As long as we're walking down memory lane... At 06:53 AM 2/1/2010, Leo Noordhuizen wrote:
You were slightly earlier and closer to the source I guess. When I started with computers in 1970, it was with a 4 KB computer, of course using magnetic core memory, and teletype and papertape as I/O devices. I very well remember having to key in about 12 instructions on the frontpanel as bootloader, to start the machine. Programming then was witchcraft in assembly language; a totally different activity compared to programming nowadays. But it was only 6 years later in 1976 that I was so lucky to visit Bell Labs and meet the likes of Dennis Ritchie and Brian Kernighan, and hear about the UNIX operating system, which I used since then. Looking at my Android-based mobile phone, based on Linux/UNIX, it is sometimes difficult to really grasp the progress what has been made in 40 years.
I often stop and smell the progress. I have some really funny stories about that, but no time this morning to convey them.
The computer you started with sounds like a DEC PDP-?? (maybe an 8). They made more computers that fit your description than anybody, but there were a lot of manufacturers at the time that didn't make as much of a name and went under fairly quickly.
Sorry I didn't know you then; we could have had lunch when you visited. I worked for Bell Labs and knew Brian Kernighan.
At 08:37 AM 2/1/2010, Martin, Sam wrote:
NASA, in the early 80s, used something called a mini-computer produced by an outfit, MODCOMP out of Ft. Lauderdale. Armed with a 64KB store using magnetic core technology, a Winchester head per track disk subsystem packing a whopping 4MB of storage, a trio of MODCOMP IIs controlled the launch sequence for the Space Shuttle.
And it was not new technology then. When you described it, my immediate reaction was "early '70s design". I checked; it was a 1972 design, with the last reissue in 1976.
But I don't doubt that NASA was using it in the early '80s. There are organizations that have far more need for reliability than for performance or economy. NASA is one, for obvious reasons. Parts of the Bell System (where I spent most of my career) were another, for different reasons. For instance, consider undersea cable (for trans-ocean phone calls, before satellite became cheap). It cost about half a megabuck to make a repair on a cable six miles under the ocean. So reliability was the single most important parameter.
Using a common set of inputs, each computer voted on any given step in the launch procedure. The odd man out on any vote was prohibited from taking further part in the decision tree.
Again, the Bell System made ample use of redundant active hardware for service-critical functions. The ESS-1 computer-controlled telephone switch went into service in the mid-1960s. It was duplicated, with error-checking hardware built in. Actually duplicated CPU, separately duplicated memory, and also separately duplicated peripherals (the switch matrix controls, etc). Any of these could be cut out if the error-checking circuits detected a problem, and the system reconfigured with only good components.
The OS was written in assembly language.
Back then, limited-performance technology, high performance requirements, and small memory often dictated assembly language. I programmed in assembler a lot early in my career, mostly for those reasons. Not so much later, but one of my most challenging assembler assignments was in 1986-87, when I was doing system programming for personal computers.
I have a devil of a time removing masking tape from the shaft when re-gripping. I looked at a tool designed to remove tape, a kind of scraper, but I can see myself gouging hell out of the graphite shaft on my driver. I havent had a lot of luck using my butane torch and I tire of peeling the stuff off by hand. Any suggestions?
If you're not willing to risk gouges, even though they'll be under the grip, then I can't help you. If you are willing to risk it, Arnie makes a great scraping tool with a curved blade that matches the curve of the shaft surface. It's a prime tool in my shop. (When it has done it job -- quickly and painlessly -- there aren't any gouges. At least not that you would notice in the presence of a previously taped surface.)
Hmm! I just thought of something else you could try. Grip tape solvent sometimes loosens the tape, so it can be scrubbed off instead of scraped.
Cheers! DaveT -- Shoptalk ** Sponsored by the new Aldila Voodoo. Learn more at http://aldilavoodoo.com/
