jobs (with appropriate changes in technologies applied), still exist today, and in greater numbers than in 1958. It's just that the other parts of the industry have grown at a much more rapid pace. This is good to bear in mind when you hear people treat "technology" as synonymous with "web site programming".