Yup. IIRC, the commonly used unit 30 years ago or so was the UTR-10 -- "University Teaching Reactor 10". Pretty much any engineering school with a nuclear engineering program back then had one of those hiding some place that was.... umm.... not well advertised. I haven't kept up, but I suspect the same unit or maybe a slightly updated design is still common. It wasn't weapons grade Uranium, though, but certainly fissionable because the whole point was learning to operate a power generation reactor as would be found at an electric utility or on a US Navy vessel.
I wouldn't have known about it at all except that my roommate's girlfriend was a NucE student who trained on it. It's existence wasn't widely known. More students could navigate the steam tunnels than knew how to find the reactor.
The problem with quantity is the cluelessness of the project leaders. People with absolutely no idea of what it takes to ramp manufacturing promise quantities of premiums that are essentially unmanufactable designs. They end up faced with hand-making irrationally large quantities. That is the key problem -- sure, you can manufacture in quantity if you know how to design for scalable manufacturing. That only reinforces my point that too many kickstarter projects are run by people that have no clue how to design for manufacturability, the funders have no clue how to sort the clueless form the savvy, and it all ends in a mess.
You comment "in large quantities, production should become cheaper" is to me a sign that you are among the clueless that has never had to try to make 100, 1000, or 100,000 of anything, and think it is trivial to replicate a prototype in quantity. It isn't. It doesn't happen by sprinkling magic pixie dust on the prototype. It takes thoughtful design to make it even remotely possible at the price points people expect today.
Problem 1: The funding crowd does a poor job of sorting out wishful thinkers from people who can deliver. Just going by the projects where I have actually met the principals, the crowd often gets it wrong. I have seen people with no track record and who are clearly clueless about how to execute a major project raise huge amounts of funds, and yet people with track records and savvy who don't make the funding deadline. It's really sad to look at the disasters that have been funded and the stuff with great potential that didn't make it -- and I am just talking about projects where I have met the people and seen the prototypes.
Problem 2: Projects have no way to cap the upper end. I've seen a couple of "success disasters" -- people set out to raise X dollars to deliver Y premiums. They end up raising 10X or 15X or more and then are faced with how to fulfill 15Y premiums. This problem often seems to visit the clueless dreamers That is a recipe for a train wreck.
Kickstarter could solve problem #2 tomorrow by simply putting in an upper funding limit. Problem #1 is tougher -- it needs some kind of reputation system, but even then, their are enough dreamers on both the buying and selling side of impractical dreams to make it a perpetual peril.
In general, you can divide the world into people who solve problems and people who execute instructions. Accounting is a problem solving profession. Bookkeeping is instruction execution. People who know how to analyze and solve problems, no matter what their discipline, are good to work with. Also, the clear thinking that goes with problem solving tends to be portable across disciplines.
I think what you have described are CPA's with excellent and portable problem solving skills. You must have been working a a company that knew how to hire the right people.
Actually he *did* brush his teeth regularly, but didn't replace his toothbrush often enough. It started growing things. Nasty, nasty, things. So regular brushing just kept putting the nasty things back into him. Lesson: replace your toothbrush regularly.
Choose a license the way you choose a screwdriver, not the way you choose a religion.
What are your goals? Are you inventing a new protocol that you want everyone, everywhere, to integrate and ship with their software, be it open or proprietary? Then you want a BSD-like or MIT-like license.
Do you want to have control over the commercial uses of your work, while keep non-commercial use low cost? Then GPL or similar is for you.
At one hardware company I worked at, we released drivers under BSD. From the company perspective, it was a great "fire and forget" license for situations where that was appropriate, and, as one company lawyer put it to me: "Don't worry, a freely available version will always be available, because somebody will take our code and fork it with a GPL license about seven seconds after we release it anyway."
Religious wars over licenses are really pointless. Choose a license that accomplishes your goals. Different situations call for different tools.
"Hello, World" uses Python. It has aged a wee bit, only because Python has moved on, so the "how to install Python" section of the book is slightly stale. Other than that, I think it is great. A Real Computer Scientist(tm) wrote it with his 10 year old son, so the book reflects the interests and questions of a young kid. I used it with my daughter, and she loved it. I've recommended the book to adults that want to learn Python, and they liked it.
As for what language to use, I say use Python. You can teach proper computer science with it, and the language doesn't get in the way. Save C for later. Pascal is of historical interest only at this point. If you don't know Python, work through "Hello, World" with your son (or on your own) -- you will be glad you did.
We have 3 Kindles, an original, the 2nd gen, and a Fire. Mostly my daughter uses the 2nd gen and the Fire. Which one gets taken along in the car depends not so much on what content is loaded on which kindle, but on things like: Will it be dark when we are driving back from the gym?
In other words, a self-lighting book is usable in places a dead-trees book is inconvenient. I would definitely like a lighted Kindle that had good battery life.
The Apple II really killed the days of soldering together your own computer from a kit. And it came with a *cabinet* even, gosh. After so many years of soldering together kits, getting an Apple II really felt like cheating.
But... I shall refrain from hollering "newb".... I was visiting with my neighbor Walt, from down the street. Walt worked at HP labs before he retired. Anyway, we were discussing the conversion to digital TV, and got into a discussion of the elegant design of the original analog B&W standard. And he said: "Oh, it took a long time to get there. They tried all sorts of things for the audio subcarrier, AM, FM, and kept moving it around. When I was a kid, there was a TV station in Philly that broadcasted 2 and 1/2 hours a week, and published plans so you could build a receiver to watch."... OK, so I'm starting to listen *REALLY* attentively... "You couldn't buy TV's, you had to build them. And they would publish updates and we would all go change our receivers."
So, yes, at this point I was impressed. And he adds a little more of his history: "Eventually I went to work for the TV station. That was before I went to Eckert & Mauchly" What is so cool about Walt, is here is a guy that was a CPU designer at Eckert & Mauchly, and he has a box running Linux now. Walt can call anybody 'newb'. I cede the ground to him.
This is why I like living in Silicon Valley -- great neighbors.
Vic 20 had a 6502 and I think 16K of SRAM. The Apple II (original) and Apple II+ (floating point basic on the mobo) had 48K of SRAM. Yes, they were earlier than the C64, but the C64 was the first Commodore to have 64 K of SRAM, so it was the first Commodore machine to approach the Apple II in functionality, even though it came later. My roommate and I both had Apple II's (original) with integer basic on the MoBo and floating point basic on add-in cards. We were among the first to get floppy drives for the Apple II when the started shipping -- getting the #2 and #3 units shipped to the local store. I had a SWTPCo 6800 with 12K of SRAM, he had some kind of Z80 single board machine (Cromemco??) with I think 64K of SRAM in a home built wooden case. After we graduated, I went off to work as a control logic designer on walk-in, refrigerated, mainframe scientific processors -- my roommate went on to become the software half of the C64 design team, and the key driver of the architecture.
In my mind, the C64 belongs in the same sentence as the Apple II, even though it came later, because it has about the same capability and was a direct effort to catch up. Since my ex-roommate was a key player, I'm pretty confident in saying that. Anyway, they were only a couple of years apart -- in the context of grouping SlashDot'ers into 4 clumps, C64 and Apple II as first machines is an appropriate clump.
I guess that falls in "some such". There were a whole bunch of computers in those days. I had a SWTPCo 6800 system before my Apple II. And before the C64 Commodore made the PET computers. TI had one... Sinclair. Fun times, actually.
Ahhh, yes.... the floor sort. Diagonal lines, good. And pranking people by collecting all the chad from the keypunches in the student keypunch area and... finding creative places to hide it.
Speaking of the floor sort, true confession time: I actually had a part time job as an "operator". Mainly feeding the card reader and filing output into pigeon holes. There was a punch card fed typesetting program that understood all the thesis requirements for margin and TOC and bibliography sites and such, and could to math and chemistry typesetting with weird escape sequences (all upper case, mind you). Think TeX, only punch-card oriented. It had one fairly serious design flaw, it pretty much insisted on reading all 80 columns of the card, so you couldn't use columns 73-80 for sequence numbers as was the usual for most programs in those days. We had a card sorter and operations would sort anyone's deck for free while-you-wait. But thesis decks were a no-go for sorting. Anyway, there was this one chemistry PhD student who's thesis deck was about 1 3/4 boxes of cards. I forget, is a box 8000 cards or there about? Anyway, he gave it to me one day. As I was loading the card reader with big fist-fulls of cards, I bumped my elbow on the reader and pretty much scattered to the wind about 1000 cards. As he silently watched I stopped the card reader, gathered all the loose cards and put them back in the box and said "Sorry." He didn't say a word -- amazing self control -- I think he was at the point of exploding. I saw him again about two weeks later -- he very quietly peeked around the door to see if it was my shift, saw me, and left. Never saw him again.
And, in fact, hybrid maize seed has been sold as an unstable cross for well over half a century. So your idea isn't exactly new. The issue here is soy beans, where you can't create an unstable hybrid cross, mother nature won't let you. Perhaps you should get a wee bit of background in agronomy and, you know, actually learn the basic science before spouting off. And so should the mods that made you "+5 insightful". But this is slashdot -- where unscientific claims about physics or security are quickly debunked, but a total lack of understanding of basic biology is totally acceptable and is rewarded.
I think your use of the word moron is..... misdirected.
I like to be able to make sketches of interesting material. The diagrams do more for me than the words. The Livescribe pen captures the audio, and can play it back in association with what I was sketching at the time. The portability is great, too.
That is the one good thing about serving on a jury in Silicon Valley. It's simply not practical to throw out anyone with an engineering degree. The two attorneys between them don't have a large enough quota of rejections to force a non-engineer jury.
Everything interesting happens on the margin. India just moved the margin 30% further out. Angels can invest anywhere. Now India is going to be getting less of it. To argue otherwise is complete nonsense. I can not be as sanguine as you. Like it or not, there is global competition for investment. If this passes, India will have put up a "please invest elsewhere" sign in their front lawn.
I suggest you retake economics. Any economist, left-leaning or right-leaning, will tell you that is how the model works. You get less of what you tax. What they argue about is whether or not you are better off, on balance, with less of that thing you are taxing. It *is* going to matter, despite your wishful thinking. Will it totally kill investment? No, but it most definitely will reduce it, drive it underground, or drive more members of their entrepreneurial class to other countries.
Agreed. Bugs are caused by poor thinking, not by lack of type checking.
If you think that then maybe you should write all your code in assembly language. Just stay sharp!
Assembly has it's place. Here's a clue: for most things, Python is better. But I happen to work in imbedded software -- ANSI C with inline asm as needed gets products shipped. On balance, I probably do more assembly than C++. For compact, high-performance code, write clean C and unleash the optimizer. For day-to-day problem solving: Python. The niche where C++ buys you anything but headaches is vanishingly small IMHO.
Agreed. Bugs are caused by poor thinking, not by lack of type checking. Strict static type checking gets in the way of more sophisticated code, especially specialized container classes. I note that people who work in statically typed languages rarely think of doing their own container classes. In Python, a container class is often a huge win for code readability and maintainability.
Where you pay the price for duck typing, IMHO, is in testing and validation. Duck typing requires good unit testing discipline. When I hear people say "static type checking finds my bugs", I think, "There is a person who is too lazy to think about the problem up front and to write good unit tests during development."
Huh? You say it doesn't matter what state you are in, and then tell me that because you are in Mass and can't get the laws enforced that it doesn't matter what state you are in? Hellooooo.... that is not logical.... it's barely even illogical.
I think what you just told me is that it matters greatly what state you are in, both in terms of the letter of the law and the enforcement of the law.
Slashdot Mirror
Yup. IIRC, the commonly used unit 30 years ago or so was the UTR-10 -- "University Teaching Reactor 10". Pretty much any engineering school with a nuclear engineering program back then had one of those hiding some place that was.... umm.... not well advertised. I haven't kept up, but I suspect the same unit or maybe a slightly updated design is still common. It wasn't weapons grade Uranium, though, but certainly fissionable because the whole point was learning to operate a power generation reactor as would be found at an electric utility or on a US Navy vessel.
I wouldn't have known about it at all except that my roommate's girlfriend was a NucE student who trained on it. It's existence wasn't widely known. More students could navigate the steam tunnels than knew how to find the reactor.
The problem with quantity is the cluelessness of the project leaders. People with absolutely no idea of what it takes to ramp manufacturing promise quantities of premiums that are essentially unmanufactable designs. They end up faced with hand-making irrationally large quantities. That is the key problem -- sure, you can manufacture in quantity if you know how to design for scalable manufacturing. That only reinforces my point that too many kickstarter projects are run by people that have no clue how to design for manufacturability, the funders have no clue how to sort the clueless form the savvy, and it all ends in a mess.
You comment "in large quantities, production should become cheaper" is to me a sign that you are among the clueless that has never had to try to make 100, 1000, or 100,000 of anything, and think it is trivial to replicate a prototype in quantity. It isn't. It doesn't happen by sprinkling magic pixie dust on the prototype. It takes thoughtful design to make it even remotely possible at the price points people expect today.
Problem 1: The funding crowd does a poor job of sorting out wishful thinkers from people who can deliver. Just going by the projects where I have actually met the principals, the crowd often gets it wrong. I have seen people with no track record and who are clearly clueless about how to execute a major project raise huge amounts of funds, and yet people with track records and savvy who don't make the funding deadline. It's really sad to look at the disasters that have been funded and the stuff with great potential that didn't make it -- and I am just talking about projects where I have met the people and seen the prototypes.
Problem 2: Projects have no way to cap the upper end. I've seen a couple of "success disasters" -- people set out to raise X dollars to deliver Y premiums. They end up raising 10X or 15X or more and then are faced with how to fulfill 15Y premiums. This problem often seems to visit the clueless dreamers That is a recipe for a train wreck.
Kickstarter could solve problem #2 tomorrow by simply putting in an upper funding limit. Problem #1 is tougher -- it needs some kind of reputation system, but even then, their are enough dreamers on both the buying and selling side of impractical dreams to make it a perpetual peril.
In general, you can divide the world into people who solve problems and people who execute instructions. Accounting is a problem solving profession. Bookkeeping is instruction execution. People who know how to analyze and solve problems, no matter what their discipline, are good to work with. Also, the clear thinking that goes with problem solving tends to be portable across disciplines.
I think what you have described are CPA's with excellent and portable problem solving skills. You must have been working a a company that knew how to hire the right people.
Actually he *did* brush his teeth regularly, but didn't replace his toothbrush often enough. It started growing things. Nasty, nasty, things. So regular brushing just kept putting the nasty things back into him. Lesson: replace your toothbrush regularly.
I love that phrase "enforcing freedom" -- have you thought deeply about that?
Choose a license the way you choose a screwdriver, not the way you choose a religion.
What are your goals? Are you inventing a new protocol that you want everyone, everywhere, to integrate and ship with their software, be it open or proprietary? Then you want a BSD-like or MIT-like license.
Do you want to have control over the commercial uses of your work, while keep non-commercial use low cost? Then GPL or similar is for you.
At one hardware company I worked at, we released drivers under BSD. From the company perspective, it was a great "fire and forget" license for situations where that was appropriate, and, as one company lawyer put it to me: "Don't worry, a freely available version will always be available, because somebody will take our code and fork it with a GPL license about seven seconds after we release it anyway."
Religious wars over licenses are really pointless. Choose a license that accomplishes your goals. Different situations call for different tools.
And what is the memory density? How many crabs does it take to make a JK flip-flop?
"Hello, World" uses Python. It has aged a wee bit, only because Python has moved on, so the "how to install Python" section of the book is slightly stale. Other than that, I think it is great. A Real Computer Scientist(tm) wrote it with his 10 year old son, so the book reflects the interests and questions of a young kid. I used it with my daughter, and she loved it. I've recommended the book to adults that want to learn Python, and they liked it.
As for what language to use, I say use Python. You can teach proper computer science with it, and the language doesn't get in the way. Save C for later. Pascal is of historical interest only at this point. If you don't know Python, work through "Hello, World" with your son (or on your own) -- you will be glad you did.
We have 3 Kindles, an original, the 2nd gen, and a Fire. Mostly my daughter uses the 2nd gen and the Fire. Which one gets taken along in the car depends not so much on what content is loaded on which kindle, but on things like: Will it be dark when we are driving back from the gym?
In other words, a self-lighting book is usable in places a dead-trees book is inconvenient. I would definitely like a lighted Kindle that had good battery life.
The Apple II really killed the days of soldering together your own computer from a kit. And it came with a *cabinet* even, gosh. After so many years of soldering together kits, getting an Apple II really felt like cheating.
But... I shall refrain from hollering "newb".... I was visiting with my neighbor Walt, from down the street. Walt worked at HP labs before he retired. Anyway, we were discussing the conversion to digital TV, and got into a discussion of the elegant design of the original analog B&W standard. And he said: "Oh, it took a long time to get there. They tried all sorts of things for the audio subcarrier, AM, FM, and kept moving it around. When I was a kid, there was a TV station in Philly that broadcasted 2 and 1/2 hours a week, and published plans so you could build a receiver to watch." ... OK, so I'm starting to listen *REALLY* attentively... "You couldn't buy TV's, you had to build them. And they would publish updates and we would all go change our receivers."
So, yes, at this point I was impressed. And he adds a little more of his history: "Eventually I went to work for the TV station. That was before I went to Eckert & Mauchly" What is so cool about Walt, is here is a guy that was a CPU designer at Eckert & Mauchly, and he has a box running Linux now. Walt can call anybody 'newb'. I cede the ground to him.
This is why I like living in Silicon Valley -- great neighbors.
Vic 20 had a 6502 and I think 16K of SRAM. The Apple II (original) and Apple II+ (floating point basic on the mobo) had 48K of SRAM. Yes, they were earlier than the C64, but the C64 was the first Commodore to have 64 K of SRAM, so it was the first Commodore machine to approach the Apple II in functionality, even though it came later. My roommate and I both had Apple II's (original) with integer basic on the MoBo and floating point basic on add-in cards. We were among the first to get floppy drives for the Apple II when the started shipping -- getting the #2 and #3 units shipped to the local store. I had a SWTPCo 6800 with 12K of SRAM, he had some kind of Z80 single board machine (Cromemco??) with I think 64K of SRAM in a home built wooden case. After we graduated, I went off to work as a control logic designer on walk-in, refrigerated, mainframe scientific processors -- my roommate went on to become the software half of the C64 design team, and the key driver of the architecture.
In my mind, the C64 belongs in the same sentence as the Apple II, even though it came later, because it has about the same capability and was a direct effort to catch up. Since my ex-roommate was a key player, I'm pretty confident in saying that. Anyway, they were only a couple of years apart -- in the context of grouping SlashDot'ers into 4 clumps, C64 and Apple II as first machines is an appropriate clump.
I guess that falls in "some such". There were a whole bunch of computers in those days. I had a SWTPCo 6800 system before my Apple II. And before the C64 Commodore made the PET computers. TI had one... Sinclair. Fun times, actually.
Ahhh, yes.... the floor sort. Diagonal lines, good. And pranking people by collecting all the chad from the keypunches in the student keypunch area and... finding creative places to hide it.
Speaking of the floor sort, true confession time: I actually had a part time job as an "operator". Mainly feeding the card reader and filing output into pigeon holes. There was a punch card fed typesetting program that understood all the thesis requirements for margin and TOC and bibliography sites and such, and could to math and chemistry typesetting with weird escape sequences (all upper case, mind you). Think TeX, only punch-card oriented. It had one fairly serious design flaw, it pretty much insisted on reading all 80 columns of the card, so you couldn't use columns 73-80 for sequence numbers as was the usual for most programs in those days. We had a card sorter and operations would sort anyone's deck for free while-you-wait. But thesis decks were a no-go for sorting. Anyway, there was this one chemistry PhD student who's thesis deck was about 1 3/4 boxes of cards. I forget, is a box 8000 cards or there about? Anyway, he gave it to me one day. As I was loading the card reader with big fist-fulls of cards, I bumped my elbow on the reader and pretty much scattered to the wind about 1000 cards. As he silently watched I stopped the card reader, gathered all the loose cards and put them back in the box and said "Sorry." He didn't say a word -- amazing self control -- I think he was at the point of exploding. I saw him again about two weeks later -- he very quietly peeked around the door to see if it was my shift, saw me, and left. Never saw him again.
Third group: Those who had Apple II or C64 or TRS-80 or some such.
Fourth (my) group: Those who carried boxes of punch cards across campus.
And, in fact, hybrid maize seed has been sold as an unstable cross for well over half a century. So your idea isn't exactly new. The issue here is soy beans, where you can't create an unstable hybrid cross, mother nature won't let you. Perhaps you should get a wee bit of background in agronomy and, you know, actually learn the basic science before spouting off. And so should the mods that made you "+5 insightful". But this is slashdot -- where unscientific claims about physics or security are quickly debunked, but a total lack of understanding of basic biology is totally acceptable and is rewarded.
I think your use of the word moron is..... misdirected.
If you are looking for a way to make long time, loyal, readers revolt, then you have found it.
My karma here has been excellent for well over 10 years. I'd like to turn off advertising, please.
Or, maybe let's change the moderation system to allow moderation of stories, not just comments, and add '-1 slashvertisement" mod option.
Or maybe, it's simply time for a slahdot clone -- just the same only having editors with both integrity and a brain.
I like to be able to make sketches of interesting material. The diagrams do more for me than the words. The Livescribe pen captures the audio, and can play it back in association with what I was sketching at the time. The portability is great, too.
That is the one good thing about serving on a jury in Silicon Valley. It's simply not practical to throw out anyone with an engineering degree. The two attorneys between them don't have a large enough quota of rejections to force a non-engineer jury.
Everything interesting happens on the margin. India just moved the margin 30% further out. Angels can invest anywhere. Now India is going to be getting less of it. To argue otherwise is complete nonsense. I can not be as sanguine as you. Like it or not, there is global competition for investment. If this passes, India will have put up a "please invest elsewhere" sign in their front lawn.
I suggest you retake economics. Any economist, left-leaning or right-leaning, will tell you that is how the model works. You get less of what you tax. What they argue about is whether or not you are better off, on balance, with less of that thing you are taxing. It *is* going to matter, despite your wishful thinking. Will it totally kill investment? No, but it most definitely will reduce it, drive it underground, or drive more members of their entrepreneurial class to other countries.
I see you are posting under the handle "slowLearner". Why, yes, I'd say you are.
Agreed. Bugs are caused by poor thinking, not by lack of type checking.
If you think that then maybe you should write all your code in assembly language. Just stay sharp!
Assembly has it's place. Here's a clue: for most things, Python is better. But I happen to work in imbedded software -- ANSI C with inline asm as needed gets products shipped. On balance, I probably do more assembly than C++. For compact, high-performance code, write clean C and unleash the optimizer. For day-to-day problem solving: Python. The niche where C++ buys you anything but headaches is vanishingly small IMHO.
Agreed. Bugs are caused by poor thinking, not by lack of type checking. Strict static type checking gets in the way of more sophisticated code, especially specialized container classes. I note that people who work in statically typed languages rarely think of doing their own container classes. In Python, a container class is often a huge win for code readability and maintainability.
Where you pay the price for duck typing, IMHO, is in testing and validation. Duck typing requires good unit testing discipline. When I hear people say "static type checking finds my bugs", I think, "There is a person who is too lazy to think about the problem up front and to write good unit tests during development."
Huh? You say it doesn't matter what state you are in, and then tell me that because you are in Mass and can't get the laws enforced that it doesn't matter what state you are in? Hellooooo.... that is not logical.... it's barely even illogical.
I think what you just told me is that it matters greatly what state you are in, both in terms of the letter of the law and the enforcement of the law.
I bet code reviews are fun for you.....