5 ms·

> Software developers are mostly to blame for obsoleting perfectly working hardware. Is it the developer, say, of some app for Android or iOS, at fault, or is

> Software developers are mostly to blame for obsoleting perfectly working hardware. Is it the developer, say, of some app for Android or iOS, at fault, or is it that the changing API surface coupled with App Store policies means that an app developer is forced to keep releasing new versions and deprecating older operating systems and hardware? I always had the impression that Google and Apple took a very dim view of apps that were stable and not refreshed.

ryandrake 2y ago
It doesn't cost anything or hurt to keep older versions available to download--versions that work on the older OS/API levels.
- vytautask 2y ago
  >It doesn't cost anything or hurt to keep older versions First, it affects my app ratings in the application store/shop. Some user with 15-year-old device gives me 1-star rating because of some annoying bug that was already fixed 10 years ago but he can not upgrade because that newer version is using a newer API. Also, what about bugfixes and support? I do not want to have to support (answer emails/calls and deal with bad reviews on all platforms) my ancient version of an app. This would make my app unaffordable for 99% of my users.
  - cassepipe 2y ago
    That could easily be fixed by preventing people using old versions to give feedback. I guess no app store currently allows that right ?
  - myaccountonhn 2y ago
    Yeah this is an important perspective for coming up with a solution, because it’s a very legitimate reason to not want to support old hardware…
  - ClumsyPilot 2y ago
    > First, it affects my app ratings in the application store/shop good example of systemic failure - we are destroying a real thing (functionality) to optimise for a fictional thing (rating). There is plenty of old windows software with no support, I have a game that is old enough to drink and it still works and I still sometimes fire it up out of nostalgia. There is no reason software cannot be like an old motorbike - a 40 year old Honda motorbike can still work, if that’s what the user wants.
    - latexr 2y ago
      > we are destroying a real thing (functionality) to optimise for a fictional thing (rating) Ratings are quite real. They influence if new people buy your app at all, which influences how viable it is as a business, which influences the app’s future, which influences current customers.
      - immibis 2y ago
        They are constructed reality as opposed to inherent reality. We invented a fake metric then created incentives (adjusting the allocation of real resources) to optimise the fake metric.
        latexr 2y ago
        It’s not a “fake metric”. It exists, measures something, and it has an effect on the world. It is real. It’s a subpar metric¹, probably even a harmful metric², but that doesn’t make it fake or fictional. ¹ At best it only measures those who rate and it can be gamed (true for every media). ² For the bad incentives you mention.
        davemp 2y ago
        Semantics/pedantry. I’m sure you’re capable of inferring what the OP means.
        2y ago
- Earw0rm 2y ago
  Until you bring the online services side into the picture. A typical app depends on both OS APIs and remote service APIs. Keeping the latter stable and supported is harder. Especially when there are third party services and frameworks involved, which there inevitably are. And whereas no-longer-fixable bugs in an app might be work-aroundable, merely annoying, or outright irrelevant to some use cases, bugs in a service/framework can easily be security stuff that can take down your whole business or land you in regulatory hot water.
hulitu 2y ago
> Is it the developer, say, of some app for Android or iOS, at fault, or is it that the changing API surface coupled with App Store policies means that an app developer is forced to keep releasing new versions and deprecating older operating systems and hardware? "Because of the release of GNOME 2.0 and 2.2, and the lack of interest in maintainership of GNOME 1.4, the gnome-core product is being closed. If you feel your bug is still of relevance to GNOME 2, please reopen it and refile it against a more appropriate component. Thanks..." "Why not be honest and resign yourself to the fact that version 0.8 is followed by version 0.8, which is then followed by version 0.8? But that's what happens when there is no incentive for people to do the parts of programming that aren't fun. Fixing bugs isn't fun; going through the bug list isn't fun; but rewriting everything from scratch is fun (because "this time it will be done right", ha ha) and so that's what happens, over and over again."

ryandrake 2y ago

It doesn't cost anything or hurt to keep older versions available to download--versions that work on the older OS/API levels.

vytautask 2y ago
>It doesn't cost anything or hurt to keep older versions First, it affects my app ratings in the application store/shop. Some user with 15-year-old device gives me 1-star rating because of some annoying bug that was already fixed 10 years ago but he can not upgrade because that newer version is using a newer API. Also, what about bugfixes and support? I do not want to have to support (answer emails/calls and deal with bad reviews on all platforms) my ancient version of an app. This would make my app unaffordable for 99% of my users.
- cassepipe 2y ago
  That could easily be fixed by preventing people using old versions to give feedback. I guess no app store currently allows that right ?
- myaccountonhn 2y ago
  Yeah this is an important perspective for coming up with a solution, because it’s a very legitimate reason to not want to support old hardware…
- ClumsyPilot 2y ago
  > First, it affects my app ratings in the application store/shop good example of systemic failure - we are destroying a real thing (functionality) to optimise for a fictional thing (rating). There is plenty of old windows software with no support, I have a game that is old enough to drink and it still works and I still sometimes fire it up out of nostalgia. There is no reason software cannot be like an old motorbike - a 40 year old Honda motorbike can still work, if that’s what the user wants.
  - latexr 2y ago
    > we are destroying a real thing (functionality) to optimise for a fictional thing (rating) Ratings are quite real. They influence if new people buy your app at all, which influences how viable it is as a business, which influences the app’s future, which influences current customers.
    - immibis 2y ago
      They are constructed reality as opposed to inherent reality. We invented a fake metric then created incentives (adjusting the allocation of real resources) to optimise the fake metric.
      - latexr 2y ago
        It’s not a “fake metric”. It exists, measures something, and it has an effect on the world. It is real. It’s a subpar metric¹, probably even a harmful metric², but that doesn’t make it fake or fictional. ¹ At best it only measures those who rate and it can be gamed (true for every media). ² For the bad incentives you mention.
        davemp 2y ago
        Semantics/pedantry. I’m sure you’re capable of inferring what the OP means.
        2y ago
Earw0rm 2y ago
Until you bring the online services side into the picture. A typical app depends on both OS APIs and remote service APIs. Keeping the latter stable and supported is harder. Especially when there are third party services and frameworks involved, which there inevitably are. And whereas no-longer-fixable bugs in an app might be work-aroundable, merely annoying, or outright irrelevant to some use cases, bugs in a service/framework can easily be security stuff that can take down your whole business or land you in regulatory hot water.

hulitu 2y ago

> Is it the developer, say, of some app for Android or iOS, at fault, or is it that the changing API surface coupled with App Store policies means that an app developer is forced to keep releasing new versions and deprecating older operating systems and hardware? "Because of the release of GNOME 2.0 and 2.2, and the lack of interest in maintainership of GNOME 1.4, the gnome-core product is being closed. If you feel your bug is still of relevance to GNOME 2, please reopen it and refile it against a more appropriate component. Thanks..." "Why not be honest and resign yourself to the fact that version 0.8 is followed by version 0.8, which is then followed by version 0.8? But that's what happens when there is no incentive for people to do the parts of programming that aren't fun. Fixing bugs isn't fun; going through the bug list isn't fun; but rewriting everything from scratch is fun (because "this time it will be done right", ha ha) and so that's what happens, over and over again."