It would be more useful to fix LibreOffice to produce output that looks as good as TeX.
One forgotten approach comes from Interleaf. Most WYSIWYG word processors today show you only the document - the markup is invisible. The old Word Perfect approach, where you could see the markup characters, or the HTML source approach, is too clunky. But Interleaf showed the output text alongside a column of annotation information. So you could see the difference between a tab indent and a paragraph indent, for example. That would be an appropriate way to present fine formatting controls.
Sonic.net offers gigabit fiber connections in Sebastopol CA now, and they're expanding next to the Sunset District in San Francisco. They may have more real paying customers on fiber than Google does. They're a small ISP and don't want to overextend themselves, so they're deploying slowly.
Basic truth: neither "social" nor "local" makes big money. Users, yes; profits, not so much. Compare Facebook's profits with Google, or Microsoft, or HP, or... And Facebook is considered a winner in "social". Failures include AOL, Geocities, Myspace, Orkut, and Google's various tries.
As for "local", Yahoo was the original "local" directory service. Where are they now? There's "local.com". Does anybody use "local.com"? Yelp is probably the leader, but loses money. If you're the industry leader for several years and are still losing money, your business model is fundamentally flawed.
If there's a winner at "local", it's going to be somebody in the phone space.
One major application for this is terminal guidance for munitions, like the Joint Direct Attack Munition and surface-to-ground missiles like the Hellfire. Those need an IMU so they can hit targets with GPS jammers. They get an initial position from the aircraft, which has a better IMU and upward-looking antennas which can probably get GPS despite ground jammers. All the small IMU has to do is keep a good position and heading for about a minute.
As this gets smaller, it becomes usable on more munitions, such as mortar rounds. Eventually, most indirect fire ammo will have this.
If the thing had good enough heading and position information, it could overlay detailed information on the real world. But it's not that good. It's just a smartphone display.
Also, I'll bet that driving with it will be prohibited after the first few hundred accidents.
There's always the possibility that the missile crew, being under orders to be ready to launch, was just doing a routine systems checkout.
During the Cuban missile crisis, US experts thought that the USSR was "sending a message". The missile sites in Cuba were easy to spot because they were laid out exactly the same way as in the USSR. Decades later, on the 40th anniversary of the crisis, there was a get-together of some of the major players from all sides. Someone asked about the layout of the sites. The Russian officer who'd been responsible for the layout said "No, we just did it that way because that was what it said to do in the field manual." Everyone in the room with military experience nodded in agreement.
Intelligence reports suggest that they can't fit a nuclear warhead on any kind of missile with decent range. Plus, their missile tests are falling way short of reaching anywhere but South Korea, China, and maybe Japan, and their test success ratio has been low.
They've achieved orbit. If you can reach orbit, you can potentially hit any place on earth. So far, they've only put a 100Kg payload in orbit. The smallest US nuclear weapons were around 50Kg, but it's not clear that North Korea can make a warhead that compact yet.
US export control on computers needs to stop. The need for it ended decades ago. All US nuclear weapons were designed with computers below 10 MIPS, and in many cases below 1 MIPS. (The most recent US nuclear weapon design is from the mid-1970s.) The problem isn't getting any harder.
From that perspective, Bitcoin is fascinating. The Bitcoin world, tiny though it is financially, has seen just about every classic scam. Ponzi/HYIP schemes, pump and dump schemes, fake bank schemes, fake exchange schemes, fake stocks, and ordinary theft - it's all there. It makes the penny stock market look legit.
Someday, assuming they haven't lost them, the people who generated Bitcoins in the early days when it was easy will cash out. Most of the early Bitcoins are not being traded. Somewhere there are early adopters with substantial value. They can't exit yet without clobbering the price, though. Not enough suckers putting in real money.
That's not insightful, it's clueless about the subject.The halting problem is very rarely a problem in program verification. It's possible to write a program for which halting is undecidable. That's a bug in the program.
The Microsoft Static Driver Verifier, in practice, runs into problems it can't decide about 5% of the time. The goal is to determine that the driver is memory-safe to be in the kernel and calls all the driver APIs with valid parameters. If it does that, the driver can't crash the rest of the OS, and any problems it has are confined to its own device. The Static Driver Verifier tries to either find a case that will crash or prove that a crash can't happen. The 5% of cases the Verifier can't decide are considered bugs. Really, if it's that hard to decide if a driver can cause a crash, the driver is broken.
I'm not sure how sound this new verification system is. There's been a lot of progress in this area in the last 30 years, since I worked on it. Formal verification is widely used in IC design, where "patch and release" doesn't work. The software world isn't as aware of the technology.
Throwing money at a problem only works if you known roughly what you want to do. The Manhattan Project had a well defined goal - 1) separate uranium isotopes or make plutonium, and 2) figure out some way to assemble them fast enough to get a fast chain reaction. They knew up front roughly what was needed. The Apollo program was a step up from the previous rocket programs, but it wasn't the first big rocket.
On the other hand, throwing money at controlled fusion has not been very successful. We don't know how to make that work. Throwing money at artificial intelligence didn't accomplish much until recent years. Interestingly, mobile robotics is now far enough along that throwing money at it works. NASA blew about $80 million on the Flight Telerobotic Servicer in the 1980s and got zip. DARPA has spent over $100 million with Boston Dynamics on BigDog, LS3, PETMAN, and ATLAS, and they're getting results.
The trouble with the BRAIN program is that they're talking about developing bigger computers to emulate a brain, but don't really know what problem they have to solve. This could turn into another supercomputer boondoggle. The comment I've made previously (once to Rod Brooks) about emulating a human brain is that you should try to emulate a mouse brain (1/1000th the mass) first. All the mammal brains have roughly the same architecture. Until you can emulate a mouse brain, you're not ready to try for a human brain. Brooks replied that "he didn't want to go down in history as the person who created the world's best robot mouse." So he tried Cog, which was an embarrassing flop, and hasn't been heard of much since.
My 2011 paper
"Social is bad for search, and search is bad for social" mentions that. It also names providers of fake "likes", fake "+1"s, fake reviews (some of which are very funny), fake accounts, fake IP addresses, and fake phone numbers for fake account verification. There's a whole ecosystem out there generating this junk.
Most of the sites identified in that paper are still in business. Some of the more blatant ones are "bulkaccounts.com" ("1000 Twitter accounts for $99") and
"pvaspot.com ("We Offer Top Quality Forwarded Phone Numbers used to create Phone Verified Accounts with a no questions asked 100% guarantee at Competitive Prices with Excellent Customer Service."). The fact that the same sites are active after two years indicates that the major social media networks can't or won't stop them.
As we point out in the paper, social media spam is cheaper and easier than link-farm spam. With a link farm, you have to set up servers, keep them up, fill them with fresh content. This gets expensive. With social media spam, the social media service hosts your spam for you, for free!
With Windows 7, Microsoft finally made it work. They developed the Static Driver Verifier, which uses proof of correctness techniques to insure that drivers won't crash the operating system, and made everybody run their drivers through it before they were signed. That eliminated about half of all crashes. Anything else was Microsoft's fault, and they knew it.
Microsoft also developed an internal tool that takes in crash dumps and matches them to other crash dumps. This made it possible to digest a huge number of crash dumps and tie them back to the cause.
With those tools, Microsoft finally had the ability to make the thing work. And they did. Windows 7 is much more reliable than previous versions of Windows.
Then, having finally produced a solid desktop system, they found they were being clobbered by the tablet industry, and came out with a desktop interface borrowed from a phone. Sigh.
I still have a copy of SuSE Pro 7 somewhere, that has a fuzzy time clock on KDE.
I built a standalone clock like that, using an AtMega CPU with an LCD display. The original idea came from a New Yorker cartoon decades ago, showing such a clock in the window of a clock shop.
Never put a PC/104 setup in a system that's going to be subjected to vibration
PC-104 is rather retro at this point, but there is something called a Can-Tainer for using PC-104 in hostile environments. "Internally, each corner of the PC/104 stack is held in place by a rubber corner system... Externally, the anodized aluminum enclosure mates with a thick rubber-mounting pad..."
We tried one of those in 2003-2005 for our DARPA Grand Challenge vehicle. (Getting board stacks into the Can-Tainer is a huge pain.) Even then, PC-104 was retro. We ended up with Tri-M industrial Pentium 4 PCs, which turned out to have an overheating problem due to really lame case design. (The CPU fan was aimed at a solid metal case bottom.) Back then there were fewer rugged computer options. Today, that situation is much better. There are good low-cost "Car PC" devices suitable for the automotive environment. Mobile hard drives are smaller and more shock-resistant. So today, getting enough compute power onto your large robot isn't a problem.
It's a good time to build robots. Many of the pieces that were hard to find or troublesome a decade ago are now mature products. Rugged computers, laser rangefinders, high precision GPS systems, attitude and heading reference systems, and servomotor controllers are all commercially available and not outrageously priced. A decade ago, you could get all of those things, but they were more expensive and didn't work as well.
I always thought of TeX as the last gasp of the RUNOFF/nroff/troff/ditroff line of document preparation; the last of the command-line oriented word processors. Having had access to Interleaf from 1985. TeX seemed so retro. (Interleaf was like Microsoft Word for Sun workstations. It was very early, very good, and very expensive.) TeX still had a compile-run-debug workflow, and without a graphic display, you had to run a hard copy on something like an electrostatic printer or a daisy wheel printer to check the results. Then you could go to the phototypesetting machine.
Once everybody got an interactive display good enough to view the output of TeX, formatting via macro calls was obsolete. Which is why almost everybody uses something like Microsoft Word now.
Since there are 3D printers controlled by four buttons and a two-line character display, the touch screen interface doesn't have to do much. Much modern industrial equipment has a touch-screen interface, simply because it's cheaper than building panels full of buttons and dials.
The emergency stop button on industrial machinery is always a physical big red button.
Painfully true. All the jerkyness in the video seems to be to conceal what a klutz the thing is. It's comparable to the OSU Hexapod from 1984. The OSU thing was supposed to have "off-road" capability, but it never did more than climb a slightly sloped dirt road. DoD cut off their funding after that.
As with the machine from 30 years ago, there seems to be an option to plant five legs and take manual control of the sixth. That's how they kicked the barrel. With this capability, the operator can (eventually) step over obstacles and ditches.
Mantis seems to have nice mechanical design. It's certainly better looking than the OSU machine. Without details of the hydraulic system, though, you can't tell how controllable it is. If it's all on-off valves with no accumulators for springiness and no force feedback, it's doomed to be clunky. If it has proportional valves, strain gauges, and accumulators in the right places, it has the potential for software upgrades to better movement.
The Mantis looks like it has pure kinematic control, like the 1980s OSU machine. There's been some progress in computing and control since the 1980s. You can do dynamic control, where balance and inertia are considered. Maybe not as good as BigDog, but better than pure kinematic. That thing should be able to go a lot faster on the flat. Hobbyist hexapod robots are moving faster and much more fluidly today.
Someone might modify the malware to still generate Bitcoins, but to record the coins generated. Then watch the blockchain to see who spends them. Bitcoins aren't anonymous. Mt. Gox has on at least one occasion frozen an account due to possession of "tainted" coins.
Bitcoin isn't as distributed as many enthusiasts think. 80% of transactions go through Mt. Gox, a/k/a Magic, the Gathering Online Exchange.
The Manhattan Project was successfully kept secret from the Germans, which was the primary goal. The German atomic program never got very far. How well the secret was kept is known, because, after the war, the major German physicists were interned in a big house in England and the house was bugged. The "Farm Hall Transcripts" record what they said. They didn't know how to enrich uranium in quantity. They didn't know how to make a workable bomb. Their calculations on assembly time for a gun bomb were way off; they didn't think a gun bomb would work. They had no clue about implosion.
I used to subscribe to the Baffler, but issues came out less and less often, then stopped completely. Now they're back. They're one of the few publications still publishing serious essays.
That said, this essay is more about Stallman vs. O'Reilly. That's a modestly interesting subject, but has little to do with government.
But what if government were "more connected"? What would it look like?
Banks used to be very disconnected internally. You could have a checking account, a savings account, a credit card, and a mortgage, and the different departments of the bank didn't know about the other accounts. Today, banks consolidate your "total relationship with the bank" into one online portal, and all the accounts can interconnect.
Suppose government did that. Federal and state income taxes, property taxes, parking fines, traffic tickets, bridge tolls, child support, welfare, social security, and Medicare, all on one convenient monthly statement. That's almost a no-brainer with today's technology. Some countries do that now; Sweden, for one.
Then integrate employment - the employer side verifies that you're in the system, and takes care of taxes, immigration status, and medical insurance premiums. Less paperwork for everybody. For casual employment, a Square reader and a smartphone app handles the paperwork.
That's "connected government". Is that what you want?
Slashdot Mirror
It would be more useful to fix LibreOffice to produce output that looks as good as TeX.
One forgotten approach comes from Interleaf. Most WYSIWYG word processors today show you only the document - the markup is invisible. The old Word Perfect approach, where you could see the markup characters, or the HTML source approach, is too clunky. But Interleaf showed the output text alongside a column of annotation information. So you could see the difference between a tab indent and a paragraph indent, for example. That would be an appropriate way to present fine formatting controls.
Sonic.net offers gigabit fiber connections in Sebastopol CA now, and they're expanding next to the Sunset District in San Francisco. They may have more real paying customers on fiber than Google does. They're a small ISP and don't want to overextend themselves, so they're deploying slowly.
Basic truth: neither "social" nor "local" makes big money. Users, yes; profits, not so much. Compare Facebook's profits with Google, or Microsoft, or HP, or ... And Facebook is considered a winner in "social". Failures include AOL, Geocities, Myspace, Orkut, and Google's various tries.
As for "local", Yahoo was the original "local" directory service. Where are they now? There's "local.com". Does anybody use "local.com"? Yelp is probably the leader, but loses money. If you're the industry leader for several years and are still losing money, your business model is fundamentally flawed.
If there's a winner at "local", it's going to be somebody in the phone space.
All the companies named are from the anal-probe sector of the tech industry.
One major application for this is terminal guidance for munitions, like the Joint Direct Attack Munition and surface-to-ground missiles like the Hellfire. Those need an IMU so they can hit targets with GPS jammers. They get an initial position from the aircraft, which has a better IMU and upward-looking antennas which can probably get GPS despite ground jammers. All the small IMU has to do is keep a good position and heading for about a minute.
As this gets smaller, it becomes usable on more munitions, such as mortar rounds. Eventually, most indirect fire ammo will have this.
If the thing had good enough heading and position information, it could overlay detailed information on the real world. But it's not that good. It's just a smartphone display.
Also, I'll bet that driving with it will be prohibited after the first few hundred accidents.
There's always the possibility that the missile crew, being under orders to be ready to launch, was just doing a routine systems checkout.
During the Cuban missile crisis, US experts thought that the USSR was "sending a message". The missile sites in Cuba were easy to spot because they were laid out exactly the same way as in the USSR. Decades later, on the 40th anniversary of the crisis, there was a get-together of some of the major players from all sides. Someone asked about the layout of the sites. The Russian officer who'd been responsible for the layout said "No, we just did it that way because that was what it said to do in the field manual." Everyone in the room with military experience nodded in agreement.
Intelligence reports suggest that they can't fit a nuclear warhead on any kind of missile with decent range. Plus, their missile tests are falling way short of reaching anywhere but South Korea, China, and maybe Japan, and their test success ratio has been low.
They've achieved orbit. If you can reach orbit, you can potentially hit any place on earth. So far, they've only put a 100Kg payload in orbit. The smallest US nuclear weapons were around 50Kg, but it's not clear that North Korea can make a warhead that compact yet.
US export control on computers needs to stop. The need for it ended decades ago. All US nuclear weapons were designed with computers below 10 MIPS, and in many cases below 1 MIPS. (The most recent US nuclear weapon design is from the mid-1970s.) The problem isn't getting any harder.
We need redundancy.
Most of the competing exchanges were scammers, incompetent, or both.
we can get real good data from this experiment.
From that perspective, Bitcoin is fascinating. The Bitcoin world, tiny though it is financially, has seen just about every classic scam. Ponzi/HYIP schemes, pump and dump schemes, fake bank schemes, fake exchange schemes, fake stocks, and ordinary theft - it's all there. It makes the penny stock market look legit.
Someday, assuming they haven't lost them, the people who generated Bitcoins in the early days when it was easy will cash out. Most of the early Bitcoins are not being traded. Somewhere there are early adopters with substantial value. They can't exit yet without clobbering the price, though. Not enough suckers putting in real money.
This might make sense for some Russian oil oligarch who has to visit oil platforms in the White Sea. For anybody else, it's kind of pointless.
At least it's more seaworthy than that boxy thing Steve Jobs had built.
Microsoft offers the Malicious Software Removal Tool (IA32 version) , (AMD64 version) which they update monthly. It's not perfect, but it's worth running on Windows machines.
If Congress wants to apply pressure to somebody, it might be worthwhile to investigate how well that's working, and what it's missing.
wow, does it solve the halting problem as well?
That's not insightful, it's clueless about the subject.The halting problem is very rarely a problem in program verification. It's possible to write a program for which halting is undecidable. That's a bug in the program.
The Microsoft Static Driver Verifier, in practice, runs into problems it can't decide about 5% of the time. The goal is to determine that the driver is memory-safe to be in the kernel and calls all the driver APIs with valid parameters. If it does that, the driver can't crash the rest of the OS, and any problems it has are confined to its own device. The Static Driver Verifier tries to either find a case that will crash or prove that a crash can't happen. The 5% of cases the Verifier can't decide are considered bugs. Really, if it's that hard to decide if a driver can cause a crash, the driver is broken.
I'm not sure how sound this new verification system is. There's been a lot of progress in this area in the last 30 years, since I worked on it. Formal verification is widely used in IC design, where "patch and release" doesn't work. The software world isn't as aware of the technology.
Throwing money at a problem only works if you known roughly what you want to do. The Manhattan Project had a well defined goal - 1) separate uranium isotopes or make plutonium, and 2) figure out some way to assemble them fast enough to get a fast chain reaction. They knew up front roughly what was needed. The Apollo program was a step up from the previous rocket programs, but it wasn't the first big rocket.
On the other hand, throwing money at controlled fusion has not been very successful. We don't know how to make that work. Throwing money at artificial intelligence didn't accomplish much until recent years. Interestingly, mobile robotics is now far enough along that throwing money at it works. NASA blew about $80 million on the Flight Telerobotic Servicer in the 1980s and got zip. DARPA has spent over $100 million with Boston Dynamics on BigDog, LS3, PETMAN, and ATLAS, and they're getting results.
The trouble with the BRAIN program is that they're talking about developing bigger computers to emulate a brain, but don't really know what problem they have to solve. This could turn into another supercomputer boondoggle. The comment I've made previously (once to Rod Brooks) about emulating a human brain is that you should try to emulate a mouse brain (1/1000th the mass) first. All the mammal brains have roughly the same architecture. Until you can emulate a mouse brain, you're not ready to try for a human brain. Brooks replied that "he didn't want to go down in history as the person who created the world's best robot mouse." So he tried Cog, which was an embarrassing flop, and hasn't been heard of much since.
My 2011 paper "Social is bad for search, and search is bad for social" mentions that. It also names providers of fake "likes", fake "+1"s, fake reviews (some of which are very funny), fake accounts, fake IP addresses, and fake phone numbers for fake account verification. There's a whole ecosystem out there generating this junk.
Most of the sites identified in that paper are still in business. Some of the more blatant ones are "bulkaccounts.com" ("1000 Twitter accounts for $99") and "pvaspot.com ("We Offer Top Quality Forwarded Phone Numbers used to create Phone Verified Accounts with a no questions asked 100% guarantee at Competitive Prices with Excellent Customer Service."). The fact that the same sites are active after two years indicates that the major social media networks can't or won't stop them.
As we point out in the paper, social media spam is cheaper and easier than link-farm spam. With a link farm, you have to set up servers, keep them up, fill them with fresh content. This gets expensive. With social media spam, the social media service hosts your spam for you, for free!
With Windows 7, Microsoft finally made it work. They developed the Static Driver Verifier, which uses proof of correctness techniques to insure that drivers won't crash the operating system, and made everybody run their drivers through it before they were signed. That eliminated about half of all crashes. Anything else was Microsoft's fault, and they knew it.
Microsoft also developed an internal tool that takes in crash dumps and matches them to other crash dumps. This made it possible to digest a huge number of crash dumps and tie them back to the cause.
With those tools, Microsoft finally had the ability to make the thing work. And they did. Windows 7 is much more reliable than previous versions of Windows.
Then, having finally produced a solid desktop system, they found they were being clobbered by the tablet industry, and came out with a desktop interface borrowed from a phone. Sigh.
I still have a copy of SuSE Pro 7 somewhere, that has a fuzzy time clock on KDE.
I built a standalone clock like that, using an AtMega CPU with an LCD display. The original idea came from a New Yorker cartoon decades ago, showing such a clock in the window of a clock shop.
Never put a PC/104 setup in a system that's going to be subjected to vibration
PC-104 is rather retro at this point, but there is something called a Can-Tainer for using PC-104 in hostile environments. "Internally, each corner of the PC/104 stack is held in place by a rubber corner system ... Externally, the anodized aluminum enclosure mates with a thick rubber-mounting pad..."
We tried one of those in 2003-2005 for our DARPA Grand Challenge vehicle. (Getting board stacks into the Can-Tainer is a huge pain.) Even then, PC-104 was retro. We ended up with Tri-M industrial Pentium 4 PCs, which turned out to have an overheating problem due to really lame case design. (The CPU fan was aimed at a solid metal case bottom.) Back then there were fewer rugged computer options. Today, that situation is much better. There are good low-cost "Car PC" devices suitable for the automotive environment. Mobile hard drives are smaller and more shock-resistant. So today, getting enough compute power onto your large robot isn't a problem.
It's a good time to build robots. Many of the pieces that were hard to find or troublesome a decade ago are now mature products. Rugged computers, laser rangefinders, high precision GPS systems, attitude and heading reference systems, and servomotor controllers are all commercially available and not outrageously priced. A decade ago, you could get all of those things, but they were more expensive and didn't work as well.
I always thought of TeX as the last gasp of the RUNOFF/nroff/troff/ditroff line of document preparation; the last of the command-line oriented word processors. Having had access to Interleaf from 1985. TeX seemed so retro. (Interleaf was like Microsoft Word for Sun workstations. It was very early, very good, and very expensive.) TeX still had a compile-run-debug workflow, and without a graphic display, you had to run a hard copy on something like an electrostatic printer or a daisy wheel printer to check the results. Then you could go to the phototypesetting machine.
Once everybody got an interactive display good enough to view the output of TeX, formatting via macro calls was obsolete. Which is why almost everybody uses something like Microsoft Word now.
Since there are 3D printers controlled by four buttons and a two-line character display, the touch screen interface doesn't have to do much. Much modern industrial equipment has a touch-screen interface, simply because it's cheaper than building panels full of buttons and dials.
The emergency stop button on industrial machinery is always a physical big red button.
Painfully true. All the jerkyness in the video seems to be to conceal what a klutz the thing is. It's comparable to the OSU Hexapod from 1984. The OSU thing was supposed to have "off-road" capability, but it never did more than climb a slightly sloped dirt road. DoD cut off their funding after that.
As with the machine from 30 years ago, there seems to be an option to plant five legs and take manual control of the sixth. That's how they kicked the barrel. With this capability, the operator can (eventually) step over obstacles and ditches.
Mantis seems to have nice mechanical design. It's certainly better looking than the OSU machine. Without details of the hydraulic system, though, you can't tell how controllable it is. If it's all on-off valves with no accumulators for springiness and no force feedback, it's doomed to be clunky. If it has proportional valves, strain gauges, and accumulators in the right places, it has the potential for software upgrades to better movement.
The Mantis looks like it has pure kinematic control, like the 1980s OSU machine. There's been some progress in computing and control since the 1980s. You can do dynamic control, where balance and inertia are considered. Maybe not as good as BigDog, but better than pure kinematic. That thing should be able to go a lot faster on the flat. Hobbyist hexapod robots are moving faster and much more fluidly today.
Someone might modify the malware to still generate Bitcoins, but to record the coins generated. Then watch the blockchain to see who spends them. Bitcoins aren't anonymous. Mt. Gox has on at least one occasion frozen an account due to possession of "tainted" coins.
Bitcoin isn't as distributed as many enthusiasts think. 80% of transactions go through Mt. Gox, a/k/a Magic, the Gathering Online Exchange.
The Manhattan Project was successfully kept secret from the Germans, which was the primary goal. The German atomic program never got very far. How well the secret was kept is known, because, after the war, the major German physicists were interned in a big house in England and the house was bugged. The "Farm Hall Transcripts" record what they said. They didn't know how to enrich uranium in quantity. They didn't know how to make a workable bomb. Their calculations on assembly time for a gun bomb were way off; they didn't think a gun bomb would work. They had no clue about implosion.
I used to subscribe to the Baffler, but issues came out less and less often, then stopped completely. Now they're back. They're one of the few publications still publishing serious essays.
That said, this essay is more about Stallman vs. O'Reilly. That's a modestly interesting subject, but has little to do with government. But what if government were "more connected"? What would it look like?
Banks used to be very disconnected internally. You could have a checking account, a savings account, a credit card, and a mortgage, and the different departments of the bank didn't know about the other accounts. Today, banks consolidate your "total relationship with the bank" into one online portal, and all the accounts can interconnect.
Suppose government did that. Federal and state income taxes, property taxes, parking fines, traffic tickets, bridge tolls, child support, welfare, social security, and Medicare, all on one convenient monthly statement. That's almost a no-brainer with today's technology. Some countries do that now; Sweden, for one.
Then integrate employment - the employer side verifies that you're in the system, and takes care of taxes, immigration status, and medical insurance premiums. Less paperwork for everybody. For casual employment, a Square reader and a smartphone app handles the paperwork.
That's "connected government". Is that what you want?