On home machines, Microsoft System Essentials. In the enterprise, Forefront. MS said that Forefront can effectively protect against the zombie horde, as well as ninja attacks in an ad campaign a few years back, and if that is true, just that ability is well worth the product's price.
If people dumped Windows for open source, there will still be a large market for AV utilities, for legal reasons.
There are a lot of companies where I had to spec out antivirus solutions for AIX, Solaris, RedHat, and OS X just for CYA reasons. Not like all the LPARs on the pSeries 795 in the server room is going to get infected, but because it is a checkbox on a contract that "all computers on the corporate network will have antivirus software on them."
+1. I've seen this also on Cydia, when looking for a command line application like bc. Why does one need calculators with sports logos dumped in the UNIX utility section? Stick that stuff in some sports memorabilia section. The Google App store, same thing... For example, a dialing program that is just a skin, have a category that it drops into separate from non-branded utilities.
Oh, for Google's store, have a way to drop in an "X/NC-17" setting. Apple states this for file managers, but for obvious pr0n apps, it would be nice if those would go to their own section, because if I'm searching for a "grep" utility, I don't really care about apps with people being "grepped".
I wonder with the ways that WPA2-PSK is being eroded, if one should just go with 30+ character long keys. TrueCrypt always recommends to go with 20+ character passphrases and since there isn't much key strengthening with WPA2-PSK, a longer key is a good thing here. My preference is to use a 63 number of letters and digits, and if it gets forgotten, just generate another string and paste it into the router from a machine on the wired network.
That is the problem with Android. Because of misconceptions, the big gaming companies are ignoring the platform. However, they are perceptions, not reality. A couple:
1: Piracy. Android has a mechanism independent of rooting as a guard against piracy, the LVL, or License Verification Library. The iPhone depends solely on if the device is jailbroken or not. Toss in JVM/DVM code obfuscation, and Android APKs are much tougher nuts to crack than iPhone apps. Click on copy-protected when the app is being uploaded and that provides another layer of shielding.
2: Faster update cycles. On Android, you upload your APK, and it is available for update almost at once. With this, even if someone cracked your 1.0.1, your 1.0.2 forces them back to the drawing board. This allows for tight development cycles and fast bug stomping.
3: Hardware/"fragmentation". This is what the manifest file is for. Don't want anything but the latest Android? Something in the uses-configuration/uses-api/uses-library/uses-feature settings in the manifest file will keep the app from being seen in the Marketplace unless someone has a device with those features [1].
4: Faster hardware development cycles than iOS.
None of this is to say that iOS is bad; it just states that Android is just as good as Apple's offerings for making apps.
The problem is that Android isn't locked down, especially if you buy a Google platform reference device (ADPs, Nexus 1, Nexus S). What is happening is that phone makers and carriers are doing the nasty stuff. Yes, it gets bypassed somehow, but as the arms race continues, the bypassing will end up becoming more and more difficult. If Motorola can make it where their devices can't be ROM-ed or rooted for a year to two years, they have won because people would have moved on.
The scenario I see that is bleak is that hardware would stop being sold to the end user. It would be leased, similar to how Ma Bell phones were ages ago. This difference means a lot when it comes to legal battles of control over root access on hardware. Now toss in ISPs refusing to permit access to the Internet unless the device was running a signed OS. There are no laws against this. Since there are only 1-2 ISPs in an area here in the US, they can effectively lock out access as they so please.
So, we combine the leasing of hardware with the fact that ISPs only will allow leased hardware that isn't on a blacklist to connect. Devices would have a "hardware WGA" chip that can scan memory memory, phone home, and brick the device if they detect a suspect utility. We will then see court cases comparing a jailbroken device to someone who cuts off the lock to their gas meter to get free utilities, judges blindly rubber-stamping what the prosecution/plaintiff wants, and people ending up in jail for long sentences.
Result: Game, set, match. People use what they are told to, pay what they are told to. We will then see devices that are leased charging for every minute a user is logged on, how much CPU hour the user uses, how many bytes transferred, per site hit, and such like that. Think the days of Compuserve/The Source/AOL/Prodigy all over again.
I am a cynic these days. It is a cool ideal to just carry around one device (or even a device with all personal details stored on a SIM-like card), but I keep thinking ways that can be abused. There are a lot of people out there who would go out of their way to cause someone else harm.
It is interesting the comparisons if you drop an iPhone and a Droid X side by side.
To me, where the iPhone wins are apps and standardization. Where the Droid X wins is almost everywhere else. I blow up the OS on the Droid X? Boot recovery, nandroid restore, load app data from Titanium Backup that have changed since the last restore.
To the average person, I'm sure their needs can be met with either device. Exchange support? If the Android device isn't up to par, Nitrodesk Touchdown can support Exchange and all the requirements without issue. Same with apps needed to open and read Word, Excel, and other files.
Those are the upsides. Actually, Intel was working on a personal server which was a HDD that used Bluetooth to communicate with things.
The downsides: Walk into the living room. Fire up a terminal. Too bad that a remote attack flashed the keyboard to log keystrokes which got sent to a botnet. Then the attacker uses a remote exploit on the house alarm to drop in some malicious code that reflashes the PCD when the owner comes near, adding some malware. One then walks past their oven, the malware leaps to that, modifying the flash of that device so it will open the gas valve, disable the pilot light, then light the pilot light 12-24 hours afterwards, causing a nice natural gas explosion. Similar code shorts out the automatic lock on the door and disables the fire alarm. Another piece of malware jumps to the car, adding in a remote override code which a local chop shop then picks up the vehicle after the owner leaves. The PCD arrives at the workplace. A brief interaction with the work printer adds a slight reflash which sends a copy of documents coming out of it to machines in the attacker's botnet for later pickup. Another brief interaction with the work's microwave causes it to flip on the magnetron when someone opens the door, causing instant cataracts and third degree burns.
I will pass. The fewer devices that are connected with each other, the better. Stuxnet showed us what can happen when embedded devices and the Internet intersect.
The trick is to get an open phone. Sadly, the only models which are open are the ones Google has for its reference models, which are the ADP line, the Nexus 1, and the Nexus S.
I'm hoping the next reference model (for Android 3.x) has a SD card slot. 24 GB just doesn't cut it, especially with large music collections. Even then, it is nice to have the apps for one task or project be on one SD card that gets removed before a vacation trip... this way even if the phone is lost/stolen, the data is definitely safe.
It isn't hard to crunch out a standard MicroSIM from a normal size one.
Only caveat with this method is, if the phone has issues, to take it to a store in Canada. The US store will hand you a replacement phone... locked to AT&T.
Of course, Java will take some overhead, but the days of Java being obnoxiously slow are far behind us, especially with Android 2.3 and the JIT compiler ability.
Want to know what the Dalvik VM brings to the table over the compiled Objective-C binaries? Security. It takes a lot more hacking to bust out of a well secured sandbox, then to try to find a hole in the userland environment than it is to just run a binary. Of course, there are examples of both, but a sandbox can be tightened up.
The VM also brings architecture independence. Say IBM made a version of the POWER7 that could execute code better, using less electricity than the ARM architecture. A Linux port later, and an Android device can use that.
You hit the nail on the head. It pretty much is finding the carrier which sucks the least. CDMA carriers are pointless if one travels unless one buys a "world phone" that has dual GSM/CDMA radios and one pays the insane roaming charges. The GSM carriers have completely incompatible 3G bands with each other, and the rest of the world.
Hard to say which sucks the least, especially with T-Mobile's new actions. Probably AT&T. At least they stab you from the front by stating their fee policy ahead of time.
Nothing wrong with a Tracfone. Less hassle in a lot of cases.
We will see if Verizon can live up to their ads and boasting about more data traffic than AT&T. Time will tell, especially the move to LTE in the upcoming months.
The one thing I would like to see is that AT&T dumps the 2GB/month plan and offer truly unlimited access again. That is the biggest thorn in my side right now. Having the Mi-Fi ability wouldn't be bad either.
1: On a call, and need to get an address from a phonebook for a friend, or a phone number. 2: Calling to doublecheck weather because a relative is heading out of town. 3: Hear about a cool new app, so flip to downloading it. 4: While yapping on a long call, flip to E-mail to see if anything is going bad at work or elsewhere. 5: While on a long call, pull up Facebook.
I am in a similar position. I was thinking of buying an unlocked Nexus S and going with T-Mobile US myself. Obviously T-Mobile US != T-Mobile UK, but policies tend to leap across the pond fairly quickly.
Now, I'm going to wait and see. Today may bring some interesting announcements when it comes to smartphones.
The situation is likely different in the US. T-Mobile US will just bury itself if it enacts similar policies, because as of now, what attracts people to T-Mobile is the fact that unlocked GSM devices are welcome there (the Android Dev Phones, the Nexus S, etc.) Their quiet policy of allowing 5GB, then dropping to EDGE is also a good one. However, if this changes, it might be a nail in their coffin.
On Windows, I like having two users for admins. Their "normal" user for receiving Exchange messages, etc. Then they get a "su" user that has the full admin rights. This way, if malware gets user context, it almost invariably will be the normal user that it gets, as opposed to the full Administrator/Domain Admin/Enterprise Admin rights.
Naming conventions vary. For example, I use "username" and "usernamesu". Other people (DoD) use "username" and "aausername".
If it is in TrueCrypt volume and accessed by not just a passphrase, but a keyfile (or keyfiles), it can be stored pretty securely (assuming secure storage of the keyfiles and non-hacked endpoints). Downside of this method is no ability have multiple users writing to it at the same time, so changes in passwords have to be propagated via different means.
By default in most UNIX operating systems (including OS X [1]), the sudo mechanism is the mechanism of choice. Directly logging on root should be denied by default on all operating systems. This is good to prevent root access by someone whose account got locked (better than trying to get all the root passwords changed.) Plus, with sudo, the root password doesn't have to be given to anyone who needs more than user access. For example, if an operator needs access to run a program with another user's permissions automatically, this is easy to do with sudo, and without requiring SGID permissions on files.
It is always good to use sudo (or at least su) if at all possible. There are cases where this might not be doable (systems and policies that forbid anything to ruin SUID root), but it really does help with reducing mistakes, adds reasonable audibility, and adds a lot of ability to fine-tune access to root.
[1]: In fact, on OS X, root is disabled by default, and takes explicitly enabling to be able to use.
The old phrase, "trust, but verify" comes into play here. At the SMB level, it is likely not worth it to employees hired to watch the other admins, other than a known good consultant who just comes in every so often. However, at the midsize to enterprise level, it doesn't hurt to have separation of duties, and not just into the usual silos (windows/network/SAN/backup/UNIX/physical security/auditing/etc.) It is good to avoid having an "internal affairs" group who is only there to second-guess the other admins. This ends up in nasty political battles.
I've seen that too. Technology helps drivers, but can't replace them, (as of now). Since I live in Texas, I see the "I have 4WD/AWD, I can do anything". An ice storm comes up, and the 4x4 vehicles are right beside the Toyota Camries in the ditches by the side of the road.
I'm not a fan of OnStar just due to the security implications. GM has to have a VERY secure system, because if a blackhat can get access, they can cause all sorts of havoc. What could happen would be similar to what took place in Austin when an ex-employee of a used car dealership (that used a "bill not paid, no start engine" black box on all cars sold and were being paid on) logged in through an employee account and shut down every single customer's car bought through that place [1]. Picture the chaos of that, scaled up a few orders of magnitude. I can see a real blackhat selling an Onstar 0-day to a group who would wait until hurricane was about to hit, people starting to evacuate, then disabling every GM car just to cause chaos.
I've seen rear-enders caused by that "simple shoulder check". The driver is checking if a lane is clear, doesn't realize someone just swooped in in front and then did a brake check. Rear-ender subsequently happens, and I end up having to sit on my ass and give the police and the parties affected my personal info as a witness.
Ideally, I'd like some display that shows a representation of vehicle, then blobs around it showing what is around the car... this way, without having to take the eyes of what's ahead, one can see that motorcycle that is hiding between the mirrors and the window. To boot, most new cars have a lot more blind spots than models made even five years ago. Having a display showing that something is occupying space in a place you want to go would be quite useful, and save quite a few fender benders.
Realistically, people are not buying cars for performance; they are buying them for MPGs, safety, and some level of creature comfort to help mitigate a long commute. With today's crowded roads, performance is pointless -- all that really matters on a pragmatic basis is MPG, safety, braking distance, reliability, handling (to swing a vehicle in a gap before it closes), and a good stereo system to make the miles go faster, even when the vehicle is still in traffic. The foreign car companies know this, and this is why the small Mazdas, Hondas, and VWs are flying off of dealership lots.
Of course MPG... this is a psychological thing, but people will trade in a vehicle and eat $10,000 worth of depreciation to save $1000 worth of gas in a year. The press beats it in viewers' heads how gas prices are always climbing. The colleges and high schools teach that a person is morally wrong if they don't drive a Prius or high MPG vehicle. Other nations make fun at the SUV driving American stereotype (which in most areas of the US outside Beverly Hills is complete BS.) MPG is beginning to sell cars more than horsepower does.
The closest analogy I can liken it to are plasma speakers. They are excellent at reproducing sound. However, using plasma as a microphone, they would take a lot of work and amplification to get any meaningful signal back from them.
Slashdot Mirror
On home machines, Microsoft System Essentials. In the enterprise, Forefront. MS said that Forefront can effectively protect against the zombie horde, as well as ninja attacks in an ad campaign a few years back, and if that is true, just that ability is well worth the product's price.
If people dumped Windows for open source, there will still be a large market for AV utilities, for legal reasons.
There are a lot of companies where I had to spec out antivirus solutions for AIX, Solaris, RedHat, and OS X just for CYA reasons. Not like all the LPARs on the pSeries 795 in the server room is going to get infected, but because it is a checkbox on a contract that "all computers on the corporate network will have antivirus software on them."
+1. I've seen this also on Cydia, when looking for a command line application like bc. Why does one need calculators with sports logos dumped in the UNIX utility section? Stick that stuff in some sports memorabilia section. The Google App store, same thing... For example, a dialing program that is just a skin, have a category that it drops into separate from non-branded utilities.
Oh, for Google's store, have a way to drop in an "X/NC-17" setting. Apple states this for file managers, but for obvious pr0n apps, it would be nice if those would go to their own section, because if I'm searching for a "grep" utility, I don't really care about apps with people being "grepped".
I wonder with the ways that WPA2-PSK is being eroded, if one should just go with 30+ character long keys. TrueCrypt always recommends to go with 20+ character passphrases and since there isn't much key strengthening with WPA2-PSK, a longer key is a good thing here. My preference is to use a 63 number of letters and digits, and if it gets forgotten, just generate another string and paste it into the router from a machine on the wired network.
That is the problem with Android. Because of misconceptions, the big gaming companies are ignoring the platform. However, they are perceptions, not reality. A couple:
1: Piracy. Android has a mechanism independent of rooting as a guard against piracy, the LVL, or License Verification Library. The iPhone depends solely on if the device is jailbroken or not. Toss in JVM/DVM code obfuscation, and Android APKs are much tougher nuts to crack than iPhone apps. Click on copy-protected when the app is being uploaded and that provides another layer of shielding.
2: Faster update cycles. On Android, you upload your APK, and it is available for update almost at once. With this, even if someone cracked your 1.0.1, your 1.0.2 forces them back to the drawing board. This allows for tight development cycles and fast bug stomping.
3: Hardware/"fragmentation". This is what the manifest file is for. Don't want anything but the latest Android? Something in the uses-configuration/uses-api/uses-library/uses-feature settings in the manifest file will keep the app from being seen in the Marketplace unless someone has a device with those features [1].
4: Faster hardware development cycles than iOS.
None of this is to say that iOS is bad; it just states that Android is just as good as Apple's offerings for making apps.
[1]: http://developer.android.com/guide/appendix/market-filters.html
The problem is that Android isn't locked down, especially if you buy a Google platform reference device (ADPs, Nexus 1, Nexus S). What is happening is that phone makers and carriers are doing the nasty stuff. Yes, it gets bypassed somehow, but as the arms race continues, the bypassing will end up becoming more and more difficult. If Motorola can make it where their devices can't be ROM-ed or rooted for a year to two years, they have won because people would have moved on.
The scenario I see that is bleak is that hardware would stop being sold to the end user. It would be leased, similar to how Ma Bell phones were ages ago. This difference means a lot when it comes to legal battles of control over root access on hardware. Now toss in ISPs refusing to permit access to the Internet unless the device was running a signed OS. There are no laws against this. Since there are only 1-2 ISPs in an area here in the US, they can effectively lock out access as they so please.
So, we combine the leasing of hardware with the fact that ISPs only will allow leased hardware that isn't on a blacklist to connect. Devices would have a "hardware WGA" chip that can scan memory memory, phone home, and brick the device if they detect a suspect utility. We will then see court cases comparing a jailbroken device to someone who cuts off the lock to their gas meter to get free utilities, judges blindly rubber-stamping what the prosecution/plaintiff wants, and people ending up in jail for long sentences.
Result: Game, set, match. People use what they are told to, pay what they are told to. We will then see devices that are leased charging for every minute a user is logged on, how much CPU hour the user uses, how many bytes transferred, per site hit, and such like that. Think the days of Compuserve/The Source/AOL/Prodigy all over again.
I am a cynic these days. It is a cool ideal to just carry around one device (or even a device with all personal details stored on a SIM-like card), but I keep thinking ways that can be abused. There are a lot of people out there who would go out of their way to cause someone else harm.
It is interesting the comparisons if you drop an iPhone and a Droid X side by side.
To me, where the iPhone wins are apps and standardization. Where the Droid X wins is almost everywhere else. I blow up the OS on the Droid X? Boot recovery, nandroid restore, load app data from Titanium Backup that have changed since the last restore.
To the average person, I'm sure their needs can be met with either device. Exchange support? If the Android device isn't up to par, Nitrodesk Touchdown can support Exchange and all the requirements without issue. Same with apps needed to open and read Word, Excel, and other files.
Those are the upsides. Actually, Intel was working on a personal server which was a HDD that used Bluetooth to communicate with things.
The downsides: Walk into the living room. Fire up a terminal. Too bad that a remote attack flashed the keyboard to log keystrokes which got sent to a botnet. Then the attacker uses a remote exploit on the house alarm to drop in some malicious code that reflashes the PCD when the owner comes near, adding some malware. One then walks past their oven, the malware leaps to that, modifying the flash of that device so it will open the gas valve, disable the pilot light, then light the pilot light 12-24 hours afterwards, causing a nice natural gas explosion. Similar code shorts out the automatic lock on the door and disables the fire alarm. Another piece of malware jumps to the car, adding in a remote override code which a local chop shop then picks up the vehicle after the owner leaves. The PCD arrives at the workplace. A brief interaction with the work printer adds a slight reflash which sends a copy of documents coming out of it to machines in the attacker's botnet for later pickup. Another brief interaction with the work's microwave causes it to flip on the magnetron when someone opens the door, causing instant cataracts and third degree burns.
I will pass. The fewer devices that are connected with each other, the better. Stuxnet showed us what can happen when embedded devices and the Internet intersect.
The trick is to get an open phone. Sadly, the only models which are open are the ones Google has for its reference models, which are the ADP line, the Nexus 1, and the Nexus S.
I'm hoping the next reference model (for Android 3.x) has a SD card slot. 24 GB just doesn't cut it, especially with large music collections. Even then, it is nice to have the apps for one task or project be on one SD card that gets removed before a vacation trip... this way even if the phone is lost/stolen, the data is definitely safe.
It isn't hard to crunch out a standard MicroSIM from a normal size one.
Only caveat with this method is, if the phone has issues, to take it to a store in Canada. The US store will hand you a replacement phone... locked to AT&T.
[Citation Needed]
Of course, Java will take some overhead, but the days of Java being obnoxiously slow are far behind us, especially with Android 2.3 and the JIT compiler ability.
Want to know what the Dalvik VM brings to the table over the compiled Objective-C binaries? Security. It takes a lot more hacking to bust out of a well secured sandbox, then to try to find a hole in the userland environment than it is to just run a binary. Of course, there are examples of both, but a sandbox can be tightened up.
The VM also brings architecture independence. Say IBM made a version of the POWER7 that could execute code better, using less electricity than the ARM architecture. A Linux port later, and an Android device can use that.
You hit the nail on the head. It pretty much is finding the carrier which sucks the least. CDMA carriers are pointless if one travels unless one buys a "world phone" that has dual GSM/CDMA radios and one pays the insane roaming charges. The GSM carriers have completely incompatible 3G bands with each other, and the rest of the world.
Hard to say which sucks the least, especially with T-Mobile's new actions. Probably AT&T. At least they stab you from the front by stating their fee policy ahead of time.
Nothing wrong with a Tracfone. Less hassle in a lot of cases.
Yes, Android 2.2 supports it. But the cellular carrier can disable it, and US carriers tend to do that.
We will see if Verizon can live up to their ads and boasting about more data traffic than AT&T. Time will tell, especially the move to LTE in the upcoming months.
The one thing I would like to see is that AT&T dumps the 2GB/month plan and offer truly unlimited access again. That is the biggest thorn in my side right now. Having the Mi-Fi ability wouldn't be bad either.
I use this daily:
1: On a call, and need to get an address from a phonebook for a friend, or a phone number.
2: Calling to doublecheck weather because a relative is heading out of town.
3: Hear about a cool new app, so flip to downloading it.
4: While yapping on a long call, flip to E-mail to see if anything is going bad at work or elsewhere.
5: While on a long call, pull up Facebook.
So, not having this available is a deal-breaker.
I am in a similar position. I was thinking of buying an unlocked Nexus S and going with T-Mobile US myself. Obviously T-Mobile US != T-Mobile UK, but policies tend to leap across the pond fairly quickly.
Now, I'm going to wait and see. Today may bring some interesting announcements when it comes to smartphones.
The situation is likely different in the US. T-Mobile US will just bury itself if it enacts similar policies, because as of now, what attracts people to T-Mobile is the fact that unlocked GSM devices are welcome there (the Android Dev Phones, the Nexus S, etc.) Their quiet policy of allowing 5GB, then dropping to EDGE is also a good one. However, if this changes, it might be a nail in their coffin.
On Windows, I like having two users for admins. Their "normal" user for receiving Exchange messages, etc. Then they get a "su" user that has the full admin rights. This way, if malware gets user context, it almost invariably will be the normal user that it gets, as opposed to the full Administrator/Domain Admin/Enterprise Admin rights.
Naming conventions vary. For example, I use "username" and "usernamesu". Other people (DoD) use "username" and "aausername".
If it is in TrueCrypt volume and accessed by not just a passphrase, but a keyfile (or keyfiles), it can be stored pretty securely (assuming secure storage of the keyfiles and non-hacked endpoints). Downside of this method is no ability have multiple users writing to it at the same time, so changes in passwords have to be propagated via different means.
By default in most UNIX operating systems (including OS X [1]), the sudo mechanism is the mechanism of choice. Directly logging on root should be denied by default on all operating systems. This is good to prevent root access by someone whose account got locked (better than trying to get all the root passwords changed.) Plus, with sudo, the root password doesn't have to be given to anyone who needs more than user access. For example, if an operator needs access to run a program with another user's permissions automatically, this is easy to do with sudo, and without requiring SGID permissions on files.
It is always good to use sudo (or at least su) if at all possible. There are cases where this might not be doable (systems and policies that forbid anything to ruin SUID root), but it really does help with reducing mistakes, adds reasonable audibility, and adds a lot of ability to fine-tune access to root.
[1]: In fact, on OS X, root is disabled by default, and takes explicitly enabling to be able to use.
The old phrase, "trust, but verify" comes into play here. At the SMB level, it is likely not worth it to employees hired to watch the other admins, other than a known good consultant who just comes in every so often. However, at the midsize to enterprise level, it doesn't hurt to have separation of duties, and not just into the usual silos (windows/network/SAN/backup/UNIX/physical security/auditing/etc.) It is good to avoid having an "internal affairs" group who is only there to second-guess the other admins. This ends up in nasty political battles.
I've seen that too. Technology helps drivers, but can't replace them, (as of now). Since I live in Texas, I see the "I have 4WD/AWD, I can do anything". An ice storm comes up, and the 4x4 vehicles are right beside the Toyota Camries in the ditches by the side of the road.
I'm not a fan of OnStar just due to the security implications. GM has to have a VERY secure system, because if a blackhat can get access, they can cause all sorts of havoc. What could happen would be similar to what took place in Austin when an ex-employee of a used car dealership (that used a "bill not paid, no start engine" black box on all cars sold and were being paid on) logged in through an employee account and shut down every single customer's car bought through that place [1]. Picture the chaos of that, scaled up a few orders of magnitude. I can see a real blackhat selling an Onstar 0-day to a group who would wait until hurricane was about to hit, people starting to evacuate, then disabling every GM car just to cause chaos.
[1]: One source on this -- http://www.wired.com/threatlevel/2010/03/hacker-bricks-cars/
I've seen rear-enders caused by that "simple shoulder check". The driver is checking if a lane is clear, doesn't realize someone just swooped in in front and then did a brake check. Rear-ender subsequently happens, and I end up having to sit on my ass and give the police and the parties affected my personal info as a witness.
Ideally, I'd like some display that shows a representation of vehicle, then blobs around it showing what is around the car... this way, without having to take the eyes of what's ahead, one can see that motorcycle that is hiding between the mirrors and the window. To boot, most new cars have a lot more blind spots than models made even five years ago. Having a display showing that something is occupying space in a place you want to go would be quite useful, and save quite a few fender benders.
Realistically, people are not buying cars for performance; they are buying them for MPGs, safety, and some level of creature comfort to help mitigate a long commute. With today's crowded roads, performance is pointless -- all that really matters on a pragmatic basis is MPG, safety, braking distance, reliability, handling (to swing a vehicle in a gap before it closes), and a good stereo system to make the miles go faster, even when the vehicle is still in traffic. The foreign car companies know this, and this is why the small Mazdas, Hondas, and VWs are flying off of dealership lots.
Of course MPG... this is a psychological thing, but people will trade in a vehicle and eat $10,000 worth of depreciation to save $1000 worth of gas in a year. The press beats it in viewers' heads how gas prices are always climbing. The colleges and high schools teach that a person is morally wrong if they don't drive a Prius or high MPG vehicle. Other nations make fun at the SUV driving American stereotype (which in most areas of the US outside Beverly Hills is complete BS.) MPG is beginning to sell cars more than horsepower does.
The closest analogy I can liken it to are plasma speakers. They are excellent at reproducing sound. However, using plasma as a microphone, they would take a lot of work and amplification to get any meaningful signal back from them.