Part of the reason why there is a move to bigger phones is engineering issues. The faster CPUs require more surface area to dissipate heat. So, the product designers might as well offer a larger screen size.
I will say that AIX is pretty good as well. In general, unless there is a show-stopper patch, or one installs a driver like EMC PowerPath that requires a reboot due to the hooks in the kernel, one can keep AIX up for a long while, only really bothering to update and reboot when the latest tech level is released, and if there are no security specific issues, even that can be ignored, although it is wise to keep up on new firmware stuff just in case.
Sorta Yes Yes No Depends. Droidwall helps a lot. So did LBE Privacy Guard, but it hasn't been updated in ages, and doesn't support 4.1 or 4.2. Use CyanogenMod or a custom ROM. Depends on app.
I'm guessing they got complaints, so had to do something in order to keep the ad providers happy, as well as the developers who get revenue from admob and other providers. Had Google not done at least a token effort, there are plenty of other places the app writers can go (Tizen is waiting in the wings, Windows 8 is extremely mature, and there is always iOS where Apple can almost guarantee people will see ads slung at them.)
The nice thing about Android is that if Google blocks the apps on their store, there is always sideloading. I can use a QR code or click a link, and go from there.
I don't think Google is really evil, except they are responding to pressure put on them, as the consequences for not doing something (developers and possibly phone makers taking their business elsewhere) would be a lot worse than having users tick a check box and grab their goodies from a different source.
Security can be done right. It is a matter of "won't", rather than "can't" in general. A lot of companies have the "security has no ROI" attitude, but if it is something that their names are on the line for, they might actually follow working practices.
Take the Mercedes DAS (drive authorization system) keys/remotes. Very few, if any, people are bypassing the anti-theft system on these vehicles. So far, there are key cloners out there at best. If there is actual money and lives on the line, even this would be much improved so "hackcidents" are something that is of fiction rather than fact.
My assumption is that self-driving cars will have actual decent security, good enough to keep a tweaker from commandeering someone's vehicle to pick up some lab "product".
Bingo. Self-driving cars would be very useful in the US for a number of reasons:
1: The passive servicing aspect. Aft night, the vehicle can haul itself to a garage for overnight servicing, and I'm sure 24 hour service stations/car dealers would sprout up for this demand. Having a car get serviced while asleep would save hours of time.
2: The ability to set up shops which are essentially reverse delivery drivers. Car drives to window, order is placed in vehicle, car drives away. This makes it possible to order a pizza from an exotic spot that would be out of the town's delivery radius. Or, pick up a part from another town.
3: Extra hours of sleep. With commutes in the hours, might as well let the car do the dirty work.
4: No need for cloverleaf intersections. Communication between stationary computers and vehicles would allow cars to drive full speed through 4-way intersections with timing no human could ever accomplish.
5: Actual respect for driving laws. A cyclist or pedestrian presses a button to cross a road, the cars -will- stop and allow them by.
6: Cars can park in a garage or parking lot by themselves.
7: Far fewer accidents (thus more people living.)
8: Better fuel economy in general. Fewer panic stops.
This is something I wish Moto would continue developing. Create a lapdock with a SD card slot for backups, allow people to have their own Linux distro, and it would be very useful as a computer for emergencies.
If Moto made the lapdock usable between phone generations, it would make a perfect place for backups, possibly migrating apps and data to a new device come upgrade time.
My last Moto phone, the Atrix 2, had great engineering, but because of the bootloader issue, and the fact that it took a good long while (about six months IIRC) just to get a flashable factory ROM (FXZ format) in case a rooting went bad.
Unless Moto gets with the competition, I'm going to patronize other places that don't mind allowing people to use their purchased device as they see fit. My HTC 1X+ does not have anywhere near the voice and Bluetooth quality that either my iPhone 5 or Atrix 2 does... but it has a custom ROM and works decently. If Motorola allowed bootloader unlocks, I'd definitely purchase another model from them. However, it looks like my next device will likely be a Galaxy 4 when it gets released.
Motorola has some things they are good at. For example, their radios tend to be top of the line for call quality in my experience, at least on par with iPhones.
MM's problems are not impossible to solve. One key that Motorola has advertised is enterprise-level friendliness and security. Google could run with that to help get their devices more entrenched into businesses.
After that, choose a niche for MM. Do they compete on the low-end with Huawei and ZTE, the high-end against Samsung and Apple, or midrange offerings? I'd say take the high/middle range if it were me.
I'd like to see Google get Motorola to make devices, not one for each provider, but devices that are either CDMA/LTE, GSM/LTE, or perhaps best of all, both. That way, it would be less about having rudimentary support for different phones, but being able to focus on a few devices at a time for ROM updates. It would be nice to see Android updates for a phone that would go past the 6-12 month mark.
Of course, doing like HTC and having source for the drivers and other parts of the ROM would be nice, perhaps even work with Cyanogen so their ROM is easily ported with full functionality.
If it were up to me, I'd be seeing about more of the computer-replacement technology that appeared in the Atrix, the Atrix 2, and other devices before it was axed. Couple that with a USB port so a keyboard and mouse could be attached, and this could function as a terminal for Citrix, RDP, or ssh if need be.
Of course, "fastboot oem unlock" would be on all devices, as well as a method of re-flashing ROMs that don't require a special program for FXZ, SHX or other files.
One of the reasons I am guessing that Google is being conservative with the Motorola division is the concern about being viewed as a monopoly. If Google does too much with MM, other Android makers (Samsung, Huawei, ZTE) will jump ship for other operating systems like Windows 8.
There is also the fact that there is the fear of being viewed as a monopoly by the EU.
Regardless, it would be nice if Motorola would do like what Sony, HTC, and others offer, and give a way to unlock the bootloader. I like Motorola phones, but I won't buy another one unless there is a way to do this.
There is a difference. The early change-muncher games actually had replayability and took some tactics to win.
Modern games, a FPS is a FPS, and a paid-for DLC rocket launcher is a rocket launcher, be it on the a Ringworld clone world, a misty abandoned town, or some other no-mans land.
The annoying thing is how iOS and Android games changed from 2010 to now. In 2010, 99 cents got you a decently playable game, such as some tower defense game. These days, the same tower defense game is tuned so you have to buy the additional powerups, the damage bonuses, the bonus XP, and other items, so a.99 cent game becomes a $15.00 game for virtually the same amount of content.
What might be an idea would be client-side HTML5 code to do the MD5/SHA/bcrypt hashing [1] on the local machine before sending it up. This doesn't get rid of the need for SSL, but it does ensure that the original passphrase is extremely hard to obtain.
If for some reason a passphrase has to be stored for some reason (perhaps it needs retrieval intact), then there is a fairly easy way to deal with that: Have a clientside app that takes the password plus a salt and encrypts it with the site's public key [2]. That way, should the key have to be recovered, it still can be, but nowhere except the client's computer does the plaintext exist without someone explicitly decrypting the encoded text.
[1]: With a salt sent from the server.
[2]: Assuming the private key part is stashed in a HSM and is of a decent length (8192 bits, ideally 32768 bits.)
I wonder if they can do a two-stage composite process... have a steel chassis set up, then have the 3D printer do its "printing" around and encapsulate the frame in the design completely. Done right, it would be similar to rebar in concrete, producing both the shape wanted, but with far more strength than the plastic alone.
Not just an open filesystem, but a LVM layer that has encryption built in. Of course, the ideal would be everyone moving to ZFS, but it would be nice to at least have a common filesystem and disk level encryption standard across platforms... preferably a FS that was made this century.
I have an Apricorn drive that handles the USB password entry with a keypad, and uses the PIN to unlock a 128 bit AES key that is randomly generated.
Should I want to erase all contents, I plug the device in with the "cancel" button in, watch for the flashing lights, then hold down "cancel" + "2" + "unlock" for ten seconds... and it will generate a new key, render all data inaccessible on it, and use the password 123456 until that gets changed.
Zero software needed in Windows whatsoever to unlock it.
Just like the parent, I like the idea of a drive performing its own authentication separate from the computer, but this isn't new territory.
Very true. However, it can be difficult to change core app stores on a device unless one "cooks" a ROM that can handle both stores from the ground up, and I don't know of any ROM that can do this.
The alternative to a two-tier system would be a ROM that had Amazon as the default, with Google's store as an alternative... but with a warning to make sure to understand what permissions are... and to read app reviews.
I will state this. Before the FCC allowed lots of stations to be owned by one player, one could always listen and end up hearing new music. After the buyouts, most stations just play the same 10-50 songs. It is a negative feedback loop because it makes local stations irrelevant. The same content can be obtained from a satellite radio station.
These days, it is hard to find one "hangout" where people can hear something new.
Maybe I'm getting old... I remember days where the people who bragged about listening to bands nobody else have heard about were mainly the club DJs.
I'd add a third variable into it, as a devil's advocate:
Music industry -- doing well. Recording industry -- meh. Artists -- dead.
Because the recording industry is not fairing well, they have only focused on markets which give them revenue, which tends to be teens/tweens. This is why we are only seeing pop acts like the Justin Beibers promoted compared to Kurt Cobains or acts that might define/push a genre outwards. Add to this radio, where most "rock" stations are living in a time warp ending around 1995, and it is impossible for an artist to "make it big" these days with a band.
There is also the fact that if people's subscriptions are cut, it will get people to do what ISPs fear, and that is to hit the VPN services. What CAS appears to do go after the "casual" user. However, if their account is on the line, they will be joining the ranks who are using encrypted tunnels for virtually all their communications..
Once VPNs become a standard thing (just like antivirus programs), ISPs now lose all passive tools, and are forced to act in an active manner. I wouldn't be surprised to see some of the following as the next step:
1: VPNs throttled or randomly disconnected. Throttling can be undetectable and it becomes a word against word argument with a quite uphill battle against the subscriber.
2: Active prohibition of VPNs who don't log, with consequences to accounts due to TOS violations.
3: A ban of VPNs altogether (Pakistan is doing this.)
4: A requirement for subscriber machines to pass a "healthcheck" before being allowed on the Net. This "healthcheck" would require software that prohibits IP forwarding/masquerading/connections to VPN addresses.
5: Similar to #4, except with mandated "anti-pirate" software. Picture an antivirus scanner, except for detecting VPN software, cracks, patches, P2P apps, and proxy software, shutting the computer down and phoning home if it is found.
6: TPMs and trusted boot paths to make #4 and #5 work better, with automatic permanent blocking of the machine off the Net if tampered with.
7: Another round of DMCA laws to support #4-6, and to further keep #1 from being provably detected.
You hit the nail on the head. I always end up rooting my Android devices to install Titanium Backup, but for a lot of people, rooting can be a very daunting task, so backups become an issue.
What would be nice from the viewpoint of an average user would be a utility that can do a nandroid type of backup, saving everything (apps, data, etc.) This image optionally could be encrypted. Restores could be done either completely (apps + data + market links), or partially (restoring single or multiple apps.) Another option that would be nice would be the ability to archive apps, so if someone is finished playing a very large game, it can be archived with the save game data, freeing up space on the device.
Then, there is music syncing. This is painless in the iOS world (buy a song or album on one device, it can propagate to the others), but doing something like copying a music library, except down-coded to 192k is very difficult on Android. The only similar thing in Android is Doubletwist's solution.
Finally, one thing iOS devices have that Android doesn't is a tight integration to a cloud provider. I can move pictures off the main camera roll to a "vault" app, which then moves the pictures to iCloud in the background. It would be nice to have similar functionality on Android, perhaps an API that would allow one to choose which cloud backend to use for each app.
You hit the nail on the head. I wish Google would split the Play Store into two tiers, one tier being the current way things are done, and another tier where rules are far stricter, and Google can flick things off that tier as they see fit.
Couple this with an option on the phone warning users about leaving the "trusted" area, and this would provide similar protection that Apple's gatekeeper does, and close one of the biggest perceived holes in Android.
The problem that this fixes is that there are users who assume that if it is on a store, that it is safe to download. Apple does a good job at keeping this perception. On Android, users will download an app, not even bothering looking at permissions, and then get stung. By having a tier of "want the app here, you do it our way", Google keeps the users who don't care about permissions from shooting themselves in the foot, while users with some knowledge can flip to the other tier and find a larger selection of stuff, check perms and reviews, and go from there.
Slashdot Mirror
Part of the reason why there is a move to bigger phones is engineering issues. The faster CPUs require more surface area to dissipate heat. So, the product designers might as well offer a larger screen size.
I will say that AIX is pretty good as well. In general, unless there is a show-stopper patch, or one installs a driver like EMC PowerPath that requires a reboot due to the hooks in the kernel, one can keep AIX up for a long while, only really bothering to update and reboot when the latest tech level is released, and if there are no security specific issues, even that can be ignored, although it is wise to keep up on new firmware stuff just in case.
Sorta
Yes
Yes
No
Depends. Droidwall helps a lot. So did LBE Privacy Guard, but it hasn't been updated in ages, and doesn't support 4.1 or 4.2.
Use CyanogenMod or a custom ROM.
Depends on app.
I'm guessing they got complaints, so had to do something in order to keep the ad providers happy, as well as the developers who get revenue from admob and other providers. Had Google not done at least a token effort, there are plenty of other places the app writers can go (Tizen is waiting in the wings, Windows 8 is extremely mature, and there is always iOS where Apple can almost guarantee people will see ads slung at them.)
The nice thing about Android is that if Google blocks the apps on their store, there is always sideloading. I can use a QR code or click a link, and go from there.
I don't think Google is really evil, except they are responding to pressure put on them, as the consequences for not doing something (developers and possibly phone makers taking their business elsewhere) would be a lot worse than having users tick a check box and grab their goodies from a different source.
Security can be done right. It is a matter of "won't", rather than "can't" in general. A lot of companies have the "security has no ROI" attitude, but if it is something that their names are on the line for, they might actually follow working practices.
Take the Mercedes DAS (drive authorization system) keys/remotes. Very few, if any, people are bypassing the anti-theft system on these vehicles. So far, there are key cloners out there at best. If there is actual money and lives on the line, even this would be much improved so "hackcidents" are something that is of fiction rather than fact.
My assumption is that self-driving cars will have actual decent security, good enough to keep a tweaker from commandeering someone's vehicle to pick up some lab "product".
Bingo. Self-driving cars would be very useful in the US for a number of reasons:
1: The passive servicing aspect. Aft night, the vehicle can haul itself to a garage for overnight servicing, and I'm sure 24 hour service stations/car dealers would sprout up for this demand. Having a car get serviced while asleep would save hours of time.
2: The ability to set up shops which are essentially reverse delivery drivers. Car drives to window, order is placed in vehicle, car drives away. This makes it possible to order a pizza from an exotic spot that would be out of the town's delivery radius. Or, pick up a part from another town.
3: Extra hours of sleep. With commutes in the hours, might as well let the car do the dirty work.
4: No need for cloverleaf intersections. Communication between stationary computers and vehicles would allow cars to drive full speed through 4-way intersections with timing no human could ever accomplish.
5: Actual respect for driving laws. A cyclist or pedestrian presses a button to cross a road, the cars -will- stop and allow them by.
6: Cars can park in a garage or parking lot by themselves.
7: Far fewer accidents (thus more people living.)
8: Better fuel economy in general. Fewer panic stops.
This is something I wish Moto would continue developing. Create a lapdock with a SD card slot for backups, allow people to have their own Linux distro, and it would be very useful as a computer for emergencies.
If Moto made the lapdock usable between phone generations, it would make a perfect place for backups, possibly migrating apps and data to a new device come upgrade time.
My last Moto phone, the Atrix 2, had great engineering, but because of the bootloader issue, and the fact that it took a good long while (about six months IIRC) just to get a flashable factory ROM (FXZ format) in case a rooting went bad.
Unless Moto gets with the competition, I'm going to patronize other places that don't mind allowing people to use their purchased device as they see fit. My HTC 1X+ does not have anywhere near the voice and Bluetooth quality that either my iPhone 5 or Atrix 2 does... but it has a custom ROM and works decently. If Motorola allowed bootloader unlocks, I'd definitely purchase another model from them. However, it looks like my next device will likely be a Galaxy 4 when it gets released.
Devil's advocate:
Motorola has some things they are good at. For example, their radios tend to be top of the line for call quality in my experience, at least on par with iPhones.
MM's problems are not impossible to solve. One key that Motorola has advertised is enterprise-level friendliness and security. Google could run with that to help get their devices more entrenched into businesses.
After that, choose a niche for MM. Do they compete on the low-end with Huawei and ZTE, the high-end against Samsung and Apple, or midrange offerings? I'd say take the high/middle range if it were me.
I'd like to see Google get Motorola to make devices, not one for each provider, but devices that are either CDMA/LTE, GSM/LTE, or perhaps best of all, both. That way, it would be less about having rudimentary support for different phones, but being able to focus on a few devices at a time for ROM updates. It would be nice to see Android updates for a phone that would go past the 6-12 month mark.
Of course, doing like HTC and having source for the drivers and other parts of the ROM would be nice, perhaps even work with Cyanogen so their ROM is easily ported with full functionality.
If it were up to me, I'd be seeing about more of the computer-replacement technology that appeared in the Atrix, the Atrix 2, and other devices before it was axed. Couple that with a USB port so a keyboard and mouse could be attached, and this could function as a terminal for Citrix, RDP, or ssh if need be.
Of course, "fastboot oem unlock" would be on all devices, as well as a method of re-flashing ROMs that don't require a special program for FXZ, SHX or other files.
Devil's advocate:
One of the reasons I am guessing that Google is being conservative with the Motorola division is the concern about being viewed as a monopoly. If Google does too much with MM, other Android makers (Samsung, Huawei, ZTE) will jump ship for other operating systems like Windows 8.
There is also the fact that there is the fear of being viewed as a monopoly by the EU.
Regardless, it would be nice if Motorola would do like what Sony, HTC, and others offer, and give a way to unlock the bootloader. I like Motorola phones, but I won't buy another one unless there is a way to do this.
There is a difference. The early change-muncher games actually had replayability and took some tactics to win.
Modern games, a FPS is a FPS, and a paid-for DLC rocket launcher is a rocket launcher, be it on the a Ringworld clone world, a misty abandoned town, or some other no-mans land.
The annoying thing is how iOS and Android games changed from 2010 to now. In 2010, 99 cents got you a decently playable game, such as some tower defense game. These days, the same tower defense game is tuned so you have to buy the additional powerups, the damage bonuses, the bonus XP, and other items, so a .99 cent game becomes a $15.00 game for virtually the same amount of content.
What might be an idea would be client-side HTML5 code to do the MD5/SHA/bcrypt hashing [1] on the local machine before sending it up. This doesn't get rid of the need for SSL, but it does ensure that the original passphrase is extremely hard to obtain.
If for some reason a passphrase has to be stored for some reason (perhaps it needs retrieval intact), then there is a fairly easy way to deal with that: Have a clientside app that takes the password plus a salt and encrypts it with the site's public key [2]. That way, should the key have to be recovered, it still can be, but nowhere except the client's computer does the plaintext exist without someone explicitly decrypting the encoded text.
[1]: With a salt sent from the server.
[2]: Assuming the private key part is stashed in a HSM and is of a decent length (8192 bits, ideally 32768 bits.)
I wonder if they can do a two-stage composite process... have a steel chassis set up, then have the 3D printer do its "printing" around and encapsulate the frame in the design completely. Done right, it would be similar to rebar in concrete, producing both the shape wanted, but with far more strength than the plastic alone.
Not just an open filesystem, but a LVM layer that has encryption built in. Of course, the ideal would be everyone moving to ZFS, but it would be nice to at least have a common filesystem and disk level encryption standard across platforms... preferably a FS that was made this century.
I have an Apricorn drive that handles the USB password entry with a keypad, and uses the PIN to unlock a 128 bit AES key that is randomly generated.
Should I want to erase all contents, I plug the device in with the "cancel" button in, watch for the flashing lights, then hold down "cancel" + "2" + "unlock" for ten seconds... and it will generate a new key, render all data inaccessible on it, and use the password 123456 until that gets changed.
Zero software needed in Windows whatsoever to unlock it.
Just like the parent, I like the idea of a drive performing its own authentication separate from the computer, but this isn't new territory.
Very true. However, it can be difficult to change core app stores on a device unless one "cooks" a ROM that can handle both stores from the ground up, and I don't know of any ROM that can do this.
The alternative to a two-tier system would be a ROM that had Amazon as the default, with Google's store as an alternative... but with a warning to make sure to understand what permissions are... and to read app reviews.
Very true. For most of the people reading this, there is the above mentioned, plus Pandora and Last.fm. However, clued /. people != Joe Sixpack.
For most of the populace, FM radio is still a player.
I will state this. Before the FCC allowed lots of stations to be owned by one player, one could always listen and end up hearing new music. After the buyouts, most stations just play the same 10-50 songs. It is a negative feedback loop because it makes local stations irrelevant. The same content can be obtained from a satellite radio station.
These days, it is hard to find one "hangout" where people can hear something new.
Maybe I'm getting old... I remember days where the people who bragged about listening to bands nobody else have heard about were mainly the club DJs.
I'd add a third variable into it, as a devil's advocate:
Music industry -- doing well.
Recording industry -- meh.
Artists -- dead.
Because the recording industry is not fairing well, they have only focused on markets which give them revenue, which tends to be teens/tweens. This is why we are only seeing pop acts like the Justin Beibers promoted compared to Kurt Cobains or acts that might define/push a genre outwards. Add to this radio, where most "rock" stations are living in a time warp ending around 1995, and it is impossible for an artist to "make it big" these days with a band.
That is the past. What they dream of is:
1: Paying for the media, which is DRM-locked to a device upon first use in a player.
2: Paying for each listen to tracks.
3: Paying for being transferred to another device.
4: Paying each year for an unlock key for the media for the locked device.
5: Paying extra for "DLC"-like ability to listen to the top songs on an album.
6: GPS device that charges if the player is playing in a public area.
7: Additional fees for playing music in more than one location in a house.
8: Additional fees for stereo, 5.1, higher quality, ability to use equalization, ability to use monitors or better speakers, or playing in a vehicle.
9: Additional fees if more than one person is in the area where the device is playing.
10: Fees to copy the tracks to and from a device.
I'm sure there are a lot more, but with DRM and devices having hardware copy-protection stacks, this all could be a reality very quickly.
What will happen is that "business" accounts won't be under those rules, just like business DSL and cable lines are not under the CAS.
There is also the fact that if people's subscriptions are cut, it will get people to do what ISPs fear, and that is to hit the VPN services. What CAS appears to do go after the "casual" user. However, if their account is on the line, they will be joining the ranks who are using encrypted tunnels for virtually all their communications..
Once VPNs become a standard thing (just like antivirus programs), ISPs now lose all passive tools, and are forced to act in an active manner. I wouldn't be surprised to see some of the following as the next step:
1: VPNs throttled or randomly disconnected. Throttling can be undetectable and it becomes a word against word argument with a quite uphill battle against the subscriber.
2: Active prohibition of VPNs who don't log, with consequences to accounts due to TOS violations.
3: A ban of VPNs altogether (Pakistan is doing this.)
4: A requirement for subscriber machines to pass a "healthcheck" before being allowed on the Net. This "healthcheck" would require software that prohibits IP forwarding/masquerading/connections to VPN addresses.
5: Similar to #4, except with mandated "anti-pirate" software. Picture an antivirus scanner, except for detecting VPN software, cracks, patches, P2P apps, and proxy software, shutting the computer down and phoning home if it is found.
6: TPMs and trusted boot paths to make #4 and #5 work better, with automatic permanent blocking of the machine off the Net if tampered with.
7: Another round of DMCA laws to support #4-6, and to further keep #1 from being provably detected.
You hit the nail on the head. I always end up rooting my Android devices to install Titanium Backup, but for a lot of people, rooting can be a very daunting task, so backups become an issue.
What would be nice from the viewpoint of an average user would be a utility that can do a nandroid type of backup, saving everything (apps, data, etc.) This image optionally could be encrypted. Restores could be done either completely (apps + data + market links), or partially (restoring single or multiple apps.) Another option that would be nice would be the ability to archive apps, so if someone is finished playing a very large game, it can be archived with the save game data, freeing up space on the device.
Then, there is music syncing. This is painless in the iOS world (buy a song or album on one device, it can propagate to the others), but doing something like copying a music library, except down-coded to 192k is very difficult on Android. The only similar thing in Android is Doubletwist's solution.
Finally, one thing iOS devices have that Android doesn't is a tight integration to a cloud provider. I can move pictures off the main camera roll to a "vault" app, which then moves the pictures to iCloud in the background. It would be nice to have similar functionality on Android, perhaps an API that would allow one to choose which cloud backend to use for each app.
You hit the nail on the head. I wish Google would split the Play Store into two tiers, one tier being the current way things are done, and another tier where rules are far stricter, and Google can flick things off that tier as they see fit.
Couple this with an option on the phone warning users about leaving the "trusted" area, and this would provide similar protection that Apple's gatekeeper does, and close one of the biggest perceived holes in Android.
The problem that this fixes is that there are users who assume that if it is on a store, that it is safe to download. Apple does a good job at keeping this perception. On Android, users will download an app, not even bothering looking at permissions, and then get stung. By having a tier of "want the app here, you do it our way", Google keeps the users who don't care about permissions from shooting themselves in the foot, while users with some knowledge can flip to the other tier and find a larger selection of stuff, check perms and reviews, and go from there.