The guy who wrote the link you posted is kind of an idiot but he's also kind of right, at least about the problems with overly-rigid frameworks.
In spite of what he says, the 1970 Royce paper is generally accepted as the earliest mention of the waterfall method (see Wikipedia) even though Royce did not call it that. However, he did diagram it quite clearly in the form that is frequently used today.
I work for a large company that produces lots of software and the waterfall method is still mentioned seriously - something I find astounding as it's been seen as a very poor method since this earliest known mention. However, Joel Spolsky apparently sticks up for it and he often seems to have a clue. However, I suspect he may be working with small waterfall-like steps inside a larger, iterative framework.
The two books mentioned should be required reading for everyone in the field. I was lucky (or unlucky) enough that these were two of the first I ever read. The "unlucky" branch was spending the following twenty years wondering why almost no other book was as good as these two.
This is the question techies never want to talk about though it's often more relevant than how quickly it runs.
Here's a solution in J in fewer than 20 lines: http://www.jsoftware.com/jwiki/Essays/Knight's_Tour .
I agree with much of what you say except for the part at the very beginning where you put the blame on rating agencies for continuing to "provide strong ratings to mortgage-backed securities without considering the ripple effect a housing-market slump would have." I work for one of these agencies and I can say that you seem to have a very mistaken idea about what they are supposed to do.
The job of rating agencies is to rate bonds, not to make guesses the possible future of the whole economy. The rating agencies do look bad here and could have done a better job, something they have all admitted. However, the basic problem with the rated instruments has a lot more to do with the sudden, covert relaxation of lending standards of the underlying mortgages than it does with anything else. Defaults are forecast based on historical data - if recent lending does not follow the historical standards, the forecasts will suffer.
Remember the housing bubble of the past few years? There was an unprecedented level of speculation by individuals trying to get rich by flipping properties. This didn't help matters either.
I think a Slashdot item "Anti-Terrorist Data Mining Doesn't Work Very Well" just a little prior to yours gives a clue how to accomplish both privacy and availability: pollute the noosphere with bad information about yourself and "fraternal twins".
For instance, something I've done for years is to subscribe to magazines, etc. with slightly different versions of my own name. As others have also done, I started by using a different middle initial for different subscriptions. As the namespace became more crowded, I branched out to using dual middle initials and variant spellings of my name and address.
Similarly, sign up for different online services with variants of your name, birthdays off by days or years, etc.
If enough of us do this for long enough, the waters will be hopelessly muddied.
The article also claims "I have even tried it with orange juice after I saw a similar device being used in the US. It didn't just make the juice taste fresher, it made it look brighter too."
If you think about it, this is a very narrow range which is why many gas explosions leave most people alive: the parts of the cloud not in the explosive range burn rather than explode.
I've recently been interviewing people for an analyst position that requires some coding skill, so I put together a simple test which I send applicants in advance. They don't have to write anything down but it provides a framework and talking points for the interview.
I deliberately made it fairly simple but it's already proved its worth. For instance, the first question is: how would you count the number of files in a directory?
Probably about 30% of the applicants could NOT give a good answer to this.
Lazy programmers?
Why go to all the trouble to maintain your own database of unique IDs when you can just use the handy one provided by the Feds? After all, we're all Americans here anyway, right?
Never mind that's it's technically not legal to require this information - people have been getting away with it for years.
OK - here's where I recommend J (jsoftware.com) as a first language, you all flame me, I call you all a bunch of clueless boobies, and it goes downhill from there.
Ready to get started?
Anyway, since we've already had a number of posts like this, it would be interesting to have some follow-up: what did people end up doing and how were the results?
Personally, I've had no luck getting my daughter interested in programming but she is very much into music composition, so I guess that's something. She did show some interest in an "Eggbert" game when she was little - it lets you design Eggbert video game landscapes/obstacles/rewards that were pretty cool. She was also interested a little bit in some of the J to do some math that could be represented graphically, like pictures of Pascal's triangle modulus some constant - see http://www.jsoftware.com/jwiki/NYCJUG/Projects/Pascal.
Well, at least they ought to be stringent - you wouldn't want a crazy person to willingly sit on hundreds of tons of highly explosive chemicals you were going to set off, would you?
Somehow I think the psych evaluations for astronauts are somewhat more stringent than those for Hollywood actors. At least that's what movies about astronauts have taught me.
Everything by Lois McMaster Bujold Ben Bova's "Dueling Machine" and others Lloyd Biggle, Jr.'s "The Still, Small Voice of Trumpets" Zelazny's "Amber" series (hard to find) novels by Jean and Jeff Sutton Keith Laumer's "Retief" series early Robert Silverberg Bradbury's "The Illustrated Man" The Kuttner and Moore books from the fifties Any short-story collection edited by H. L. Gold (also 1950s-60s)
Later on, Norman Spinrad, especially "The Last Hurrah of the Golden Horde".
My daughter loved Heinlein's "Starship Troopers" when she was, maybe, ten. I re-read it after she had finished it and was surprised at all the Libertartian politics, none of which I remembered from reading it as a youth.
APL is still around - there's at least four companies that sell their own version of an interpreter - and it's been dying for at least the last thirty years.
Slashdot Mirror
This is a good point.
Especially if the source code has definitions that can be set at compile time:
#ifdef SCREWYOU
RandomResult()
#else
RealTest()
#endif
The guy who wrote the link you posted is kind of an idiot but he's also kind of right, at least about the problems with overly-rigid frameworks.
In spite of what he says, the 1970 Royce paper is generally accepted as the earliest mention of the waterfall method (see Wikipedia) even though Royce did not call it that. However, he did diagram it quite clearly in the form that is frequently used today.
I work for a large company that produces lots of software and the waterfall method is still mentioned seriously - something I find astounding as it's been seen as a very poor method since this earliest known mention. However, Joel Spolsky apparently sticks up for it and he often seems to have a clue. However, I suspect he may be working with small waterfall-like steps inside a larger, iterative framework.
The two books mentioned should be required reading for everyone in the field. I was lucky (or unlucky) enough that these were two of the first I ever read. The "unlucky" branch was spending the following twenty years wondering why almost no other book was as good as these two.
This is the question techies never want to talk about though it's often more relevant than how quickly it runs. Here's a solution in J in fewer than 20 lines: http://www.jsoftware.com/jwiki/Essays/Knight's_Tour .
The story is that the "God particle" is a shortened version of the original description: http://solapanel.org/article/comments/god_in_a_particle/ .
I agree with much of what you say except for the part at the very beginning where you put the blame on rating agencies for continuing to "provide strong ratings to mortgage-backed securities without considering the ripple effect a housing-market slump would have." I work for one of these agencies and I can say that you seem to have a very mistaken idea about what they are supposed to do.
The job of rating agencies is to rate bonds, not to make guesses the possible future of the whole economy. The rating agencies do look bad here and could have done a better job, something they have all admitted. However, the basic problem with the rated instruments has a lot more to do with the sudden, covert relaxation of lending standards of the underlying mortgages than it does with anything else. Defaults are forecast based on historical data - if recent lending does not follow the historical standards, the forecasts will suffer.
Remember the housing bubble of the past few years? There was an unprecedented level of speculation by individuals trying to get rich by flipping properties. This didn't help matters either.
The Economist magazine surveyed hundreds of economists and found that they overwhelmingly preferred Obama's economic policies to McCain's: http://www.economist.com/world/unitedstates/displaystory.cfm?story_id=12342127 .
...I think they're targeting that vast group of Americans who think it's unpatriotic to drink water.
Only if this "evidence-based" reasoning thing catches on again. I'm not holding my breath either.
I think a Slashdot item "Anti-Terrorist Data Mining Doesn't Work Very Well" just a little prior to yours gives a clue how to accomplish both privacy and availability: pollute the noosphere with bad information about yourself and "fraternal twins".
For instance, something I've done for years is to subscribe to magazines, etc. with slightly different versions of my own name. As others have also done, I started by using a different middle initial for different subscriptions. As the namespace became more crowded, I branched out to using dual middle initials and variant spellings of my name and address.
Similarly, sign up for different online services with variants of your name, birthdays off by days or years, etc.
If enough of us do this for long enough, the waters will be hopelessly muddied.
The article also claims "I have even tried it with orange juice after I saw a similar device being used in the US. It didn't just make the juice taste fresher, it made it look brighter too."
Definitely hokum.
J, from jsoftware.com is also free and good for doing numerical analysis.
If you think about it, this is a very narrow range which is why many gas explosions leave most people alive: the parts of the cloud not in the explosive range burn rather than explode.
The wiki at http://www.jsoftware.com/jwiki is a very good resource and getting better as more people contribute.
I've recently been interviewing people for an analyst position that requires some coding skill, so I put together a simple test which I send applicants in advance. They don't have to write anything down but it provides a framework and talking points for the interview.
I deliberately made it fairly simple but it's already proved its worth. For instance, the first question is: how would you count the number of files in a directory?
Probably about 30% of the applicants could NOT give a good answer to this.
You forgot about her attempts at censorship: http://news.bostonherald.com/news/2008/view.bg?articleid=1117009&srvc=2008campaign&position=15 or perhaps you don't have a good analogue on the Obama side.
and APL in the 60s.
Lazy programmers? Why go to all the trouble to maintain your own database of unique IDs when you can just use the handy one provided by the Feds? After all, we're all Americans here anyway, right? Never mind that's it's technically not legal to require this information - people have been getting away with it for years.
OK - here's where I recommend J (jsoftware.com) as a first language, you all flame me, I call you all a bunch of clueless boobies, and it goes downhill from there.
Ready to get started?
Anyway, since we've already had a number of posts like this, it would be interesting to have some follow-up: what did people end up doing and how were the results?
Personally, I've had no luck getting my daughter interested in programming but she is very much into music composition, so I guess that's something. She did show some interest in an "Eggbert" game when she was little - it lets you design Eggbert video game landscapes/obstacles/rewards that were pretty cool. She was also interested a little bit in some of the J to do some math that could be represented graphically, like pictures of Pascal's triangle modulus some constant - see http://www.jsoftware.com/jwiki/NYCJUG/Projects/Pascal .
x= a*b;
Somehow I think the psych evaluations for astronauts are somewhat more stringent than those for Hollywood actors. At least that's what movies about astronauts have taught me.
Private corporations could run most countries better.
Everything by Lois McMaster Bujold
Ben Bova's "Dueling Machine" and others
Lloyd Biggle, Jr.'s "The Still, Small Voice of Trumpets"
Zelazny's "Amber" series
(hard to find) novels by Jean and Jeff Sutton
Keith Laumer's "Retief" series
early Robert Silverberg
Bradbury's "The Illustrated Man"
The Kuttner and Moore books from the fifties
Any short-story collection edited by H. L. Gold (also 1950s-60s)
Later on, Norman Spinrad, especially "The Last Hurrah of the Golden Horde".
My daughter loved Heinlein's "Starship Troopers" when she was, maybe, ten. I re-read it after she had finished it and was surprised at all the Libertartian politics, none of which I remembered from reading it as a youth.
This one goes to eleven.
APL is still around - there's at least four companies that sell their own version of an interpreter - and it's been dying for at least the last thirty years.