Y2K retrospective / was Re: Algol vs ...

--------------- I get back to my point that a lot of the comments and criticisms concerning the absence of the century in system design comes from people ill equipped and knowledgeable to be able to pass judgment. By 1985 I had been working in Data Processing for 20 years! Hardware constraints dictated what we did and didn't do. The hardware constraints were our main obstacle and were twofold: 1) Cost - Memory was incredibly expensive (Read The mythical man month) 2) Hardware Technology - Early systems were mag tape only. The bottom line to this is that there is always a price point. Added to which, and factored in, is the life expectancy of any system. Back then it was considered around 10-15 years plus. A lot of systems had no upward compatability and applications needed to be modified to run. In the mid 60's only large companies had systems with greater than 16K memory and disc drives. Mag tape 800 and 1600 bpi if you were lucky was the norm. -------------------- I fully agree with Russ. In many files we used julian date, expressed as YYDDD. In COMP-3 (packed decimal), this would occupy 3 bytes. A full date, CCYYMMDD, would occupy 5 bytes. You might laugh, but in many financial systems this was a PITA. The smalles mainframe I used in the 70's, was an IBM 360/25. It was equipped with a whopping. When I developped a program for statistics, the first compilation wanted 80K of memory (double buffering, blocked records, etc). After having cut the fat away, I couldnt get it below 32K, so the boss had to go to his boss, and ask for money so we could RENT 4K of memory ! When that was in place, I had 4 (four) bytes left. Nico