Putting users at a low priority is part of Microsoft's corporate DNA. And it makes sense. Microsoft products aren't ones users select and buy; they're selected for the users.
You know iPhone model after the one we're talking about is much, much harder for governments to break into because of its crypto coprocessor, and such things may become both more common and even harder to penetrate in the future.
The FBI gets a lot of secondary "benefits" (I used that term advisedly) if they win here. They get a legal precedent that says vendors are obligated to help them break security. That's huge. They get an authentically signed insecure version of iOS -- think what they could do with that.
I actually think so to; I was just summarizing the FBI's ostensible case, which is based on what they admit is possible.
I suspect there might be some legitimate reasons to prefer Apple to do it, but I know there are illegitimate reasons. But of course you can't prove anything either way. But it may be possible to address the legitimate needs while thwarting the illegitimate ones.
For example I have heard the suggestion that Apple make the weakened version of the OS, but never give it to the government; instead if the FBI had a court order Apple would take physical custody of the phone, extract the data in a restricted area of its campus.
I understand this, but again under specific limited circumstances. For one example, suppose they saw you shooting at passing cars from your bedroom window. They wouldn't need a warrant to barge into your house, and because their reason for entering the house were legitimate they'd be able to bust you for the pot plants they find there.
On the other hand if they simply believe you're growing pot in your house, even with probable cause, they can't just bust in. And if they do any evidence they find is not usable (it's so-called "fruit of the poisoned tree").
In the case of the San Bernadino phone, that is in the FBI's lawful possession. They have the lawful authority to search the phone, but not the technical ability or (very likely) the legal authority to compel Apple to provide them with the technical ability.
It's very important to get all this stuff straight, because if you get it wrong you either hamper the government in the exercise of its important legitimate duties, OR you open the door to all kinds of illegitimate activities.
The government has all kinds of powers that are very easy to abuse; but generally (and this is a key point) it is constrained in using those powers. The police can bash down your door and invade your house with drawn weapons -- but only if they have a warrant. Now you may argue that even with warrants they're often abusing their power, and I'd agree with you. That doesn't mean I don't think they should ever be able to do that.
Well, that's how I would do it too, but it's evidently not how this particular reactor was designed back in the 50s. If the reactor failed in any kind of planned way, we'd have some idea what that was. What we're learning is that we don't have any idea how this particular design behaves when it's operating outside its normal operating envelope.
I think it's important to stipulate I'm talking about this design. You can't really talk about the "safety of nuclear power plants" as if all designs are the same. Clearly the old Soviet RBMKs are dangerous pieces of shite; and concerns about the GE Mark 1s were raised all the way back in the early 70s. Fukushima, in my opinion, probably means it's a high priority not to extend the operational licensing of any Mark 1s. That doesn't necessarily mean the Germans should shut down reactors whose design is 20 years newer. And it certainly doesn't mean Japan should shut down it's brand new Gen 3 reactors it just put into service, although maybe they should look at how those plants are managed and sited.
I'll go one better. I have a small chip of uranium ore in my desk drawer right this very moment. I got it to test a geiger counter gizmo I got for testing whether old watches I was working on had radium pigment. Naturally I had to make sure the thing works before I trusted a "normal" reading.
I'm not afraid of handling this bit of ore, not in the least. But I wouldn't feel the same about handling the same amount of refined fuel, or the random by-products of a reactor disaster. Clearly I'm not radiation phobic, but extrapolating from the safety of handling ore to the products of a reactor accident is just plain stupid.
I'd have no fear of wearing a watch with radium pigment by the way; I've measured the radiation from them and what you get, even on the face of the watch, isn't a big concern especially if you don't wear it every day; maybe 3x background radiation in my neck of the woods. But I don't work on old radium watches because the binder in the pigment breaks down. If you open the case it'll release radium dust into the room. Would I freak out if I opened a radium watch by accident? No, I just wouldn't voluntarily put myself in that situation. There's a world of difference between carrying around intact radium pigment in a sealed case and breathing loose radium dust, just like there's no comparison between handling a piece of low grade uranium ore and exposing yourself to a radiological disaster.
In particular, it's worth noting that there is a rupture disk here precisely to prevent the reactor pressure vessel from experiencing a catastrophic rupture
This seemed wrong to me, since the RPV in Unit 2 was already breached. I believe AmiMoJo was talking about is pressure in the the primary containment vessel (PCV), not the RPV. Just to be clear the reactor core is inside the metal RPV and the RPV is inside the reinforced concrete PCV. The in the Mark 1 reactor design the PCV is the outer wall of the "dry well".
I also looked up some design diagrams for the venting system. While venting system rupture disk is indeed designed to protect the PCV it is not built into the PCV itself. Operators have to open two sets of valves in order to transfer pressure from inside the PCV to the disk. So while it's true the purpose of the rupture disk is to prevent catastrophic failure, that can only happen if the venting system is activated and works as expected.
Now the report you linked to is four years old and assumes that the venting system worked correctly, delivering overpressure to the the rupture disk. AmiMoJo is referring to evidence which came out later which indicates that the venting system almost certainly failed.
In a way I do agree with you. It's not mysterious why the PCV didn't explode; it didn't explode because it failed in some other, unknown way. Under the circumstances that was a very good thing in comparison to the alternative, but it takes a rather determined optimism to construe it as an endorsement of the reactor's design. It was more like a stroke of good luck.
This article has both a detailed diagram of the Oyster Creek reactor, which is the same design, and a schematic of the venting system.
Well, to a programmer a "bug" is an error in logic. To a user a "bug" is an error in results. A single result may stem from many errors, particularly if we're talking about some kind of resource starvation, which this sounds like. If the starvation was caused by a single line memory leak, well that wouldn't take that long to find and fix. But if it were a faulty approach to managing memory used extensively over several years of development, you'd have a substantial body of logic errors that would take some serious spadework to fix.
One of the thing I've noticed is that some programs exhibit misbehaviors that seem to persist no matter how much fixing you do. These programs don't suffer from a single error, but rather a a small set of sloppy or inattentive coding habits that are repeated over and over. Fortunately sloppy code tends to be redundant, so you can often refactor and score big reductions in code size along with improved reliability and performance.
Just to play devil's advocate here for a second, clearly you don't need a human to run a lab test you planned to do right from the drawing board, but you might find it helpful to have a human available if an unexpected problem or opportunity arises. Humans are versatile, machines are not -- not in comparison at least. When the heater for your assay breaks down a human can look at it and scavenge parts from another piece of equipment. If a Martian strolls up and says "hi", a human can say "hi" back.
Take geology. I have no doubt whatsoever that a geologist actually on Mars with his full toolkit and a general purpose lab would get more geology done than some robot carrying out a few specially designed experiments. I wouldn't be surprised to learn that there are a few geology things we just can't do yet with robots yet. I just don't think we can get more geology done with humans for what we have to spend, even supposing that's enough to pay for a manned mission.
In time that may change. At some point we may experience diminishing returns from each new robotic mission, and then it will be time to prioritize manned missions over robotic ones. But we're far from that point ow.
Thank you, Captain Obvious. However I think the poster's point may have been that it's been historically difficult to hold corporations responsible for the messes they leave behind, and when you can't do that it means the public has to pay to clean them up.
100% false. We can study a lot in orbit but there are limits. You cannot study much about biological issues on other planetary bodies in orbit. For that you have to go there.
Alright. Name one issue of human biology that can only be studied by putting a human on Mars that is not specifically limited to the question of whether humans can live on Mars.
Setting aside human biology is akin to dumping the whole point of a manned space program. Can't do it.
Well, "whole point" is an exaggeration, but there's a reason I said "for the moment". I'm inviting you to consider the opportunity cost that a manned mission to Mars would present -- it's impact on other research priorities.
(1) Robotic exploration alone will never develop the technology needed to sustain life in space. (2) It would be very easy to over-focus on robotic exploration and starve the (harder and more expensive) manned exploration.
I agree with (1), but I wasn't suggesting abandoning manned space flight entirely, just prioritizing robotic spaceflight for exploration beyond the immediate vicinity of Earth. I disagree in a qualified way with (2). I think the most cost-effective way of extending manned space travel to Mars is get really good at all the Mars-specific aspects of getting stuff there and landing it on the surface by practicing with robotics, while continuing to pursue the generic aspects of manned spaceflight in orbit, or perhaps with Lunar missions.
Any reason to establish a base on the Moon or Mars or elsewhere will need to have some sort of economic benefit.
Agreed, but with the stipulation that it's the most cost effective way of obtaining that benefit. If it isn't, you're doing it for anthropological or political reasons.
There are things that cannot be easily or productively studied with robotic probes, not the least of which is human physiology and biology.
All of which could be studied in orbit. Now if you set human biology aside just for a moment, it would be more precise to say "some things are easier to study with human-run experiments than robots." The key point is that that set of things is a function of mission duration. Some things that are most efficiently done by humans on the Moon would be more efficiently done by robots on the Mars. Some things that are most efficiently done by humans on Mars would be more efficiently done by robots on Europa.
The problem is that the things that can only be done by humans on Mars would cost beyond what anyone wants to pay for a realistic mission, and a half-assed program that gets axed when the true cost of success becomes undeniable is a waste of time and money.
I'd like to see humans on Mars, but I think the shortest path to that happening starts with a sustained and regular program of robotic exploration. With experience we'll get better at getting things there cheaply and landing them there reliably. Then at some point an acceptably risky manned mission will become a financial possibility. But we won't do it because it's the most efficient way of doing Mars research, robotics will continue to advance as fast or faster than our space capabilities. The reasons for doing a manned mission will be a question for anthropology and political science, not economics or engineering.
On the other hand the relationship between a system's reliability and the reliability of the system's components isn't one-to-one. You can build unreliable systems out of reliable components, and more surprisingly, you can build reliable systems out of unreliable components. That latter principle is the basis for the Internet, which provides end-to-end communication that is more reliable than any of the possible paths between those endpoints.
Every component is unreliable to some degree; as it becomes increasingly reliable it moves from unusable, to something you can work around, to something whose potential failure you can ignore in many applications.
And everyone else has been saying for months that part of the key is stored in the cpu in a region that is not readable. So all you need to do to back up the key is disassemble the cpu and hope you don't destroy it in the process.
Then "everyone else" would be wrong in this case. The "secure enclave" was introduced in Apple's A7 processor. The phone in question is an iPhone 5c which has an A6 processor. There is no crypto coprocessor to store anything in. The Apple docs even describe the process used to erase the keys in question from flash.
Now as for the iPhone 5S and later, there's no question they're tougher to hack -- that's the whole point of the new hardware features. But I still wouldn't put absolute faith in the inability of the new features to keep the government out if they had physical access to your phone.
Because people will listen to him, because he's Snowden.
I've been saying the same thing for months: the key is stored (albeit encrypted itself) in flash, so all you need to do is to back up the flash chip and you've got as many goes as you need, which with a four digit PIN chosen by humans isn't that many. But even though you can point this out the relevant details in Apple's documentation, people just refuse to believe that the government can get into an iPhone without Apple's help.
That actually kinds of mystifies me. Why would anyone believe that a government that (in part at least) created Stuxnet would be stymied by an iPhone? Whatever the reason, Snowden's imprimatur seems to help them get over it.
And yet RDBMs use ISAMs as part of their design to create and search records.
Using this to claim that RDBMSs haven't changed significantly in sixteen years is like claiming cars didn't change much between 1975 and 1990. Yeah, they still had cylinders and spark plugs, and the basic UI still had the same controls, but in terms of efficiency, reliability, performance and amenities the differences were stark.
Yeah, maybe if you're using your RDBMS as a backing store for your VB app, none of the changes that have happened in the last fifteen years are relevant to you, but believe me things like replication, clustering, MVCC, native geospatial processing, etc. are very relevant to a lot of people, and they didn't happen on their own. Even today there's a lot of things that could be improved, like transaction log forensics.
The boom is not a thing in itself; it is the product of economic conditions. So is the end of the boom. When spending an additional dollar on software development returns $0.99 in value instead of $1.00, people stop spending on software development. If nothing else changes, but suddenly that $0.99 of benefit only costs $0.98, the spending resumes.
Now that effect doesn't matter why the return is only $0.99. Perhaps you've saturated the market, and what once would have brought in $1.10 now only brings in $0.99. Sure that means that spending is done for, but only because you can't sell for a profit at that price.
You seem to be more concerned with words than reality. I actually remember things like the ISAM and network databases of the 70s. Trust me, they aren't up to the job of handling modern transaction volumes. Someone had to create new software, because you can run friggin' models.
Really, your response is full of straw men it's hard to respond to, so I'll limit myself to just this point: you seem to be under the impression that we have all the kinds of software we'll ever need already; in which case you're right: it doesn't matter if our jobs are taken by AI's; there's no more work.
I guess time will tell which of us is right, but I think the idea that all the kinds of software we need have already been invented could only be true if we've already imagined all the way there are to process information, and I doubt that's the case. If we'd been having this discussion fifteen years ago you might well be excused if you took the same position, but in fact there are many kinds of software developed since then that are either novel or radically different from how we thought things should be done back in 2000. For example I don't think that Oracle's RDBMS is really all that much better than it was back then, but it has some clear incremental improvements and useful new features; but that also misses the whole NoSQL movement.
Slashdot Mirror
Third option: indifferent to users.
Putting users at a low priority is part of Microsoft's corporate DNA. And it makes sense. Microsoft products aren't ones users select and buy; they're selected for the users.
It is. I'm not saying I think that's a good thing, it's just the way our rules work.
You know iPhone model after the one we're talking about is much, much harder for governments to break into because of its crypto coprocessor, and such things may become both more common and even harder to penetrate in the future.
The FBI gets a lot of secondary "benefits" (I used that term advisedly) if they win here. They get a legal precedent that says vendors are obligated to help them break security. That's huge. They get an authentically signed insecure version of iOS -- think what they could do with that.
I actually think so to; I was just summarizing the FBI's ostensible case, which is based on what they admit is possible.
I suspect there might be some legitimate reasons to prefer Apple to do it, but I know there are illegitimate reasons. But of course you can't prove anything either way. But it may be possible to address the legitimate needs while thwarting the illegitimate ones.
For example I have heard the suggestion that Apple make the weakened version of the OS, but never give it to the government; instead if the FBI had a court order Apple would take physical custody of the phone, extract the data in a restricted area of its campus.
What? The negation of a negation of a reified proposition? I should think that would be child's play for this crowd.
How about this: I might think they should sometimes be able to do this.
I understand this, but again under specific limited circumstances. For one example, suppose they saw you shooting at passing cars from your bedroom window. They wouldn't need a warrant to barge into your house, and because their reason for entering the house were legitimate they'd be able to bust you for the pot plants they find there.
On the other hand if they simply believe you're growing pot in your house, even with probable cause, they can't just bust in. And if they do any evidence they find is not usable (it's so-called "fruit of the poisoned tree").
In the case of the San Bernadino phone, that is in the FBI's lawful possession. They have the lawful authority to search the phone, but not the technical ability or (very likely) the legal authority to compel Apple to provide them with the technical ability.
It's very important to get all this stuff straight, because if you get it wrong you either hamper the government in the exercise of its important legitimate duties, OR you open the door to all kinds of illegitimate activities.
The government has all kinds of powers that are very easy to abuse; but generally (and this is a key point) it is constrained in using those powers. The police can bash down your door and invade your house with drawn weapons -- but only if they have a warrant. Now you may argue that even with warrants they're often abusing their power, and I'd agree with you. That doesn't mean I don't think they should ever be able to do that.
Well, that's how I would do it too, but it's evidently not how this particular reactor was designed back in the 50s. If the reactor failed in any kind of planned way, we'd have some idea what that was. What we're learning is that we don't have any idea how this particular design behaves when it's operating outside its normal operating envelope.
I think it's important to stipulate I'm talking about this design. You can't really talk about the "safety of nuclear power plants" as if all designs are the same. Clearly the old Soviet RBMKs are dangerous pieces of shite; and concerns about the GE Mark 1s were raised all the way back in the early 70s. Fukushima, in my opinion, probably means it's a high priority not to extend the operational licensing of any Mark 1s. That doesn't necessarily mean the Germans should shut down reactors whose design is 20 years newer. And it certainly doesn't mean Japan should shut down it's brand new Gen 3 reactors it just put into service, although maybe they should look at how those plants are managed and sited.
I'll go one better. I have a small chip of uranium ore in my desk drawer right this very moment. I got it to test a geiger counter gizmo I got for testing whether old watches I was working on had radium pigment. Naturally I had to make sure the thing works before I trusted a "normal" reading.
I'm not afraid of handling this bit of ore, not in the least. But I wouldn't feel the same about handling the same amount of refined fuel, or the random by-products of a reactor disaster. Clearly I'm not radiation phobic, but extrapolating from the safety of handling ore to the products of a reactor accident is just plain stupid.
I'd have no fear of wearing a watch with radium pigment by the way; I've measured the radiation from them and what you get, even on the face of the watch, isn't a big concern especially if you don't wear it every day; maybe 3x background radiation in my neck of the woods. But I don't work on old radium watches because the binder in the pigment breaks down. If you open the case it'll release radium dust into the room. Would I freak out if I opened a radium watch by accident? No, I just wouldn't voluntarily put myself in that situation. There's a world of difference between carrying around intact radium pigment in a sealed case and breathing loose radium dust, just like there's no comparison between handling a piece of low grade uranium ore and exposing yourself to a radiological disaster.
In particular, it's worth noting that there is a rupture disk here precisely to prevent the reactor pressure vessel from experiencing a catastrophic rupture
This seemed wrong to me, since the RPV in Unit 2 was already breached. I believe AmiMoJo was talking about is pressure in the the primary containment vessel (PCV), not the RPV. Just to be clear the reactor core is inside the metal RPV and the RPV is inside the reinforced concrete PCV. The in the Mark 1 reactor design the PCV is the outer wall of the "dry well".
I also looked up some design diagrams for the venting system. While venting system rupture disk is indeed designed to protect the PCV it is not built into the PCV itself. Operators have to open two sets of valves in order to transfer pressure from inside the PCV to the disk. So while it's true the purpose of the rupture disk is to prevent catastrophic failure, that can only happen if the venting system is activated and works as expected.
Now the report you linked to is four years old and assumes that the venting system worked correctly, delivering overpressure to the the rupture disk. AmiMoJo is referring to evidence which came out later which indicates that the venting system almost certainly failed.
In a way I do agree with you. It's not mysterious why the PCV didn't explode; it didn't explode because it failed in some other, unknown way. Under the circumstances that was a very good thing in comparison to the alternative, but it takes a rather determined optimism to construe it as an endorsement of the reactor's design. It was more like a stroke of good luck.
This article has both a detailed diagram of the Oyster Creek reactor, which is the same design, and a schematic of the venting system.
Yet strangely enough we don't find it necessary today to bury the by-products of the chlor-alkalai process.
Well, to a programmer a "bug" is an error in logic. To a user a "bug" is an error in results. A single result may stem from many errors, particularly if we're talking about some kind of resource starvation, which this sounds like. If the starvation was caused by a single line memory leak, well that wouldn't take that long to find and fix. But if it were a faulty approach to managing memory used extensively over several years of development, you'd have a substantial body of logic errors that would take some serious spadework to fix.
One of the thing I've noticed is that some programs exhibit misbehaviors that seem to persist no matter how much fixing you do. These programs don't suffer from a single error, but rather a a small set of sloppy or inattentive coding habits that are repeated over and over. Fortunately sloppy code tends to be redundant, so you can often refactor and score big reductions in code size along with improved reliability and performance.
Just to play devil's advocate here for a second, clearly you don't need a human to run a lab test you planned to do right from the drawing board, but you might find it helpful to have a human available if an unexpected problem or opportunity arises. Humans are versatile, machines are not -- not in comparison at least. When the heater for your assay breaks down a human can look at it and scavenge parts from another piece of equipment. If a Martian strolls up and says "hi", a human can say "hi" back.
Take geology. I have no doubt whatsoever that a geologist actually on Mars with his full toolkit and a general purpose lab would get more geology done than some robot carrying out a few specially designed experiments. I wouldn't be surprised to learn that there are a few geology things we just can't do yet with robots yet. I just don't think we can get more geology done with humans for what we have to spend, even supposing that's enough to pay for a manned mission.
In time that may change. At some point we may experience diminishing returns from each new robotic mission, and then it will be time to prioritize manned missions over robotic ones. But we're far from that point ow.
This was actually a case of the PUBLIC SCHOOL BOARD going for the cheap land and plopping a school on top of a declared toxic dump site.
Which shouldn't have been there in the first place.
Love Canal wasn't a mining site.
Thank you, Captain Obvious. However I think the poster's point may have been that it's been historically difficult to hold corporations responsible for the messes they leave behind, and when you can't do that it means the public has to pay to clean them up.
100% false. We can study a lot in orbit but there are limits. You cannot study much about biological issues on other planetary bodies in orbit. For that you have to go there.
Alright. Name one issue of human biology that can only be studied by putting a human on Mars that is not specifically limited to the question of whether humans can live on Mars.
Setting aside human biology is akin to dumping the whole point of a manned space program. Can't do it.
Well, "whole point" is an exaggeration, but there's a reason I said "for the moment". I'm inviting you to consider the opportunity cost that a manned mission to Mars would present -- it's impact on other research priorities.
(1) Robotic exploration alone will never develop the technology needed to sustain life in space. (2) It would be very easy to over-focus on robotic exploration and starve the (harder and more expensive) manned exploration.
I agree with (1), but I wasn't suggesting abandoning manned space flight entirely, just prioritizing robotic spaceflight for exploration beyond the immediate vicinity of Earth. I disagree in a qualified way with (2). I think the most cost-effective way of extending manned space travel to Mars is get really good at all the Mars-specific aspects of getting stuff there and landing it on the surface by practicing with robotics, while continuing to pursue the generic aspects of manned spaceflight in orbit, or perhaps with Lunar missions.
Any reason to establish a base on the Moon or Mars or elsewhere will need to have some sort of economic benefit.
Agreed, but with the stipulation that it's the most cost effective way of obtaining that benefit. If it isn't, you're doing it for anthropological or political reasons.
There are things that cannot be easily or productively studied with robotic probes, not the least of which is human physiology and biology.
All of which could be studied in orbit. Now if you set human biology aside just for a moment, it would be more precise to say "some things are easier to study with human-run experiments than robots." The key point is that that set of things is a function of mission duration. Some things that are most efficiently done by humans on the Moon would be more efficiently done by robots on the Mars. Some things that are most efficiently done by humans on Mars would be more efficiently done by robots on Europa.
The problem is that the things that can only be done by humans on Mars would cost beyond what anyone wants to pay for a realistic mission, and a half-assed program that gets axed when the true cost of success becomes undeniable is a waste of time and money.
I'd like to see humans on Mars, but I think the shortest path to that happening starts with a sustained and regular program of robotic exploration. With experience we'll get better at getting things there cheaply and landing them there reliably. Then at some point an acceptably risky manned mission will become a financial possibility. But we won't do it because it's the most efficient way of doing Mars research, robotics will continue to advance as fast or faster than our space capabilities. The reasons for doing a manned mission will be a question for anthropology and political science, not economics or engineering.
On the other hand the relationship between a system's reliability and the reliability of the system's components isn't one-to-one. You can build unreliable systems out of reliable components, and more surprisingly, you can build reliable systems out of unreliable components. That latter principle is the basis for the Internet, which provides end-to-end communication that is more reliable than any of the possible paths between those endpoints.
Every component is unreliable to some degree; as it becomes increasingly reliable it moves from unusable, to something you can work around, to something whose potential failure you can ignore in many applications.
No, I just read the documentation.
And everyone else has been saying for months that part of the key is stored in the cpu in a region that is not readable. So all you need to do to back up the key is disassemble the cpu and hope you don't destroy it in the process.
Then "everyone else" would be wrong in this case. The "secure enclave" was introduced in Apple's A7 processor. The phone in question is an iPhone 5c which has an A6 processor. There is no crypto coprocessor to store anything in. The Apple docs even describe the process used to erase the keys in question from flash.
Now as for the iPhone 5S and later, there's no question they're tougher to hack -- that's the whole point of the new hardware features. But I still wouldn't put absolute faith in the inability of the new features to keep the government out if they had physical access to your phone.
Because people will listen to him, because he's Snowden.
I've been saying the same thing for months: the key is stored (albeit encrypted itself) in flash, so all you need to do is to back up the flash chip and you've got as many goes as you need, which with a four digit PIN chosen by humans isn't that many. But even though you can point this out the relevant details in Apple's documentation, people just refuse to believe that the government can get into an iPhone without Apple's help.
That actually kinds of mystifies me. Why would anyone believe that a government that (in part at least) created Stuxnet would be stymied by an iPhone? Whatever the reason, Snowden's imprimatur seems to help them get over it.
And yet RDBMs use ISAMs as part of their design to create and search records.
Using this to claim that RDBMSs haven't changed significantly in sixteen years is like claiming cars didn't change much between 1975 and 1990. Yeah, they still had cylinders and spark plugs, and the basic UI still had the same controls, but in terms of efficiency, reliability, performance and amenities the differences were stark.
Yeah, maybe if you're using your RDBMS as a backing store for your VB app, none of the changes that have happened in the last fifteen years are relevant to you, but believe me things like replication, clustering, MVCC, native geospatial processing, etc. are very relevant to a lot of people, and they didn't happen on their own. Even today there's a lot of things that could be improved, like transaction log forensics.
The boom is not a thing in itself; it is the product of economic conditions. So is the end of the boom. When spending an additional dollar on software development returns $0.99 in value instead of $1.00, people stop spending on software development. If nothing else changes, but suddenly that $0.99 of benefit only costs $0.98, the spending resumes.
Now that effect doesn't matter why the return is only $0.99. Perhaps you've saturated the market, and what once would have brought in $1.10 now only brings in $0.99. Sure that means that spending is done for, but only because you can't sell for a profit at that price.
You seem to be more concerned with words than reality. I actually remember things like the ISAM and network databases of the 70s. Trust me, they aren't up to the job of handling modern transaction volumes. Someone had to create new software, because you can run friggin' models.
Really, your response is full of straw men it's hard to respond to, so I'll limit myself to just this point: you seem to be under the impression that we have all the kinds of software we'll ever need already; in which case you're right: it doesn't matter if our jobs are taken by AI's; there's no more work.
I guess time will tell which of us is right, but I think the idea that all the kinds of software we need have already been invented could only be true if we've already imagined all the way there are to process information, and I doubt that's the case. If we'd been having this discussion fifteen years ago you might well be excused if you took the same position, but in fact there are many kinds of software developed since then that are either novel or radically different from how we thought things should be done back in 2000. For example I don't think that Oracle's RDBMS is really all that much better than it was back then, but it has some clear incremental improvements and useful new features; but that also misses the whole NoSQL movement.