Each side [...] has reclaimed 60% of the labor (unemployment). [...] With that additional money, the consumers can now buy more products
If you're unemployed, you can buy zero products. You try to cover that:
In short: the labor freed up from the farm is repurposed. [...] we make more doctors and stuff like Netflix and cell phone networks.
These are also more highly skilled professions, for which most of "the labor freed up from the farm" is likely unqualified. Who covers the cost of retraining?
I'd only use Java for Android and would stick with native Objective C for iOS, and keep as much of the logic in JSON RESTful webservices as I could get away with.
You won't be able to "get away with" much if an app needs to run offline, such as if it's for tablets that drift out of Wi-Fi often, phones whose data plan has been used up for the month, and devices carried on airplanes. You'll have to replicate most of the logic in an app that accesses a cached view of the data retrieved from your "JSON RESTful webservices".
Let me rephrase so as not to be "Offtopic": Has Google clarified whether "Official Android Support" for RPi includes Google Play Store or is only AOSP?
In the end its mostly the PBX end that needs upgrading if their excuse is valid, so it's not some insurmountable task.
No "PBX end" upgrade will keep speeds up if all subscribers run BitTorrent 24/7 over the air. There exist fundamental physical limits in how much throughput a given amount of spectrum and land for cell towers can support.
if your network capacity is limited you lower speeds
Cut speeds and add QoS favoring VoIP, and watch your revenue collapse as your customers flock to a competing carrier that advertises higher speeds for non-VoIP apps made possible by caps.
So because an ISP has declined to update their networks over the last 10 years
Or has attempted to do so for ten years but has been constrained by red tape that it can document.
they should be fined continuously until their network is upgraded.. if hat means no profit for the next 3 years TFB.
No profit for three years would result in a shareholder lawsuit. Would it be preferable to dissolve one of the major cellular carriers and sell its assets at auction to the other major carriers, resulting in yet another contraction of the U.S. cellular industry?
So why do metered cell phone plans measured in megabytes even still exist?
For the same reason that metered cell phone plans measured in minutes (rather than hours) exist: some carriers want to compete for the lowest-usage subscribers.
Aiming for sustained peak load simply means you statistically predict how much actual load you'll get for x customers at the busiest moments and you build for that +5% or +10% based on your datasource.
If an ISP lacks the finances at the moment to "build for that" throughout its service area, how else should it reduce peak load?
This isn't exactly rockey science, every self respecting engineer can do these calculations.
Engineers say one thing. Finance says another. An executive needs to balance the two, and it may come down to ordering the engineers to come up with a solution that reduces peak load. What such solution do you recommend?
Wirth's law may be blocking these devices from getting an upgrade to the next major version of Android if the device's RAM size or disk performance does not meet the minimum system requirements of the next major version. I can speak from experience that for some devices, though an update was issued, it never should have been. Android 4.4 "KitKat" on the Nexus 7 (2012) "grouper" tablet is fine; Android 5 "Lollipop" on the same hardware is a jank-fest, with the UI often freezing for five or ten seconds at a time. All reports that I've read imply that the jankiness is caused by ASUS cheaping out on slow NAND flash memory. Clearing the cache helps some but not all users in this situation. It's like the last version of iOS that Apple releases for a given iPhone model: its increased system demands often bog down the UI to the point where some users end up assuming it is an intentional measure to get people to replace otherwise working hardware.
PHP7 is a huge improvement over old versions. If you haven't looked at it, do so. PHP sucks in 2016 is mostly FUD now.
To see how much of an improvement PHP 7 is, let's check the 7 things wrong with PHP 5 that I mentioned in this article (and previously in a Slashdot comment):
Things such as switch use the byzantine semantics of operator ==
Not fixed to my knowledge.
Parse errors and undefined functions are fatal
PHP 7 makes them "engine exceptions" which are catchable Error instances.
Inconsistent naming and argument order conventions for standard library functions
Not fixed completely for back-compat reasons, but PHP 7 has started to drop some duplicate functionality.
Backwards associativity for operator ?:
Not fixed for back-compat reasons, but at least the new null coalesce operator ?? has its associativity on the more useful side.
Letting the server operator change program semantics in annoying ways
Safe mode and magic quotes were already gone in PHP 5.4. cURL would not follow redirects if open_basedir was set until September 2013. But upload_max_filesize still defaults to 2MB, which I see as unnecessarily small.
Semantics breakage in minor (5.x) versions encouraging shared hosting operators to delay making new versions available to subscribers
Not fixed until the majority of shared hosts offer PHP 7. Has someone run a survey about this?
No keyword arguments
Not fixed.
your network should be able to handle peak load without negative traffic routing interactions.
Allowing all subscribers to transmit at peak rate simultaneously "without negative traffic routing interactions" would require such an investment in capacity, including the purchase of land and equipment for cell towers, that most subscribers would consider the service unaffordable. For this reason, most ISPs oversell their networks under the assumption that not all subscribers will reach peak load simultaneously. Metering during congested periods is ostensibly a way to reduce peak load to the point where it can be "handle[d] without negative traffic routing interactions" as a way of delaying the need to invest in capacity. But I agree with you that metering during uncongested periods is a cash grab.
Currently data can be transferred on your phone by apps that may be running in the background.
Android logs the amount of data transferred by each app running on a device and lets the user sort a list of apps by decreasing data volume. Is this not enough?
Voice calls were never made without the customers knowledge
Almost never. (Search for butt dialing.)
It is not really feasible or helpful to keep logs of every IP address and size of every packet transferred to/from your phone.
So let me get this straight: Unlike satellite ISPs, cellular ISPs have failed to offer a reward for moving away from congested times of day, and you're rationalizing it as because the device communicates with more distinct destinations than when voice calls dominated cellular usage. Let me see if I can think of a way to interact with Internet data transfer accounting mechanisms already present in mobile operating systems.
Windows 8, Windows 10, and Android operating systems allow users to mark particular SSIDs as metered, and an application is expected to respect that by querying cost whenever it receives a notification of change in the device's Internet connection. I guess the real solution is for carriers to signal to a device when the connection has become unmetered or has become metered, so that applications can switch to more conservative download schedules.
Cats can be trained to use a toilet. Many prefer it because they no longer have to spend time burying their waste. The biggest drawback is you have to leave the lid up and the seat down, but that drawback wouldn't quite apply to hypothetical cat people.
Telecom customers (and probably producers as well) as a whole are just not savvy enough to deal with a complicated pricing system
Back before the majority of postpaid cellular plans came with unlimited voice minutes, a lot of cellular carriers and landline long-distance companies offered discounts on night and weekend calls. Are telecom customers less tech savvy than they used to be?
Networks are measured in maximum throughput over a given period of time, not total number bytes transferred to a given place.
Data caps measure each user's contribution to average throughput, which is (imperfectly) correlated to each user's contribution to peak throughput.
Data caps are a solution to the wrong problem.
Agreed. If the problem is peak throughput, the solution is to run the meter when the network is congested, that is, when the user's packets are passing through a segment at its maximum throughput. Satellite ISPs have the right idea: run the meter only during congested hours to encourage users to move their transfers away from times of maximum demand for throughput.
the "problem" they're really solving is that of people using the Internet the way it was intended--as a communications medium that doesn't really care if what's being communicated is voice, text, images, videos, etc.
That or the "problem" is one of using the Internet as a real-time communications medium, as opposed to an asynchronous application where usage can be shifted away from . People download multi-gigabyte HD movies and Steam games during prime time, contributing to dropped packets at times of maximum throughput, when they could be setting the computer to download them at 0-dark-thirty and not affecting the maximum.
You just need to pick the right combination of frameworks and programming language. Everything can be converted to everything else nowadays.
Unless the web application is a workaround for a platform owner's policy that specifically forbids a set of native apps. For example, applications doing any of these things cannot be converted to a native iOS application intended for distribution to the public.
Now there may be valid reasons to reimplement in another language (better matching your available talent pool, better performance, you would just prefer it)
Especially when the deployment platform dictates the choice of language. Examples include Xbox Live Indie Games and Windows Phone 7 requiring verifiably type-safe CIL for the.NET Compact Framework (in practice, requiring C#), or the web requiring either ECMAScript 5 or a language that transpiles to ECMAScript 5. Then "better matching your available talent pool" means "we would have to hire someone to write a transpiler to the only supported language", and "better performance" means "the alternative is writing an interpreter".
but if it's just to 'make it clean', that can pretty much always be accomplished within the existing language choice.
Even if "the existing language choice" is assembly language, as in the case of porting a classic video game to a modern platform? Can assembly language be made as clean as, say, Python?
Slashdot Mirror
Each side [...] has reclaimed 60% of the labor (unemployment). [...] With that additional money, the consumers can now buy more products
If you're unemployed, you can buy zero products. You try to cover that:
In short: the labor freed up from the farm is repurposed. [...] we make more doctors and stuff like Netflix and cell phone networks.
These are also more highly skilled professions, for which most of "the labor freed up from the farm" is likely unqualified. Who covers the cost of retraining?
traffic isn't constant
It'll become a lot closer to constant if people start running BitTorrent 24/7.
I'd only use Java for Android and would stick with native Objective C for iOS, and keep as much of the logic in JSON RESTful webservices as I could get away with.
You won't be able to "get away with" much if an app needs to run offline, such as if it's for tablets that drift out of Wi-Fi often, phones whose data plan has been used up for the month, and devices carried on airplanes. You'll have to replicate most of the logic in an app that accesses a cached view of the data retrieved from your "JSON RESTful webservices".
Let me rephrase so as not to be "Offtopic": Has Google clarified whether "Official Android Support" for RPi includes Google Play Store or is only AOSP?
In the end its mostly the PBX end that needs upgrading if their excuse is valid, so it's not some insurmountable task.
No "PBX end" upgrade will keep speeds up if all subscribers run BitTorrent 24/7 over the air. There exist fundamental physical limits in how much throughput a given amount of spectrum and land for cell towers can support.
if your network capacity is limited you lower speeds
Cut speeds and add QoS favoring VoIP, and watch your revenue collapse as your customers flock to a competing carrier that advertises higher speeds for non-VoIP apps made possible by caps.
So because an ISP has declined to update their networks over the last 10 years
Or has attempted to do so for ten years but has been constrained by red tape that it can document.
they should be fined continuously until their network is upgraded.. if hat means no profit for the next 3 years TFB.
No profit for three years would result in a shareholder lawsuit. Would it be preferable to dissolve one of the major cellular carriers and sell its assets at auction to the other major carriers, resulting in yet another contraction of the U.S. cellular industry?
So why do metered cell phone plans measured in megabytes even still exist?
For the same reason that metered cell phone plans measured in minutes (rather than hours) exist: some carriers want to compete for the lowest-usage subscribers.
Aiming for sustained peak load simply means you statistically predict how much actual load you'll get for x customers at the busiest moments and you build for that +5% or +10% based on your datasource.
If an ISP lacks the finances at the moment to "build for that" throughout its service area, how else should it reduce peak load?
This isn't exactly rockey science, every self respecting engineer can do these calculations.
Engineers say one thing. Finance says another. An executive needs to balance the two, and it may come down to ordering the engineers to come up with a solution that reduces peak load. What such solution do you recommend?
Wirth's law may be blocking these devices from getting an upgrade to the next major version of Android if the device's RAM size or disk performance does not meet the minimum system requirements of the next major version. I can speak from experience that for some devices, though an update was issued, it never should have been. Android 4.4 "KitKat" on the Nexus 7 (2012) "grouper" tablet is fine; Android 5 "Lollipop" on the same hardware is a jank-fest, with the UI often freezing for five or ten seconds at a time. All reports that I've read imply that the jankiness is caused by ASUS cheaping out on slow NAND flash memory. Clearing the cache helps some but not all users in this situation. It's like the last version of iOS that Apple releases for a given iPhone model: its increased system demands often bog down the UI to the point where some users end up assuming it is an intentional measure to get people to replace otherwise working hardware.
You are able to side load on a firestick.
That's fine if the app's publisher makes an APK available. A lot of applications are available only through Google Play Store.
PHP7 is a huge improvement over old versions. If you haven't looked at it, do so. PHP sucks in 2016 is mostly FUD now.
To see how much of an improvement PHP 7 is, let's check the 7 things wrong with PHP 5 that I mentioned in this article (and previously in a Slashdot comment):
Things such as switch use the byzantine semantics of operator == Not fixed to my knowledge. Parse errors and undefined functions are fatal PHP 7 makes them "engine exceptions" which are catchable Error instances. Inconsistent naming and argument order conventions for standard library functions Not fixed completely for back-compat reasons, but PHP 7 has started to drop some duplicate functionality. Backwards associativity for operator ?: Not fixed for back-compat reasons, but at least the new null coalesce operator ?? has its associativity on the more useful side. Letting the server operator change program semantics in annoying ways Safe mode and magic quotes were already gone in PHP 5.4. cURL would not follow redirects if open_basedir was set until September 2013. But upload_max_filesize still defaults to 2MB, which I see as unnecessarily small. Semantics breakage in minor (5.x) versions encouraging shared hosting operators to delay making new versions available to subscribers Not fixed until the majority of shared hosts offer PHP 7. Has someone run a survey about this? No keyword arguments Not fixed.Does it mean my webpage won't either die with no error or run fine with unhelpful warnings anymore?
If you want exception semantics instead of "error reporting", use ErrorException .
your network should be able to handle peak load without negative traffic routing interactions.
Allowing all subscribers to transmit at peak rate simultaneously "without negative traffic routing interactions" would require such an investment in capacity, including the purchase of land and equipment for cell towers, that most subscribers would consider the service unaffordable. For this reason, most ISPs oversell their networks under the assumption that not all subscribers will reach peak load simultaneously. Metering during congested periods is ostensibly a way to reduce peak load to the point where it can be "handle[d] without negative traffic routing interactions" as a way of delaying the need to invest in capacity. But I agree with you that metering during uncongested periods is a cash grab.
Currently data can be transferred on your phone by apps that may be running in the background.
Android logs the amount of data transferred by each app running on a device and lets the user sort a list of apps by decreasing data volume. Is this not enough?
Voice calls were never made without the customers knowledge
Almost never. (Search for butt dialing.)
It is not really feasible or helpful to keep logs of every IP address and size of every packet transferred to/from your phone.
So let me get this straight: Unlike satellite ISPs, cellular ISPs have failed to offer a reward for moving away from congested times of day, and you're rationalizing it as because the device communicates with more distinct destinations than when voice calls dominated cellular usage. Let me see if I can think of a way to interact with Internet data transfer accounting mechanisms already present in mobile operating systems.
Windows 8, Windows 10, and Android operating systems allow users to mark particular SSIDs as metered, and an application is expected to respect that by querying cost whenever it receives a notification of change in the device's Internet connection. I guess the real solution is for carriers to signal to a device when the connection has become unmetered or has become metered, so that applications can switch to more conservative download schedules.
I think the idea is to leave insert mode to paste using a dedicated paste command rather than using your terminal's keypress simulation.
If I ever paste anything into terminal I always paste it into a text editor first.
Then gosh help you if what you paste contains an exploit for your text editor. There's a vi exploit in one of the examples.
Cats can be trained to use a toilet. Many prefer it because they no longer have to spend time burying their waste. The biggest drawback is you have to leave the lid up and the seat down, but that drawback wouldn't quite apply to hypothetical cat people.
Telecom customers (and probably producers as well) as a whole are just not savvy enough to deal with a complicated pricing system
Back before the majority of postpaid cellular plans came with unlimited voice minutes, a lot of cellular carriers and landline long-distance companies offered discounts on night and weekend calls. Are telecom customers less tech savvy than they used to be?
In my apartment complex, the previous owners entered into an illegal "exclusivity contract" many years ago.
How long do you plan to live there, compared to the difference between what you'd pay for Business DSL and how much it'd cost to hire a lawyer?
Networks are measured in maximum throughput over a given period of time, not total number bytes transferred to a given place.
Data caps measure each user's contribution to average throughput, which is (imperfectly) correlated to each user's contribution to peak throughput.
Data caps are a solution to the wrong problem.
Agreed. If the problem is peak throughput, the solution is to run the meter when the network is congested, that is, when the user's packets are passing through a segment at its maximum throughput. Satellite ISPs have the right idea: run the meter only during congested hours to encourage users to move their transfers away from times of maximum demand for throughput.
the "problem" they're really solving is that of people using the Internet the way it was intended--as a communications medium that doesn't really care if what's being communicated is voice, text, images, videos, etc.
That or the "problem" is one of using the Internet as a real-time communications medium, as opposed to an asynchronous application where usage can be shifted away from . People download multi-gigabyte HD movies and Steam games during prime time, contributing to dropped packets at times of maximum throughput, when they could be setting the computer to download them at 0-dark-thirty and not affecting the maximum.
You just need to pick the right combination of frameworks and programming language. Everything can be converted to everything else nowadays.
Unless the web application is a workaround for a platform owner's policy that specifically forbids a set of native apps. For example, applications doing any of these things cannot be converted to a native iOS application intended for distribution to the public.
Now there may be valid reasons to reimplement in another language (better matching your available talent pool, better performance, you would just prefer it)
Especially when the deployment platform dictates the choice of language. Examples include Xbox Live Indie Games and Windows Phone 7 requiring verifiably type-safe CIL for the .NET Compact Framework (in practice, requiring C#), or the web requiring either ECMAScript 5 or a language that transpiles to ECMAScript 5. Then "better matching your available talent pool" means "we would have to hire someone to write a transpiler to the only supported language", and "better performance" means "the alternative is writing an interpreter".
but if it's just to 'make it clean', that can pretty much always be accomplished within the existing language choice.
Even if "the existing language choice" is assembly language, as in the case of porting a classic video game to a modern platform? Can assembly language be made as clean as, say, Python?
Assuming everyone uses up to the 250 GB cap
The 250 GB per month cap is for wired Internet. Caps for wireless are much smaller, on the order of 3 to 10 GB per month.
It has turned into basic cable.
But with the equivalent of a lot more public access channels than any basic cable operator has ever had.
I assume "megabytes" was used because that's the unit in which carriers state their data transfer tariffs.