[HN Gopher] My iPhone 8 Refuses to Die: Now It's a Solar-Powered...
___________________________________________________________________
My iPhone 8 Refuses to Die: Now It's a Solar-Powered Vision OCR
Server
Author : hemant6488
Score : 421 points
Date : 2025-06-18 15:49 UTC (1 days ago)
(HTM) web link (terminalbytes.com)
(TXT) w3m dump (terminalbytes.com)
| jdon wrote:
| Soon you'll also be able to do speech to text locally, as Apple
| is adding a SpeechAnalyzer API [0] which is apparently faster
| than whisper [1].
|
| [0]: https://developer.apple.com/videos/play/wwdc2025/277/
|
| [1]: https://www.macrumors.com/2025/06/18/apple-transcription-
| api...
| jkmcf wrote:
| Tangentially, https://github.com/finnvoor/yap A
| CLI for on-device speech transcription using Speech.framework
| on macOS 26
|
| The MacStories article made it seem about 2x as fast as
| Whisper, but there's no network or shared servers involved, so
| it's effectively faster.
| xd1936 wrote:
| Faster, but... quality? I would take something 10x _slower_
| than Whisper 3 if it meant a 5% increase in quality.
| redundantly wrote:
| I love projects like this, doing things because you can.
| Especially low power, off-grid projects.
|
| However I did not love the writing style of this article. Lots of
| repetition. Asking questions to stress a funny point. Lots of
| repetition.
|
| I don't mean to sound like a jerk, even though I've succeeded at
| it. The author is cool, what they did is just as cool.
| rbinv wrote:
| It's AI slop. In fact, most (if not all) of this blog's recent
| posts are AI slop.
| CaptainFever wrote:
| That's not what slop means. This is anything but low-effort
| or low-quality.
| hagbard_c wrote:
| I see your fruitPhone 8 and raise my Motorola MB525 'Defy',
| Motorola MB526 'Defy+' and Samsung J3 which are in use as Wifi-
| enabled trailer camera. The phones provide a Wifi hotspot through
| which the camera's images are accessed. Hook up the trailer,
| connect to the Wifi network and voila, you can see what's
| happening in the trailer behind you. The oldest device in this
| list is from 2010, all of them run either Cyanogenmod (MB525 and
| MB526) or its successor LineageOS (J3). I replaced the batteries
| in the Motorola's, the J3 runs on its original battery. Oh, all
| of them run without a screen since that is not visible anyway and
| was broken in 2 of the 3. Android runs just fine without a screen
| and using the things this way takes a little less power.
| FlyingSnake wrote:
| That's pretty impressive. I love when people give old devices a
| new life and save them from being eWaste. True to the hacker
| spirit.
| ideashower wrote:
| I'm confused. What are you OCR'ing that requires a solution like
| this? What images are you processing?
| theandrewbailey wrote:
| My guess is that he wanted to use that Apple OCR framework and
| that iPhone was whatever he had handy. I went to his blog's
| homepage hoping to find some article as to what he's
| processing, but I didn't find anything. Is he scanning all of
| his novels?
| wing-_-nuts wrote:
| I loved the 'it turns out I'm an indoor cat with outdoor
| aspirations'. I often joke I'm an 'avid indoorsman'
| frereubu wrote:
| You might like this song, The Outdoor Type by The Lemonheads:
| https://www.youtube.com/watch?v=Ijlk0GTQbB4
| mmmlinux wrote:
| Maybe i'm missing something. Where are these thousands of users
| coming from? is this some service you offer?
| leakycap wrote:
| Yes, the author offers a public-facing website that allows free
| OCR of uploads.
| FlyingSnake wrote:
| Wonderful story!
|
| We don't give enough credit to Apple for keeping these old
| devices alive and kicking.
|
| I have a similar story wherein I repurposed my ancient OG iPhone
| SE and gave it a new life.
|
| https://samkhawase.com/blog/dumb-smartphone/
| tclancy wrote:
| >We don't give enough credit to Apple for keeping these old
| devices alive and kicking.
|
| I'm not sure I follow. It feels exceedingly hard to find new
| uses for old iPads without doing a lot of heavy lifting. Has
| that changed?
| brailsafe wrote:
| My iPad 3 is only unusable because anything beyond iOS 9
| isn't installable, most of the like 5 Apps I did have
| installed on it didn't survive a "backup", and obvs nobody's
| going out of their way to support ancient platforms.
|
| Otherwise, it still functions as an epub reader as long as
| iBooks continues functioning, but it's lame that I can't
| really use it for much else unless I made it a hobby.
| tech234a wrote:
| As a counterexample, VLC surprisingly still supports iOS
| 9.0
| brailsafe wrote:
| That's a great counterexample, since built-in video
| playback capability is awful. It's one of the few I still
| have installed if memory serves. It think I also have
| "The Room" and a few Google apps. Hardware-wise I always
| thought it was pretty solid, the software and general
| utility not so much, but I look at newer versions hat
| have come out since 2013 and don't really see how they're
| fundamentally any more capable than mediocre content
| consumption devices, and while that does do something for
| me, I would have hard time rationalizing the purchase of
| another one in the future.
| leakycap wrote:
| I was able to regain access to a lot of ipsw app backups
| from old Time Machine drives, in case you are wanting
| apps that are easy to use on your device. Any files from
| your backups will work, since they'll have your Apple ID
| in them.
|
| Can you not install apps from the Purchased section in
| the App Store? I was able to download the new version of
| an app on my iOS 18 iPhone, then reload the App Store on
| iOS 9 and download from the "Purchased" section, assuming
| the app existed back in the iOS 9 days or had a version
| targeting this old OS.
| leakycap wrote:
| To be fair, your iPad 3 is an iPad 2 with a retina screen;
| I remember buying an iPad 3 and it was glacially slow even
| at launch on the original iOS.
|
| I would imagine the best use of this device post eReader is
| a photo stand given the gorgeous screen... or something
| else that wouldn't need any interaction (it will be too
| slow to want to have touch interactions with).
|
| I use an iPad 2 as a IPcam monitor. The battery doesn't
| last long, but I'm able to grab it off the charger and take
| it around the house if I'm watching something going on. It
| doesn't support my new AI smart cams, but it still
| functions.
| jerlam wrote:
| For me, iPads (base model, non-Air/Pro) and iPhones seem to
| exist on opposite ends of the longevity spectrum. Never had
| an iPad last over 2-3 years without feeling sluggish and
| ready for an upgrade. Never had an iPhone since the 4 that
| felt sluggish when Apple stopped supporting it (5+ years).
| criddell wrote:
| My iPad is a 2018 iPad Pro and it still works great. It's
| my most used computer by far. AFAIK, it's still supported
| by Apple.
|
| My phone is an iPhone 13 (2021) and I'll probably upgrade
| in the next 24 months to get a better camera.
| prmoustache wrote:
| 2018 is still fairly new.
|
| I own a laptop from 2011 and it runs the latest fedora
| perfectly and is not limited at all performance wise as
| long as you don't try to run AAA games.
| dncornholio wrote:
| I don't agree with this take at all. I had to give up my iPhone
| 7 because I couldn't update iOS and my banking app refused to
| work on the older version.
|
| Apple would also gladly throttle your phone, see Batterygate.
| leakycap wrote:
| Informed technical users should know that the alternative to
| Batterygate is that iPhones would randomly be turning OFF
| with no warning in user's hands.
|
| When a battery is old and has low state of charge (under
| 25%), it is easy for a device to request more power than the
| battery can provide and BOOM, the screen is black.
|
| Apple mitigated and avoided that experience for users by
| programming the phone to slow down when a user's old battery
| could not support the power needs of the device at full
| speed. It makes sense when you take the time to be informed
| about it.
| jiqiren wrote:
| HomePods perform real-time vision processing on multiple camera
| streams for HomeKit. However, the primary quality challenge lies
| in the video quality of HomeKit-enabled doorbell cameras that can
| consistently stream to Wi-Fi. For instance, my doorbell operates
| on 2.4 GHz Wi-Fi, resulting in highly compressed video streams.
| This compression likely impacts the results.
| cosmic_cheese wrote:
| The range of HomeKit-enabled doorbells and cameras is
| disappointing to begin with and even worse when removing
| options that require a proprietary adapter box and/or
| subscription. The best option at the moment seems to be a
| Ubiquiti setup that integrates into HomeKit by way of
| Homebridge or other similar solutions rather than anything that
| supports HomeKit specifically.
| dzhiurgis wrote:
| At this point I'd just avoid HomeKit entirely.
|
| Any sort of automation in Home app besides 2-3 line demo is
| quickly turning into nightmare, you are locked in bunch of
| annoying limitations and devices are always costing more than
| open source alternative.
| cosmic_cheese wrote:
| It's the smart home ecosystem that the FOSS world has kind
| of coalesced around, though (see HomeBridge, HomeAssistant,
| etc). The others are all much more centered around someone
| else's servers and subscriptions and offer little to no
| possibility of running things locally.
| dzhiurgis wrote:
| Yes I run Home Assistant too. Also got quite a bit of
| devices on Aqara's platform, and a device each on
| Ewelink, Tuya, Meross which all technically are a
| platforms. There's probably another 5 devices with their
| own apps. Tasmota + Home Assistant is the only one I'm
| happy about.
|
| Home Assistant (with all its dumb quirks) at least makes
| an attempt to integrate them. Some FOSS devices I've
| exposed to HomeKit for presence automation, but seeing
| Siri is going nowhere I don't think I'll continue.
| laurensr wrote:
| In my browser the ads cover the actual content.
|
| User agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64)
| AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0
| Safari/537.36
| nickburns wrote:
| I found the page quite clean (with cloudflareinsights.com,
| googlesyndication.com, and googletagmanager.com blocked of
| course).
| Dansvidania wrote:
| uMatrix?
| nancyminusone wrote:
| >I'm saving approximately $84-120 CAD annually.
|
| I suppose most of this is eaten up by the need to pay apple $99
| per year just to run your own app on your own phone for longer
| than a week.
| procinct wrote:
| I believe you only have to pay to put your app on the App
| Store. I've made apps for my iPhone before and never had to
| pay.
| mcpherrinm wrote:
| It's the "for longer than a week" bit - Unless you have a
| paid developer account, you can only sign apps to sideload
| that last one week.
|
| There's some tools to automate "refreshing" the app, but that
| requires you have some other computer that pushes a new app
| every week.
|
| The "1 week" restriction is usually fine when you're
| developing (as you typically are continually rebuilding and
| updating when actively working on an app) but is clearly
| intended to avoid being a way to sideload apps without the
| developer account "nearby".
| tech234a wrote:
| If you trust it, SideStore manages to do it on device by
| using a local VPN to make an on-device server appear to be
| an external device on the network.
| weaksauce wrote:
| is there a reason to be wary of sidestore? first time
| i've heard of that but seems like a legit opensource
| project. doesn't seem like it would be a project that is
| all that lucrative for bad actors... people that don't
| want to pay $99/year to load apps on their personal
| iphones/ipads doesn't seem like a big score.
| sheepscreek wrote:
| I'm not a 100% on this, but I believe you need to pay them to
| "sign" your app. For iOS, that means there is no way anyone
| else will be able to use your app unless they side-load it
| themselves (and we all know how cumbersome that is, Apple
| doesn't want to make it easy).
| notnmeyer wrote:
| correct
| behnamoh wrote:
| This Apple fee is one of the most absurd things they do. Like,
| how is it even justified--does Apple really spend $99 on infra
| maintenance and server costs to host your app?
|
| When I buy a device I want to know that I own it, but Apple
| keeps pushing the narrative that "we LET you use this device in
| ways we see fit". So basically the customer is just borrowing a
| device from Apple while paying the full price.
|
| I'm a longtime Apple user but can't shake off this love-hate
| relationship with the company.
| aerostable_slug wrote:
| I think it's fair to also cover the fairly rigorous testing
| that occurs for each app store submission. I'm not sure a
| hundred bucks is the right number, but it's not fair to say
| all they do is host the file.
| rahimnathwani wrote:
| You have to pay $99/year even if you only want to use the
| app on your own device.
|
| You can only sideload for free if you are willing to
| reinstall every X days.
|
| They don't need to test an app if you're not asking them to
| distribute it through their store.
| mitemte wrote:
| What's worse is it used to be 90 days. Apple changed it
| to 7 days years ago.
| demosthanos wrote:
| 90 days is still absurd. I have custom apps I install on
| my Android phones _once per phone_. I go years without
| bothering to rebuild them.
| akutlay wrote:
| I would guess they do it because they want to minimize
| the chance that someone will install an unapproved app to
| someone's phone and cause harm. I know it's already
| pretty hard but Apple seems to be very particular when it
| comes to this.
| prmoustache wrote:
| That is not their job.
| Someone wrote:
| That's an opinion. Apple's take is that they sell
| "everything that runs on your phone has gone through our
| reviews, so you can trust it isn't malware"
|
| That, in their opinion, makes it their job to prevent
| people from permanently installing software on other
| people's phones. I'm sure they would remove the
| "permanently" if they could, but developers have to test
| builds so frequently that they can't review them all.
| phanimahesh wrote:
| Popup on app open that warns app is sideloaded?
|
| There are simpler and more usable options that are more
| defensible than what they do today.
| fmbb wrote:
| > You can only sideload for free if you are willing to
| reinstall every X days.
|
| Does this mean you lose data, or is data retained when
| reinstalling?
| neilv wrote:
| > _I think it 's fair to also cover the fairly rigorous
| testing that occurs for each app store submission._
|
| By "fairly rigorous", do you mean "fickle, random"?
| lm411 wrote:
| Not even close compared to Google Play and their review
| and appeal process.
|
| Recently I had an app for a customer. Approved easily by
| Apple. Rejected by Google.
|
| The reason given by Google was completely meaningless in
| the context of the app. When this happens, I usually make
| a bullshit change, increment the version, and submit
| again. That was also rejected in this case. I asked for
| more info and they provided a meaningless screenshot of
| the app - that was all. So I appealed. That was also
| useless! They provided no info to help.
|
| Eventually I just created a new Google Play account and
| re-submitted a new version of the app, and it was
| accepted near immediately.
|
| I've had some annoying experiences with Apples review
| process but it is gold compared to Google Play.
| bigyabai wrote:
| "fair" would be letting me sideload if I didn't want to go
| through Apple's vetting. Their expensive review process is
| only required because _they_ decide it 's arbitrarily
| necessary and unavoidable.
| 8note wrote:
| apple could easily pay that with its money printer
| commision on app sales or on its money printer iphone
| sales, both of which are inpart because of the app
| developers.
|
| whats the value add of rigourously validating an app that
| youre only running on your own phone?
| wpm wrote:
| OK, then don't charge me until I submit something to the
| App Store.
|
| I should be able to self-sign an app for longer than a week
| on a free developer account.
| notnmeyer wrote:
| i'd guess it's more to keep extremely low effort submissions
| out of the app store.
| Gigachad wrote:
| Which is not unreasonable for something listed in the App
| Store. It is unreasonable that you can't sideload though.
| phire wrote:
| I'm pretty sure the $99 fee is explicitly there to
| prevent "normal" users from side-loading.
| eddythompson80 wrote:
| It could be playing 2 roles, acting as a limiting gate
| for the App Store spam and preventing a simple 2 step
| tutorial to enable side loading.
| notnmeyer wrote:
| yeah, this also makes sense
| rkagerer wrote:
| This is why I switched to Android 10 years ago. Unfortunately
| the grass isn't looking much greener over there these days.
|
| I'd love to hear from individuals who worked at these
| companies whether it disgusts them as much as it does me, and
| ideas (from a business perspective as much as technical) on
| how a new platform might wrest control back into the hands of
| users/owners.
| sampullman wrote:
| The Android fee is only $25, but in my experience
| everything around the submission process is at least 4x
| worse, so it evens out.
|
| At least Apple has humans doing review and support.
| prmoustache wrote:
| But you don'have to pay to sideload your app and have it
| stay forever on your device.
| yjftsjthsd-h wrote:
| In this very narrow case, the grass on the Android side is
| much greener: You can install your own APKs on an Android
| device without paying anyone at all, without having to
| upload anything anywhere, and without requiring any
| particular device to build the APK in the first place. You
| don't even need to touch the bootloader or root it; you
| just toggle a setting to allow the installation and it
| works.
| freedomben wrote:
| For now at least. There have been articles recently about
| how Google is looking to change that
| asimovfan wrote:
| what is it that you "love" about Apple?
| theshackleford wrote:
| Not op but...
|
| It started out originally that I just needed a UNIX/Linux
| _like_ but I also needed at the time better support for
| some propietary stuff than linux had, which is how I
| entered the fold.
|
| What has kept me a customer has been their quality of
| service over the 15 years I have been a customer, which has
| more than made up for the extra cost of their hardware.
|
| I get an OS I find reasonable to use, in a hardware package
| I like (give or take quite a few years there) and generally
| at this point still know that if something goes wrong the
| apple of today (but maybe not tomorrow) will look after me
| as a customer. If this changes, i'll go elsewhere, shunt
| OSX off and just go back to linux on the desktop I suppose.
| I'm not wedded to them. If they had'nt released the silicon
| variants when they did I was already getting to jump ship
| over to Lenovo/Dell land (at the time.)
|
| Phones are a bit different, i've still received brilliant
| service from them in that regard, but I tend to flip back
| and forwards between android and iOS depending on my mood
| at the time.
| lmm wrote:
| > how is it even justified
|
| Money is nice, they can charge it and people will pay them.
| Would be letting their shareholders down not to charge it
| really. I'm surprised they haven't tried bumping it up yet.
| cortesoft wrote:
| > Like, how is it even justified--does Apple really spend $99
| on infra maintenance and server costs to host your app?
|
| How much something costs is not what determines how much a
| company charges for something.
|
| A company sets prices based on what will make it the most
| money. A company only lowers prices if they think doing so
| will generate higher total profits in the long run.
|
| Apple seems to think charging $99 a year for developers will
| help its long term bottom line the most.
|
| There are probably many reasons for that, some of them
| already mentioned in sibling comments - keeping low effort
| apps out, preventing spammers from constantly buying new
| accounts to bypass bans, reducing the workload for approvers,
| generating revenue from the fees, etc.
|
| Prices aren't justified or not, you choose to pay them or
| not.
| sigmoid10 wrote:
| >How much something costs is not what determines how much a
| company charges for something.
|
| It actually does - in a free market. That's, like, one of
| the main arguments why capitalism is good for the
| population and not evil. But in a gate-kept oligopoly like
| phones, actors can abuse the system to squeeze more money
| out of consumers, leaving the corporations as sole
| beneficiaries. That's why this kind of stuff usually gets
| curbed in functioning democracies.
| keerthiko wrote:
| I'm pretty sure in a free market, how much someone is
| willing to pay for something is what determines how much
| a company charges for something, not how much it cost to
| provide. We wouldn't have inflation of _most_ goods
| /services if it was based on how much it cost to
| produce/provide.
| freedomben wrote:
| You're right, but generally in a free market competition
| will force prices down until they are close enough to
| production costs that going lower risks loss. In practice
| this rarely happens because we don't really have "free"
| markets, but rather a weird hybrid plus legal landmines
| all over the place.
| bitdivision wrote:
| True - how much someone is willing to pay matters.
| However in a competitive market, companies can't just
| charge whatever people will pay. Competitors will
| undercut them, so prices should eventually align with the
| cost of production plus a reasonable margin.
| jxjnskkzxxhx wrote:
| > It actually does - in a free market
|
| Meaningless sentence.
| phanimahesh wrote:
| > capitalism is good for the population and not evil
|
| This is the biggest lie that we keep telling ourselves.
| Capitalism is destroying the only place in the universe
| we can survive, and with the absurdly unequal wealth
| distribution and centralisation it enables, has caused
| more collective misery than any other idea in human
| history, in my opinion.
| ptaffs wrote:
| I agree and piling on. Capitalism is good for those with
| capital, the wealthy few. Then wonder where they got the
| capital, and mostly it's something environmentally bad,
| like the extraction industry such as coal and oil.
| FranzFerdiNaN wrote:
| Free markets have absolutely nothing to do with
| capitalism. You can have markets without capitalism. You
| can have free trade without capitalism, and you can have
| unfree trade with capitalism too.
|
| It's one of the great achievements of capitalism that it
| managed to convince people that trade == capitalism and
| that without capitalism you are reduced to the Soviet
| Union, because no other options are possible.
| robertlagrant wrote:
| > It's one of the great achievements of capitalism that
| it managed to convince people that trade == capitalism
| and that without capitalism you are reduced to the Soviet
| Union, because no other options are possible.
|
| Never heard anyone say this before, although it may be
| pretty much the case[0].
|
| [0] https://en.wikipedia.org/wiki/Foreign_trade_of_the_So
| viet_Un...
| FranzFerdiNaN wrote:
| If you criticise capitalism one of the most likely
| responses you're going to get is ''so you want to become
| communist like the SU"?
|
| And that wkkipedia article is of course not proving that
| trade equals capitalism (or are you saying that America
| stops being capitalistic if Trumps dream of a self-
| sufficient nation somehow succeeds?). Trade is trade.
| There was trade in the past when capitalism did not yet
| exist and there will be trade in the future when
| capitalism no longer exists.
| robertlagrant wrote:
| > Trade is trade. There was trade in the past when
| capitalism did not yet exist and there will be trade in
| the future when capitalism no longer exists.
|
| Indeed. I don't think anyone thinks otherwise. Fuedal
| lords traded. Totalitarian states traded. We know there
| was and is trade.
| kaonwarb wrote:
| Only for commodities, and even then only sometimes.
| cortesoft wrote:
| Even in a free market, not every product has perfect
| competition. Luxury brands always charge a lot more than
| it costs to make a product, because there are other
| factors that go into price.
| kaptainscarlet wrote:
| Yeah companies charge as much as they can getaway with
| irrational wrote:
| There can't be that many iOS developers that the $99 really
| affects their bottom line. I always assumed it was a
| barrier to entry to help discourage low effort apps.
| KeplerBoy wrote:
| Keeping low effort apps out of the store helps their
| bottom line. It's a second order effect.
| kccqzy wrote:
| Yes but the $99 fee doesn't just allow selling apps on
| the App Store. It is also required for testing the app
| such as on TestFlight.
|
| Apple should long ago make the $99 an App Store fee, not
| tied to any provisioning certificates or code signing.
| engcoach wrote:
| Without a fee, people would make new accounts and
| circumvent distribution restrictions.
| leakycap wrote:
| The fee could be less and have a similar deterrent on the
| type of activity you describe. The real question isn't
| what Apple is gaining from this fee, but what they are
| losing.
|
| Apple's $99 fee is annoying and feels like a waste of
| time and one more thing to manage.
|
| The paid ADC program has kept me from sharing projects
| with other developers who would have otherwise been able
| to contribute (but they aren't paid devs because they'd
| rather have a year of Costco hotdogs than pay Apple to
| help me with my app for a week)
| kccqzy wrote:
| But it's asinine for developers to have to pay $99 in
| order to test their app, such as on TestFlight. When you
| have an app idea, when you are far from deciding on
| monetization, you just want to test out the central
| features of the app among friends, it's wrong to require
| payment for that.
|
| Remember all apps have once been low effort apps: the
| first few weeks when you begin working on them. Polish
| comes later.
| iwontberude wrote:
| They can test and iterate using simulator without
| spending $99
| kccqzy wrote:
| I said test among friends, i.e. potential but real users.
| The gulf between the simulator and TestFlight is so large
| that they are better considered completely different
| stages of testing.
|
| Furthermore, there are so many things that can't
| realistically tested by the developer on the simulator.
| cortesoft wrote:
| You aren't paying $99 per app, you have to pay that once
| per year and you can develop as many apps as you want.
| $99 isn't a huge amount.
| kccqzy wrote:
| > $99 per app
|
| Meaningless distinction. Most starting indie developers
| don't have more than one app anyway. It's like going to a
| fancy steakhouse and being offered a $99 all-you-can-eat
| where the only menu item is a 18oz porterhouse.
|
| > $99 isn't a huge amount
|
| It isn't if this is your main job. It could be if this is
| merely a hobby.
| leakycap wrote:
| $99 is a show-stopping barrier for more people than you
| can possibly imagine.
|
| Please, if you are of the mindset $99 is not a life-
| changing amount for someone else, I implore you to widen
| your world and at least stay in touch with what the
| average human experience is like.
|
| The person working McDonald's who has an app idea now
| needs an iOS device, a Mac, and $99 of available funds.
| Then, remember that person is richer than many people in
| other countries.
|
| $99 is a huge amount, especially given that you get
| nothing except a privilege that has no inherent value.
| rollcat wrote:
| Of course there are. Many browser extensions are
| available for all platforms except Apple's, because you
| need that $99/y ( _and_ a Mac) to wrap ( _and_ fix up) a
| bunch of JS you already wrote and tested everywhere else.
|
| I applaud the authors of the few good extensions who went
| the extra 20.000 leagues. (But I still reluctantly
| switched to Ungoogled Chromium.)
| encom wrote:
| >discourage low effort apps
|
| Well that obviously didn't work. I got rid of my Iphone,
| but I remember the app store as being an absolute
| wasteland of garbage, and discoverability was awful. I
| don't know if it was a slogan, or an ad campaign once,
| but there was this thing with "there's an app for that".
| Yea I guess maybe there is, but good luck finding it, and
| finding one that isn't riddled with ads and scammy in-app
| purchases, and then further good luck that the developer
| of it keeps paying apple 99$ dollars every year so the
| app isn't delisted.
|
| I'm not saying Google is any better. I've pretty much
| given up on apps and app stores at this point. If I find
| something new, it's something I'm made aware of via other
| channels (or unavoidable bullshit like mandatory app
| based car parking etc.).
|
| --love Ted K.
| wobfan wrote:
| I mean you're right and you've said it yourself already,
| but in comparison to try Play Store there apps from the
| App Store are like double the quality on average. Because
| most of the extremely low effort bs is kept out. I still
| hate the fee though, dont get me wrong.
| timewizard wrote:
| > A company sets prices based on what will make it the most
| money.
|
| No company does this. Prices are set based upon demand.
| This does provide opportunities to make more money during
| some periods than others. If you have a monopoly then you
| can ignore this and just pick what makes you the most.
|
| > Apple seems to think charging $99 a year for developers
| will help its long term bottom line the most.
|
| It's absolutely a bespoke filter to prevent spam and
| automated misbehavior. Admittedly there does seem to be a
| resulting overall quality difference between iOS apps and
| other platforms.
|
| > Prices aren't justified or not, you choose to pay them or
| not.
|
| Business models are legal or not. You choose to play by the
| rules or you don't play.
| ndr42 wrote:
| >> A company sets prices based on what will make it the
| most money.
|
| > No company does this. Prices are set based upon demand.
|
| I read an interview a long long time ago (with Jobs,
| Schiller or Cook - I don't remember) where they were
| saying explicitly that Apple charge the amount that get
| them the most money not marketshare. I remember the times
| when analysts where obsessed with market share and that
| apple had to lose because they were to expensive. I don't
| hear that opinion that often today.
| timewizard wrote:
| That's what they say. Anyways it would be a clever way of
| rephrasing "many of our products have very low demand and
| high lock in."
| II2II wrote:
| At the time, eroding marketshare was a legitimate
| concern. It takes money to develop products, and without
| continuous development they would not remain competitive.
| Whether they liked it or not, marketshare is a factor in
| making the most money since you need to spread out the
| cost of development. Many companies were failing at the
| time, including those who made high end workstations
| because of that. Many years ago, I read an article about
| how the development of Alpha processors could not keep up
| simply because Intel could invest far more into R&D.
| realusername wrote:
| > No company does this. Prices are set based upon demand.
|
| In a market without competition (such as the mobile
| duopoly), that's how it works. The customer has no choice
| anyways so no price comparison can happen.
| cortesoft wrote:
| Demand is a factor in determining what price will
| generate them the most money in the long term, but it is
| not the only factor. Competition is another factor, like
| you mentioned.
|
| They want to prevent spam and automated misbehavior
| because that will maximize their long term profit.
|
| Business models can be illegal, but not your pricing.
| AnthonyMouse wrote:
| > There are probably many reasons for that, some of them
| already mentioned in sibling comments
|
| Those reasons don't really make a lot of sense:
|
| > keeping low effort apps out
|
| "Low effort" apps are critical to establishing demand.
| Small developers can't justify spending a large amount of
| resources on something you're not sure anybody wants. If
| you post the MVP and get a lot of downloads, now you know
| it's worth your time to make it better. If you can't post
| the MVP then you don't post it at all and neither the MVP
| nor the polished version ever exists.
|
| That's the recipe for having an app store full of loot box
| games and similar trash which is known to be profitable to
| the developers while losing thousands of apps people might
| actually want to the uncertainty of not knowing that ahead
| of time. Which is exactly what we see. How is that in their
| interest?
|
| > keeping low effort apps out, preventing spammers from
| constantly buying new accounts to bypass bans, reducing the
| workload for approvers
|
| These are things that would imply an account creation fee
| rather than an annual fee, and also have nothing to doing
| app _development_ where you 're only installing the app on
| your own device.
|
| > generating revenue from the fees
|
| This is the thing people are complaining about. They feel
| as though a troll has jumped out from under a bridge to
| demand money without providing anything of value in return.
| You've already paid for the phone, now it's _your_ phone,
| what gives them the right to double dip?
|
| > Prices aren't justified or not, you choose to pay them or
| not.
|
| That's true in a competitive market. If you don't like
| Apple's prices then go use one of the other app
| distribution services for your iPhone. Unless there isn't
| one, right?
| 7speter wrote:
| Whats seemingly more absurd is you already paid for the phone
| AND the Mac you have to develop for iOS devices for
| latexr wrote:
| > Apple keeps pushing the narrative that "we LET you use this
| device in ways we see fit".
|
| No, they do not. That is how _you are interpreting their
| actions_. It's obviously not the narrative they are pushing,
| that would be utterly absurd. The narrative Apple pushes over
| and over is that it's _your_ device, and that what you do
| with it is private and stays with it. Outright saying the
| device is theirs and they only let you do what they choose
| would be incredibly stupid, and their marketing is not
| incompetent.
|
| Mind you, this doesn't mean your interpretation (which is
| shared by many people) is wrong. On the contrary, it has
| merit. But it makes no sense to say Apple is pushing it as a
| narrative, that's not what the expression means.
| carlhjerpe wrote:
| I recall seeing a lot of "in your hand", "on you device",
| "tailored for you" and such in their keynotes and press
| material.
| rfoo wrote:
| Apple pushes a narrative that their devices are _secure_
| (not private, but secure). And my less tech-savvy friends
| sincerely believe that it 's due to it being a walled
| garden, with curated software only.
|
| Apple made no attempt clarifying this.
| notnullorvoid wrote:
| I believe they are talking about Apple's anti-trust legal
| defense narrative. Not the marketing narrative, which is in
| direct conflict, and maybe false advertising.
| KolibriFly wrote:
| Feels even sillier in an era where people are trying to find
| creative, sustainable uses for older hardware
| carlosjobim wrote:
| What other serious business to business agreements can you
| enter into without spending at least $100? The fee is not to
| cover technical costs, but administration costs.
|
| Welcome to the world of having a small business. Be happy
| it's only $100. Your fees for cost-of-doing business is many
| times higher for a hot dog stand or any other thing you can
| come up with.
| dzhiurgis wrote:
| EcoFlow batteries are pretty expensive too.
|
| Also that's about 500kWh of power annually which averages to
| 50W. There is just no way iPhone uses that much.
| winter_blue wrote:
| The author has a mini PC plugged into the EcoFlow as well.
| That uses the bulk of the power.
| slg wrote:
| There is also the roughly $1k in costs for the solar and
| battery hardware even if we consider the iPhone itself free
| since it is so old.
| HenryBemis wrote:
| I was just checking the combo he is using [0] (River 2 Pro +
| 220W solar generator) and it's currently at USD 619. In the
| post, the author sums it at USD 780. I assume price dropped
| because of newer models, etc.
|
| [0]: https://us.ecoflow.com/products/river-2-pro-portable-
| power-s...
| slg wrote:
| There were also $280 of other vague miscellaneous costs
| listed among the initial investments that I was including
| as part of that "roughly $1k"
| nico_h wrote:
| Also you can only run the compile-sign-deploy from a mac AFAIK.
| ajross wrote:
| That was exactly my thought. Out of the whole universe of
| development platforms we have to choose from to do an off-label
| maker-think gadget hack, iOS _is inarguably, and by a huge
| margin, the worst_.
|
| There are literally home appliances with more customizable app
| development and deployment stories than iPhones.
| nfriedly wrote:
| The iPhone 8 has the unpatchable checkm8 bootrom vulnerability,
| so while it doesn't say this in the article, the author could
| have jailbroken the device to run whatever software they want
| without paying any Apple fees.
|
| That vulnerability was a huge win. It just recently stopped,
| with the final vulnerable device (7th gen iPad) not getting the
| iPad OS 26 update.
| selcuka wrote:
| Isn't it an in-memory exploit, though? I believe it would
| stop working if the phone restarts for some reason.
| nfriedly wrote:
| Yes, that's correct, it's "tethered" meaning you have to
| basically redo it each time the phone restarts.
| callbacked wrote:
| surely on an iPhone that has the checkm8 hardware vulnerability
| available, one could jailbreak the device, install a
| codesigning bypass plugin on it, then develop and sideload
| their app without the whole "pay apple $99/yr to keep your
| sideloaded app on your phone" thing?
| joshstrange wrote:
| I wonder if someone will make a LLM farm from older (probably not
| too old) iPhones using Apple's new foundation models. I know they
| won't hold a candle to SOTA models, they are much smaller for
| one, but when they announced API access that's the first thing I
| thought of, a sort of "folding @ home" but routing queries to a
| phone and spitting back the results.
|
| It's silly and probably makes no sense at all based on how weak
| the model will probably be but it's a fun thing to think about.
| romain_batlle wrote:
| nop probably a very bad idea even if you had enough iPhones and
| you could parallelise them, it would be 10x less electricity
| efficient
| lucb1e wrote:
| Manufacturing newer CPUs makes sense only if the device it's
| meant to replace is like 25 years old:
|
| "The emissions from production of computing devices far
| exceed the emissions from operating them" [...] "the European
| Environmental Bureau [7] makes the scale of the problem very
| clear. For laptops and similar computers, manufacturing,
| distribution and disposal account for 52% of their Global
| Warming Potential [...]. For mobile phones, this is 72%. The
| report calculates that the lifetime of these devices should
| be at least 25 years"
| https://wimvanderbauwhede.codeberg.page/articles/frugal-
| comp...
| piperswe wrote:
| Used Mac Minis are probably cheaper and more energy efficient
| etra0 wrote:
| This reminded me of the guy that built a meme database using
| iPhone's OCR as well [1].
|
| I find incredible the idea of giving these devices another life.
| I wonder how hard is to host a sort-of vps on an abandoned
| android phone these days... I guess as long as you can put
| ethernet + docker you'd have a very capable device.
|
| [1] https://news.ycombinator.com/item?id=34315782
| leakycap wrote:
| This was a great read!
|
| I'd never heard of a website hosted in any way on an iOS
| device... makes me wish it was an option.
| tootie wrote:
| I have an ancient ipad that is still functional but stuck on iOS
| 9. Xcode doesn't let you target that version anymore. Is it still
| possible to compile an ipa for devices out of support?
| daneel_w wrote:
| It's a painfully sluggish alternative, but you can run older
| versions of OS X (and thus Xcode) in VirtualBox.
| WalterGR wrote:
| On Apple x86 hardware: Running Windows in VMWare Fusion works
| very, very well. I can't see a reason why that wouldn't also
| be the case for old versions of OS X, though admittedly I
| haven't tried.
|
| It's curious to me that OS X in VirtualBox is sluggish. Both
| VMWare Fusion and VirtualBox use virtualization...
| daneel_w wrote:
| Software framebuffer. Remaining devices are also emulated.
| namuol wrote:
| Interesting tech but there's zero explanation of the actual
| application, so it's all a little abstract.
| unangst wrote:
| Agreed. Came to the comments thinking the same thing.
| nerbert wrote:
| A little detail in the otherwise great write up! I'm curious
| too.
| troupo wrote:
| > Welcome to my corner of the internet! I'm Hemant, a Senior
| Software Engineer based in Canada . I'm passionate about cloud
| computing, DevOps, and building robust distributed systems.
|
| Somehow you're also passionate about selling user data to
| hundreds of data brokers with no easy way to opt-out
| yegle wrote:
| This still requires a mini PC to bridge the API call and the iOS
| app.
|
| I wonder if the new Android 16 terminal app would allow combining
| both.
| ubercow13 wrote:
| I think this wouldn't work with any iPhone that's on a version of
| iOS new enough to have the 'feature' where it automatically
| restarts after a few days without being used?
| neilv wrote:
| Nice hacker effort and writeup, but I want to comment on a
| general HN pattern of what tech people promote implicitly with
| hacker network effects...
|
| For every HN blog post of "I accomplished ___ despite a hacker-
| hostile platform, and now you can use what I built, and be
| hopelessly tied to the platform"... Baby Jesus Linus sheds a
| tear.
|
| In this case, it's a bit odd, since the writer has an entire
| section, "Why This Actually Matters", of unusually good hacker
| and social values.
| rtaylorgarlock wrote:
| THANK YOU. For example, I'm currently a user of an android app
| (installed thru Play Store) which I found through front page;
| cool. The headline: "I developed <insert FOSS app which meets
| regular value prop> and didn't use <insert commonly used
| framework>". Tragically, I care less about framework than I do
| about functionality, and ever since installing, i've been left
| wondering how many of the hundreds of upvoters tried running
| what I describe as the single buggiest app on my phone. I've
| rage uninstalled multiple times in hopes of fixing issues which
| are sometimes only fixed by clean installing. My point: Guiding
| philosophies are important, and evaluating them at scale is
| critical* work.
|
| *see what i did there
| neilv wrote:
| Yes, but usually it doesn't come down to something that works
| vs. doesn't work.
|
| And some of the times that it does, it's because someone
| earlier didn't think about values before establishing network
| effects that stuffed a bad-values thing while starving a
| good-values thing.
| kennywinker wrote:
| Repurposing an old device is good. If the closed platform
| bothers you, don't buy and iPhone - but regardless of what you
| do there are millions of old iPhones that could be saved from
| the landfill by projects like this
| deadbabe wrote:
| The privacy obsession and the fact he never mentions what kind of
| images the service is processing or what they're for just kinda
| gives me the creeps, especially for the amount of requests he
| gets. There is a non-zero chance this is for illicit purposes.
| nottorp wrote:
| Maybe he's ashamed it's not done in Rust...
| CaptainFever wrote:
| It's not your business. This is just the old nothing to hide
| argument:
| https://en.wikipedia.org/wiki/Nothing_to_hide_argument
| deadbabe wrote:
| The idea of him looking outside a window at peaceful birds
| feeding, while his phone also sits in the foreground of the
| window crunching whatever horrifying OCR workloads may be
| hitting the device, is a juxtaposition worthy of cinema.
| 0xdeadbeefbabe wrote:
| More like a juxtaposition worthy of a thriller novel
| trilogy. Where's the OCR in cinema?
| bpiroman wrote:
| this is so cool! Is it possible to boot linux on an old iphone?
| hackyhacky wrote:
| I'm not sure if there's a FOSS OCR package of equivalent
| quality to Apple Vision. I'm happy to be corrected otherwise.
| pabs3 wrote:
| It has definitely been done before, but probably not recently.
| Maybe if you use the checkra1n bootROM exploit you could do it
| with lots of dev work.
|
| https://linuxoniphone.blogspot.com/ https://checkra.in/
| xydac wrote:
| Mine bailed out on a Baseband error due to which i am not even
| able to boot it anymore :(
| eddieh wrote:
| The iPhone 8 was peak iPhone. I'm on my second iPhone 8 and am
| posting from it now.
|
| I did also like the original iPhone SE mostly because of the
| size, but the haptics make the the iPhone 8, along with having a
| bezel, square screen, and home button.
| gregoriol wrote:
| Had the iPhone 8 and now on an SE3, which is the same but
| better, peak device forms and features indeed.
| nikolayasdf123 wrote:
| interesting, who (why?) is using and even paying for this
| service?
|
| 1. if you are on device, then use on device OCR (e.g. use Apple
| Vision directly)
|
| 2. if you are on cloud, then self-deployed OCR models
|
| 3. if you are on browser, then WASM/local self-deployed OCR
| models
| nottorp wrote:
| > One unexpected discovery: the phone performs OCR faster when
| slightly warm (but not hot). Cold Canadian mornings mean slower
| processing times - something I never would have noticed with wall
| power.
|
| Interesting. Apple throttles on cold too?
|
| In my experience it would shut down on cold, but I don't think I
| noticed throttling. But then I don't run anything important
| enough to benchmark on a phone...
| victorantos wrote:
| I donated my iPhone 8 a few years ago and it's still going
| strong, at least from what I heard last time earlier this year.
| Honestly impressive how long these older iPhones keep up, both in
| performance and battery life (still original battery)
| CaptainFever wrote:
| A classic related article, also using iPhones as OCR servers:
| https://findthatmeme.com/blog/2023/01/08/image-stacks-and-ip...
| seanalltogether wrote:
| I have about 7 old android phones/tablets that I would love to
| put to use as some kinda makeshift server farm, I just can't
| think of a good workflow that could take advantage of them
| elchangri wrote:
| Run webservers with Cloudflare Tunnels or ngrok. Free compute
| gganley wrote:
| Even in death, it still serves.
| crazygringo wrote:
| > _The phone's battery health held up reasonably well. After over
| a year of constant operation, it's at 76% capacity._
|
| I have an iPhone SE that I've tried keeping plugged in all the
| time and its battery has turned into a spicy pillow three times,
| first with Apple replacing the whole device (since they won't
| touch it with a swollen battery), then using third-party
| replacement kits.
|
| This isn't going to work for long if the battery is usually at
| 100%.
|
| My #1 wish for being able to repurpose old phones is to operate
| without touching the battery, and/or keeping the battery at 50%.
| Newer Apple phones have an 80% limit option which is an
| improvement, but I'm not sure how much. And unfortunately the
| option isn't there on any but the most recent phones, even on up-
| to-date iOS.
| progbits wrote:
| Most of these devices can't run "without touching the battery"
| because the external supply can't provide the required peak
| current, so during some CPU burst it would shut off.
|
| I've seen hacks that replace the battery with a supercapacitor
| though.
| crazygringo wrote:
| Couldn't the power management simply throttle the CPU to
| never go above supplied power in a battery-free mode? Don't
| they already implement a power threshold for degraded
| batteries? It seems like that would just be part of the
| feature I'm asking for, and easy to implement.
|
| It really seems like, if it weren't for the battery part,
| these phones could run for decades... but right now you have
| to replace the battery every couple years because it swells
| when constantly kept at 100% which it is not designed for.
| rollcat wrote:
| > since they won't touch it with a swollen battery
|
| Interesting. I've had a spicy pillow on a 2017 MBP, they fixed
| the poor thing, and while at it: replaced the cursed keyboard,
| and left some kind of tape to reinforce the loosened USB-C
| ports.
|
| Unfortunately, they didn't do the thermal paste - I had to do
| DIY, which is something I will never touch again. It did pay
| off though, it's cooler by some 10degC under load, and runs
| faster too. It's still loved and in everyday use.
| Eric_WVGG wrote:
| Plug your charger to any Homekit-compatible "smart plug," and
| create a shortcut that turns the the plug _on_ when the battery
| reaches 45%, and _off_ when it reaches 55%.
|
| This will of course require a Homekit hub.
| jacktheturtle wrote:
| This
| kccqzy wrote:
| A timer is sufficient. No need to be precisely 45% or 55%.
| lucb1e wrote:
| I can't imagine that a timer wouldn't quickly drift and
| either drain it to zero or charge it fully
| crazygringo wrote:
| That's an intriguing idea, I had no idea that was a
| possibility.
|
| Unfortunately it wouldn't work for my particular usage, which
| was keeping it plugged into an old but expensive smart
| speaker as a music player via its lightning port. A smart
| plug would turn off the speaker along with the phone... But I
| appreciate the suggestion, as complicated as it is!
| KolibriFly wrote:
| It's frustrating that Apple doesn't offer a proper "battery
| bypass" mode or even let you set charge limits
| kccqzy wrote:
| I don't believe a battery bypass mode is physically feasible,
| since the user could be charging the phone with a cheap 5V 1A
| charger, and yet the peak power consumption of an iPhone
| could very well exceed that.
| DinoNuggies45 wrote:
| This feels like a modern-day Ship of Theseus -- how much of the
| original phone is still actually in use? Genuinely curious what
| it struggles with most today.
| The_President wrote:
| I have an iPhone 8 still in service and compared to an equally
| old Android device, the Android (some kind of Motorola eX series
| low end phone) runs circles around the iPhone. Even playing
| background video or audio streamed from wifi and output over
| bluetooth with the screen off, the Android will burn 15% in an
| hour while the iPhone will burn over 60%. Both are the same age
| but the iPhone feels subjectively obselete while the low end
| Motorola feels like a mid-2010s computer. Even for it's age the
| Android will last two weeks in Airplane mode.
| martey wrote:
| This just suggests that the battery in your iPhone 8 is more
| degraded than your low end Motorola. This could easily occur if
| you have used the iPhone more over its lifetime and isn't a
| good measure of relative performance.
| KolibriFly wrote:
| Love the mix of "just because I can" engineering and actual
| practical benefits
| arjie wrote:
| This is very cool! I have a similar EcoFlow battery hooked up to
| a Pixel 4. With Termux this is a very capable tool. I have it on
| Tailscale (using my own headscale server) and I use it for speech
| to text and text to speech using the native APIs (which are well
| supported with Termux's APIs and helper functions). It's a very
| capable computer. The one thing that I haven't quite figured out
| is how to run a high-quality wake-word tool.
|
| I initially intended to use it with a ReSpeaker speaker/mic
| system so that I could use it as a smart home assistant / Q&A bot
| since Google Home constantly frustrates me with its inability to
| answer questions that LLMs answer flawlessly but the mic/speaker
| on the phone is good enough. The only problem is the wake word
| functionality. I'm going to try Porcupine next and see.
___________________________________________________________________
(page generated 2025-06-19 23:01 UTC)