Here in Austin, unless seeing a matinee, it is going to be well above $10. One theater has tickets of $30 per person.
I am sure that people will pay a couple bucks more to not wade through commercials. Just look at how well the Alamo Drafthouse is doing with their business model over other theaters, not to mention the customer loyalty.
+1 for the Alamo Drafthouse. If it doesn't show there, I don't bother seeing it. Especially the little things, such as being able to duck under a table and zoom out the aisle if someone needs to use the restroom (as opposed to stomping on people's feet at other places), getting real food, real beer (not Bud Light, but actual stuff worth drinking), and ticket prices actually cheaper than the other corporate chains.
I have yet to try the Gold Class Cinemas which just opened up a theater, but there is just no competition to the Alamo as of now.
This is only going to get worse. Even if the website has a privacy guarantee, there are plenty of other parties that can steal private info and sling ads with potential browser exploits:
The ad serving company. If it pays them to have blackhats put on exploits randomly, they will do it, assuming they don't get caught.
Ad clients who have their own servers.
ISPs using Phorm-esque MITM appliances to modify the connection in flight. This is a very lucrative thing for ISPs, because they can not just replace ads, but capture click-throughs and other data. And since there are zero laws preventing this, queries that were typed in on one webpage can be redirected to another.
Moral: A legal solution won't be happening anytime soon.
Expectations of privacy. Ad companies want people to think that anything on the Web is public for any to see. This way, in court cases, the judge just finds the ad company not guilty because the plaintiff has no reasonable expectation of privacy on the Web, as per what other people think.
The only good thing is that there are so many sheep out there that ad companies are not reacting to people who use AdBlock or other means of security. However, I'm sure it will only be a matter of time before most sites become like Hulu and offer anything other than a "Go away" message on the front page. Of course, the next step will be add-ons or Greasemonkey scripts that fake cookies and identifying info, and the arms race will go on.
Time to register your copy of sandboxie and VMWare Workstation. I'm sure we will get more pretty feel good crap like this coupled with more invasive shared objects stored by add-ons in our future.
It is just advertising PR. It also is a wakeup call to the more technologically adept people that they are far better served by using products like AdBlock, NoScript, Sandboxie, BetterPrivacy, and other items as opposed to thinking that the website or its advertisers will do anything else but sell any and every bit of info they get to the highest bidder.
Of course, the opt out button is there, but is the private info still present on their servers? Even in saved database snapshots? All this information can be easily mined and used for profit the second that advertising company gets bought out, or declares bankruptcy, and there is nothing a person can do about it.
Didn't we go through this before with the TRUSTe logo of showing if the site only used the information in-house versus sharing with others?
This just seems like more feel good PR fluff, like the P3P stuff about a decade ago. We don't need more "assurances" about privacy. We need the data not to be collected in the first place. No Flash shared objects. No shared objects in Quicktime or other add-ons. No using tricks in a browser to "personalize/individualize" content.
What would be great is some service that you drop in a torrent link, upload a.torrent file, or pass it a magnet link, it does the packet passing, seeds for 2-3x as long as the download (for courtesy reasons), then offers you a URL to grab the file when done.
This right here would solve a lot of BT's issues and make money for an enterprising soul with spare disk space and bandwidth. The file can even be cached so if multiple people pass the same torrent, they instantly get the file.
For the love of $DEITY, don't use Tor for P2P. It is not designed for it, and it only gets people to throttle or even just yank their exit node off when they get handed a whopping bandwidth bill from their ISP.
Tor was designed for people browsing the Web, not slamming the chain of nodes with traffic so someone can grab the latest Justin Bieber release.
There are other alternatives. Anonymous proxies come to mind, although it is an exercise left up to the reader to find one that doesn't log/report your activities to whomever comes a calling. Renting a VM from a VPS might be an answer as well.
This is mitigated somewhat because Blizz drops the large download of new models a week or two before the patch. This way, eventually it gets downloaded over a number of days.
However, I really wish Blizz just had a.torrent available without having to rip it out of the executable. That, and the availability of a complete (not a diff) patch so a reinstall only requires one update. Thankfully WoW doesn't require any real Registry entries, so if it does get corrupted, I can just restore the directory from a previous version or run the Repair tool to search and fix anything scrozzled.
This is a nasty feedback loop. I can see laws being passed in the UK out of knee-jerk reaction which will make life worse for every UK citizen. Perhaps nationwide NAC forcing people to type in their ID number before they get the ability to send packets out? Couple this with large prison sentences for tampering/removing the software.
There are universities which not just block traffic, but use NAC to force software to be installed on any computers connected. This software not just blocks creation of VPNs and port forwardings, but allows for scanning for software and documents at anytime by the upstream provider.
I can see this on a nationwide level. Green Dam may not have worked well, but neither did the first generation of spyware. Things improve, and I can see lawmakers mandating a "lawful compliance" package which not just adds more spying capabilities, but locks out all non mainstream operating systems. All "for the children" or "for national security", of course.
That is when people get an anonymous VPN account. I don't care about movies or warez, but I like having the VPN available for a couple reasons. The first is so I don't have to worry about geolocation trackers. If I wanted a site to have where I live, I will happily enter in a ZIP code.
Another way to have the advantages of P2P is to rent a VM from a VPS or cloud hosting company. Make sure of the fees because some will charge you a lot of cash for relatively small amounts of bandwidth and/or CPU time. This way, you do your P2P stuff in the VM that sits on a top tier Internet connection, sftp the downloaded files back, and assuming the bandwidth doesn't get in the way, you can keep seeding without issue.
You just summed up one of MS's great strengths -- availability of expertise for the platform.
Take Exchange. If you get stuck in any phase of Exchange use, be it planning, deployment, expanding, security, backups, archive, availability, or failover, one can find consultants and books with relative ease. If someone needs an Exchange server for home, that is quite easy to do. Similar if someone wants a hub/edge configuration with incoming mail, outgoing mail, OWA, mobile device, and on the inside, multiple mail hubs for redundancy.
This doesn't mean Exchange is the be all and end all for messaging. It just means that being able to get Exchange working is easy for a lot of businesses. Perhaps Domino might be a solution, but trying to find the Notes experience is significantly more difficult than Exchange brainpower.
There is no magic bullet. For a small company with 5 people that want E-mail hosted securely, a PC with RedHat Enterprise Linux and POP/SMTP/IMAP might be the solution of choice. For another SMB, a machine running Microsoft's SBS might be the answer. Still another SMB might just use a hosted Exchange provider so they do not have to bother with an always on network connection.
I worry though... Minnesota pretty much jumped off the diving board and it seems that they didn't even check if there is water below them. If they were moving to a new platform, there are plenty of other options to explore on the spectrum before just going whole-hog with a relatively nascent technology architecture.
You also can't look at the piece from different angles to see the raised brush strokes.
Digital photography has come a long way. Higher end cameras have more than 8 bits for colors, megapixels are getting to a reasonable amount for even 11x17 prints viewed close up.
However, nuances are still lost, and it probably will take a major advance in cameras, perhaps with a lens that moves around and takes a number of snapshots from slightly different angles, then uses the pictures to reconstruct as close to a gestalt of the image as possible.
And RedHat doesn't have legal representation? I'm sure if a patent troll started hammering RedHat, other people in the supply chain whose futures are tied to the OS will join the fray, such as IBM.
Personally, I'd worry about other factors such as if the OS and platform are up to the task at hand. This generally is far more important than worrying about patent trolls as the primary reason to choose an application stack.
What would be ideal is functionality that KeePass has. It has the option to scale the amount of rounds to one second of your hardware's CPU time with the ability to edit the rounds up and down to preference. For BB users who don't want this detail, this can be a semi-hidden option and the device can compute how many rounds it does to suck up a second or two of CPU times automatically.
It is understandable why TrueCrypt doesn't do this (because it has to guess a number of times with various combinations of hashes, algorithm combinations, and header variations before it can mount a volume), but for something that it doesn't matter if it is obviously encrypted (where it can have an obvious header), this should be an option available.
It is still a hole though, and one that is completely preventable. Most serious crypto products around uses key strengthening, be it KeePass with its variable number of rounds that are user selectable, TrueCrypt with its 1000 rounds, or iOS 4's 10,000 rounds. Heck, even the venerable crypt(3) mechanism had a number of rounds to slow down people running Crack over 20 years ago back before passwords were stored in/etc/shadow.
How can this be fixed? Use a reasonable amount of rounds (enough so it slows down brute forcing, but not too many that it kills day to day normal operation.) Also, use a salt, so rainbow table pre-computation of keys is impossible.
In the meantime, the parent poster probably has the best solution. For maximum security, add a cryptographic token and store a TC keyfile on that. This way, if someone tries to brute force the token's passphrase, they have 3-20 tries before the token permanently fries itself.
Brick and mortar stores are not relevant to this discussion.
Say I have an app I am actively developing and maintaining. One central distribution mechanism means that I can focus my time on writing and updating it, upload the.apk file, and go on my merry way. I eyeball the reviews, and make sure to take heed of any constructive critiques for the next revision of the app.
With multiple app stores, each of which has different rules, each of which are present or absent on different phones, in order to have my app available to as many users, I have to jump through every store's hoops. I also have to pay each store's ticket to entry. Google's store is very reasonable, just pay your $25 and you can play. However, with other stores in the mix, they can set prices any way they feel like. They can also set many restrictive conditions.
Want to know where the shit will really hit the fan? When stores demand exclusivity. If store "A" demands I only can use them, then any Android device that ships with store "B" and only store "B" on their device, my app is locked out of that market. This definitely will fragment Android far worse than it is now.
Don't forget that as of now, one can sideload and install via ADB on almost all devices. However, both of those abilities can easily be removed in a new model of phone forcing people to either get their apps from the store or do without.
Of course, there is the slippery slope: What happens if cellular carriers want to hop in this pool? More stores are not better in this choice, because I'm sure some carriers would only allow access to their specific store and no others.
The Amazon store brings some walled garden stuff, but my fear is that cell phone makers and providers will end up locking phones only to Amazon's store. Of course, they can be rooted so one could use the Google App Store.
In any case, this isn't a boon for developers, mainly because they now have at least two places they must keep their apps updated in.
It doesn't warn you, but I do recommend an app for rooted phones called DroidWall. I use this for apps which have zero need to be connecting to the Internet.
I saw earlier today on another news site a post about something similar saying that no OS commercially made can support more than 32 cores.
One of the followup postings was someone with an IBM 780 doing a prtconf|grep proc and showing 64 virtual processors on an LPAR. AIX supports up to 256 CPUs (physical or virtual.) I'm sure Solaris can do similar without breaking a sweat.
Actually neither. Its just that to an observer like me, FB is trying to reinvent the wheel on a problem that already has been solved.
Obviously, IBM is not cheap. Nor is Oracle/Sun hardware. However, the time and money spent developing a large scale framework on the application layer is not a trivial expense either. It might be that the time FB puts in trying to deploy something uncharted like this may cost them more in the long run.
Very true. However, by moving the redundancy to the top of the application stack, doesn't that become inefficient after a while?
For example, I have an application that sits on a mainframe on a big LPAR with the database server on another LPAR on the same machine. Because the redundancy is handled by everything below it, the application can be a lot simpler, with fewer bugs. It does not have to worry about being consistent among a number of boxes, just run and let the rest of the stack below it do
On a more traditional non-mainframe platform such as Solaris or AIX, the whole stack shares in the high availability. The machines have a heartbeat monitor, the database is bifurcated and knows when to let the secondary machine take over, the OS has multiple IO paths, and the application either doesn't care, or it is able to know when to spin up on the secondary hardware and when to spin back down, letting the primary machine take the lead. (This is assuming an active/passive failover scenario, of course.)
With only the application handling the redundancy, it is up to the custom application writers to handle consistency issues, deal with dead servers by failing away and failing back to the hardware when it comes up. Of course, this means that more hardware has to be thrown at the application. There is a point with parallelism of diminishing returns, but this depends on what the app is doing. It also means that a lot of additional functionality has to be coded in at the app layer such as when to consider a machine or a data center failed and to shunt to another, when to fail back, and so on.
This doesn't mean that Big Iron is the only solution out there by any means. However, it means that if anything fails in the application, there will be a world of hurt, as opposed to a failure in a piece of the stack of a more traditional system being able to be worked around by other means.
Of course, it would get misused. You will have patrolmen seeing people in supermarkets texting, then pulling them over as soon as they drive out with a ticket for texting because the time was so close. However, if it is another revenue stream for municipalities looking for cash and can't raise taxes, I'm sure they will try to squeeze cellular providers into providing them a method for real time access to call logs by staff. Very simple method of trade -- no real time access, no right of way for new LTE towers or wiring. Fines and seized vehicles are a great income source, so as time goes on, we see the laws for the average person becoming more and more complicated, and the penalties for infractions get greater and greater. This isn't to protect the common citizen, it is to keep milking a cash cow.
Texting sucks and people doing that are clear and present dangers on the road, but there is a point where too much enforcement only benefits the attorneys in the justice system, and is of little to no benefit for citizens.
That is a good point, but to use a car analogy, isn't it like strapping a ton of motorcycles together with duct tape and having people on staff to keep them all maintained so the contrivance can pull a 18-wheeler load? Why not just buy an 18-wheeler which is designed and built from the ground up for this exact task?
Yes, you have to use the 18-wheeler's shipping crates (to continue the analogy), but even with the vendor lock-in, it might be a lot better to do this as opposed to trying to cobble a suboptimal solution that does work, but takes a lot more man-hours, electricity, and hardware maintaining as opposed to something built from the factory for the task at hand.
Plus, zSeries machines and pSeries boxes happily run Linux LPARs. That is as open as you can get. It isn't like it would be moving the backend to CICS.
What we might start seeing are servers using multiple ports for one IP and incoming NAT, perhaps with a hack to DNS (perhaps another record) to support ports as well as IPs.
This would mean that if someone wanted to go to www.foo.com, DNS would hand them not just 9.0.0.1, but the port number, so www.foo.com would resolve to 9.0.0.1:80, and www.bar.com would resolve to 9.0.0.1:8000
Slashdot Mirror
Here in Austin, unless seeing a matinee, it is going to be well above $10. One theater has tickets of $30 per person.
I am sure that people will pay a couple bucks more to not wade through commercials. Just look at how well the Alamo Drafthouse is doing with their business model over other theaters, not to mention the customer loyalty.
+1 for the Alamo Drafthouse. If it doesn't show there, I don't bother seeing it. Especially the little things, such as being able to duck under a table and zoom out the aisle if someone needs to use the restroom (as opposed to stomping on people's feet at other places), getting real food, real beer (not Bud Light, but actual stuff worth drinking), and ticket prices actually cheaper than the other corporate chains.
I have yet to try the Gold Class Cinemas which just opened up a theater, but there is just no competition to the Alamo as of now.
This is only going to get worse. Even if the website has a privacy guarantee, there are plenty of other parties that can steal private info and sling ads with potential browser exploits:
The ad serving company. If it pays them to have blackhats put on exploits randomly, they will do it, assuming they don't get caught.
Ad clients who have their own servers.
ISPs using Phorm-esque MITM appliances to modify the connection in flight. This is a very lucrative thing for ISPs, because they can not just replace ads, but capture click-throughs and other data. And since there are zero laws preventing this, queries that were typed in on one webpage can be redirected to another.
Moral: A legal solution won't be happening anytime soon.
Expectations of privacy. Ad companies want people to think that anything on the Web is public for any to see. This way, in court cases, the judge just finds the ad company not guilty because the plaintiff has no reasonable expectation of privacy on the Web, as per what other people think.
The only good thing is that there are so many sheep out there that ad companies are not reacting to people who use AdBlock or other means of security. However, I'm sure it will only be a matter of time before most sites become like Hulu and offer anything other than a "Go away" message on the front page. Of course, the next step will be add-ons or Greasemonkey scripts that fake cookies and identifying info, and the arms race will go on.
Time to register your copy of sandboxie and VMWare Workstation. I'm sure we will get more pretty feel good crap like this coupled with more invasive shared objects stored by add-ons in our future.
It is just advertising PR. It also is a wakeup call to the more technologically adept people that they are far better served by using products like AdBlock, NoScript, Sandboxie, BetterPrivacy, and other items as opposed to thinking that the website or its advertisers will do anything else but sell any and every bit of info they get to the highest bidder.
Of course, the opt out button is there, but is the private info still present on their servers? Even in saved database snapshots? All this information can be easily mined and used for profit the second that advertising company gets bought out, or declares bankruptcy, and there is nothing a person can do about it.
Didn't we go through this before with the TRUSTe logo of showing if the site only used the information in-house versus sharing with others?
This just seems like more feel good PR fluff, like the P3P stuff about a decade ago. We don't need more "assurances" about privacy. We need the data not to be collected in the first place. No Flash shared objects. No shared objects in Quicktime or other add-ons. No using tricks in a browser to "personalize/individualize" content.
What would be great is some service that you drop in a torrent link, upload a .torrent file, or pass it a magnet link, it does the packet passing, seeds for 2-3x as long as the download (for courtesy reasons), then offers you a URL to grab the file when done.
This right here would solve a lot of BT's issues and make money for an enterprising soul with spare disk space and bandwidth. The file can even be cached so if multiple people pass the same torrent, they instantly get the file.
For the love of $DEITY, don't use Tor for P2P. It is not designed for it, and it only gets people to throttle or even just yank their exit node off when they get handed a whopping bandwidth bill from their ISP.
Tor was designed for people browsing the Web, not slamming the chain of nodes with traffic so someone can grab the latest Justin Bieber release.
There are other alternatives. Anonymous proxies come to mind, although it is an exercise left up to the reader to find one that doesn't log/report your activities to whomever comes a calling. Renting a VM from a VPS might be an answer as well.
This is mitigated somewhat because Blizz drops the large download of new models a week or two before the patch. This way, eventually it gets downloaded over a number of days.
However, I really wish Blizz just had a .torrent available without having to rip it out of the executable. That, and the availability of a complete (not a diff) patch so a reinstall only requires one update. Thankfully WoW doesn't require any real Registry entries, so if it does get corrupted, I can just restore the directory from a previous version or run the Repair tool to search and fix anything scrozzled.
This is a nasty feedback loop. I can see laws being passed in the UK out of knee-jerk reaction which will make life worse for every UK citizen. Perhaps nationwide NAC forcing people to type in their ID number before they get the ability to send packets out? Couple this with large prison sentences for tampering/removing the software.
There are universities which not just block traffic, but use NAC to force software to be installed on any computers connected. This software not just blocks creation of VPNs and port forwardings, but allows for scanning for software and documents at anytime by the upstream provider.
I can see this on a nationwide level. Green Dam may not have worked well, but neither did the first generation of spyware. Things improve, and I can see lawmakers mandating a "lawful compliance" package which not just adds more spying capabilities, but locks out all non mainstream operating systems. All "for the children" or "for national security", of course.
That is when people get an anonymous VPN account. I don't care about movies or warez, but I like having the VPN available for a couple reasons. The first is so I don't have to worry about geolocation trackers. If I wanted a site to have where I live, I will happily enter in a ZIP code.
Another way to have the advantages of P2P is to rent a VM from a VPS or cloud hosting company. Make sure of the fees because some will charge you a lot of cash for relatively small amounts of bandwidth and/or CPU time. This way, you do your P2P stuff in the VM that sits on a top tier Internet connection, sftp the downloaded files back, and assuming the bandwidth doesn't get in the way, you can keep seeding without issue.
You just summed up one of MS's great strengths -- availability of expertise for the platform.
Take Exchange. If you get stuck in any phase of Exchange use, be it planning, deployment, expanding, security, backups, archive, availability, or failover, one can find consultants and books with relative ease. If someone needs an Exchange server for home, that is quite easy to do. Similar if someone wants a hub/edge configuration with incoming mail, outgoing mail, OWA, mobile device, and on the inside, multiple mail hubs for redundancy.
This doesn't mean Exchange is the be all and end all for messaging. It just means that being able to get Exchange working is easy for a lot of businesses. Perhaps Domino might be a solution, but trying to find the Notes experience is significantly more difficult than Exchange brainpower.
There is no magic bullet. For a small company with 5 people that want E-mail hosted securely, a PC with RedHat Enterprise Linux and POP/SMTP/IMAP might be the solution of choice. For another SMB, a machine running Microsoft's SBS might be the answer. Still another SMB might just use a hosted Exchange provider so they do not have to bother with an always on network connection.
I worry though... Minnesota pretty much jumped off the diving board and it seems that they didn't even check if there is water below them. If they were moving to a new platform, there are plenty of other options to explore on the spectrum before just going whole-hog with a relatively nascent technology architecture.
You also can't look at the piece from different angles to see the raised brush strokes.
Digital photography has come a long way. Higher end cameras have more than 8 bits for colors, megapixels are getting to a reasonable amount for even 11x17 prints viewed close up.
However, nuances are still lost, and it probably will take a major advance in cameras, perhaps with a lens that moves around and takes a number of snapshots from slightly different angles, then uses the pictures to reconstruct as close to a gestalt of the image as possible.
And RedHat doesn't have legal representation? I'm sure if a patent troll started hammering RedHat, other people in the supply chain whose futures are tied to the OS will join the fray, such as IBM.
Personally, I'd worry about other factors such as if the OS and platform are up to the task at hand. This generally is far more important than worrying about patent trolls as the primary reason to choose an application stack.
What would be ideal is functionality that KeePass has. It has the option to scale the amount of rounds to one second of your hardware's CPU time with the ability to edit the rounds up and down to preference. For BB users who don't want this detail, this can be a semi-hidden option and the device can compute how many rounds it does to suck up a second or two of CPU times automatically.
It is understandable why TrueCrypt doesn't do this (because it has to guess a number of times with various combinations of hashes, algorithm combinations, and header variations before it can mount a volume), but for something that it doesn't matter if it is obviously encrypted (where it can have an obvious header), this should be an option available.
It is still a hole though, and one that is completely preventable. Most serious crypto products around uses key strengthening, be it KeePass with its variable number of rounds that are user selectable, TrueCrypt with its 1000 rounds, or iOS 4's 10,000 rounds. Heck, even the venerable crypt(3) mechanism had a number of rounds to slow down people running Crack over 20 years ago back before passwords were stored in /etc/shadow.
How can this be fixed? Use a reasonable amount of rounds (enough so it slows down brute forcing, but not too many that it kills day to day normal operation.) Also, use a salt, so rainbow table pre-computation of keys is impossible.
In the meantime, the parent poster probably has the best solution. For maximum security, add a cryptographic token and store a TC keyfile on that. This way, if someone tries to brute force the token's passphrase, they have 3-20 tries before the token permanently fries itself.
Brick and mortar stores are not relevant to this discussion.
Say I have an app I am actively developing and maintaining. One central distribution mechanism means that I can focus my time on writing and updating it, upload the .apk file, and go on my merry way. I eyeball the reviews, and make sure to take heed of any constructive critiques for the next revision of the app.
With multiple app stores, each of which has different rules, each of which are present or absent on different phones, in order to have my app available to as many users, I have to jump through every store's hoops. I also have to pay each store's ticket to entry. Google's store is very reasonable, just pay your $25 and you can play. However, with other stores in the mix, they can set prices any way they feel like. They can also set many restrictive conditions.
Want to know where the shit will really hit the fan? When stores demand exclusivity. If store "A" demands I only can use them, then any Android device that ships with store "B" and only store "B" on their device, my app is locked out of that market. This definitely will fragment Android far worse than it is now.
Don't forget that as of now, one can sideload and install via ADB on almost all devices. However, both of those abilities can easily be removed in a new model of phone forcing people to either get their apps from the store or do without.
Of course, there is the slippery slope: What happens if cellular carriers want to hop in this pool? More stores are not better in this choice, because I'm sure some carriers would only allow access to their specific store and no others.
Just what Android needs, more fragmentation.
The Amazon store brings some walled garden stuff, but my fear is that cell phone makers and providers will end up locking phones only to Amazon's store. Of course, they can be rooted so one could use the Google App Store.
In any case, this isn't a boon for developers, mainly because they now have at least two places they must keep their apps updated in.
It doesn't warn you, but I do recommend an app for rooted phones called DroidWall. I use this for apps which have zero need to be connecting to the Internet.
I saw earlier today on another news site a post about something similar saying that no OS commercially made can support more than 32 cores.
One of the followup postings was someone with an IBM 780 doing a prtconf|grep proc and showing 64 virtual processors on an LPAR. AIX supports up to 256 CPUs (physical or virtual.) I'm sure Solaris can do similar without breaking a sweat.
Actually neither. Its just that to an observer like me, FB is trying to reinvent the wheel on a problem that already has been solved.
Obviously, IBM is not cheap. Nor is Oracle/Sun hardware. However, the time and money spent developing a large scale framework on the application layer is not a trivial expense either. It might be that the time FB puts in trying to deploy something uncharted like this may cost them more in the long run.
Very true. However, by moving the redundancy to the top of the application stack, doesn't that become inefficient after a while?
For example, I have an application that sits on a mainframe on a big LPAR with the database server on another LPAR on the same machine. Because the redundancy is handled by everything below it, the application can be a lot simpler, with fewer bugs. It does not have to worry about being consistent among a number of boxes, just run and let the rest of the stack below it do
On a more traditional non-mainframe platform such as Solaris or AIX, the whole stack shares in the high availability. The machines have a heartbeat monitor, the database is bifurcated and knows when to let the secondary machine take over, the OS has multiple IO paths, and the application either doesn't care, or it is able to know when to spin up on the secondary hardware and when to spin back down, letting the primary machine take the lead. (This is assuming an active/passive failover scenario, of course.)
With only the application handling the redundancy, it is up to the custom application writers to handle consistency issues, deal with dead servers by failing away and failing back to the hardware when it comes up. Of course, this means that more hardware has to be thrown at the application. There is a point with parallelism of diminishing returns, but this depends on what the app is doing. It also means that a lot of additional functionality has to be coded in at the app layer such as when to consider a machine or a data center failed and to shunt to another, when to fail back, and so on.
This doesn't mean that Big Iron is the only solution out there by any means. However, it means that if anything fails in the application, there will be a world of hurt, as opposed to a failure in a piece of the stack of a more traditional system being able to be worked around by other means.
Of course, it would get misused. You will have patrolmen seeing people in supermarkets texting, then pulling them over as soon as they drive out with a ticket for texting because the time was so close. However, if it is another revenue stream for municipalities looking for cash and can't raise taxes, I'm sure they will try to squeeze cellular providers into providing them a method for real time access to call logs by staff. Very simple method of trade -- no real time access, no right of way for new LTE towers or wiring. Fines and seized vehicles are a great income source, so as time goes on, we see the laws for the average person becoming more and more complicated, and the penalties for infractions get greater and greater. This isn't to protect the common citizen, it is to keep milking a cash cow.
Texting sucks and people doing that are clear and present dangers on the road, but there is a point where too much enforcement only benefits the attorneys in the justice system, and is of little to no benefit for citizens.
That is a good point, but to use a car analogy, isn't it like strapping a ton of motorcycles together with duct tape and having people on staff to keep them all maintained so the contrivance can pull a 18-wheeler load? Why not just buy an 18-wheeler which is designed and built from the ground up for this exact task?
Yes, you have to use the 18-wheeler's shipping crates (to continue the analogy), but even with the vendor lock-in, it might be a lot better to do this as opposed to trying to cobble a suboptimal solution that does work, but takes a lot more man-hours, electricity, and hardware maintaining as opposed to something built from the factory for the task at hand.
Plus, zSeries machines and pSeries boxes happily run Linux LPARs. That is as open as you can get. It isn't like it would be moving the backend to CICS.
What we might start seeing are servers using multiple ports for one IP and incoming NAT, perhaps with a hack to DNS (perhaps another record) to support ports as well as IPs.
This would mean that if someone wanted to go to www.foo.com, DNS would hand them not just 9.0.0.1, but the port number, so www.foo.com would resolve to 9.0.0.1:80, and www.bar.com would resolve to 9.0.0.1:8000