Problem is that it isn't unheard of for firmware updates to come with other things that are useless, be it some sort of ad mechanism, some way to invade privacy, or even code to disallow the TV to work if it doesn't have an always-on connection.
I wish TVs would just have a SD card, and accept signed firmware updates through that. Online updates is just a vector for attackers, similar to how CCleaner was hijacked.
This is not surprising in the least. On a physical level, the person who likely has the most access is the janitor or cleaning staff.
Almost any access can be abused, if someone feels vindictive enough. An electrical worker can toss a dead rat in an opened panel, and the arc flash likely would take out a good amount of power in the building.
Having access controls to minimize things are critical, but even with those in place, there is a point where the problem changes from a technological issue to a HR issue, of why someone is that pissed and vindictive in the first place... and why they were cleared for access.
I wouldn't mind an OS-less phone, provided that it is easy to install one's ROM of choice. It would be nice to have LineageOS supported from the start, for example.
Here is the ironic thing, when I see a report about remarking about climage change: It must be something very notable and significant, because it is definitely in the financial interest of the powers that be to play it down.
So, if scientists that will have Hell to pay for climate change are stepping forward with these results, the actual damage being done may be far, far worse than what we see now. Especially areas like the Sahel in Africa where when resources dry up, conflicts start, mainly because it turns to fighting or starving.
Right now, the view in a lot of places may be "who cares about Africa?", but that view only makes groups like Daesh stronger. The world's problems cannot all be solved by bullets (Iraq and Afghanistan have shown the US and USSR that), so it might be in the interest of civilization to at least find ways to mitigate desertification, work on desalination and effective irrigation, and find ways to reclaim arable land from the sea.
Bubbles may come and go, but stuff does remain from them. We will see blockchains rise and fall... and over the long haul, wind up a useful tool, just like SANs, Linux, and many other technologies.
The reason we are seeing another bubble is that with the GDPR going in effect next year, ad companies are going to have to make a lot of changes, and this will result in some revenue losses due to this. So, people are taking their cash and going somewhere else that has a possibility of growth.
AI especially. AI has a lot of promise to add an advantage, be it military or competitive to any company that is advanced enough. Even an all-out war can be handled by an AI good enough that it alone can replace generals on the battlefield, with strategies and tactics changing in real time from views and casualties taken from the troops.
Robots as well. Why risk people's lives when you can send robots onto a battlefield to find and destroy mines? Plus, a robot can be equipped with sensors to help an AI gauge how to change tactics/strategy.
The cloud? Definitely not going anywhere. The promise of NoOps and being able to fire the "rack and stack" guys, the OS guys, the DBAs, and keep an offshored dev team is very tempting. O365 is making big inroads here, and Microsoft's "we do all IT" offering for SMBs announced recently will also change this, to where all but the biggest companies are using some cloud provider, even if it is a security item.
There are bubbles, but there are markets where there is true growth happening. Automation, especially.
Say a blockchain is used as part of making a widget. Everything on the widget's bill of materials [1] would be signed for and part of the ledger. On one hand, this may be a good thing, as it ensures every part is signed off (the maker stated they used grade 8 bolts, so if they used grade 2, it would be obvious). On the other hand, I doubt companies want to guarantee every single part's quality, especially if they just buy and spec what's the absolute cheapest thing from China.
[1] I hate saying the abbreviation because in a previous job, the words, "that is a crazy-ass BOM" got the cops called to the accounting office.
It is a balancing act. One one hand, if someone uses weak (but memorable) passwords, that can be brute-forced, that is far more likely than a password manager getting compromised, especially a password manager with 2FA.
However, selecting a password manager is critical. LastPass is one that has had security intrusions succeed... but were mitigated. Some other PW managers which have, as of their latest versions, required cloud access (1Password, mSecure) not just don't have a proven track record... but don't even give any details on what security they actually bother with. For all we know, they could stash everything on a public S3 bucket.
I like PW managers which piggyback on existing cloud providers and have decent encryption [1], like Enpass or Codebook. That way, not all eggs are in one basket, and Google Drive provides adequate 2FA protection.
[1]: The idea would be separating the passphrase protecting the database on the cloud provider versus the encrypted copy, or even better, using public key encryption and "introducing" new devices, to make the copy sitting on the cloud provider as brute force resistant as possible.
I wonder how one would protect against keyboard loggers. Since they are totally passive, an ID on a keyboard would do little at all.
The only way I can really see it happening is with a separate protocol from USB (perhaps fiber optic, a la S/PDIF), where the keyboard and the computer are paired, the keyboard uses epoxy potting and tamper-evident wiring and enclosures, and some form of cryptographic handshaking is done. The instruct users that no "secure" light on the keyboard, no typing.
Of course, this also shows how important 2FA is, especially with regards to grades. One ideal would be having the info changed on the computer, then confirmations showing the changes appear on someone's smartphone, similar to how the old IBM Zurich ZTIC would show proposed transactions and ask to allow or deny them. That way, someone would have to get ahold of the access token as well as get the username/password pairing.
KeePassXC might be a suitable replacement. I like KeePass's password generator, especially with the fact that it can generate via templates and use input from the keyboard/mouse to supplement the RNG. However, KeePass isn't the only game in town.
It would be nice to have some added security with data sitting on a cloud provider, so someone who grabs the password database can't just brute-force a password. With some password managers, one can have a sync password that is different from the one used to access the DB, so one can have a 64 character password for that, and a shorter one for access on the local machine. Other password managers require endpoints to be "introduced", and store the database encrypted, with the master key to the DB encrypted to each endpoint's private key. That way, there is no password that can be guessed.
What I really detest is how utilities like mSecure (and to a lesser extent, 1Password) have moved. Want to use them? You have to use their cloud specific cloud, which has no certifications other than "we use AES-256". If I want to use someone's proprietary cloud for passwords, I would use LastPass which at least has been proven to mitigate attacks with its structure. Who knows how these guys store their DB... it could be stashed in a publically accessible S3 bucket, for all we know.
If a 2FA device has some means of communication to the site that is authenticating, 2FA is trivial. Just like with Google, Blizzard, or Duo... when you log on, your phone pops up (login attempt detected... Allow/Deny), you hit "allow", and you are in.
It would be nice if there were an open standard for this, with the site wanting authenticating storing a public key, and the 2FA device generating and storing a private key onboard. Right now, we have an open standard for shared secrets, but it would be nice to move to a public/private key standard, so a compromise of the server requesting authenticating gives little help to an attacker.
Normal people may not know how exactly the mechanism of their deadbolt at home works, but they turn the key and ensure it is locked. They may not know how their car's remote does a handshake with the vehicle's computer, but they at least know how to press the lock button.
Computer encryption is insanely easy. You don't have to know about S-boxes or shifting stuff around to click on a file, click "encrypt", type in "correct horse battery staple" and be on your way.
This. It is trivial to ensure data on a USB flash drive is encrypted:
1: $50 gets you an Iepin hardware encrypting USB drive that has a keypad on it. Ten wrong guesses, and you have a blank USB hard drive. You can get an IronKey drive for a bit more that has actual epoxy potting and physical destruction of circuits if one tries to guess the password too often. 2: BitLocker, FileVault, LUKS, and VeraCrypt are common and easy to use. If you have a keyfile at home and at work, and you use VeraCrypt, an attacker has no way to decrypt the drive, except for a brute force against the entire 256 bit length, or physically breaking into the two locations. 3: If policies lock out encryption on the drive level, WinZIP, WinRAR, Acrobat and Microsoft Word can all AES-128 encrypt any files and encapsulate them in their own format. 4: Any big company has some security/compliance education to teach people the dangers of unencrypted media, be it tape, CDs, or whatnot. 5: One's smartphone can carry data. Both iOS and Android support the ability to copy files to the device, and both have apps that can actually create and open VeraCrypt containers on the device for further security. A rooted Android phone can even mount the VC container so you can copy stuff directly into it. 6: There are commercial utilities like Boxcryptor which make encryption extremely transparent. 7: If policies lock out software solutions, then that is what a hardware encrypted drive is for (see #1.)
Ten years ago, there might have been an excuse. These days, with encryption so easy to access and use, there isn't an excuse for this. I almost wonder if the USB flash drive being lost was a deliberate act of sabotage.
Rooting not just allows you to rip crap like that out, but gives you the ability to firewall, guaranteeing that if it ever came back, it wouldn't be able to phone home. I wish DonkeyGuard or xPrivacy were maintained. Those programs were top notch at not just keeping an app from getting to goodies, but giving it fake info, so it thinks it has contacts, photos, camera, mic... but in reality, it is pulling random numbers and repeatedly uploading a black screen.
The Indian smartphone market being large makes sense because for a lot of people, that is their main computing device. Other countries, the smartphone is a secondary device and usually paired with a laptop or desktop, perhaps with a tablet as well.
What would be interesting is designing a smartphone from the ground up as a primary computing device. This would mean far more sophisticated backups (which are encrypted), ease of recovery over the air, even if one's ROM was corrupted, choices of ROMs, more flexibility in storage (one MicroSD card for internal secondary storage, one that would be external for backups, and at least 64-128 gigs of fast internal storage.) Battery replacement may be useful, but I'd assert that IP68 resistance would be more important, and a way to do an external battery case would be a better idea.
Software-wise, designing a phone around VMs, so one can have private stuff, work stuff, and general items in separate areas, switchable by varying authentication modes (fingerprint, face print, PIN, password, or some combination of the above.) That way, if one's "social network" VM got compromised, it is easily dumped, or restored back to a known safe snapshot. Ideally, it would be nice to have some deduplication capability present on the/system partition (if looking at an Android model), so multiple VMs take up as little space as possible, while still providing encryption and separation.
There is a lot of research a company could do, if making a phone as a primary computing device. Perhaps a wireless standard to allow charging, as well as streaming video, and using a HID, so the phone can be tossed on top of a dock, and it can function as a desktop, similar to the old Motorola Atrix line, but wirelessly.
Microsoft still makes money hand-over-fist in the mobile department with their patents. They really don't need to do much in that field other than sit back and take the licensing fees. Windows Mobile was a decent OS... but developers just won't bother with a third platform.
Ironically, before Android and iOS, WM was the most popular smartphone OS out there.
How about a college system like a lot of European countries? When I was in school, my German friend had his tuition paid for by the Fatherland. The Russian student? By the Motherland. The Chinese engineer? By his country. The guy from Chile? Paid for by his government. Compare that to Americans which have to mortgage their future and have to earn significantly more to maintain the same lifestyle, and it is no wonder why there are economic issues in the US.
The problem with STEM here in the US is that the government and businesses make a mistake that even the most brain-dead farmer would never do. If you expect to have a harvest, you plant a crop. You have to fund R&D, fund colleges and universities, and give Joe Sixpack Jr. a reason to go into engineering or science, and not law or finance. Because there isn't any interest in plowing a field, there are very few returns, and it is no wonder why other countries (like Germany or China) who offer university education for a reasonable cost are reaping rewards, while here in the US, many people consider having roads and a power grid "socialism"... and then wonder why prosperity has left this country.
Tenure has/had a place, mainly to ensure that a professor that states an unpopular viewpoint won't get run out as a knee-jerk reaction. Is it still relevant? With the insane political divisions in the US, where stating something about a hot button issue can have grave consequences, tenure is definitely a must.
Maybe it needs redone, but it seems to work, and it is better than just having universities just be an echo chamber for a small set of political beliefs.
For laptops like this where battery life is a major factor, I wonder how the AMD side of the house compares to Intel's offerings. Having the GPUs on the same chip might help things.
There is a way to have enough data for a transaction, but no more. A certificate based system, where one's ID card just validates the cardholder is whom they claim to be, and is a repository for certificates. For example, a certificate showing the person is over age 21. That way, they can go to a bar in the US, and the cert provides what the bar needs to know to comply with the law. The bar doesn't need names, ages, or anything else. Just that the bearer is over 21.
This could be extended to a lot of other things, and to reduce fraud, short-lived certs should be used. For example, a cert that lasts 1-2 days that is done by a police department certifying someone has no entries on their RAP sheet, and has no pending charges. This way, stuff can be done, but relevant info can be sequestered in small, scattered databases, so a breach would be of limited damage.
With the system we have, those watchdogs will fall to regulatory capture, and at best, be a rubber-stamping department.
The only thing that really can break this trend is Europe's GDPR. Time will tell if it actually will get companies to do something about security, or if it winds up being a joke, like SOX (where it was used to jail a guy who ran over his fishing bag limit at its best.) I'm sure BRICS will have similar laws on the books soon, because they want to stick it to US companies, so even that might be a help.
My cynical self feels like security issues will be all too common. It is too easy for an exec to hear that their firm was breached, they sell all their stock and short it, then make out like bandits when the issue hits the front pages.
Diesel can be synthesized, or if one is willing to use a process that takes a lot of energy (which is doable near a hydroelectric plant or somewhere where geography permits), one could take plastic trash, then use thermal depolymerization in order to get a usable diesel oil. It is energy intensive, but it removes plastic from the environment.
Of course, there is biodiesel and all the waste oil that comes from restaurants, as well as motor oil. Run that (B100) in a diesel engine, and the engine will run extremely clean.
Hybrids are nice. I have not gotten why we don't just have some type of engine that is happy at a low RPM that keeps a battery bank charged up, and the vehicle runs on electric motors. This way, for day to day use, the battery bank keeps things going, but there is always the IC engine with fuel sitting ready to charge the battery when it gets low. The closest we have to this would be a plug in Prius or a Chevy Volt.
Designing an engine to run at a certain RPM is a lot easier than figuring out power bands and a transmission. You could even use a turbine engine, which is content to burn a wide variety of fuels, because it is always running at a specific RPM to keep the battery bank charged.
A mantrap is similar, but this is more of a vestibule for someone to put stuff in a secure spot without having access to the rest of the place. Perhaps it could be done by having a little hallway with a door at each end. It wouldn't need to have one door be locked for the other to be opened, but is intended to be a space to ensure stuff is secure from the outside, while keeping the rest of the home secure.
One could always get away with a dumb waiter type of mechanism with a multiple story house as well.
Slashdot Mirror
Problem is that it isn't unheard of for firmware updates to come with other things that are useless, be it some sort of ad mechanism, some way to invade privacy, or even code to disallow the TV to work if it doesn't have an always-on connection.
I wish TVs would just have a SD card, and accept signed firmware updates through that. Online updates is just a vector for attackers, similar to how CCleaner was hijacked.
This is not surprising in the least. On a physical level, the person who likely has the most access is the janitor or cleaning staff.
Almost any access can be abused, if someone feels vindictive enough. An electrical worker can toss a dead rat in an opened panel, and the arc flash likely would take out a good amount of power in the building.
Having access controls to minimize things are critical, but even with those in place, there is a point where the problem changes from a technological issue to a HR issue, of why someone is that pissed and vindictive in the first place... and why they were cleared for access.
I wouldn't mind an OS-less phone, provided that it is easy to install one's ROM of choice. It would be nice to have LineageOS supported from the start, for example.
Here is the ironic thing, when I see a report about remarking about climage change: It must be something very notable and significant, because it is definitely in the financial interest of the powers that be to play it down.
So, if scientists that will have Hell to pay for climate change are stepping forward with these results, the actual damage being done may be far, far worse than what we see now. Especially areas like the Sahel in Africa where when resources dry up, conflicts start, mainly because it turns to fighting or starving.
Right now, the view in a lot of places may be "who cares about Africa?", but that view only makes groups like Daesh stronger. The world's problems cannot all be solved by bullets (Iraq and Afghanistan have shown the US and USSR that), so it might be in the interest of civilization to at least find ways to mitigate desertification, work on desalination and effective irrigation, and find ways to reclaim arable land from the sea.
Bubbles may come and go, but stuff does remain from them. We will see blockchains rise and fall... and over the long haul, wind up a useful tool, just like SANs, Linux, and many other technologies.
The reason we are seeing another bubble is that with the GDPR going in effect next year, ad companies are going to have to make a lot of changes, and this will result in some revenue losses due to this. So, people are taking their cash and going somewhere else that has a possibility of growth.
AI especially. AI has a lot of promise to add an advantage, be it military or competitive to any company that is advanced enough. Even an all-out war can be handled by an AI good enough that it alone can replace generals on the battlefield, with strategies and tactics changing in real time from views and casualties taken from the troops.
Robots as well. Why risk people's lives when you can send robots onto a battlefield to find and destroy mines? Plus, a robot can be equipped with sensors to help an AI gauge how to change tactics/strategy.
The cloud? Definitely not going anywhere. The promise of NoOps and being able to fire the "rack and stack" guys, the OS guys, the DBAs, and keep an offshored dev team is very tempting. O365 is making big inroads here, and Microsoft's "we do all IT" offering for SMBs announced recently will also change this, to where all but the biggest companies are using some cloud provider, even if it is a security item.
There are bubbles, but there are markets where there is true growth happening. Automation, especially.
Say a blockchain is used as part of making a widget. Everything on the widget's bill of materials [1] would be signed for and part of the ledger. On one hand, this may be a good thing, as it ensures every part is signed off (the maker stated they used grade 8 bolts, so if they used grade 2, it would be obvious). On the other hand, I doubt companies want to guarantee every single part's quality, especially if they just buy and spec what's the absolute cheapest thing from China.
[1] I hate saying the abbreviation because in a previous job, the words, "that is a crazy-ass BOM" got the cops called to the accounting office.
It is a balancing act. One one hand, if someone uses weak (but memorable) passwords, that can be brute-forced, that is far more likely than a password manager getting compromised, especially a password manager with 2FA.
However, selecting a password manager is critical. LastPass is one that has had security intrusions succeed... but were mitigated. Some other PW managers which have, as of their latest versions, required cloud access (1Password, mSecure) not just don't have a proven track record... but don't even give any details on what security they actually bother with. For all we know, they could stash everything on a public S3 bucket.
I like PW managers which piggyback on existing cloud providers and have decent encryption [1], like Enpass or Codebook. That way, not all eggs are in one basket, and Google Drive provides adequate 2FA protection.
[1]: The idea would be separating the passphrase protecting the database on the cloud provider versus the encrypted copy, or even better, using public key encryption and "introducing" new devices, to make the copy sitting on the cloud provider as brute force resistant as possible.
I wonder how one would protect against keyboard loggers. Since they are totally passive, an ID on a keyboard would do little at all.
The only way I can really see it happening is with a separate protocol from USB (perhaps fiber optic, a la S/PDIF), where the keyboard and the computer are paired, the keyboard uses epoxy potting and tamper-evident wiring and enclosures, and some form of cryptographic handshaking is done. The instruct users that no "secure" light on the keyboard, no typing.
Of course, this also shows how important 2FA is, especially with regards to grades. One ideal would be having the info changed on the computer, then confirmations showing the changes appear on someone's smartphone, similar to how the old IBM Zurich ZTIC would show proposed transactions and ask to allow or deny them. That way, someone would have to get ahold of the access token as well as get the username/password pairing.
KeePassXC might be a suitable replacement. I like KeePass's password generator, especially with the fact that it can generate via templates and use input from the keyboard/mouse to supplement the RNG. However, KeePass isn't the only game in town.
It would be nice to have some added security with data sitting on a cloud provider, so someone who grabs the password database can't just brute-force a password. With some password managers, one can have a sync password that is different from the one used to access the DB, so one can have a 64 character password for that, and a shorter one for access on the local machine. Other password managers require endpoints to be "introduced", and store the database encrypted, with the master key to the DB encrypted to each endpoint's private key. That way, there is no password that can be guessed.
What I really detest is how utilities like mSecure (and to a lesser extent, 1Password) have moved. Want to use them? You have to use their cloud specific cloud, which has no certifications other than "we use AES-256". If I want to use someone's proprietary cloud for passwords, I would use LastPass which at least has been proven to mitigate attacks with its structure. Who knows how these guys store their DB... it could be stashed in a publically accessible S3 bucket, for all we know.
If a 2FA device has some means of communication to the site that is authenticating, 2FA is trivial. Just like with Google, Blizzard, or Duo... when you log on, your phone pops up (login attempt detected... Allow/Deny), you hit "allow", and you are in.
It would be nice if there were an open standard for this, with the site wanting authenticating storing a public key, and the 2FA device generating and storing a private key onboard. Right now, we have an open standard for shared secrets, but it would be nice to move to a public/private key standard, so a compromise of the server requesting authenticating gives little help to an attacker.
Normal people may not know how exactly the mechanism of their deadbolt at home works, but they turn the key and ensure it is locked. They may not know how their car's remote does a handshake with the vehicle's computer, but they at least know how to press the lock button.
Computer encryption is insanely easy. You don't have to know about S-boxes or shifting stuff around to click on a file, click "encrypt", type in "correct horse battery staple" and be on your way.
This. It is trivial to ensure data on a USB flash drive is encrypted:
1: $50 gets you an Iepin hardware encrypting USB drive that has a keypad on it. Ten wrong guesses, and you have a blank USB hard drive. You can get an IronKey drive for a bit more that has actual epoxy potting and physical destruction of circuits if one tries to guess the password too often.
2: BitLocker, FileVault, LUKS, and VeraCrypt are common and easy to use. If you have a keyfile at home and at work, and you use VeraCrypt, an attacker has no way to decrypt the drive, except for a brute force against the entire 256 bit length, or physically breaking into the two locations.
3: If policies lock out encryption on the drive level, WinZIP, WinRAR, Acrobat and Microsoft Word can all AES-128 encrypt any files and encapsulate them in their own format.
4: Any big company has some security/compliance education to teach people the dangers of unencrypted media, be it tape, CDs, or whatnot.
5: One's smartphone can carry data. Both iOS and Android support the ability to copy files to the device, and both have apps that can actually create and open VeraCrypt containers on the device for further security. A rooted Android phone can even mount the VC container so you can copy stuff directly into it.
6: There are commercial utilities like Boxcryptor which make encryption extremely transparent.
7: If policies lock out software solutions, then that is what a hardware encrypted drive is for (see #1.)
Ten years ago, there might have been an excuse. These days, with encryption so easy to access and use, there isn't an excuse for this. I almost wonder if the USB flash drive being lost was a deliberate act of sabotage.
Rooting not just allows you to rip crap like that out, but gives you the ability to firewall, guaranteeing that if it ever came back, it wouldn't be able to phone home. I wish DonkeyGuard or xPrivacy were maintained. Those programs were top notch at not just keeping an app from getting to goodies, but giving it fake info, so it thinks it has contacts, photos, camera, mic... but in reality, it is pulling random numbers and repeatedly uploading a black screen.
The Indian smartphone market being large makes sense because for a lot of people, that is their main computing device. Other countries, the smartphone is a secondary device and usually paired with a laptop or desktop, perhaps with a tablet as well.
What would be interesting is designing a smartphone from the ground up as a primary computing device. This would mean far more sophisticated backups (which are encrypted), ease of recovery over the air, even if one's ROM was corrupted, choices of ROMs, more flexibility in storage (one MicroSD card for internal secondary storage, one that would be external for backups, and at least 64-128 gigs of fast internal storage.) Battery replacement may be useful, but I'd assert that IP68 resistance would be more important, and a way to do an external battery case would be a better idea.
Software-wise, designing a phone around VMs, so one can have private stuff, work stuff, and general items in separate areas, switchable by varying authentication modes (fingerprint, face print, PIN, password, or some combination of the above.) That way, if one's "social network" VM got compromised, it is easily dumped, or restored back to a known safe snapshot. Ideally, it would be nice to have some deduplication capability present on the /system partition (if looking at an Android model), so multiple VMs take up as little space as possible, while still providing encryption and separation.
There is a lot of research a company could do, if making a phone as a primary computing device. Perhaps a wireless standard to allow charging, as well as streaming video, and using a HID, so the phone can be tossed on top of a dock, and it can function as a desktop, similar to the old Motorola Atrix line, but wirelessly.
Microsoft still makes money hand-over-fist in the mobile department with their patents. They really don't need to do much in that field other than sit back and take the licensing fees. Windows Mobile was a decent OS... but developers just won't bother with a third platform.
Ironically, before Android and iOS, WM was the most popular smartphone OS out there.
How about a college system like a lot of European countries? When I was in school, my German friend had his tuition paid for by the Fatherland. The Russian student? By the Motherland. The Chinese engineer? By his country. The guy from Chile? Paid for by his government. Compare that to Americans which have to mortgage their future and have to earn significantly more to maintain the same lifestyle, and it is no wonder why there are economic issues in the US.
The problem with STEM here in the US is that the government and businesses make a mistake that even the most brain-dead farmer would never do. If you expect to have a harvest, you plant a crop. You have to fund R&D, fund colleges and universities, and give Joe Sixpack Jr. a reason to go into engineering or science, and not law or finance. Because there isn't any interest in plowing a field, there are very few returns, and it is no wonder why other countries (like Germany or China) who offer university education for a reasonable cost are reaping rewards, while here in the US, many people consider having roads and a power grid "socialism"... and then wonder why prosperity has left this country.
Tenure has/had a place, mainly to ensure that a professor that states an unpopular viewpoint won't get run out as a knee-jerk reaction. Is it still relevant? With the insane political divisions in the US, where stating something about a hot button issue can have grave consequences, tenure is definitely a must.
Maybe it needs redone, but it seems to work, and it is better than just having universities just be an echo chamber for a small set of political beliefs.
For laptops like this where battery life is a major factor, I wonder how the AMD side of the house compares to Intel's offerings. Having the GPUs on the same chip might help things.
There is a way to have enough data for a transaction, but no more. A certificate based system, where one's ID card just validates the cardholder is whom they claim to be, and is a repository for certificates. For example, a certificate showing the person is over age 21. That way, they can go to a bar in the US, and the cert provides what the bar needs to know to comply with the law. The bar doesn't need names, ages, or anything else. Just that the bearer is over 21.
This could be extended to a lot of other things, and to reduce fraud, short-lived certs should be used. For example, a cert that lasts 1-2 days that is done by a police department certifying someone has no entries on their RAP sheet, and has no pending charges. This way, stuff can be done, but relevant info can be sequestered in small, scattered databases, so a breach would be of limited damage.
With the system we have, those watchdogs will fall to regulatory capture, and at best, be a rubber-stamping department.
The only thing that really can break this trend is Europe's GDPR. Time will tell if it actually will get companies to do something about security, or if it winds up being a joke, like SOX (where it was used to jail a guy who ran over his fishing bag limit at its best.) I'm sure BRICS will have similar laws on the books soon, because they want to stick it to US companies, so even that might be a help.
My cynical self feels like security issues will be all too common. It is too easy for an exec to hear that their firm was breached, they sell all their stock and short it, then make out like bandits when the issue hits the front pages.
Diesel can be synthesized, or if one is willing to use a process that takes a lot of energy (which is doable near a hydroelectric plant or somewhere where geography permits), one could take plastic trash, then use thermal depolymerization in order to get a usable diesel oil. It is energy intensive, but it removes plastic from the environment.
Of course, there is biodiesel and all the waste oil that comes from restaurants, as well as motor oil. Run that (B100) in a diesel engine, and the engine will run extremely clean.
Hybrids are nice. I have not gotten why we don't just have some type of engine that is happy at a low RPM that keeps a battery bank charged up, and the vehicle runs on electric motors. This way, for day to day use, the battery bank keeps things going, but there is always the IC engine with fuel sitting ready to charge the battery when it gets low. The closest we have to this would be a plug in Prius or a Chevy Volt.
Designing an engine to run at a certain RPM is a lot easier than figuring out power bands and a transmission. You could even use a turbine engine, which is content to burn a wide variety of fuels, because it is always running at a specific RPM to keep the battery bank charged.
A mantrap is similar, but this is more of a vestibule for someone to put stuff in a secure spot without having access to the rest of the place. Perhaps it could be done by having a little hallway with a door at each end. It wouldn't need to have one door be locked for the other to be opened, but is intended to be a space to ensure stuff is secure from the outside, while keeping the rest of the home secure.
One could always get away with a dumb waiter type of mechanism with a multiple story house as well.