One VAX at Apple in Cupertino (Bandley One) had a major chunk of ceiling fall on it during a major quake. I think there were a couple of IO retrys, couple of syserrs logged, but otherwise continued without a glitch. A rack of drives rocked over and fell against a wall. They had to shut down the system because they were afraid a disk crash might happen if they just pushed them back upright.
I really doubt if there was anything in history quite as unreliable as the Xerox Sigma series. I worked for XDS for about 6 years. The disk drives (2nd rate CDC's) leaked hydraulic fluid.
Good engineers were a little hard to come by -- one gentleman from a certain Asian subcontinent with many letters after his name was asked to clean one of the extremely precious spare RAMAC drives (remember them?) for NASA's DSN. He took MEK and four-ought steel wool and got that brown stuff off two platters before they caught him.
I remember that a lot of the timesharing system engineers (the good ones) left after the Sigma 9 fiasco to a place back east, Maynard Mass. to work for a new minicomputer company that was going to do something better than UTS.
When the company was failing, we joked that Fairchild and Honeywell were buying us out, and the restructured company would be named Farewell Honeychild.
So there is at least some common thread from UTS through VMS to WNT. Don't think Dave Cutler was there at XDS though.
Ahh, I f$miss("''DCL'")... loved the syntax, dreamed in it.
Way more interesting is the SDS 940 emulator; first machine I ever played with. Discrete transistor and diode logic. My old friend Bob Long had written an assembler and an application for it - half of the 8k word core tank was used for his "calculator", an infinite precision calculator that worked in any base between 2 and 32. When I first typed "9**81" and watched the ASR 33 typing out three rows of numbers, I knew what my career would be right then and there. It had room to store one constant; taking the 81'st root of the result took about two hours, followed by a bell, the bang of the teletype and the number 9.
Bob had an old AM transistor radio tuned to the end of the dial, sitting on top of the M register (a couple of large, heavy cards) and we could hear the calculation's progress. Handled fractional roots, too. Computing in 1969; Them Waz The Dayz.
Make them cheap and replaceable; go for redundancy for coverage and reliability; if one blows away send up another. Switch to hydrogen to improve cost and lift. And since they carry no passengers, it doesn't matter whether you cover them in flash powder or not a'la Hindenburg.
Would depend on such things for emergency coverage in a hurricane though?
I remember when an integrated circuit meant the transistors were soldered really close to each other, disk drives leaked hydraulic fluid, and plated wire was the next best thing. Those war thee days, laddie!
Still thinking on this -- I'd imagine the landing stage to be something of a large flying wing, optimised for takeoff and landing. The centre back -- the receiving socket -- could be shaped in such a way to form a shaped air pocket to assist in landing, perhaps by raising baffles before docking. If air flow was shaped just so it could form a partial vacuum to assist in docking.
May never be practical, but it's fun to speculate, no?
No, just bring it down to subsonic speeds before docking, cruise speed for large jet. You'd get versatility in landing options, the use of heavier landing gear, and you wouldn't need all that systems infrastructure for landing. If the landing stage was purpose-designed, you could socket the suckers together -- use a chined surface instead of wings for the spacecraft & avoid their attendant potential for leading edge burn-through. You could use simple structures for control (vectored vents instead of ailerons or other drag surfaces perhaps) and the whole thing could be a lot cleaner. Hypersonic control surfaces may act the same as low speed control surfaces, but you don't need as much surface area either. Lots more control than a parachute for final approach, and oversea approaches wouldn't offer quite the opportunities for impromptu oceanography.
If by "crash-worthy" you mean structures for absorbing the energy of speed differentials at mating point, those could be carried by the landing stage; no reason to carry additional structure on the spacecraft.
We did things differently back in the 70's when I was playing with spacecraft systems; I would have crawled through broken glass to get at the computational power & control systems we have available today. I believe the docking manouvre could be pretty much bump-free with the stuff available today, and there's heaps of optimisation possible once you remove one or two subsystems -- the follow-on benefits tend to be nonlinear.
Using a large, efficient aircraft as a booster stage is known to work.
Innocent question -- would it help to design a large, efficient aircraft to assist in landing?
I was wondering if a two-stage landing scheme would help to remove some of the design compromises required for low-speed flight. I'd think you could cut back on a lot of subsystems, a fair amount of weight, and possibly even reduce risk. Mid-flight capture has been done before in early surveillance satellite days, although in a much simplified format (bomber hooks parachute of probe out of the air). Could a specialised large flying wing dock reliably with the spacecraft on the way down? That way you'd only need to worry about controlling the trajectory & surface areas designed for dumping heat down to subsonic speeds. You wouldn't need flaps, landing gear, or any low-speed control surfaces, and could cut down wing area etc. couldn't you?
I've hired about 100 programmers/analysts/dba's etc over the last thirty years; track record counts a bit, certs count a bit (but not hugely -- if someone claims a doctorate, I'll ask to read the thesis) -- screening is more intuitive to me than analytical. What counts is evidence of intelligence, enthusiasm and involvement, plus clarity in communication (and of course the ability to communicate their subject knowledge). If they can't allude to that in the application, the alphabet soup will count for exactly zero. And if I can't get an applicant waxing lyrical during their interview about a pet bit of work they've done, it's a good indication you're talking to a boat anchor. There is no substitute for brains, either in engineering or management.
Nothing compared with the west coast of Tasmania -- eleven or twelve feet of rain per year. People in Strahan call a fortnight without rain a "major drought" (pronounced "draft" locally).
I'd include some form of graphics drawing package with a graphics tablet. Let 'em draw, scrawl a note, mail the images to a friend.
Re:Hardcore Quiche
on
Hardcore Java
·
· Score: 2, Interesting
My old departed friend and LGP-30 programmer once caught Cray filing down a resistor before soldering it back into the discrete circuitry. Something about tuning a pipeline cascade. Cray pretty much invented the instruction pipeline, and was known for tweaking his designs personally.
Agreed. I once sent a less than lauditory note to the LOP people who had installed their doubly-distilled, quadruply qualified and unmentionable (so ok, I read Doc Smith!) toolbar on my daughter's browser. Their response was "you agreed to it". Where? When? At no point did she click on a link that clearly indicated this would happen. Hate LOP. Hate, hate hate hate.
Slashdot Mirror
Problem was, when you did that you got a different SYSUAF, and they weren't transitive.
One VAX at Apple in Cupertino (Bandley One) had a major chunk of ceiling fall on it during a major quake. I think there were a couple of IO retrys, couple of syserrs logged, but otherwise continued without a glitch. A rack of drives rocked over and fell against a wall. They had to shut down the system because they were afraid a disk crash might happen if they just pushed them back upright.
Good engineers were a little hard to come by -- one gentleman from a certain Asian subcontinent with many letters after his name was asked to clean one of the extremely precious spare RAMAC drives (remember them?) for NASA's DSN. He took MEK and four-ought steel wool and got that brown stuff off two platters before they caught him.
I remember that a lot of the timesharing system engineers (the good ones) left after the Sigma 9 fiasco to a place back east, Maynard Mass. to work for a new minicomputer company that was going to do something better than UTS.
When the company was failing, we joked that Fairchild and Honeywell were buying us out, and the restructured company would be named Farewell Honeychild.
So there is at least some common thread from UTS through VMS to WNT. Don't think Dave Cutler was there at XDS though.
Way more interesting is the SDS 940 emulator; first machine I ever played with. Discrete transistor and diode logic. My old friend Bob Long had written an assembler and an application for it - half of the 8k word core tank was used for his "calculator", an infinite precision calculator that worked in any base between 2 and 32. When I first typed "9**81" and watched the ASR 33 typing out three rows of numbers, I knew what my career would be right then and there. It had room to store one constant; taking the 81'st root of the result took about two hours, followed by a bell, the bang of the teletype and the number 9.
Bob had an old AM transistor radio tuned to the end of the dial, sitting on top of the M register (a couple of large, heavy cards) and we could hear the calculation's progress. Handled fractional roots, too. Computing in 1969; Them Waz The Dayz.
I just so want some of your drugs. What a hoot!
Yes -- that's why they called it a Pentium instead of an Intel 586; they added 100 to 486 and kept getting 585.913343251...
I support Windows users. When all my hair fell out, I discovered my skull was blue.
Would depend on such things for emergency coverage in a hurricane though?
How about another code to show how many times the VIN has been ground down and re-stamped?
Why not? SF has been doing it for decades.
Time to dust off the old VE novels, perhaps -- substitute "Power" for "Communications" and they will probably read just as well...
Paul Simon had something to say about that...
I remember when an integrated circuit meant the transistors were soldered really close to each other, disk drives leaked hydraulic fluid, and plated wire was the next best thing. Those war thee days, laddie!
"Clear!" (presses paddles to chest) "Reboot!"
May never be practical, but it's fun to speculate, no?
If by "crash-worthy" you mean structures for absorbing the energy of speed differentials at mating point, those could be carried by the landing stage; no reason to carry additional structure on the spacecraft.
We did things differently back in the 70's when I was playing with spacecraft systems; I would have crawled through broken glass to get at the computational power & control systems we have available today. I believe the docking manouvre could be pretty much bump-free with the stuff available today, and there's heaps of optimisation possible once you remove one or two subsystems -- the follow-on benefits tend to be nonlinear.
Innocent question -- would it help to design a large, efficient aircraft to assist in landing?
I was wondering if a two-stage landing scheme would help to remove some of the design compromises required for low-speed flight. I'd think you could cut back on a lot of subsystems, a fair amount of weight, and possibly even reduce risk. Mid-flight capture has been done before in early surveillance satellite days, although in a much simplified format (bomber hooks parachute of probe out of the air). Could a specialised large flying wing dock reliably with the spacecraft on the way down? That way you'd only need to worry about controlling the trajectory & surface areas designed for dumping heat down to subsonic speeds. You wouldn't need flaps, landing gear, or any low-speed control surfaces, and could cut down wing area etc. couldn't you?
I've hired about 100 programmers/analysts/dba's etc over the last thirty years; track record counts a bit, certs count a bit (but not hugely -- if someone claims a doctorate, I'll ask to read the thesis) -- screening is more intuitive to me than analytical. What counts is evidence of intelligence, enthusiasm and involvement, plus clarity in communication (and of course the ability to communicate their subject knowledge). If they can't allude to that in the application, the alphabet soup will count for exactly zero. And if I can't get an applicant waxing lyrical during their interview about a pet bit of work they've done, it's a good indication you're talking to a boat anchor. There is no substitute for brains, either in engineering or management.
Nothing compared with the west coast of Tasmania -- eleven or twelve feet of rain per year. People in Strahan call a fortnight without rain a "major drought" (pronounced "draft" locally).
I'd include some form of graphics drawing package with a graphics tablet. Let 'em draw, scrawl a note, mail the images to a friend.
My old departed friend and LGP-30 programmer once caught Cray filing down a resistor before soldering it back into the discrete circuitry. Something about tuning a pipeline cascade. Cray pretty much invented the instruction pipeline, and was known for tweaking his designs personally.
I google, you google, he/she/it googles, they have googled, we have googled...
The next challenge is to become an irregular verb.
Agreed. I once sent a less than lauditory note to the LOP people who had installed their doubly-distilled, quadruply qualified and unmentionable (so ok, I read Doc Smith!) toolbar on my daughter's browser. Their response was "you agreed to it". Where? When? At no point did she click on a link that clearly indicated this would happen. Hate LOP. Hate, hate hate hate.
...Screaming Lord Such
Kept most of them in view for a while, but the force was strong with some of them, and I lost them in the trench.
Not everybody thought it was boring. Maybe I'm just a born suit and tie fighter.