Re: Ministry of...
"yard as exactly 0.9144 meters"
This was changed to 1 inch = exactly 2.54 cm
552 posts • joined 6 Jun 2015
"her monitor has a proper shimmy on"
At work our group was moved to a couple of trailers during some remodelling work. Most of the CRT computers exhibited a shimmy. We noticed 4 amps through the safety ground. Further checking finally revealed a bad ground wiring in one of the fluorescent lights. Lots of tracing with a clamp ammeter and magnetic field detector.
In the early days of time sharing Dartmouth got a time sharing system working on a GE-225/DN-30. The hardware was upgraded to a GE-235/DN-30 system that was 3 times faster. It also had more instructions. So, of course, the person writing the GE-235 exec used these extra instructions.
Since this was all on an NFS grant, GE took the software and sold it to customers. But the software no longer worked on a GE-225. So they sold the customer a GE-235 with a wait loop in the exec to slow down the machine. This code was clearly marked. The customer was happy - they got what they wanted at the quoted price.
It was fairly common in the 1960's to ship a computer with many instructions disabled. For a price, wires were cut that disabled certain instructions.
The IBM 1401 had an instruction "read tape binary" that required extra money. But the diagnostics required this instruction so all 1401's had this instruction - paying extra money to enable this instruction caused no changes to the hardware.
We had a 407 accounting machine we used to list cards. It ran at 100 cards per minute. I noticed a couple of relays that were disabling every third cycle. removing this relay caused the machine to print at 150 cards per minute. I told this to the customer engineer and he said, "Yes, but you are paying for a 100 card per minute machine." So I left the relay in place.
"It was a real-time BASIC program - and the user had decided to add a subroutine starting at line 1000000."
The original BASIC only used a maximum of 5 digits for the line number. Users using longer line numbers were continually surprised at the results:
Deleting line 123456 wouldn't work - you had to delete line 12345 etc.
I believe that on an early version of MULTICS some programmer cleared a large array by columns instead of rows. This caused a page fault for just about every memory access.
On the IBM 1620 using FORTRAN copying the value of one uninitialized variable to another variable would sometimes clear all of memory because the flag on the uninitialized variable wasn't set so it would copy until it hit a digit with a flag set. But all of the set flags were cleared out by the copy so it never stopped. A program like:
I = J (J not initialized)
(depending on the order of I and J in memory) was all that was needed.
The IBM 704 had a clear memory button that was active only when the machine was running!
Although, when I worked at Xerox, I designed and built a laser printer driver for a new product. The developers claimed I was sending a bogus character so I put in microcode a logger that logged every character sent and received. Nope, no extra characters.
So I went to the developers lab using my hardware and software and they had put a terminal in parallel with the laser printer. Sure enough, there was an extra character. After some head scratching i noticed that the character was sent when the software terminated. A few minutes later I discovered that resetting the UART was causing the extra character to be sent (after my microcoded logger had shut down). Easy fix was to output a MARK before resetting the UART.
Sometimes the customer is right and it pays to keep an open mind.
When i was teaching computer science we would run all submissions through a "cheating checker". It would compare every submission against every other submission and print out the top 15 high scorers. I remember one pair of submissions (which were quite lengthy) where the "diff" output was less than a page consisting of the student's name and class number.
Another time two students in my sections got the high score on this cheating checker so I printed out both submissions and discovered that they were using totally different approaches to solving the problem. Copying wouldn't have done them any good - but why the high score? I looked at the raw data and discovered that the checker ignored the contents of strings and that it had matched up many "System.out.println()'s". Oh well.
At the college where I taught one of the professors had a sig with 20 lines of "do not forward, check with your attorney etc.", followed by even more bizarre cautions, warnings, threats etc. It was a total spoof of the required sig that the university insisted on.
I was working in the college computer center about 2 AM around 1970 over the Christmas break. I was the only one in the building except for the operator. I heard a funny "clunk" and was puzzled as to what it could be. I investigated and discovered the operator lying in a corridor breathing but otherwise totally unresponsive. Who to call?
I called the campus police who arrived quickly in their cruiser. The campus policeman said to get the stretcher from the cruiser. While I was fiddling with the stretcher a dean of the college, walking home, spotted me and thought I was ripping off the cruiser. I figured the best way to get him to hurry was to ignore him. When he showed up I told him he could carry the other end of the stretcher into the building. The operator was carted off to the hospital about a block away.
Then we discovered that the operator had all of the keys on his person and we had to do some fast talking to get the keys out of the hospital safe.
The operator had had a stroke that was so sudden that he fell, bumping his head on the wall on the was down. He did recover fully though.
"Interesting, the object had around the speed of an airgun slug so to do any damage would need to have a bit more mass."
Actually, you only have to aim at something in a different orbit and you might have many miles per second velocity difference when it hit the target. Just launching ball bearings straight up to intersect a satellite would result in lots of damage.
On the old telephone 4A crossbar office (a telephone router with no subscribers) they had metal cards to control the routing of calls in and out of the exchange. They once had incorrectly punched some metal cards and caused a routing loop that tied up all of the lines from Chicago to Milwaukee.
I always wondered about 89724102172714182892114I7551670349743096734346773478647892349863592355648544996312855148587659264921 so (after changing the I to a 1) I found the (4) factors:
393649 657481 × 6335 348305 575195 179179 076633 × 69466 777282 071413 430494 981273 × 51 790677 862630 192029 633722 348449
I once saw a couple of operators working a cord board exchange. You wouldn't believe the complexity of the cable tangle. The operators would be gossiping while interrupting with "Number Please." They would occasionally have to carefully unthread a cable to tear down a connection.
The Dartmouth Time Sharing system went on line in 1964. It was financed with an NSF grant which meant anyone could have the source code. GE cloned the system and started selling commercial time sharing. We (college students) thought it was unfair that no one in the college got credit so we hacked their system, set up another account etc. There was a field I didn't understand in the accounting record which I assumed granted permissions. So I set all of the bits to one. Turns out it was an accounting field keeping track of "A" time, "B", time, and "C" time. There was no B or C time but when they ran their monthly accounting our bogus account had accumulated B and C time.
Then there was the time that the college employed student workers in the data center. This stopped when one of the workers flunked his roommate, And it was discovered that Nikita Khrushchev was enrolled in three freshman seminars.
At the Palo Alto Research Center they brought up a new machine and needed everyone's password. Rather than asking everyone they used a net sniffer that got 99% of the passwords to load into the new machine.
Our sysadmin at college had a tar pit for spam mail. For selected senders he would set the maximum packet data length to one byte (which really slowed them down). After receiving the email it would be rejected with a code to resend it.
His theory was that he was tying up resources of the sender as a public service.
Talk about rate limiting.
When I was in college writing an executive I would sketch out ideas on a huge blackboard in my office. Yes, a real blackboard of Italian slate with white chalk. A real blackboard has a great tactile feel. If there was any code it would be a short sequence of instructions to see if the sequence was optimum. The actual code was written on backs of listings or other scraps of paper. I would then keypunch it onto cards and assemble it.
I would never let anyone else keypunch the code as this was my final check and I discovered many errors while keypunching.
I love SQLite. When I had to maintain a website for thousands of square dances etc. I thought I would learn SQL for the fun of it. So I put all of the information in an SQLite database and generated web pages from the database.
Best decision I ever made. If something needs to change or page headers need to be replaced I just modify the appropriate place in the database and regenerate the website. When one of our round dance cuers got married I changed one entry and, presto, everywhere she showed up on the website was changed.
When we were writing the Dartmouth Time Sharing System there was one programmer that churned out voluminous amounts of buggy code. There was another programmer that didn't like to write code, just find and fix bugs.
When we discovered this we just put the second programmer on the trail of the first and produced lots of debugged code.
"Anyone remember NASA not doing a full up integration test of Hubble before it was launched?"
I believe that the main mirror of the Hubble telescope was figured with very precise laser interferometric testing. Unfortunately there was a .01" in the dimension of part of the test setup. This was cross checked with a knife edge test that showed the error but, since the laser interferometer was much more precise, it was assumed that the laser interferometer was correct.
You really can't do a "full up integration test" on this kind of system.
When model 35 TTY's were used to connect to time sharing mainframes our head secretary got a call from a user. She heard a loud buzzing noise in the background caused by low paper in the TTY so she said press the "buzzer release" button on the far left to stop the noise so they could talk. The buzzing stopped and she asked the user what the problem was. The user sheepishly said, "That was the problem."
I was sitting at home when it felt like a really heavy truck was passing and the lights dimmed a couple of times. I thought a truck had run into an electrical pole. Turns out a few railroad cars were dropping off of an elevated track a few blocks away taking out power wires, crushing parked cars etc., and narrowly missing a car driving under a railroad bridge. Fortunately, no one was injured or killed.
A lot of the old strowager exchanges had a mechanism to "absorb digits. In one exchange with 60,000 subscribers) dialing "2" would just be ignored; dialing "4" would also be ignored but would take the next digit (even 2 or 4). So dialing 242-xxxx would result in 2 (ignored), 4 (ignored but prime for the next digit), 2 (would actually step to the second level and go on to the next bank of switches).
This could cause connections to numbers only loosely related to what was dialed.
I actually saw this exchange (I think in Albuquerque, NM, about 50 years ago) and it was an amazing sight and sound. Continuous clickity-clacks and alerts etc. Amazing that they could keep this acre of equipment working.
A physicist, engineer, and a mathematician examine the statement that all odd numbers are prime.
Physicist: 3's a prime, 5's a prime, 7's a prime, 9 - experimental error, 11's a prime, 13's a prime - all odd numbers are prime.
Engineer: 3's a prime, 5's a prime, 7's a prime, 9 - well, 9 is the exception, 11's a prime, 13's a prime - all odd numbers are prime.
Mathematician: 3's a prime, therefore, by induction, all odd numbers are prime.
"Similar vintage a friend got the Fortran control characters wrong so it threw a new page instead of a new line."
This was quite common as the first column was generally used as a page control. If there was a number there it could cause this one line per page problem. Or even worse, overprinting on the same line until the paper tore through.
I had a windows XP machine that was rock solid for 15 years. Last month one of the drives stopped working so I debated getting the machine repaired. However WINE ran all of my XP programs I needed on my UBUNTU machine so I think I will not bother fixing up the XP machine. I then discovered that my XP software ran considerably faster under WINE (because of a faster processor) so I am abandoning the XP machine.
I have a windows 10 laptop and am currently trying to figure out how to stop all updates.
Biting the hand that feeds IT © 1998–2020