SQL server is a database server, and some applications require it... but at least there are others, and one doesn't have to run their business on it. There are alternatives, from MySQL/MariaDB to Oracle, and the nice thing about Oracle is that there are no license keys to manage, so if there is a disaster, getting your RAC cluster back operable isn't dependent on licensing/activation.
This isn't to say SQL server is bad, but if one wants to move from Windows, there are RDBMS products which are just as good available. If you like NoSQL, but still want ACID... there is always MarkLogic.
This. I'd love the ability to provision a Windows box, toss a SSH key on it and have it ready to be managed via Ansible.
On the development side, being able to Vagrant up a Windows box as easily as I do other boxes would be nice, and make life a -lot- easier when it comes to testing. If I need to create a Windows box to make sure a certain set of Registry settings works, it would be nice to create a base box, boot it, have Vagrant provision it, and have it ready to go. Then, when I want to prove my stuff works to another developer, I point them to the repository with my Vagrantfile and provisioning scripts.
Vagrant is a wonderful tool for testing in the UNIX environment. It (pretty much) guarantees that I will have the exact same environment for testing as the developer, and if their code works in a Vagrant box, it will work in mine. I'd love to have the same ease of use on the Windows side. The closest I can come to this is a WIM image and a directory full of MSI files.
I'm sort of reminded of the early 1990s, pre-Linux, where if one wanted an OS to run on their computer, be it a UNIX flavor, DOS, or OS/2, it cost, and wasn't cheap. It makes me wonder if there would be a niche for a company that produced farm equipment to charge a tad more, as they are not using the cheapest stuff from China, but circuits would be diagrammed, parts would be available, and the equipment would be designed from the ground up for serviceability. Unlike phones and tablets where shaving off a few fractions of a millimeter is critical, a 1950s-era tractor does the job just as well as a modern one.
Of course, there is reliability. A closed source, locked-down ECU might allow something to run for a longer time between servicings, at the cost of more expensive upkeep (since parts only come from the maker.) Would customers mind dealing with a more frequent maintenance cycle, in return for the fact that parts would be cheaper and easy to get ahold of 10-20 years from now, or is the mindset of "use it until it breaks, pitch it, replace it, repeat" too firmly ingrained in the mind of consumers?
It may take some time before this happens. I'm just waiting for "consolization" of cars, where vehicles come with the same engines across the board, but you have to pay license fees to enable the turbos, unlock all horsepower, use the BlueTooth functionality on the audio head... and none of those licenses will transfer with the vehicle, which guarentees that car makers make a significant, tidy sum when a vehicle is sold. Similar with farm equipment. Want to use the PTO? That is a licensed feature and even though the transmission supports it, the TCM won't enable it unless the manufacturer gets $2000 for a license key. Want to use a combine attachment? Another $2000, and it is only good for this harvest season, but you can pay $5000 to have it enabled for five seasons.
How hot will the water get before the frog jumps out?
Android also had this problem. A few years ago, there was AirPush, which eventually forced Google into putting in a mechanism to disable apps from making notifications due to the spam.
I just wonder how long it will be until the advertising bubble bursts, especially if the economy tanks and people are not spending money on gewgaws. Even with apps gathering "god mode" data, there is a limit on how much stuff that can be slurped down and sold. Especially with both malvertising becoming a constant issue , and the general pushback against ads that demand interaction for 30 seconds, demand someone take a survey, demand access to FB and E-mail, and if on a phone, demand an app be downloaded, or any/all of the above. Since stock prices are based on "growth" not actual earnings, when a market hits saturation, it hits the wall, hard. Just like companies did in the first dot.com era.
What will companies like Whatsapp do after that? Join the list of dead companies on the successor of fuckedcompany.com? Charge for their product to end users? Even Google is pushing their way to get territory that isn't ad related so they survive if that bubble pops. People will say that the Internet will grind to a halt without ads that require full screen access... but the Internet existed for decades without that. The economy may move from ads to clearinghouses, but the Internet won't disappear because AdChoices isn't getting their telemetry data from an ever expanding pool each quarter.
The EFF is one party which gets my respect. Signal looks interesting. It reminds me of the old standby app on Android, called TextSecure, which not just was a decent app for texting, but stashed the messages somewhere encrypted, as a secondary layer of protection. I wish the iOS version had a PIN or the ability to use the fingerprint scanner, just for additional security for messages on the app.
Personally, my ideal app would be one that piggybacks of of existing protocols, but uses OpenPGP for its endpoint encryption. This way, it provides a standard for adding keys, it would be compatible with a ton of existing code, OpenPGP is a known, secure quantity with decades of debugging behind it, and it would make having to worry about transport encryption less of an issue. Messages could be stored with a NNTP-like protocol (where one server would store, forward, then expire when the server's disk space hit a high water mark), or a more direct protocol could be used. Since there are so many transport protocols to choose from, separating the endpoint encryption from the protocol would allow for a lot of flexibility.
I would actually purport that isn't the case. Facebook, for example. It is used instead of E-mail, NNTP, news websites, discussion forums, chat sites, meeting scheduling, appointments, and many other items. People are too interconnected with it to leave it. In fact, not having a FB ID can be a negative in general, just because FB is used for so much.
People won't be leaving FB like they bailed MySpace. There is just too much tying them to that social network. Heck, there are a number of websites use FB for authentication now... no FB account, no access.
The sad thing is that a password manager isn't a tough thing. However, it requires some thought to do it right.
For example, stashing a syncable database on a cloud provider. Most PW managers either use the same password one uses for the local storage.
However, the database on the cloud provider is where security needs to be tight, and, if possible, not brute-forcable. Ideally, the database would be protected by a randomly generated key, which is then encrypted by each device's private key. If the user wants to add a new device, the new device's key is slapped on a keyserver, one of the other devices shows the user that device's fingerprint and asks the user to compare and approve that the key is the same, then would allow the new device to have its decrypting key entry. If all devices are lost, a recovery mechanism can be as simple as having a password on the chain that unlocks the key, or a shared secret. All solved problems -- this functionality is all native to the OpenPGP format (where one encrypted file can be decoded by any key on the list, or a passphrase.)
For Android and iOS, both have secure modes to store sensitive data. These should be used in combination with the app's encryption, so there is both the device's hardware protection, and the app's protection.
For desktop usage, the app's encryption is likely enough, since desktops are less likely to be stolen.
Backups? Again, there is an easy, secure way to do this. Use a similar encryption mechanism to what Titanium Backup (a must have for Android) uses. It has a public key, and encrypts backups with that. When a restore is needed, each backup file has a private key, which is encrypted with the user's passphrase. This results in being able to do backups without having to hold a key or password indefinitely in memory between sessions, but allows the user to restore/decrypt without worrying about having the proper key... just the right passphrase.
Of course, there is the password generator. Yes,/dev/random or Windows's equivalent is "good enough", but having a password generator which can take user input (keystrokes, using a high speed timer, as well as mouse movements) provides additional randomness, which would be useful if a bug happened, and/dev/random just outputted zeroes or some other glitch happened. The ideal would be a combination of Apple's and Keepass's, where one can use memorable words with a number or two, or generate custom passwords [1].
I just wish someone would do it "right". KeePass has everything nailed, except good syncing with a cloud provider, and if each instance of the password manager would use a PGP/gpg key, then store the database as an encrypted file, this would provide excellent resistance to brute-force attacks, should the cloud provider get compromised, as there would be no passwords to guess.
[1]: Sometimes I was asked to send a user a password over one channel, and data over another. I liked sending passwords in a standard format (like 2-3 Windows CD keys for more sensitive stuff, or phone numbers for less sensitive items) so the receiver knew they were not totally lost when typing in a long password. Thus, having templates come in handy. Same with generating a large amount of starting passwords for an AD domain, where I wanted the passwords to fit the criteria, but be of a certain format so the user knows they are typing in the right thing, and there wouldn't be any "0/O", "1/l" mixups.
Ideally, the best fuel would be something with a high energy per volume, such as the Audi-made synthetic diesel (e-diesel) from CO2 in the air, or perhaps ethanol. Something that doesn't need anything more significant than a liquid storage tank, as opposed to what is needed for CNG, or even LP gas. Cars get into wrecks, and who knows what might puncture the gas tank, so having a complex system is nice, but if it takes out a city block if the vehicle using it gets rear-ended, it isn't workable.
Then comes the engine. Moving to an IC engine design means that you get 1/3 of the energy coming out as torque, and the rest as heat or exhaust. Engine design isn't an easy task either, because it will see environments that the engineers have never even anticipated. So, it would be generations of engines before a H2 motor became as reliable as what moves an average gasser.
What would be the ideal, is what the parent poster mentions -- high capacity batteries. Get a type of battery that is portable, fairly safe (think LiFePO4 batteries), can work in the temperature extremes that vehicles are put to, and has a high discharge/recharge cycle rate. If this gets within a tenth of what gasoline has for energy by volume, this changes everything. The IC engine can be tossed, and electric motors used.
For the average Joe Sixpack, what does the hydrogen economy bring? It means more energy used (hydrolysis is very energy wasteful), so utility prices go up and more base load power plants needed. Hydrogen is very explosive, so expensive means have to be used so mitigate damage to the tank and lines. This means vehicles cost more. There has to be an ecosystem put in place to fill up on H2 or replace fuel storage cells... and that is expensive, which gets passed to the consumer.
Even more insidious, is the fact that the H2 economy will be tied to a limited group that provides the H2 creation and distribution. With an electric vehicle, it doesn't care where the electricity comes from. It can come from a solar array, a water turbine, a wind turbine... there are many ways from the ground up to make usable electricity that it provides robustness. We really do not need another monopoly/cartel.
Hybrids need to go past the novelty stage, to becoming something every vehicle has, just like power door locks. For example, a hybrid pickup, if coupled with an inverter that can handle heavy loads, would make the need for a PTO generator moot. Since a vehicle engine has a lot more in the way of pollution controls than most generators, it is better, environmentally, to have that engine do the work. It also means one less motor to worry about. GM has tried attempts at a hybrid pickup, but the ideal would be to have the technology in the 2500 and 3500 vehicles, which the beefed up electrical storage would be a lot more useful. Even more useful would be a hybrid diesel. If FCGEN (an European group making a fuel cell that runs on diesel) gets something useful out there, then the vehicle could be made entirely electric, but a diesel fuel cell (generating 3-10kw) able to keep the batteries powered, eliminating range anxiety.
tl;dr, the absolute best thing are electric vehicles. Second to that, diesels and synthesizing diesel fuel which doesn't go boom due to vapors. In the interim, it would be nice to see hybrid diesels. Hydrogen is a nice dog and pony show which really is a distraction, and would cause far more expense to every party (except the H2 makers.) Not worth it.
I notice Intel enterprise SSDs have capacitors on them, and I'm guessing it is there so the drive has enough power to complete any in-flight writes (or at least find a stable, consistent point on the block/page level to stop and power down.) This makes me guess that for the hard-power off issue, this is a solved problem.
SSD do fail, but they fail in different ways than HDDs. It is wise to have backups regardless of what one's primary media is, but SSDs are nice in the fact that they can take environmental issues than a HDD. It is still not good to drop one, but if one drops an external SSD while it is plugged in, it is almost certain to continue working. Drop a HDD, who knows.
I would say that the benefits for using a SSD as primary storage well overcome their drawbacks. Just the fact that don't have that bottleneck of waiting until the set of heads aligns up with the data needed (so a virtualized OS and a host OS don't have to fight for the head array), data doesn't have to be shoved on the inner or outer tracks for performance regions, and other HDD specific issues goes a long way with performance, especially random I/O. Caching and smart HDD controllers have helped some, but SSD just gets rid of the problem entirely.
Of course, this doesn't make HDD entirely pointless. It is cheap, and holds a good amount of storage. Which makes it useful for backups or secondary storage, a place where tape once reigned. I wouldn't be surprised to see HDD cartridges used in a silo, like Sphere 3D's RDX format (formerly Imation).
The future of hard drives are also long term archival storage. Even though tape has an archival rating, while HDDs don't, it wouldn't be surprising for HDD makers to go this route, especially coupled with HAMR, SMR, patterned media, and other technologies which add storage, but trade off with I/O performance and the need for TRIM-like commands or garbage collection by the HDD controller. A good start in this direction are the NAS drives WD has. I wouldn't be surprised to see more drives go this route, but likely in the 2.5" form factor so RAID enclosures can be smaller.
Of course, the downside of that is that when the car has major trouble, such as an electrical malfunction, or a dead battery bank, there isn't any fixing of the vehicle... the vehicle just will have to be replaced. Similar with car wrecks... more than just a "love tap" can cause the insurance company to just total the vehicle and call it done.
Electric cars have their place, and for a lot of people are useful... but until there is a better infrastructure for long distance trips, range anxiety will be an issue, and people will still continue to buy vehicles with IC motors, just because they will start and run even if there is a power blackout. I wouldn't say dealers and service stations will be gone anytime soon. In fact, with all the proprietary gizmos, they will make even more money. A clogged particulate filter? $3200 right there. An air filter with a built in MAF sensor? Even a battery replacement on some vehicle models require one to haul the vehicle to a service station to have the battery "registered" with the ECM, and the new battery will only last a year before failing.
I fear that we are in for more lock-in. With the fact that ECMs are getting more controlled, with more exotic defenses against chipping, combined with the talk about having cars require software upgrades, not to mention having the -software- not transfer when a vehicle is sold (so one has to pay for a relicense to have the ability to start the vehicle, much less use stuff like climate control or the audio system), it gets worrisome. Think consoles and DLC as an example.
I don't know what is worse. Vehicles that stop getting upgrades after a year, or vehicles that do get upgrades... but they come with more ways to reach in your pocketbook. Want 4x4? The front diff is there, but you have to pay $3000 for a license key so the TCM will use it, and that key doesn't transfer to the new owner, so if the automaker finds the vehicle under a new title, they can remove all the keys via a remote "kill switch" system (oh, you accepted the EULA when you drove the vehicle, including the arbitration clause, good luck suing.) Of course, you can "rent" 4x4 capability with another license for $100 per 24 hour period... This is a revenue stream automakers will enjoy, but not much good for the consumer.
What my fear is, is that hobbyist CNC mills wind up like CarveWright machines. You wind up paying a few grand for the machine... but, you have to pay dearly for additional software, DRM-protected templates, special memory cards, special readers, and other proprietary crap.
This may be a working model for IBM mainframes, but as consumers, we need to fight tooth and nail so this doesn't happen in other industries.
Back in the early 1990s where CNC mills were not cheap, but inexpensive enough for smaller companies to spend money for, there was a time where there were many, many bicycle parts made out of CNC machined aluminum, usually anodized very funky colors as well. This lasted for a few years, but then Shimano and the other brands went to drop-forging, which allowed for more long-wearing components, especially on the drivetrain. This, plus the move to titanium and carbon fiber caused CNC machining to fade away in that industry.
However, with the resurgence of interest, I wouldn't be surprised if people would be interested in custom-building their own mountain bikes with this technology. Their bikes may a couple grams heavier than with CF technology, but the biggest advantage is that parts would be easy to build and replace in a scenario where one can't get a new XTR set shipped to them.
There is also the independent, "prepper" mindset starting to take hold, where having something that one can maintain oneself is a lot better than the latest and greatest gadget which can only be repaired by a specific shop... or the device can't be repaired at all, and has to be tossed. Having a bike where most of the "consumable" parts like the chainrings are easily made on a CNC mill may be what people want, especially if mills become more common in hobbyist installations.
I hope CNC mills become more common. 3D printing is nice, but there are only so many plastic pieces one can make, and the plastic used is quite toxic to the environment. If one can do metal work, it gives a lot more options... if only to allow one to make molds and use those for an injection molding machine for plastic parts that are decently strong.
It seems to come and go in waves, and just seems like a solution in search of a problem. Even back in the mid 1990s, there were 3D headsets which worked with Duke Nukem 3D, but other than at Egghead, I never saw one purchased and "in the wild."
Eventually we will get past the VR stage of the goggles and the PowerGlove, but until we get to where equipment worn is light to none and we have something similar to a holodeck, VR will be in the fringes with cool stuff happening here and there, but tending to be too expensive and specialized for mainstream use. However, things can change. The same thing was said about MP3 players (expensive/specialized/only for a few people) until Apple got into that market, so I could be proven completely and utterly wrong about this in the coming year.
I wonder about lubrication. A lot of the mills I see need to hose down the piece with water or some other lubricant to keep the metal cool. Is disposing of the water with all the metal bits in it something that is a major concern?
With a machine like that, I am surprised to see USB 2.0 on Type C connectors. Most everyone else has moved to USB 3.0 or USB 3.1. Since this machine can do a desktop role, it should have the ability for someone to plug in an external HDD or SDD because the storage on those devices tends to be pretty small, or just as a way to back the thing up via a program or wbadmin. Wireless access is one thing, but until someone designs a wireless storage protocol that is up to par for the task.
What HP needs to do is make a business grade model of this. It should have 16-64 gigs of RAM, a minimum 512GB SSD, and at least two USB 3.1, type C connectors which can also work for Thunderbolt. It also should have a docking station which uses the Thunderbolt bus and have some decent GPU power in that (or at the minimum, a PCIe slot), a number of USB ports (with multiple I/O buses), and a 10gigE Ethernet port. This way, the tablet can be used for business functions and can double as a desktop with ease. Oh, and have Windows 10 Enterprise pre-activated, just so a company wouldn't have to blow a KMS license on the machine. Done right, it would sell, just because it is able to do multiple roles, from being able to display PowerPoint slides with an adapter getting video from the USB-C connector, to doing day to day work in a dock at the office.
[1]: Wi-Fi Direct isn't usable in this case, unless one doesn't mind switching SSIDs often to go from the Internet connection to the drive's SSID, or putting the HDD on the same Wi-Fi segment and having it available for anyone there to poke at. The ideal would be a protocol that piggybacks off of BlueTooth's encrypted pairing, and can work at a decent speed at a short range.
If HDD capacities could go up, but HDD makers pivot from just shoveling more bits into less space to redundancy and reliability, HDDs will take the niche that tapes have now. Especially if a HDD maker could guarantee an archive life of a tape, perhaps having a dual-head mechanism (I remember seeing some older drives which actually had two sets of heads, each independent of the other and in an active/active configuration.) Doing this would keep HDDs around as backup media or media to stick in the NAS (for a low-end NAS for stashing archived files or backups, a 5200 RPM WD Red or two are just fine.)
I can see HDD makers also "smartening" up their controllers, by going with more SSHD style drives... mainly by adding SSD as a cache or write landing zone... where the I/O first lands on the SSD, then gets shuffled off to the spinny disks. Of course, there is the slowdown once the SSD winds up full... but done right, it can be a way to help with disk I/O for all but the worst sustained random writes. While doing this, HDD makers might be able to add more features, such as autotiering (think the Apple Fusion drives, except done on the controller) and block level deduplication (although this changes a drive from just a being a raw device into presenting the disk as a LUN, as opposed to just a raw chunk of storage.)
I do agree that SSDs, even now, are pretty much a "why not?" choice for primary storage. Even a USB 3 attached SSD with its limited throughput has an advantage over a HDD, just because multiple tasks don't have to fight for control of the drive heads.
In a twisted way, this makes sense. One is oftentimes better off by mining a currency like AltCoin, DASH, or another cryptocurrency, then trading that for BTC, as opposed to trying to just mine BitCoins, just due to the requirement of having a huge investment in ASICs (not to mention a cheap energy source) to mine BTC oneself.
It wasn't that long ago when Windows Mobile was the dominant OS in the smartphone arena. To boot, MS and hardware makers did a pretty good job. The HTC Wizard comes to mind, with a week's battery life. It might be laughable by our standards now, since it only puttered at EDGE speeds, and didn't have the latest rev of BlueTooth, not to mention the use of MiniSD cards... but for the time, it was a very nice phone.
Microsoft dropped the ball when the iPhone came out. The biggest problem is that Windows Mobile was designed around a stylus, while iOS and Android were designed for finger gestures. The shift in the UI expectations left MS in the dust, just because all their device apps were not "finger friendly". MS was encumbered with an existing solution, and having to either figure out how to retrofit the latest UI style, or to just toss everything out and start anew.
MS did a good job with starting anew, and they have a competitive device.
As for a spiritual successor for the Surface... this gets me wondering... are they going to try for an Intel x86 type of computer running W10 in a smartphone form factor? If they could pull this off running a real x86 version of Windows 10 and all Windows applications, it would be a game-changer.
Of course, other companies tried this, such as Motorola with the Atrix and Atrix 2... but if a phone could be tossed in a stand or have a USB-C cable connected, and it take the role of a desktop machine with an x86 version of W10, Microsoft would be breaking new ground. It would mean that one wouldn't need to have anything other than a keyboard, monitor, and USB-C hub in order to have a functioning desktop.
There is one missing piece of the puzzle, and that is getting a phone to handle heavy GPU tasks. This is easy. Since MS has their own graphics standard, it would be trivial for them to make something like a LAN version of OnLive, and have smartphones and tablets send the DirectX commands over the network to a render box, and the render box send back streaming video. Since 4k res of streaming video is about 10Mbps, a wireless LAN can easily handle this. If one uses a newer graphics protocol like ZPEG which gets even better compression for the same quality, it would be even less.
If Microsoft pulls this off, where the only GPU needed for a device would be for the basic UI... they would have a major breakthrough market that would be in high demand.
It wouldn't be worth the time, since I found a far better place to be at anyway, job-wise.
As for FB/whatever, I decided to make an account, and keep them around. I now use Twitter for announcing GitHub releases I make. That way, the account is of actual use.
As for IoT, whining about is not going to do much. However, there are a few ways to actually make IoT truly secure... not secure as in the sense of "locking it down" secure... but secure as in resisting unauthorized intrusions, modifications, deletions... the classic sense.
Three ways to make it work:
1: Get some people who know what they are doing, such as Bruce. Make a UL type independent organization whose job it is to check security of products in both white-box testing and black-box testing. Security such as resisting attacks via the network, ease of resetting the device, should the owner lose the password, how firmware updates are handled [1], how the device reacts to intrusion attempts, internal security like chrooting, signed executables, SELinux, ASLR, and other methods. Have the independent organizations's approval a must for the device to be sold. Of course, this invites regulatory capture, and genuine security can easily be perverted into "keeping the user out" security... but anything in IoT is better than nothing.
2: Move to a different topology for IoT devices than having the devices connecting directly to the Internet via a 3G/4G connection or using a Wi-Fi access point. Instead, the devices should communicate on the LAN basis to a hardened appliance... and that appliance does the sending and receiving for the devices. This way, the "smart toaster" communicating to the hub via BlueTooth will be extremely difficult to hack because it sends the user's toaster preferences up through the BT hub, which then relays it through the Internet. Going with a hub/spoke, with redundant hubs possible, would significantly decrease the attack surface of IoT devices.
3: Use the principle of least privilege. If an Internet connection isn't needed (say for a device to work as a remote), use Bluetooth. If the device has to have an Internet connection for updates, have documentation that describes the sites it connects to [2], and what ports that it should be allowed. Anything else should be blocked. The device should even enforce this in its OS firewall (netfilter for Linux, for example) to protect against unauthorized processes trying to get out. If "smart" functionality isn't needed, don't bother with it.
Take the "smart" refrigerator. If appliance companies wanted to make something expensive, why not a fridge with two cooling mechanisms... the standard compressor that plugs into the wall, and an absorption mechanism which can be powered by electricity, natural gas, or propane. This way, if there is a power blackout, the fridge still retains cooling capacity, and with a thermalelectric generator (think a Peltier running in reverse), would have enough power to keep the core circuit board running. I'm sure there would be more demand for a fridge that keeps the food cold if power goes out, than a fridge which can display ads 24/7 on the screen.
[1]: I believe in the old school idea of a physical button or switch that is used before flashing firmware... but this isn't something that can be done if the device is not physically accessible, so maybe a fallback would be some other mechanism. That way if the RSA key is compromised, the vendor can use a different, but still secure, way to get the updates to devices.
[2]: Ideally, it should just fetch a signed manifest via SSL, and go from there. If the embedded OS is Linux, it could even use an existing package manager like Yum or apt so that wheel doesn't have to be reinvented.
Here is where things get nefarious. IoT are like social networks. In the past, you could just tell people where to stick it when they talked about their livejournal, MySpace, or Orkut stuff. However, if one doesn't have a LinkedIn account, FB account, and a Twitter account, you will be turned down for jobs.
I know this firsthand. Had a job interviewer tell me that I was too old for IT work and show me the door because he wanted to read/follow my Twitter account, and I told him that I didn't have one.
IoT has the potential for being just like that. For example, the Bluetooth deadbolt. It might be that apartment managers and other landlords install IoT security devices because it makes their job easier to lock out tenants being evicted, know who is going into a tenant's place, or to let maintenance in on a schedule regardless if the tenant wants it or not. For more flexible for the property owner, and the tenant would have no choice in the matter.
Insurance can also demand IoT devices, say CCTV monitoring and file storage, or IoT deadbolts and other devices so they can be assured that a property is secured when the owners are away. If this isn't done, they won't renew the policy.
Then, there is the phone home aspect. Pull the internet connection on a modern console, it halts. I wouldn't be surprised if a future HDCP spec that requires all devices to authenticate with a central server for a healthcheck every so often, would require that all TVs and such be always on and in communication. As per the EULA of the TV, video and audio would also be sent back for "IP enforcement purposes". If someone disagrees with that... well, good luck with the no-sue arbitration agreement they agreed to...
Next comes devices. Take the refrigerator for instance. Good luck trying to find a completely mechanical one with a thermostat and compressor that runs for decades. Most have various computer controls. It wouldn't be surprising that IoT functionality is important, and no network connection means the device does not function, especially if the fridge maker starts demanding license keys to activate the ice maker, crisper section, and such.
The key is to not just avoid buying IoT shit, but make it -damn well known- that you will never buy that because you don't want another route an intruder can trespass into your home. Because IoT security is so weak, and there is zero incentive for companies to actually do something about it, it needs to die on the vine.
TFA didn't have many details, but I wonder how the files were watermarked. Some metadata perhaps, that added atime/ctime/ and the computer's install ID?
Perhaps a modification to the filesystem to allow for alternate data streams (a la NTFS) which would have a linked list of machines the file has sat on, which would be hidden from the user, and would move with the file?
That is definitely possible. It would require a few steps though.
First, I can see someone showing an ID somewhere, and picking a blank cryptographic token. Something like a CAC (if military), PIV card (if civilian), or just a generic USB token like an Aladdin/SafeNet eToken if one uses that for secure storage for PGP keys. It is picked at random, just like how paper ballots are laid out, so the government knows the key is valid... but by default doesn't know whose key it belongs to.
Second, said key is used as a client certificate. Only PW used is the PIN to unlock the token for the browsing session.
If a user wants to change identities, they just bring the old card back, someone checks that it was a valid card, they carve it into little bitty pieces, and allow the citizen to grab a replacement card with a fresh, new identity.
The result is solid pseudo-anonymity, but because every person only can have one ID at a time, it keeps the lid down on sock puppets, shills, and general trolls.
Of course, this assumes the CAC/PIV maker can be trusted, as well as the distribution channel... but that is what government controls are for. Plus, since this isn't as much tied to the future of a country directly, it would be less of a target for the bad guys than pwning voting machines.
Of course, the problem how to prove a user only has one card, needs to be solved. This way, if the user loses or has their card destroyed, they can get another one, have the cert revoked on the old key. However, if the government keeps track of user to public key mappings, then the whole anonymity concept goes into the shitter.
I can see someone designing software that watches someone screen and ensures they vote for the "proper" candidate, either what the employer wants, or what a criminal organization wants.
I'm with you. Voting by paper and snail mail as a last resort, but voting booths and locations give people to be able to vote and have their vote be truly anonymous, so they can vote for the candidate they so choose and be resistant to coercion.
Slashdot Mirror
SQL server is a database server, and some applications require it... but at least there are others, and one doesn't have to run their business on it. There are alternatives, from MySQL/MariaDB to Oracle, and the nice thing about Oracle is that there are no license keys to manage, so if there is a disaster, getting your RAC cluster back operable isn't dependent on licensing/activation.
This isn't to say SQL server is bad, but if one wants to move from Windows, there are RDBMS products which are just as good available. If you like NoSQL, but still want ACID... there is always MarkLogic.
This. I'd love the ability to provision a Windows box, toss a SSH key on it and have it ready to be managed via Ansible.
On the development side, being able to Vagrant up a Windows box as easily as I do other boxes would be nice, and make life a -lot- easier when it comes to testing. If I need to create a Windows box to make sure a certain set of Registry settings works, it would be nice to create a base box, boot it, have Vagrant provision it, and have it ready to go. Then, when I want to prove my stuff works to another developer, I point them to the repository with my Vagrantfile and provisioning scripts.
Vagrant is a wonderful tool for testing in the UNIX environment. It (pretty much) guarantees that I will have the exact same environment for testing as the developer, and if their code works in a Vagrant box, it will work in mine. I'd love to have the same ease of use on the Windows side. The closest I can come to this is a WIM image and a directory full of MSI files.
I'm sort of reminded of the early 1990s, pre-Linux, where if one wanted an OS to run on their computer, be it a UNIX flavor, DOS, or OS/2, it cost, and wasn't cheap. It makes me wonder if there would be a niche for a company that produced farm equipment to charge a tad more, as they are not using the cheapest stuff from China, but circuits would be diagrammed, parts would be available, and the equipment would be designed from the ground up for serviceability. Unlike phones and tablets where shaving off a few fractions of a millimeter is critical, a 1950s-era tractor does the job just as well as a modern one.
Of course, there is reliability. A closed source, locked-down ECU might allow something to run for a longer time between servicings, at the cost of more expensive upkeep (since parts only come from the maker.) Would customers mind dealing with a more frequent maintenance cycle, in return for the fact that parts would be cheaper and easy to get ahold of 10-20 years from now, or is the mindset of "use it until it breaks, pitch it, replace it, repeat" too firmly ingrained in the mind of consumers?
It may take some time before this happens. I'm just waiting for "consolization" of cars, where vehicles come with the same engines across the board, but you have to pay license fees to enable the turbos, unlock all horsepower, use the BlueTooth functionality on the audio head... and none of those licenses will transfer with the vehicle, which guarentees that car makers make a significant, tidy sum when a vehicle is sold. Similar with farm equipment. Want to use the PTO? That is a licensed feature and even though the transmission supports it, the TCM won't enable it unless the manufacturer gets $2000 for a license key. Want to use a combine attachment? Another $2000, and it is only good for this harvest season, but you can pay $5000 to have it enabled for five seasons.
How hot will the water get before the frog jumps out?
Android also had this problem. A few years ago, there was AirPush, which eventually forced Google into putting in a mechanism to disable apps from making notifications due to the spam.
I just wonder how long it will be until the advertising bubble bursts, especially if the economy tanks and people are not spending money on gewgaws. Even with apps gathering "god mode" data, there is a limit on how much stuff that can be slurped down and sold. Especially with both malvertising becoming a constant issue , and the general pushback against ads that demand interaction for 30 seconds, demand someone take a survey, demand access to FB and E-mail, and if on a phone, demand an app be downloaded, or any/all of the above. Since stock prices are based on "growth" not actual earnings, when a market hits saturation, it hits the wall, hard. Just like companies did in the first dot.com era.
What will companies like Whatsapp do after that? Join the list of dead companies on the successor of fuckedcompany.com? Charge for their product to end users? Even Google is pushing their way to get territory that isn't ad related so they survive if that bubble pops. People will say that the Internet will grind to a halt without ads that require full screen access... but the Internet existed for decades without that. The economy may move from ads to clearinghouses, but the Internet won't disappear because AdChoices isn't getting their telemetry data from an ever expanding pool each quarter.
The EFF is one party which gets my respect. Signal looks interesting. It reminds me of the old standby app on Android, called TextSecure, which not just was a decent app for texting, but stashed the messages somewhere encrypted, as a secondary layer of protection. I wish the iOS version had a PIN or the ability to use the fingerprint scanner, just for additional security for messages on the app.
Personally, my ideal app would be one that piggybacks of of existing protocols, but uses OpenPGP for its endpoint encryption. This way, it provides a standard for adding keys, it would be compatible with a ton of existing code, OpenPGP is a known, secure quantity with decades of debugging behind it, and it would make having to worry about transport encryption less of an issue. Messages could be stored with a NNTP-like protocol (where one server would store, forward, then expire when the server's disk space hit a high water mark), or a more direct protocol could be used. Since there are so many transport protocols to choose from, separating the endpoint encryption from the protocol would allow for a lot of flexibility.
I would actually purport that isn't the case. Facebook, for example. It is used instead of E-mail, NNTP, news websites, discussion forums, chat sites, meeting scheduling, appointments, and many other items. People are too interconnected with it to leave it. In fact, not having a FB ID can be a negative in general, just because FB is used for so much.
People won't be leaving FB like they bailed MySpace. There is just too much tying them to that social network. Heck, there are a number of websites use FB for authentication now... no FB account, no access.
I'd love to be proven wrong on this, though.
The sad thing is that a password manager isn't a tough thing. However, it requires some thought to do it right.
For example, stashing a syncable database on a cloud provider. Most PW managers either use the same password one uses for the local storage.
However, the database on the cloud provider is where security needs to be tight, and, if possible, not brute-forcable. Ideally, the database would be protected by a randomly generated key, which is then encrypted by each device's private key. If the user wants to add a new device, the new device's key is slapped on a keyserver, one of the other devices shows the user that device's fingerprint and asks the user to compare and approve that the key is the same, then would allow the new device to have its decrypting key entry. If all devices are lost, a recovery mechanism can be as simple as having a password on the chain that unlocks the key, or a shared secret. All solved problems -- this functionality is all native to the OpenPGP format (where one encrypted file can be decoded by any key on the list, or a passphrase.)
For Android and iOS, both have secure modes to store sensitive data. These should be used in combination with the app's encryption, so there is both the device's hardware protection, and the app's protection.
For desktop usage, the app's encryption is likely enough, since desktops are less likely to be stolen.
Backups? Again, there is an easy, secure way to do this. Use a similar encryption mechanism to what Titanium Backup (a must have for Android) uses. It has a public key, and encrypts backups with that. When a restore is needed, each backup file has a private key, which is encrypted with the user's passphrase. This results in being able to do backups without having to hold a key or password indefinitely in memory between sessions, but allows the user to restore/decrypt without worrying about having the proper key... just the right passphrase.
Of course, there is the password generator. Yes, /dev/random or Windows's equivalent is "good enough", but having a password generator which can take user input (keystrokes, using a high speed timer, as well as mouse movements) provides additional randomness, which would be useful if a bug happened, and /dev/random just outputted zeroes or some other glitch happened. The ideal would be a combination of Apple's and Keepass's, where one can use memorable words with a number or two, or generate custom passwords [1].
I just wish someone would do it "right". KeePass has everything nailed, except good syncing with a cloud provider, and if each instance of the password manager would use a PGP/gpg key, then store the database as an encrypted file, this would provide excellent resistance to brute-force attacks, should the cloud provider get compromised, as there would be no passwords to guess.
[1]: Sometimes I was asked to send a user a password over one channel, and data over another. I liked sending passwords in a standard format (like 2-3 Windows CD keys for more sensitive stuff, or phone numbers for less sensitive items) so the receiver knew they were not totally lost when typing in a long password. Thus, having templates come in handy. Same with generating a large amount of starting passwords for an AD domain, where I wanted the passwords to fit the criteria, but be of a certain format so the user knows they are typing in the right thing, and there wouldn't be any "0/O", "1/l" mixups.
Ideally, the best fuel would be something with a high energy per volume, such as the Audi-made synthetic diesel (e-diesel) from CO2 in the air, or perhaps ethanol. Something that doesn't need anything more significant than a liquid storage tank, as opposed to what is needed for CNG, or even LP gas. Cars get into wrecks, and who knows what might puncture the gas tank, so having a complex system is nice, but if it takes out a city block if the vehicle using it gets rear-ended, it isn't workable.
Then comes the engine. Moving to an IC engine design means that you get 1/3 of the energy coming out as torque, and the rest as heat or exhaust. Engine design isn't an easy task either, because it will see environments that the engineers have never even anticipated. So, it would be generations of engines before a H2 motor became as reliable as what moves an average gasser.
What would be the ideal, is what the parent poster mentions -- high capacity batteries. Get a type of battery that is portable, fairly safe (think LiFePO4 batteries), can work in the temperature extremes that vehicles are put to, and has a high discharge/recharge cycle rate. If this gets within a tenth of what gasoline has for energy by volume, this changes everything. The IC engine can be tossed, and electric motors used.
For the average Joe Sixpack, what does the hydrogen economy bring? It means more energy used (hydrolysis is very energy wasteful), so utility prices go up and more base load power plants needed. Hydrogen is very explosive, so expensive means have to be used so mitigate damage to the tank and lines. This means vehicles cost more. There has to be an ecosystem put in place to fill up on H2 or replace fuel storage cells... and that is expensive, which gets passed to the consumer.
Even more insidious, is the fact that the H2 economy will be tied to a limited group that provides the H2 creation and distribution. With an electric vehicle, it doesn't care where the electricity comes from. It can come from a solar array, a water turbine, a wind turbine... there are many ways from the ground up to make usable electricity that it provides robustness. We really do not need another monopoly/cartel.
Hybrids need to go past the novelty stage, to becoming something every vehicle has, just like power door locks. For example, a hybrid pickup, if coupled with an inverter that can handle heavy loads, would make the need for a PTO generator moot. Since a vehicle engine has a lot more in the way of pollution controls than most generators, it is better, environmentally, to have that engine do the work. It also means one less motor to worry about. GM has tried attempts at a hybrid pickup, but the ideal would be to have the technology in the 2500 and 3500 vehicles, which the beefed up electrical storage would be a lot more useful. Even more useful would be a hybrid diesel. If FCGEN (an European group making a fuel cell that runs on diesel) gets something useful out there, then the vehicle could be made entirely electric, but a diesel fuel cell (generating 3-10kw) able to keep the batteries powered, eliminating range anxiety.
tl;dr, the absolute best thing are electric vehicles. Second to that, diesels and synthesizing diesel fuel which doesn't go boom due to vapors. In the interim, it would be nice to see hybrid diesels. Hydrogen is a nice dog and pony show which really is a distraction, and would cause far more expense to every party (except the H2 makers.) Not worth it.
I notice Intel enterprise SSDs have capacitors on them, and I'm guessing it is there so the drive has enough power to complete any in-flight writes (or at least find a stable, consistent point on the block/page level to stop and power down.) This makes me guess that for the hard-power off issue, this is a solved problem.
Non-enterprise drives, who knows.
SSD do fail, but they fail in different ways than HDDs. It is wise to have backups regardless of what one's primary media is, but SSDs are nice in the fact that they can take environmental issues than a HDD. It is still not good to drop one, but if one drops an external SSD while it is plugged in, it is almost certain to continue working. Drop a HDD, who knows.
I would say that the benefits for using a SSD as primary storage well overcome their drawbacks. Just the fact that don't have that bottleneck of waiting until the set of heads aligns up with the data needed (so a virtualized OS and a host OS don't have to fight for the head array), data doesn't have to be shoved on the inner or outer tracks for performance regions, and other HDD specific issues goes a long way with performance, especially random I/O. Caching and smart HDD controllers have helped some, but SSD just gets rid of the problem entirely.
Of course, this doesn't make HDD entirely pointless. It is cheap, and holds a good amount of storage. Which makes it useful for backups or secondary storage, a place where tape once reigned. I wouldn't be surprised to see HDD cartridges used in a silo, like Sphere 3D's RDX format (formerly Imation).
The future of hard drives are also long term archival storage. Even though tape has an archival rating, while HDDs don't, it wouldn't be surprising for HDD makers to go this route, especially coupled with HAMR, SMR, patterned media, and other technologies which add storage, but trade off with I/O performance and the need for TRIM-like commands or garbage collection by the HDD controller. A good start in this direction are the NAS drives WD has. I wouldn't be surprised to see more drives go this route, but likely in the 2.5" form factor so RAID enclosures can be smaller.
Of course, the downside of that is that when the car has major trouble, such as an electrical malfunction, or a dead battery bank, there isn't any fixing of the vehicle... the vehicle just will have to be replaced. Similar with car wrecks... more than just a "love tap" can cause the insurance company to just total the vehicle and call it done.
Electric cars have their place, and for a lot of people are useful... but until there is a better infrastructure for long distance trips, range anxiety will be an issue, and people will still continue to buy vehicles with IC motors, just because they will start and run even if there is a power blackout. I wouldn't say dealers and service stations will be gone anytime soon. In fact, with all the proprietary gizmos, they will make even more money. A clogged particulate filter? $3200 right there. An air filter with a built in MAF sensor? Even a battery replacement on some vehicle models require one to haul the vehicle to a service station to have the battery "registered" with the ECM, and the new battery will only last a year before failing.
I fear that we are in for more lock-in. With the fact that ECMs are getting more controlled, with more exotic defenses against chipping, combined with the talk about having cars require software upgrades, not to mention having the -software- not transfer when a vehicle is sold (so one has to pay for a relicense to have the ability to start the vehicle, much less use stuff like climate control or the audio system), it gets worrisome. Think consoles and DLC as an example.
I don't know what is worse. Vehicles that stop getting upgrades after a year, or vehicles that do get upgrades... but they come with more ways to reach in your pocketbook. Want 4x4? The front diff is there, but you have to pay $3000 for a license key so the TCM will use it, and that key doesn't transfer to the new owner, so if the automaker finds the vehicle under a new title, they can remove all the keys via a remote "kill switch" system (oh, you accepted the EULA when you drove the vehicle, including the arbitration clause, good luck suing.) Of course, you can "rent" 4x4 capability with another license for $100 per 24 hour period... This is a revenue stream automakers will enjoy, but not much good for the consumer.
What my fear is, is that hobbyist CNC mills wind up like CarveWright machines. You wind up paying a few grand for the machine... but, you have to pay dearly for additional software, DRM-protected templates, special memory cards, special readers, and other proprietary crap.
This may be a working model for IBM mainframes, but as consumers, we need to fight tooth and nail so this doesn't happen in other industries.
Back in the early 1990s where CNC mills were not cheap, but inexpensive enough for smaller companies to spend money for, there was a time where there were many, many bicycle parts made out of CNC machined aluminum, usually anodized very funky colors as well. This lasted for a few years, but then Shimano and the other brands went to drop-forging, which allowed for more long-wearing components, especially on the drivetrain. This, plus the move to titanium and carbon fiber caused CNC machining to fade away in that industry.
However, with the resurgence of interest, I wouldn't be surprised if people would be interested in custom-building their own mountain bikes with this technology. Their bikes may a couple grams heavier than with CF technology, but the biggest advantage is that parts would be easy to build and replace in a scenario where one can't get a new XTR set shipped to them.
There is also the independent, "prepper" mindset starting to take hold, where having something that one can maintain oneself is a lot better than the latest and greatest gadget which can only be repaired by a specific shop... or the device can't be repaired at all, and has to be tossed. Having a bike where most of the "consumable" parts like the chainrings are easily made on a CNC mill may be what people want, especially if mills become more common in hobbyist installations.
I hope CNC mills become more common. 3D printing is nice, but there are only so many plastic pieces one can make, and the plastic used is quite toxic to the environment. If one can do metal work, it gives a lot more options... if only to allow one to make molds and use those for an injection molding machine for plastic parts that are decently strong.
It seems to come and go in waves, and just seems like a solution in search of a problem. Even back in the mid 1990s, there were 3D headsets which worked with Duke Nukem 3D, but other than at Egghead, I never saw one purchased and "in the wild."
Eventually we will get past the VR stage of the goggles and the PowerGlove, but until we get to where equipment worn is light to none and we have something similar to a holodeck, VR will be in the fringes with cool stuff happening here and there, but tending to be too expensive and specialized for mainstream use. However, things can change. The same thing was said about MP3 players (expensive/specialized/only for a few people) until Apple got into that market, so I could be proven completely and utterly wrong about this in the coming year.
I wonder about lubrication. A lot of the mills I see need to hose down the piece with water or some other lubricant to keep the metal cool. Is disposing of the water with all the metal bits in it something that is a major concern?
With a machine like that, I am surprised to see USB 2.0 on Type C connectors. Most everyone else has moved to USB 3.0 or USB 3.1. Since this machine can do a desktop role, it should have the ability for someone to plug in an external HDD or SDD because the storage on those devices tends to be pretty small, or just as a way to back the thing up via a program or wbadmin. Wireless access is one thing, but until someone designs a wireless storage protocol that is up to par for the task.
What HP needs to do is make a business grade model of this. It should have 16-64 gigs of RAM, a minimum 512GB SSD, and at least two USB 3.1, type C connectors which can also work for Thunderbolt. It also should have a docking station which uses the Thunderbolt bus and have some decent GPU power in that (or at the minimum, a PCIe slot), a number of USB ports (with multiple I/O buses), and a 10gigE Ethernet port. This way, the tablet can be used for business functions and can double as a desktop with ease. Oh, and have Windows 10 Enterprise pre-activated, just so a company wouldn't have to blow a KMS license on the machine. Done right, it would sell, just because it is able to do multiple roles, from being able to display PowerPoint slides with an adapter getting video from the USB-C connector, to doing day to day work in a dock at the office.
[1]: Wi-Fi Direct isn't usable in this case, unless one doesn't mind switching SSIDs often to go from the Internet connection to the drive's SSID, or putting the HDD on the same Wi-Fi segment and having it available for anyone there to poke at. The ideal would be a protocol that piggybacks off of BlueTooth's encrypted pairing, and can work at a decent speed at a short range.
If HDD capacities could go up, but HDD makers pivot from just shoveling more bits into less space to redundancy and reliability, HDDs will take the niche that tapes have now. Especially if a HDD maker could guarantee an archive life of a tape, perhaps having a dual-head mechanism (I remember seeing some older drives which actually had two sets of heads, each independent of the other and in an active/active configuration.) Doing this would keep HDDs around as backup media or media to stick in the NAS (for a low-end NAS for stashing archived files or backups, a 5200 RPM WD Red or two are just fine.)
I can see HDD makers also "smartening" up their controllers, by going with more SSHD style drives... mainly by adding SSD as a cache or write landing zone... where the I/O first lands on the SSD, then gets shuffled off to the spinny disks. Of course, there is the slowdown once the SSD winds up full... but done right, it can be a way to help with disk I/O for all but the worst sustained random writes. While doing this, HDD makers might be able to add more features, such as autotiering (think the Apple Fusion drives, except done on the controller) and block level deduplication (although this changes a drive from just a being a raw device into presenting the disk as a LUN, as opposed to just a raw chunk of storage.)
I do agree that SSDs, even now, are pretty much a "why not?" choice for primary storage. Even a USB 3 attached SSD with its limited throughput has an advantage over a HDD, just because multiple tasks don't have to fight for control of the drive heads.
In a twisted way, this makes sense. One is oftentimes better off by mining a currency like AltCoin, DASH, or another cryptocurrency, then trading that for BTC, as opposed to trying to just mine BitCoins, just due to the requirement of having a huge investment in ASICs (not to mention a cheap energy source) to mine BTC oneself.
It wasn't that long ago when Windows Mobile was the dominant OS in the smartphone arena. To boot, MS and hardware makers did a pretty good job. The HTC Wizard comes to mind, with a week's battery life. It might be laughable by our standards now, since it only puttered at EDGE speeds, and didn't have the latest rev of BlueTooth, not to mention the use of MiniSD cards... but for the time, it was a very nice phone.
Microsoft dropped the ball when the iPhone came out. The biggest problem is that Windows Mobile was designed around a stylus, while iOS and Android were designed for finger gestures. The shift in the UI expectations left MS in the dust, just because all their device apps were not "finger friendly". MS was encumbered with an existing solution, and having to either figure out how to retrofit the latest UI style, or to just toss everything out and start anew.
MS did a good job with starting anew, and they have a competitive device.
As for a spiritual successor for the Surface... this gets me wondering... are they going to try for an Intel x86 type of computer running W10 in a smartphone form factor? If they could pull this off running a real x86 version of Windows 10 and all Windows applications, it would be a game-changer.
Of course, other companies tried this, such as Motorola with the Atrix and Atrix 2... but if a phone could be tossed in a stand or have a USB-C cable connected, and it take the role of a desktop machine with an x86 version of W10, Microsoft would be breaking new ground. It would mean that one wouldn't need to have anything other than a keyboard, monitor, and USB-C hub in order to have a functioning desktop.
There is one missing piece of the puzzle, and that is getting a phone to handle heavy GPU tasks. This is easy. Since MS has their own graphics standard, it would be trivial for them to make something like a LAN version of OnLive, and have smartphones and tablets send the DirectX commands over the network to a render box, and the render box send back streaming video. Since 4k res of streaming video is about 10Mbps, a wireless LAN can easily handle this. If one uses a newer graphics protocol like ZPEG which gets even better compression for the same quality, it would be even less.
If Microsoft pulls this off, where the only GPU needed for a device would be for the basic UI... they would have a major breakthrough market that would be in high demand.
It wouldn't be worth the time, since I found a far better place to be at anyway, job-wise.
As for FB/whatever, I decided to make an account, and keep them around. I now use Twitter for announcing GitHub releases I make. That way, the account is of actual use.
As for IoT, whining about is not going to do much. However, there are a few ways to actually make IoT truly secure... not secure as in the sense of "locking it down" secure... but secure as in resisting unauthorized intrusions, modifications, deletions... the classic sense.
Three ways to make it work:
1: Get some people who know what they are doing, such as Bruce. Make a UL type independent organization whose job it is to check security of products in both white-box testing and black-box testing. Security such as resisting attacks via the network, ease of resetting the device, should the owner lose the password, how firmware updates are handled [1], how the device reacts to intrusion attempts, internal security like chrooting, signed executables, SELinux, ASLR, and other methods. Have the independent organizations's approval a must for the device to be sold. Of course, this invites regulatory capture, and genuine security can easily be perverted into "keeping the user out" security... but anything in IoT is better than nothing.
2: Move to a different topology for IoT devices than having the devices connecting directly to the Internet via a 3G/4G connection or using a Wi-Fi access point. Instead, the devices should communicate on the LAN basis to a hardened appliance... and that appliance does the sending and receiving for the devices. This way, the "smart toaster" communicating to the hub via BlueTooth will be extremely difficult to hack because it sends the user's toaster preferences up through the BT hub, which then relays it through the Internet. Going with a hub/spoke, with redundant hubs possible, would significantly decrease the attack surface of IoT devices.
3: Use the principle of least privilege. If an Internet connection isn't needed (say for a device to work as a remote), use Bluetooth. If the device has to have an Internet connection for updates, have documentation that describes the sites it connects to [2], and what ports that it should be allowed. Anything else should be blocked. The device should even enforce this in its OS firewall (netfilter for Linux, for example) to protect against unauthorized processes trying to get out. If "smart" functionality isn't needed, don't bother with it.
Take the "smart" refrigerator. If appliance companies wanted to make something expensive, why not a fridge with two cooling mechanisms... the standard compressor that plugs into the wall, and an absorption mechanism which can be powered by electricity, natural gas, or propane. This way, if there is a power blackout, the fridge still retains cooling capacity, and with a thermalelectric generator (think a Peltier running in reverse), would have enough power to keep the core circuit board running. I'm sure there would be more demand for a fridge that keeps the food cold if power goes out, than a fridge which can display ads 24/7 on the screen.
[1]: I believe in the old school idea of a physical button or switch that is used before flashing firmware... but this isn't something that can be done if the device is not physically accessible, so maybe a fallback would be some other mechanism. That way if the RSA key is compromised, the vendor can use a different, but still secure, way to get the updates to devices.
[2]: Ideally, it should just fetch a signed manifest via SSL, and go from there. If the embedded OS is Linux, it could even use an existing package manager like Yum or apt so that wheel doesn't have to be reinvented.
Here is where things get nefarious. IoT are like social networks. In the past, you could just tell people where to stick it when they talked about their livejournal, MySpace, or Orkut stuff. However, if one doesn't have a LinkedIn account, FB account, and a Twitter account, you will be turned down for jobs.
I know this firsthand. Had a job interviewer tell me that I was too old for IT work and show me the door because he wanted to read/follow my Twitter account, and I told him that I didn't have one.
IoT has the potential for being just like that. For example, the Bluetooth deadbolt. It might be that apartment managers and other landlords install IoT security devices because it makes their job easier to lock out tenants being evicted, know who is going into a tenant's place, or to let maintenance in on a schedule regardless if the tenant wants it or not. For more flexible for the property owner, and the tenant would have no choice in the matter.
Insurance can also demand IoT devices, say CCTV monitoring and file storage, or IoT deadbolts and other devices so they can be assured that a property is secured when the owners are away. If this isn't done, they won't renew the policy.
Then, there is the phone home aspect. Pull the internet connection on a modern console, it halts. I wouldn't be surprised if a future HDCP spec that requires all devices to authenticate with a central server for a healthcheck every so often, would require that all TVs and such be always on and in communication. As per the EULA of the TV, video and audio would also be sent back for "IP enforcement purposes". If someone disagrees with that... well, good luck with the no-sue arbitration agreement they agreed to...
Next comes devices. Take the refrigerator for instance. Good luck trying to find a completely mechanical one with a thermostat and compressor that runs for decades. Most have various computer controls. It wouldn't be surprising that IoT functionality is important, and no network connection means the device does not function, especially if the fridge maker starts demanding license keys to activate the ice maker, crisper section, and such.
The key is to not just avoid buying IoT shit, but make it -damn well known- that you will never buy that because you don't want another route an intruder can trespass into your home. Because IoT security is so weak, and there is zero incentive for companies to actually do something about it, it needs to die on the vine.
TFA didn't have many details, but I wonder how the files were watermarked. Some metadata perhaps, that added atime/ctime/ and the computer's install ID?
Perhaps a modification to the filesystem to allow for alternate data streams (a la NTFS) which would have a linked list of machines the file has sat on, which would be hidden from the user, and would move with the file?
That is definitely possible. It would require a few steps though.
First, I can see someone showing an ID somewhere, and picking a blank cryptographic token. Something like a CAC (if military), PIV card (if civilian), or just a generic USB token like an Aladdin/SafeNet eToken if one uses that for secure storage for PGP keys. It is picked at random, just like how paper ballots are laid out, so the government knows the key is valid... but by default doesn't know whose key it belongs to.
Second, said key is used as a client certificate. Only PW used is the PIN to unlock the token for the browsing session.
If a user wants to change identities, they just bring the old card back, someone checks that it was a valid card, they carve it into little bitty pieces, and allow the citizen to grab a replacement card with a fresh, new identity.
The result is solid pseudo-anonymity, but because every person only can have one ID at a time, it keeps the lid down on sock puppets, shills, and general trolls.
Of course, this assumes the CAC/PIV maker can be trusted, as well as the distribution channel... but that is what government controls are for. Plus, since this isn't as much tied to the future of a country directly, it would be less of a target for the bad guys than pwning voting machines.
Of course, the problem how to prove a user only has one card, needs to be solved. This way, if the user loses or has their card destroyed, they can get another one, have the cert revoked on the old key. However, if the government keeps track of user to public key mappings, then the whole anonymity concept goes into the shitter.
I can see someone designing software that watches someone screen and ensures they vote for the "proper" candidate, either what the employer wants, or what a criminal organization wants.
I'm with you. Voting by paper and snail mail as a last resort, but voting booths and locations give people to be able to vote and have their vote be truly anonymous, so they can vote for the candidate they so choose and be resistant to coercion.