8 ms·

Yeah, you really have three options now. Either (1) never introduce features that require more processing power but newer phones are capable of handling, (2) ha

Yeah, you really have three options now. Either (1) never introduce features that require more processing power but newer phones are capable of handling, (2) have those new features but lock older phones out of being able to use them, or (3), stop supporting older phones with your updates. The problem with (1) is now you've completely stagnated the market, and if your competitors choose either of the other options you're left lagging behind and you lose sales. The problem with (2) is now customers of older phones will be upset that they didn't get the full upgrade for reasons they don't understand. The iPhone 4S and Siri or the iPhone X and Animoji show the reaction you should expect. The problem with (3) is now you have planned obsolescence and old phones just become throw-away leading to an ecological disaster. None of these are good options.

simion314 8y ago
2 is a good option, give users security fixes, if you promised to deliver new features to them then do so but make sure you did not made the OS and core apps worse.
garmaine 8y ago
Option (4): give the user a choice to opt into the new features. A checkbox isn't that hard.
- kodablah 8y ago
  > A checkbox isn't that hard A naive assumption. In fact, each checkbox exponentially increases the number of possible runtime environments you have to maintain and test.
  - smichel17 8y ago
    Alternatively, a choice of whether to install the update. "Installing this update will give you access to feature X, but will slow down your phone." For comparison, my AT&T Samsung Galaxy S6 forces me to install updates, which I hate. Wasn't my choice to get this particular phone, wouldn't do it again.
    - candiodari 8y ago
      Assuming some users choose to install, and some don't, doesn't change the fact that updates will become incredibly expensive support-wise. Exponetially expensive even. This will mean: almost never updates at all. Well, I have to admit: my mother will be happy about it.
  - garmaine 8y ago
    Not if checking the obscure and hidden "Don't throttle me!" box comes with a warning that it is unsupported. Then you don't have to do any serious maintenance and testing.
woah 8y ago
4) Don’t ship updates to phones in the EU
- pmontra 8y ago
  And maybe get sued because of unfixed vulnerabilities.
  - twtw 8y ago
    Then ship security updates independently of bells and whistles updates that make old phones dog slow?
    - dcbadacd 8y ago
      Then you get something like spectre and the fix is making things slower.
      - sangnoir 8y ago
        There are thousands of use-after-free (or generally more pedestrian) bugs for each spectre-like bug. Let's not make perfect the enemy of good.
      - Fnoord 8y ago
        Then you kindly ask Intel to reimburse the costs. Security > performance.
tomatotomato37 8y ago
Where's (4) Don't write slow software? Modern phone's have better specs than 2000-era PCs with software that needs to do less; they shouldn't be having these performance problems in the first place
- kodablah 8y ago
  > Where's (4) Don't write slow software? Says who, the government? We are talking about court-imposed fines here, not consumer wishes. Do we want a literal slow-software police?
  - c0nducktr 8y ago
    > Do we want a literal slow-software police? That obvious answer is no, but as I think back on how many times I've been frustrated by some poorly performing piece of shit application over the past year, it's starting to not seem so bad of a solution.
  - ginko 8y ago
    >Do we want a literal slow-software police? Yes please!
  - Darthy 8y ago
    > Do we want a literal slow-software police? This court case shows that we kinda now already have one. The phones from 5 years ago already had amazing processors with billions of cycles per second. They can probably do nearly everything. But you know us software engineers: "make it faster till it runs wells on my test kit, then oh well ship it!" I think this ruling is a nudge into a better way. Yes deplay updates for more than 2 years, but don't stop optimizing new stuff when the latest phone can run it pretty well - keep optimizing until devices from the last 5 years run it very well.
    - geezerjay 8y ago
      > This court case shows that we kinda now already have one. That's patently wrong. The point of the fine was not that the phones were noticeably slower after the upgrade. The point of the fine was that the slowness was motivated as a way of inducing forced obsolescence. The manufacturers forced the updates with no rollback mechanism to render phones unusable and thus force their users to purchase replacements. If users could roll back the changes then the manufacturers wouldnt be fined.
twtw 8y ago
> (3) is now you have planned obsolescence and old phones just become throw-away leading to an ecological disaster. I don't see how planned obsolescence via missing bells and whistles present in new phones is any worse than planned obsolescence via all functionality getting slower. Written on my iPhone 5S (2013) running iOS 8 (2014) - it is still a phone and is just as fast as it was in 2014. Is that planned obsolescence?
- sangnoir 8y ago
  I'm OK with manufactures not providing full-version/new-feature updates to old phones, as long as they provide frequent security fixes - which is one thing Google is doing right with Android security fixes. Samsung could trivially provide security updates on old version as Google does most of the heavy-lifting so they have no excuse for recommending upgrades on older, slower phones. Apple doesn't presently have a mechanism for this on iOS, to my knowledge.
  - sxg 8y ago
    Yes, but the problem is that the casual user will view this as planned obsolescence rather than a delicate tradeoff between performance and new features for older devices. Slowing down old iPhones was done to preserve battery life, but users did not see it that way, which led to massive backlash. Sure, Apple, should've communicated its reasoning and decision making before (which they have now done), but the casual user just doesn't care enough about the nuances here.
    - sangnoir 8y ago
      I think slow phone is more planned-obsolencence-y compared to "no OS upgrade, but phone works just as it did since the last supported upgrade - features and speed". I don't think the majority of users would know there is a new version out if they didn't get the upgrade nag-screens.
      - zepto 8y ago
        That sounds good in principle, except that 3rd party apps would also be frozen and start to fail.
Youden 8y ago
I don't think the problem you're addressing (new features being slow) is the problem the regulator had. I believe the problem was that the updates made existing functionality slower. e.g. the same Facebook app you had before the update scrolls slower afterwards or the old launcher is replaced with one that's more jittery, without letting you use the old one. Adding new features is fine, just don't break the old ones.
- azinman2 8y ago
  Except if part of the new features is a rewrite of display output that takes advantage of new hardware features / assumptions in ram characteristics, then that can be both a new feature and slow down older phones at the same time.
  - emn13 8y ago
    Given the actual changes apple+samsung made it's hard to argue this was in any way necessary - they chose to prioritize making the old phones look more like the new ones, which makes sense from their point of view because they want to encourage upgrades - so maximizing (visual) consistency at the cost of sluggishness is much better than maximizing quality for old phones at the cost of making phone differences more apparent. And in any case: they're free not to offer feature upgrades to old phones; I just doubt that's a wise choice in the market. If they really can't provide a few updates and reasonable performance - good riddance, their successors surely will.
  - vokep 8y ago
    Then you don't include that in the build of the update for the older models it doesn't apply to, or you include both approaches to output and choose which is appropriate at runtime.
- pbhjpbhj 8y ago
  Could they have saved themselves by giving users an informed choice: "this update will harm performance but improves security and gives access to new features; your choice" (imagined release note on update clickthrough) ?
empath75 8y ago
4) notify users that updates may slow down their model of phone.
imgabe 8y ago
2 seems the reasonable choice. If old hardware can't support a new feature, don't add it. I don't think customers will have trouble understanding "Your phone is old and it can't handle this new thing". That isn't difficult to understand. For older phones just provide necessary security updates for the older version of the OS.
- greglindahl 8y ago
  Apple already doesn't turn on new features on old phones that "can't support" them. The new features being talked about are ones that make an old phone slower. I owned an iPhone 5s and updated to the latest iOS version for 4 years and never noticed any slowness. So, basically, (2) means I would miss out on new features because other people complain about slowness that I don't notice.
  - ummonk 8y ago
    Yeah, some people seem to have much higher expectations of speed than I do. Having lived through dial up and actual hard disks (rather than SSDs), I really don't notice any slowness in my iPhone SE. I just notice all the awesome new features.
PhasmaFelis 8y ago
Or just make high-powered features optional and don't fuck up your core OS. That seems like the obvious choice. I've had several experiences with OS "upgrades" making a system much slower even though I wasn't using any of the new features. That's poor design if not actually malicious, and it doesn't have to be that way.
newnewpdro 8y ago
(4) you open everything and let the community keep old devices significant numbers of people still use current and secure.

simion314 8y ago

2 is a good option, give users security fixes, if you promised to deliver new features to them then do so but make sure you did not made the OS and core apps worse.

garmaine 8y ago

Option (4): give the user a choice to opt into the new features. A checkbox isn't that hard.

kodablah 8y ago
> A checkbox isn't that hard A naive assumption. In fact, each checkbox exponentially increases the number of possible runtime environments you have to maintain and test.
- smichel17 8y ago
  Alternatively, a choice of whether to install the update. "Installing this update will give you access to feature X, but will slow down your phone." For comparison, my AT&T Samsung Galaxy S6 forces me to install updates, which I hate. Wasn't my choice to get this particular phone, wouldn't do it again.
  - candiodari 8y ago
    Assuming some users choose to install, and some don't, doesn't change the fact that updates will become incredibly expensive support-wise. Exponetially expensive even. This will mean: almost never updates at all. Well, I have to admit: my mother will be happy about it.
- garmaine 8y ago
  Not if checking the obscure and hidden "Don't throttle me!" box comes with a warning that it is unsupported. Then you don't have to do any serious maintenance and testing.

woah 8y ago

4) Don’t ship updates to phones in the EU

pmontra 8y ago
And maybe get sued because of unfixed vulnerabilities.
- twtw 8y ago
  Then ship security updates independently of bells and whistles updates that make old phones dog slow?
  - dcbadacd 8y ago
    Then you get something like spectre and the fix is making things slower.
    - sangnoir 8y ago
      There are thousands of use-after-free (or generally more pedestrian) bugs for each spectre-like bug. Let's not make perfect the enemy of good.
    - Fnoord 8y ago
      Then you kindly ask Intel to reimburse the costs. Security > performance.

tomatotomato37 8y ago

Where's (4) Don't write slow software? Modern phone's have better specs than 2000-era PCs with software that needs to do less; they shouldn't be having these performance problems in the first place

kodablah 8y ago
> Where's (4) Don't write slow software? Says who, the government? We are talking about court-imposed fines here, not consumer wishes. Do we want a literal slow-software police?
- c0nducktr 8y ago
  > Do we want a literal slow-software police? That obvious answer is no, but as I think back on how many times I've been frustrated by some poorly performing piece of shit application over the past year, it's starting to not seem so bad of a solution.
- ginko 8y ago
  >Do we want a literal slow-software police? Yes please!
- Darthy 8y ago
  > Do we want a literal slow-software police? This court case shows that we kinda now already have one. The phones from 5 years ago already had amazing processors with billions of cycles per second. They can probably do nearly everything. But you know us software engineers: "make it faster till it runs wells on my test kit, then oh well ship it!" I think this ruling is a nudge into a better way. Yes deplay updates for more than 2 years, but don't stop optimizing new stuff when the latest phone can run it pretty well - keep optimizing until devices from the last 5 years run it very well.
  - geezerjay 8y ago
    > This court case shows that we kinda now already have one. That's patently wrong. The point of the fine was not that the phones were noticeably slower after the upgrade. The point of the fine was that the slowness was motivated as a way of inducing forced obsolescence. The manufacturers forced the updates with no rollback mechanism to render phones unusable and thus force their users to purchase replacements. If users could roll back the changes then the manufacturers wouldnt be fined.

twtw 8y ago

> (3) is now you have planned obsolescence and old phones just become throw-away leading to an ecological disaster. I don't see how planned obsolescence via missing bells and whistles present in new phones is any worse than planned obsolescence via all functionality getting slower. Written on my iPhone 5S (2013) running iOS 8 (2014) - it is still a phone and is just as fast as it was in 2014. Is that planned obsolescence?

sangnoir 8y ago
I'm OK with manufactures not providing full-version/new-feature updates to old phones, as long as they provide frequent security fixes - which is one thing Google is doing right with Android security fixes. Samsung could trivially provide security updates on old version as Google does most of the heavy-lifting so they have no excuse for recommending upgrades on older, slower phones. Apple doesn't presently have a mechanism for this on iOS, to my knowledge.
- sxg 8y ago
  Yes, but the problem is that the casual user will view this as planned obsolescence rather than a delicate tradeoff between performance and new features for older devices. Slowing down old iPhones was done to preserve battery life, but users did not see it that way, which led to massive backlash. Sure, Apple, should've communicated its reasoning and decision making before (which they have now done), but the casual user just doesn't care enough about the nuances here.
  - sangnoir 8y ago
    I think slow phone is more planned-obsolencence-y compared to "no OS upgrade, but phone works just as it did since the last supported upgrade - features and speed". I don't think the majority of users would know there is a new version out if they didn't get the upgrade nag-screens.
    - zepto 8y ago
      That sounds good in principle, except that 3rd party apps would also be frozen and start to fail.

Youden 8y ago

I don't think the problem you're addressing (new features being slow) is the problem the regulator had. I believe the problem was that the updates made existing functionality slower. e.g. the same Facebook app you had before the update scrolls slower afterwards or the old launcher is replaced with one that's more jittery, without letting you use the old one. Adding new features is fine, just don't break the old ones.

azinman2 8y ago
Except if part of the new features is a rewrite of display output that takes advantage of new hardware features / assumptions in ram characteristics, then that can be both a new feature and slow down older phones at the same time.
- emn13 8y ago
  Given the actual changes apple+samsung made it's hard to argue this was in any way necessary - they chose to prioritize making the old phones look more like the new ones, which makes sense from their point of view because they want to encourage upgrades - so maximizing (visual) consistency at the cost of sluggishness is much better than maximizing quality for old phones at the cost of making phone differences more apparent. And in any case: they're free not to offer feature upgrades to old phones; I just doubt that's a wise choice in the market. If they really can't provide a few updates and reasonable performance - good riddance, their successors surely will.
- vokep 8y ago
  Then you don't include that in the build of the update for the older models it doesn't apply to, or you include both approaches to output and choose which is appropriate at runtime.
pbhjpbhj 8y ago
Could they have saved themselves by giving users an informed choice: "this update will harm performance but improves security and gives access to new features; your choice" (imagined release note on update clickthrough) ?

empath75 8y ago

4) notify users that updates may slow down their model of phone.

imgabe 8y ago

2 seems the reasonable choice. If old hardware can't support a new feature, don't add it. I don't think customers will have trouble understanding "Your phone is old and it can't handle this new thing". That isn't difficult to understand. For older phones just provide necessary security updates for the older version of the OS.

greglindahl 8y ago
Apple already doesn't turn on new features on old phones that "can't support" them. The new features being talked about are ones that make an old phone slower. I owned an iPhone 5s and updated to the latest iOS version for 4 years and never noticed any slowness. So, basically, (2) means I would miss out on new features because other people complain about slowness that I don't notice.
- ummonk 8y ago
  Yeah, some people seem to have much higher expectations of speed than I do. Having lived through dial up and actual hard disks (rather than SSDs), I really don't notice any slowness in my iPhone SE. I just notice all the awesome new features.

PhasmaFelis 8y ago

Or just make high-powered features optional and don't fuck up your core OS. That seems like the obvious choice. I've had several experiences with OS "upgrades" making a system much slower even though I wasn't using any of the new features. That's poor design if not actually malicious, and it doesn't have to be that way.

newnewpdro 8y ago

(4) you open everything and let the community keep old devices significant numbers of people still use current and secure.