Google will develop Android OS behind closed doors starting next week | Hacker News Reader

Google will develop Android OS behind closed doors starting next week

9to5google.com

320 pointsjosephcsible230 comments

This really reminds me of Open Solaris.

> We will no longer distribute source code for the entirety of the Solaris operating system in real-time

In the case of Open Solaris, the code never came out from that point onwards. For Android, the likely end goal is to do the bare minimum of distributing only copyleft code that they don't own copyright to. Until those get replaced with a closed alternative.

https://www.youtube.com/watch?v=-zRN7XLCRhc&t=2482s

Can someone just fork it? Like MySQL was forked into MariaDB?

Yes: https://illumos.org/

This doesn’t appear mobile oriented, or is this an open Solaris fork?

I suppose lineageOS is the closest thing for Android tho from what I understand it is built upon AOSP

Yes, an OpenSolaris fork. I thought that was the "it" you were referring to, as that was what the parent was talking about.

There are quite a few AOSP forks, like LineageOS and GrapheneOS to cite the most famous ones, but with this change in process by Google, they will have to wait a year of no changes, then scramble when the next release of Android drops.

But copyleft is bad and everyone should use BSD ...

Code is turned proprietary by huge corporate with massive development resources. No one could have predicted this.

"curse your sudden and inevitable betrayal"

Stallman was right again.

and yet, corporate drones will make fun of him, and they will keep missing the point...

Maybe they are not missing the point? Maybe the making fun is some sort of targeted hostility and opposition?

The dude is solid as it gets on all this.

Yet people ignore him and then realize that was one of those too late types

none of us have economic incentives to build global infrastrcture. Let me know what governments decide among the people that there's value in open source and it's software

Agreed, but I doubt governments would be good stewards of open source either. They engage in illegal mass surveillance, cyber warfare, and is constantly trying to undermine encryption.

in the abstract, governments are indistinguishable from any other collective of people.

So yeah, if open source orgs can keep going, there's no reason to think government can't do the same. It's about public goods.

> in the abstract, governments are indistinguishable from any other collective of people.

Sure, but in reality they have a legal monopoly on the use of violence, which is a very big deal and makes them qualitatively different from any other collective of people.

AOSP staying open to community patches in practice is "global infrastructure"?

There was an attempt in Oregon back in the early 00's. [I was part of this attempt]

The basic argument is we all either benefit or struggle more when the work of the people, for the people, is locked up in proprietary closed tools and data formats.

When open tools and data are employed, anyone can step up to improve those tools, and the State can fund the creation and maintenance of necessary tools and when that is done the savings over longer dpans of time.

A simple example:

PDX needed water billing. A 40 million bid was tendered and a private company created a system that they own and the State Basically pays them to use. And the State pays for fixes etc too.

What happens when a 40 million dollar investment is made in people using open code to do the same thing?

Well, the State owns the tool, and the data is open meaning anyone needing access can use open tools for that purpose.

When it matures, the org that got it done can fade away, leaving a small crew to maintain

, or

Maybe that org approaches other municipalities interested in similar savings. Over time, that problem is solved and most of the nation is enjoying a great savings and developers make a fine living, etc..

Wash rinse and repeat to reduce the cost of government and the work of the people is lean, mean, effective.

Everyone enjoys the benefit of a lower cost environment too.

This had broad bipartisan support and who pushed it away?

Big companies paid lobby shut the effort down hard.

They could very easily see the popular appeal and chose to spend huge now to shit it down, paying the house speaker to block it all from having a vote.

It was going to pass easily.

I think many governments can see how to think this way.

Big companies do not want it.

> Stallman was right again.

He usually is, no matter how many times people write him off.

Yep. An open source license doesn't mean anything if a project is dependent on a single company for ongoing development.

The weird thing is, why is that the case?

I'm trying to think of a mobile OS feature addition that has made me say "I need to upgrade my phone" and it just hasn't happened recently. It's more like, damn it, the dastardly thing stopped receiving security updates and now I have to replace it for no good reason.

Isn't Android done yet? What further development is required that couldn't be done by the community?

The point is it doesn't have a developer community that could maintain the project without Google.

Porting lineage to new hardware mostly includes development teams that have nothing to do with Google.. Convincing no one to leave the thin veneer of mediocre code and "free" storage is a game that is harder to play than Apple makes it look with OsX.

It doesn't have one because Google is doing it and then publishing the code so there is no point in someone else redoing what's already been done.

If they were to stop, the demand for someone to do it would still be there, and that demand wouldn't be getting met anymore, which creates the incentive for others to do it.

Meanwhile the point is that most of "it" doesn't actually need to be done anyway. You don't need to do everything Google is currently doing. Adding support for new hardware is important, but that has an obvious source of someone to do it because the hardware vendors want their new hardware to be widely supported so they can sell more of it. So all you really need is security updates, and a community can handle that as evidenced by the many instances of it actually happening for other code.

What stops the thing that makes Debian work from making this work?

Debian doesn't "work" like Android works. Almost no end-user runs Debian on any of their devices because no one wants Debian over anything else. If you want to achieve Debian's stunning success of having almost no consumer adoption, you should follow its model of community development.

You're right, if Google steps away from Android completely then there would be incentive for others to do it, another megacorp will step in. Maybe Facebook or Microsoft or Samsung.

Millions of people run Debian, even though it has zero marketing budget and is competing primarily with Microsoft Windows (infamous for anti-competitive practices) and macOS (a largely competent operating system with the backing of another multi-trillion dollar megacorp that itself increasingly uses anti-competitive practices).

Meanwhile there are hardly any devices that come with it, because Macs come with macOS and Microsoft exerts pressure on PC OEMs, so all of the people running it are people who explicitly did want Debian over anything else, as opposed to many millions of Windows users who have no real preference or an active dislike of their operating system but got it by default with the hardware and may not even realize that anything else is available.

If Google stopped developing Android, it would still be one of the two major incumbent platforms and people would continue to use it. It might even get better because third party apps would have to stop depending on proprietary Google APIs/services and then the community could strip out the Google spying code without worrying about losing access to those APIs. So then the question isn't how to get a critical mass of users -- that's already there -- you just need basic maintenance of a stable code base, which is a thing the community can demonstrably do.

Yeah millions of people out of billions, I was talking in relative terms. Good to know that Debian could have a fighting chance in a world where everyone had zero marketing budget and there weren't any rich corporate backers that used anti-competitive practices.

Anyway in the world we both live in, if Google abandoned Android then most people would instantly switch to a megacorp fork of Android with that megacorp's own proprietary APIs and services because people will follow the proprietary things they care about like for eg fast & battery-efficient centralized notifications, an out-of-the-box app store with popular apps like Instagram, and tap to pay.

But I didn't have to explain this to you, you already know this. You know this because "millions of people" is derived from you knowing that the peak Linux desktop marketshare is like 4% out of billions of people. You know this because you said users are "worrying about losing access to those [proprietary, spying] APIs" which is why megacorps who provide these proprietary spying APIs will actually win over users. You know this because you know friends or family or colleagues who are aware of Debian and still don't choose it because they rely on some proprietary service or API that Debian's community developers have never given a rat's ass about.

> Yeah millions of people out of billions, I was talking in relative terms.

In relative terms, Linux market share is increasing and Windows market share is declining.

> Good to know that Debian could have a fighting chance in a world where everyone had zero marketing budget and there weren't any rich corporate backers that used anti-competitive practices.

We could enforce the antitrust laws, yes.

> Anyway in the world we both live in, if Google abandoned Android then most people would instantly switch to a megacorp fork of Android with that megacorp's own proprietary APIs and services because people will follow the proprietary things they care about like for eg fast & battery-efficient centralized notifications, an out-of-the-box app store with popular apps like Instagram, and tap to pay.

None of that requires anything proprietary in the operating system.

Centralized notifications are implemented as a lock-in mechanism. Idle TCP connections don't consume battery unless they need keepalives, and in the latter case you provide applications with a non-proprietary API to have the OS handle keepalives by sending them together for any open connections that need them. Then the radio only has to wake up the same number of times it does with a single connection and there is no real advantage to centralization.

App stores are likewise only glued to operating systems for anti-competitive reasons. Spending 30 seconds once to install one that didn't come with the OS is such a low barrier that it can't be the thing preventing anyone from choosing an OS, and apps can be listed in more than one store, so there is no reason to expect any one store to dominate the market in the absence of anti-competitive practices.

The way tap to pay ought to work is you tap to get a payment request from the merchant which is then passed to your bank app using a standard protocol to make the payment, and then money is transferred from your bank to the merchant's bank with no intermediaries leeching a percentage. In the absence of sane regulations allowing this, you could also use any existing payment processors, but this is still something that an app does and not something that the OS does and the app doesn't have to be from the same entity as the OS.

> You know this because "millions of people" is derived from you knowing that the peak Linux desktop marketshare is like 4% out of billions of people.

Once again, Debian isn't what came with their computer. "Most people keep the defaults" works the other way when the default is Android.

> You know this because you said users are "worrying about losing access to those [proprietary, spying] APIs" which is why megacorps who provide these proprietary spying APIs will actually win over users.

The proprietary APIs don't provide anything good, they exist for the purpose of lock-in, because then third party developers use them without realizing or caring that it creates a dependency on proprietary code or services, since the existing installed base of phones that don't provide them is negligible.

Linux often does provide implementations of these things (e.g. wine), and certainly provides its own non-proprietary alternatives to them, but because the purpose of those things is lock-in the incumbent takes measures to prevent interoperability.

If there was no one providing proprietary APIs to begin with, or the antitrust laws were being enforced as they ought to be, that wouldn't be an issue. As it is, Linux market share keeps going up, but slowly, because the incumbents fight tooth and nail to keep the users in their cages.

Only if they stop pushing code changes to AOSP...

I hope they make it closed source and make us much money out of it as possible for shareholders that is their job and their duty, why are they giving this away for free they have already captured market share by claiming opensource and building a community now all they have to do is make it proprietary and the old opensource version slowly wither and become unstable, then they can charge money for the operating system just like mircrosoft but this time on phone millions and billions of phones, $$$$$. /S (I obviously do not agree with this)

It's a bit disturbing that that might be the plan though.

But they can already do that, because most people and apps (outside China and maybe some other regions) expect Google Play/Play Services to be present.

They're under antitrust scrutiny. You have to boil the frog slowly.

What apps? I think only android auto requires it. All the rest works just fine with occasional glitches. Of course that most probably greatly depends of what kind of apps you are using but usually microG is just enough.

No, most "geo" applications depend on Google services and will crash without them. Local authority transit apps, etc.

I've been running GrapheneOS without any Google Services. Most things work without issue.

In the past 2 years I've encountered a total of 3 apps that wouldn't run without Google.

How and why, any idea? Do they have some kind of backend check verified with Google that might return “naah, not a Googled device”? By the way which apps?

Many banking apps will outright refuse to run on rooted devices, much less google-free forks. There are ways around that but those ways are unreliable and could break at any given time -- I am not risking losing the ability to pay for things online just for the sake of running an android fork. (in my country, all the banks switched to requiring 2FA from their app if you make online purchases with your visa or mastercard)

Right, but if Google's Android becomes closed source and a well-funded FOSS fork becomes available, that changes the situation. If Samsung, Xiaomi, Motorola, HONOR, OPPO, etc. all agree to use a new FOSS fork as their base, well Google's new closed-source Android becomes irrelevant. Samsung alone is large enough to be able to maintain a fork of Android, and a large enough percentage of the smartphone market that whatever OS it's running will be supported by banking apps.

> Samsung alone is large enough to be able to maintain a fork of Android

Samsung would rather not - they threated this card once before, while negotiating for Google to get rid of Motorola, and their bluff got called. Samsung tried to prop up Tizen as an Android alternative. Samsung since closed a number of its US OS offices - why sacrifice profits when they have a cozy arrangement: Samsung & other Android partners will continue to get the Android previews before anyone else: open source or not.

Samsung recently deprecated their built in SMS/texting app and put an advertisement to tell us to switch for Google Messages in their app. They threw in the towel and not only will not maintain an android fork, they don't even want to maintain their own apps anymore.

I agree with you: they'd rather not. If Google's Android became closed source to the public, but Google gave Samsung and other OEMs the right to modify it (including a hard fork later on if they desired), then I agree with you that they'd almost certainly continue with Google's Android — they'd lose nothing and eliminate future competitors.

My comment was in part addressing the higher up comments in the thread stating OEMs couldn't do a hard fork. My thoughts are that they have the marketshare that if Google's terms were bad enough, they could. They'd love to take some of the Play Store revenue, but currently dropping the Play Store would tank hardware sales as competitors would keep it. But if Google's terms were to get bad enough that multiple OEMs wanted to hard fork, that calculation could change. I don't foresee Google ever putting forth that bad of terms though, in part because of the option to hard fork.

FWIW, my banking app works very well on /e/OS, which is an AOSP derivative.

No Email or TOTP? (No, I am not suggesting sms)

In my country one bank had a 2fa app. Then they backtracked on security, but kept policy: they included the 2fa in the regular bank app. Now you don't have to use 2fa if you are using the bank app, because the bank app generates its own authorization, in the same device (app) without user interaction!

Fake admiration off. We also don't have to use any 2fa when we access banking through the website. (works on FF on Android)

> No Email or TOTP? (No, I am not suggesting sms)

Nope. And accessing the website/online banking portal is not allowed without going through the smartphone app 2fa too.

I think banks requiring you to buy hardware you don't actually own (a blessed android phone) to run software you can't control (this banking app) so you can access money you worked to make is absolutely ridiculous and dystopian. Why do you allow a bank to do this to you?

I sincerely hope there is some alternative option in your country. In mine, I can still perform banking activity by going to a physical branch, by calling in, or by using the website with a physical 2FA token (i.e. not my phone). The bank keeps trying to get me to switch over to their app but I will continue to protest this until it's no longer possible to not use an app at which point I will likely switch banks.

That is sad. I do not have a problem with my 2 banking apps and Revolut. But I am running MicroG version of Lineage OS.

So no google, but still works. I think this is worth a try, considering how many adds you have to see on Android running full Google (which I have just one to be able to use Android Auto inside my car).

What fork is that? Also, how easy peasy will it be to run it on a Pixel 5?

Amazon maintains their Fire devices with their Android fork.

Huawei maintains their Android fork that runs without the Play store.

Google will not be able to close Android/AOSP without triggering a well-funded fork.

The Fire devices seem like they are on life support from a technical point of view (even if Amazon is making money from them). Their fork is really out of date.

Also Meta Horizon OS is based on AOSP. This will probably also affect their development.

Why would a second company freeloading kernel and services from Google change Google's OSdev roadmap? Google could just say to Meta 'fine, fork and maintain it yourself, you're a grown-up company' and continue on with whatever nefarious rug-pull closed source pivot to Fuschia gambit that they might be plotting.

I doubt at this point corps are not looking at that option. But then I don’t know what kind of contract these corps have with Google. I have worked at Samsung and they are such a large incompetent but aware software clusterfuck that they would rather want an almost readymade OS on which they will smear paint and sell. Can’t speak of other OEMs.

Isn't the plan to move Android to Fuchsia eventually (and seamlessly)?

Hasn't this been the plan for like a decade now? I know some of their products run Fuchsia, like the Nest displays, but no word on main Android devices

If they did ever have such a plan, they've clearly long since abandoned it. I bet they entertained the idea at one point but it was never seriously put into action.

As far as I can tell, Google has never stated that as a goal. It had always been a hypothesis.

The plan was initially to merge Android & Chrome and create "Andromeda" - https://arstechnica.com/gadgets/2016/09/android-chrome-andro...

Then they decided that Fuchsia was going to be the way forward - https://www.thurrott.com/mobile/android/117587/google-allege...

Now the latest is that they're once again merging Android & Chrome - https://www.androidauthority.com/chrome-os-becoming-android-...

But Fuchsia is still being actively developed, not sure to what end - https://fuchsia-review.googlesource.com/q/status:open+-is:wi...

As far as I remember, Fuchsia is used for all Googles smart devices now. So maybe just for them?

if it really is, they should take note of exactly what happened to Oracle when they acquired solaris and mysql and turned them into proprietary applications.

Whatever investment they had made in them literally evaporated in a week as mariadb and galera showed up. OpenIndiana basically made continued solaris development at Oracle a moot point, not that it wasnt already with Linux on the scene.

RedHat has tried something similar with CentOS, Encumbering it to try and drive sales, which backfired just as predictably. Rocky is a treat to run.

Rolling up Android into a proprietary walled garden would be a disaster. This isnt apple. What you could expect is a massive developer exodus from the open community to other friendlier projects. If your interest is western security/hegemony in technology then it would be a shame to see all that intellectual capital suddenly captured by a FOSS project from a marxist leninist country thats all too happy to give it away for free (DeepSeek anyone?)

>What you could expect is a massive developer exodus from the open community to other friendlier projects

You act as if anyone to a first approximation cares about indie developers. Most of the popular apps on Android and iOS are from the big corporations and pay to win games. They could care less about ideology.

As much as I - a Pixel 7 (GrapheneOS) user - would hate that, doing so would essentially put Android on equal footing as iOS and would give me a serious reason to consider switching to an iPhone. Apple's hardware is just so much sleeker, faster, and better than Google's mediocre Pixel line.

Maybe I'm just a tech philistine but I find my Pixel 8a and my wife's iPhone <N> to be basically indistinguishable in any meaningful way.

They are completely different. One of them lets you have the sexy blue text boxes in iPhone group chats, and the other makes everyone see you with the green box of shame.

Unless you live outside of the US in which nobody gives a damn and they all use WhatsApp. After moving, I only get sms messages from the doctors office and my friends back in America.

Not true at all. This effect is pretty common in countries outside of US. It makes sense. WhatsApp is just another messaging app thats popular depending on a local whim. Where as iMessages get sent automatically between iOS users.

What countries other than US is this so prevalent?

According to https://worldpopulationreview.com/country-rankings/iphone-ma... iPhone also has > 50% market share in Canada, Norway, Sweden, Australia and Japan. FWIW I live in Australia and most group chats are via WhatsApp vs SMS/iMessage.

I didnt say this is so prevalent. Just that it exists.

My experience from few european countries is that middle class - tech/business/law people have iOS. Go to tech or business conference and its all iphones.

So its really easy to be in such circles. I live in EU country and its all iMessage or Signal. Nobody uses WhatsApp if something its Facebook Messenger or Instagram messages.

Outside the US, and a handful of other countries, Android is far more dominant than iOS (75% vs 25%). [1]

WhatsApp has an estimated active userbase of approximately 3 billion. [2] The number of iMessage users is estimated to be about 1 billion. [3]

[1] https://explodingtopics.com/blog/iphone-android-users#iphone...

[2] https://www.statista.com/statistics/1306022/whatsapp-global-...

[3] https://usesignhouse.com/blog/imessage-stats/#:~:text=iMessa...

On [1], If you only look at countries with decent GDP per Capita. iOS will have anywhere from 40 to 65% market share. Even in China iOS is well over 30%.

When you consider Africa + ASEAN + India has 3.5B population and has very low iPhone market share that sort of Skew the figures.

Yeah, weighting the figures by GDP makes sense. Indeed, poor people aren't really people at all.

Because it doesn't show an accurate representation. iOS is growing at a faster rate than ever before in developing countries, Apple achieved 11% shipment market share for the first time in India in Q424.

Sorry but how that invalidates that in my country nobody uses WhatsApp? It doesn't matter at all if they are on Android or iOS.

And since iMessages are seamless they are used very often between iOS users.

Y'all bitches gotta get on Signal. ;-)

I hear even the US government trusts Signal

I'm sure there are dozens of people in Germany who actively prefer iMessage but I haven't met one yet. Whatsapp achieved pretty much total market capture here back when SMS still were costly and the network effects that arose from that are among the strongest I've ever seen. I'm pretty sure if someone was to do a survey, almost everyone would say Messages is for SMS only and I think most of them wouldn't know it can do more than that.

In Berlin everyone seems to be on Telegram these days. People leaning more left and/or are privacy-aware and/or hate FB/Meta/big tech use Signal.

WhatsApp seems only to be used by the elderly/old in my circle of friends lately.

The transition has been gradual; started during the pandemic, I'd say.

It doesn't surprise me. Telegram is surerior both technologically and UX side.

I have impression that WhatsApp team just dont do anything for years.

Yet i write with most of my German friends using iMessages because its automatic and i dont have Telegram or WhatsApp because nobody uses it in my european country.

> Unless you live outside of the US [...] and they all use WhatsApp

This is such a ridiculously incorrect over-generalisation. China, Japan & Russia are obvious counter-examples, plus many others.

i don't think China uses SMS though.

In China WeChat is the dominant messaging platform, SMS is only used for delivery notifications, spam, etc.

My understanding is Line is on top in Japan.

I understand that. My point was that the original comment was grossly overstating the supposed 'dominance' of WhatsApp.

The real point of the original comment was that outside the US iMessage is not dominant and so nobody cares about what color a chat bubble is.

I appreciate that I was a bit too confrontational in my first reply and should have just added that WeChat, Line and Telegram are also used (plus many others), not just WhatsApp.

Why I think I (over) reacted is that it was, to me, an example of only partial escape from US American insularity. They understood that ppl outside the USA don't use SMS much, but only suggested a US American messaging platform as what was used instead.

Did we read the same thing? There is nothing about color of chat bubble but there is "they all use WhatsApp". No they don't. Not all. More like half.

Whatsapp is dominant in most countries in world, it's not overstating. Your 3 country example mean nothing in comparison.

If by most you mean 50% then sure. But the other 50% countries prefer to use different app. Be it Facebook Messenger or Telegram... In my eu country nobody has WhatsApp and its not uncommon. Network effects are at play so it's what became popular first.

True, but God does WhatsApp suck.

They've got a billion-strong userbase and yet the Android app still dumped literally every attachment I received into my camera roll until I manually added '.nomedia' files in the right places.

And, oh man, the API for businesses is a Kafkaesque nightmare. Maybe it was good before it got Zucked, but I had to fight with their support for over a week to get an automated ban-hammer overturned... only for it to get auto-banned again two days later. We hadn't even deployed the damn thing yet!

Dont know about Andriod but iOS you can select not to save images into camera roll.

That's a feature, not a bug Users are conditioned to look there for any kind of image on their device

> literally every attachment I received into my camera roll

There's a global setting for this: media visibility: off.

Then you can enable that per chat.

I have all chat images contained to only Whatsapp by default. Then I only enable some chats (family, friends) to expose images/videos to the phone.

Google Photos scoops those and backsup all family pictures shared to whatsapp to my Google Photos.

I'm the family's reliable source of truth when it comes to family photos.

What's "camera roll"? WA images go into the "WhatsApp Images" folder. They're separate from "Camera".

Every Android "gallery"-type app I've ever used defaults to a view that mixes together every known album/folder on the device.

That works fine (and is in fact easier than remembering the exact album) when I get to pick and choose exactly what I photograph/save, but it became borderline unusable once WhatsApp started vomiting hundreds of stupid GIFs and throwaway screenshots into it.

Neither of them causes anyone else to see your messages as green. Or blue.

The camera - the only reason I would buy an iPhone is for the better camera.

Sure, the Pixel 8a camera is not bad for the price but it's still noticeably worse. The kind of difference you notice when someone with an iPhone shares photos with you.

Apps and the whole phone experience are a sh*tshow on both sides and I hate both with a passion. I'm still waiting for a decent linux experience on a phone - possibly with stupid banking apps support.

Whenever they do these blind tests for different phone cameras, the iPhone never wins. https://www.youtube.com/watch?v=VRoTOE3FqT0

Pixels have been beating iPhones in camera quality for years now on pretty much every blind test. Even the cheap ones.

I play videogames and I can tell the difference between 60 and 120herz screen. Always amused me that a so called premium device had worse hardware than a flagship killer. Apple apparently finally caught up with their latest phone?

High refresh on its own is not enough to constitute a better screen. While Apple could certainly stand to bump the minimum refresh rate to 120hz across the board, I wouldn’t want that to come at the expense of the other more important specs as has been common outside of the Apple sphere, particularly on budget devices. If one has to choose, a color accurate screen is preferable to a fast but less accurate one in most circumstances for instance.

Almost anyone can tell the difference between 60 and 120 Hz.

Seriously, if you haven't upgraded your desktop PC to a higher refresh rate screen yet: It's the biggest "feels like a new computer!" upgrade since we swapped HDDs for SSDs and the days when your new CPU was 2.5x as fast as the old one. There is no turning back after having experienced the buttery smoothness, and the impact is IMHO higher during regular usage than during games.

I upgraded to 144 hz on my primary monitor in 2016 and it's absolutely noticeable and makes things feel so smooth.

Last week I went to 240 hz and while it's noticeably even smoother, it wasn't nearly the upgrade, so there's certainly diminishing returns. Though I did go from IPS to OLED and THAT is really nice.

This is usually the point where someone will chime in and say something dumb like "The human eye can't see more than X frames per second" which is just hogwash. It's not about individual frames, but the fluidity of motion. At 60 fps, an object moving across the screen is moving 4x as many pixels per frame as 240 fps. When you get used to 240 fps, 60 fps feels like it's strobing.

I recently bought a 240 Hz OLED monitor as well, but to my dismay the 80 Gbps data rate mode is actually optional in DisplayPort 2.1a. Even though I have a DP2.1-capable GPU (9070 XT) and monitor, it's not possible to do 4k 10bit HDR at 240 Hz with DSC off. Since I don't want compression to be on I've compromised to sticking with 180 Hz, which as you say due to the diminishing returns above 120 is still plenty.

Still quite frustrating that the display industry did it again in specifying a standard that makes most of what's interesting about it optional, so everyone can print it on their boxes without delivering the expected value.

Why do you need DSC off?

Psychology more than anything, to be honest. I know it's supposed to be visually lossless, but it does touch and nudge many pixels, and I just want the raw image. I do dabble in graphics and UI frontend at times, and I don't want to take chances. It's something that spending a little more time digging into the algorithmic details of DSC might address for me, but I haven't so far.

DSC has Rate Control, so the very minimal compression needed to go from from 180 Hz to 240 Hz will be negligible except in artificial scenarios like random noise. DSC is intended to be visually lossless up to 3x compression levels.

Meh, I turned back. While 120+hz is nice, there are more compelling attributes of a monitor that I wouldn't compromise on in the name of refresh rate. Aspect ratio, size, resolution, picture quality, and viewing angles are all things I personally value more than refresh rate. If I could get my exact monitor that quite nicely meets all of those criteria, but with 120+hz, I still wouldn't value it so highly that I'd pay more than $500 to replace it, but I would pick it over 60hz if needing a new screen and everything else was already matched or better.

It's an impactful and noticeable upgrade in addition to everything else being awesome, but for me it doesn't come close to being the the most important. If all else was equal or better, and I had to pick between 6k resolution or high refresh rate, I'd have a hard time picking refresh rate, but I'd prefer both.

Apple phones have had ~120hz screens for many years now. Only on the Pro models though, not the cut down economy models, that’s how they force people with disposable income to buy the Pro.

In my understanding, animations cannot be fully disabled on iOS (please tell me if I am wrong) while they can on Android. This leads to much better usability since you don't have to wait after every interaction.

You can enable reduce motion[1] which effectively replaces all transitions with a brief fade.

https://support.apple.com/en-us/guide/iphone/iph0b691d3ed/io...

Parity will only be achieved for me when the iPhone supports FDroid, and allows the replacement of default apps with apps of my choosing.

I'm in a similar spot. There are a few pixel exclusive features that I would certainly miss but I spent a few decades not having a personal assistant robot screen my calls and texts, I can survive without it in the future.

Tasker used to be in a class of its own but I believe shortcuts is now as powerful and it even has a user experience that isn't hostile! That might be a net benefit...

I hate the iOS keyboard and method of text selection but I could adopt.

I'll have to re buy some apps or find alternatives but that's not an impossible hill to climb.

The biggest pain points are file management and notifications. Having spent a decade plus on a blackberry before going Android full-time, neither dominant platform is even close to good with respect to notifications but Android is far less crappy than iOS.

File management is probably a deal breaker. Every time I have to download a file on my iPad and try to use it in another app or even just get it off the damn thing, I spend 5 minutes swearing before I just give up and attach the file to an email and then go to a PC to pull the file out of the draft folder...

my chief grief with Apple is the same as yours, I solve it by using a home server w/ self hosted pwas to do everything this pocketable "computer" should be able to do. Sometimes I wonder about flying to the EU to reactivate my phone there if that's possible but knowing Apple there'll be some terrifying kafkaesque twist

> Sometimes I wonder about flying to the EU to reactivate my phone there if that's possible but knowing Apple there'll be some terrifying kafkaesque twist

IIRC you needed to be like in the region like every 30 days, else it updates to your current location (but don't quote me on it, I might be really misremembering the company/product)

And fixes their garbage autocorrect and notification UI

  Apple's hardware is just so much sleeker, faster, and better than Google's

On the other hand I was recently testing a friends pixel phone and was shocked by the speed and integration of Gemini.

It’s important to compare apples to apples. Certain phone models have different CPU’s etc. I’d love to see a benchmark of iPhone 16 Pro Max vs Pixel’s top of the line model.

It's not going to be possible because Apple hasn't come out with an equivalent to Gemini. It would be like if Siri also had full LLM conversation capabilities, and access to all your documents/email/etc, and direct integration to run web searches, analyze web pages, etc. And it's fast enough to have a real-time conversation[1], as if you were speaking to a human on the phone. Oh, and that's on my Pixel 7, which is two years behind.

[1] https://www.youtube.com/watch?v=6Ks21fpjH9M

The Pixel line has never, IMO, been the best hardware in the Android ecosystem. It might be the best hardware/software combo, but Samsung's hardware has always seemed better.

> doing so would essentially put Android on equal footing as iOS

No it wouldn't. Google as an org is bad at product and the fact AOSP exists is not why.

I've built AOSP based products multiple times over the years, and closed source Google Play Services has spent years picking off ever increasing swaths of the user-facing functionality covered by AOSP. I mean the writing was on the wall with Doze, but we don't even have a calculator anymore last I checked.

Google just can't make good products like Apple can.

Apple's worst products come from moments where they act like Google (becoming developer driven with weak top down direction), and vice versa. Fortunately for iOS users, neither org defaults to acting like the other.

>and better than Google's mediocre Pixel line

Pixel 7, or any android in that era would definitely be slower than iPhone. ( Google Pixel itself uses mediocre SoC ) But the recent ones are catching up fast and latest Samsung is Snapdragon Elite is actually faster than iOS.

I think that is partly because Google had to optimise the hell out of its software due to slower CPU performance. And partly just Apple's iOS has fallen a lot in quality.

On the other hand, this would breathe new life into the world of open source cellphone software that isn't Android. It has existed in a limbo of free and corporate for over a decade and everyone gave up on alternatives like FirefoxOS.

If you actually read the article, they explain that it isn't actually that big of change to how it was done.

" For a while now, Google has been developing most parts of Android behind closed doors in its “internal branches,” with the “AOSP branch” only having certain other aspects of Android’s framework (including Bluetooth, kernel, and some other core components). As such, it’s been quite a while since the current state of AOSP is at the same level as Google’s internal builds, leaving developers and others to wait on Google to make a public release to get all of the new changes.

With this change to move everything to its internal, private development branch, Google isn’t changing the speed at which these new builds arrive. Rather, this will potentially streamline the process and prevent conflicts when merging the branch"

Hmm while this could be sold as just to prevent leaks (which occasionally happen), I think this is more likely a first step to closing the source, in light of the EU messing up their open source monetisation strategy.

Well, more like the 5th step really. They already moved a ton of functionality into Google Play Services, and discontinued a load of the open source stock apps like Calendar.

Google kinda irked me when they purchased Motorola - I was hopeful we'd get nice hardware with the android team focused on making android better but they instead gutted the company and sold it off and android has been more solid on non Google devices from what I can tell. From pixels not having an operating system shipping to people to pixels 4 and 7 having notable worse camera features (I should note I'm biased to how well the camera works since it's one of my three main uses of a smartphone) compared to say the Nexus 6 phones.

I've wondered if android and pixel breaking off Google would be a good thing or not honestly since there is talk about cutting chrome off.

The real hidden gem in that deal would have been keeping Motorola's cable set-top box. It owned more than half of the cable box market, and would have been an opportunity to push YouTube TV and Google TV into the hands of the majority of the market overnight.

A classic Ars story I always like linking in discussions like these: https://arstechnica.com/gadgets/2018/07/googles-iron-grip-on...

This headline really is misleading. The source will still be released, it just means that the work leading up to a release will be in private. IMO, there's nothing wrong with that, since it's likely that a lot of the intermediate com mediates are likely just noise.

I think the determinant will be how transparently this process is maintained.

Right - Google has been less than stellar with Chromium with ad revenue motivations behind manifest v3. One can only assume that other changes are being pushed and it'll be too late to fight when the OS ships before or same day as source.

As others say, it breaks contributions and any chance that other forks will keep up.

First they came....(Explains the strategy)

If I understand correctly, it sounds a bit like Valve and SteamOS. They publish the source, but they optimize for their internal developers and tools, not for easy source access by anonymous members of the public. Valve has been publishing its source as tarballs, not git commits (hence, projects like Jovian and Evlav reconstructing it in GitHub).

Which is bad since all the forks will be behind at least few weeks, months maybe.

And think about all the merge conflicts you have to resolve after 3 months of code changes.

"likely just noise"

It's always noise to people that don't care, but matters to people who rely on AOSP, including third-party ROM developers.

You are saying that non-Google parties are reliant on AOSP being developed in the open on a public Gerrit instance? In what way?

We can be less blindsided when the latest wave of anti-consumer features are announced.

Where do you think LineageOS gets their feature updates and security patches?

I don't understand the difference between what is described in the article and how Android has been developed from the start. They have always developed new versions internally and then dumped into AOSP right before release. https://groups.google.com/g/android-building/c/T4XZJCZnqF8/m...

Some projects developed directly in AOSP. Those projects will now develop on the internal branch like the rest. So it's not as big of a change as some are making it out to be.

Additionally, that's already the case for much of the Android projects. The remaining projects that developed directly in AOSP will develop on the internal branch like the rest.

They don't publish the really interesting sources anyway, this doesn't really change anything.

>The source will still be released

Lol nice pipe dream. The source isn't even fully released today.

The source for AOSP. Individual Android devices have never been open source (minus some very few exceptions), and they have no obligation to do so as AOSP is Apache/permissively licensed.

Note that the demand for secrecy is primarily driven by device manufacturers, not Google. Manufacturers want to keep their "secret sauce" from their competitors.

Google has gradually been moving things from AOSP (open) to GApps (closed). Some of the things that have been moved are fairly essential for a mobile operating system (like a location provider and a SMS app). Projects building on AOSP now have to provide replacements, or declare them out of scope and punt.

This is technically how open source is supposed to work. There never was any obligation to develop in public, or technically to release the source code publicly either. There is no obligation to communication, bug requests, etc.

The obligation is specifically to provide the source code (without certain usage restrictions) for binary releases when requested, and no more.

Right, they’re making the Gerrit instance private.

Developing in the open would make contributions from the community way more viable, would give the public the ability to see what's coming and prepare for it, would increase the likelihood that security vulnerabilities or other bad things are discovered and prevented early on. It would make the project more likely to serve the interests of its users.

Both The Verge and 9to5 are citing this as the original source:

https://www.androidauthority.com/google-android-development-...

Thanks for linking it.

I’m glad they’re dropping the pretense of being an open platform. Maybe this will create space for a truly open mobile platform that respects our privacy even better than Apple. I never liked Android anyway, felt half baked and unpolished even on several flagship phones I used it on some years ago before giving up on it.

what did you switch to?

The only reasonable alternative for now, Apple. Only thing worse about their devices than the competition is the price. One day they’ll turn as evil as Google, but hopefully other viable options will exist by then.

How come Android, that you can build yourself, is worse than Apple, which had plans to use CSAM?

I've seen that one before. Next thing you know licenses start changing, features locked out and completely removed and development slowly starts creeping towards full proprietary prison. Time to throw some money at Jolla and Sailfish OS and migrate I suppose.

Android was really cool for the first couple years, but now it's optimized for brand presence and ad delivery. Every Android device I've purchased for the last 10 years there's been a common expectation of the manufacturer trying to totally capture, record, sell, and leverage my digital experience in ways you never wanted. The only escape was LineageOS, and the technical barrier was plenty hard to make it work just right.

I'm currently on iOS and as many other commenters said, I'm still waiting for a truly open and privacy-respecting OS that we can install on open hardware. It worked for Unix in the computer world.

More and more Apple has been adding ads into iOS. Always have another Apple credit card ad in wallet, or Apple Music ad in settings, etc.

And lots of people wants to yield full control of the web to google by letting then be the sole web browser engine...

Just do what Apple does, less likely to be sued by DOJ

I bet they'll regret turning their back on grapheneOS https://grapheneos.social/@GrapheneOS/114230671644378529

the definition of open: “mkdir android ; cd android ; repo init -u git://android.git.kernel.org/platform/manifest.git ; repo sync ; make” - Andy Rubin

This has really stood the test of time...

It's funny because this only works on a single device, the Android emulator, and not even first-party Nexus and Pixel devices without first downloading hundreds of megabytes of proprietary blobs.

I was there and applaud when Google shared their "Don't be evil" motto. Everybody used to absolutely hate "M$" at the time.

How naive we were. We never realized that "Don't be evil" was not a choice, it something that natural happens to a public company. Today it would feel so much of quixotism for a company were to to come out and say "Don't be evil".

It will remain open source

> This does not mean that Google is making Android a closed-source platform, but rather that the open-source aspect will only be released when a new branch is released to AOSP with those changes, including when new full versions or maintenance releases are finished.

Too bad. There should be a proper open Linux + Wayland for mobile, not Android Frankenstein anyway.

That's what https://plasma-mobile.org is supposed to be.

But the problem is there is yet to be a single device that has every hardware feature working %100...see https://wiki.postmarketos.org/wiki/Devices#Main

Right, plus it needs a better backing to make a nice UI. Plasma mobile is more of a POC, not really a polished user experience by far.

One good thing Google could do is demand upstream open drivers from hardware makers.

Help me understand what this means to e.g. GrapheneOS, please. Will it be able to exist and just have to wait longer for updates or will it be in real jeopardy?

Potentially longer gaps between major updates, since they can no longer follow upstream closely.

However, Google might have planned to slowly make Android proprietary, which means death to 3rd party forks.

Most of those forks twiddle a lot of low-level knobs, and if Google does not want to support those then the forks will have a hard time anyway.

The big problem is that aaaaaalll these forks are still just a tiny tiny tiny drop in the bug mobile phone OS bucket.

If the forks want to be sustainable they need to cater to the market a bit. (Of course that's much harder said than done, but we see - for example with Nothing Tech - that there are new successful upstarts from time to time.)

Discussion on the GrapheneOS forum: https://discuss.grapheneos.org/d/21231-google-will-develop-t...

https://grapheneos.social/@GrapheneOS/114230629374651118

> not much will change overall. It's a major step in the wrong direction but without a large direct impact on us

They might be cooking up their own silicon, like Apple did when transitioning from the A-series chips to the M1. If that's the case, they'd definitely want to keep any low-level system tweaks or new APIs under wraps.

This whole situation also reminds me of the Manifest V3 drama with Chromium. They might be planning something similar and want to avoid the early headaches of public criticism.

Or maybe they're doing a bigger overhaul of Android's core, changing how apps can communicate and control each other (e.g., computer-user, Operator)!

Idk, I dnt want to be cynical fingers crossed.¯\_(ツ)_/¯

Don't they already have their own silicon (Tensor)?

Kind of. If you look at the specs of the tensor chips they appear to essentially copy exactly what Qualcomm is doing. They use off the shelf ARM reference design cores packaged together the same way QC does. They’re also about a year behind QCs latest stuff.

In comparison Apple (and the very very latest QC chipsets) use custom ARM cores. Google has yet to do this.

Keeping android open source would be a helpful antitrust defence when some of Alphabet / Googles other products dominate markets (chrome, search). I wonder if they're now less worried about anti trust under the current administration.

Google has had a strategy of keeping some parts of Android open source for the benefit of the world, and other parts closed source inside Google Play Services. However European regulators continue to rule that Google Play Services itself is the leverage Google uses to lessen competition. If regulators have antitrust concerns with just one proprietary blob, there's really no incentive to make Android truly open. Opening up the rest of the Android literally doesn't help with antitrust.

Will they still push all the native commits once they release?

I think it's OK if they do the dev internally, but if they just release source snapshots and we lose visibility of the dev process that occurred to get there, that will be a big loss.

>This does not mean that Google is making Android a closed-source platform, but rather that the open-source aspect will only be released when a new branch is released to AOSP with those changes, including when new full versions or maintenance releases are finished.

I'll believe it when I see it. These days, words are basically meaningless from these large tech companies. Actions are what matter.

at the very least, I'm not convinced this internal branch and AOSP will be close to feature parity if they do throw some stuff out.

The good thing is that at least the kernel and the VM/runtime need to be open source

the VM/runtime need to be open source

Isn't ART Apache 2-licensed? They don't have to provide the source. I don't see why they would completely close it though. It would lead to forks, plus they have a lot of control through Google Play Services etc. already.

Parts are GPLv2 with Classpath exception:

https://cs.android.com/android/platform/superproject/main/+/...

yeah it’s mostly the runtime which is basically openjdk bits.

Until they swap out linux for fuschia!

....any day now...

No OEM is interested in AOSP builds, because no one is going to buy an Android phone without Google Play Services, which is required for many popular apps to work (banking, games). There were some Chinese OEMs that ran an Android build and have unofficial ways to get Play Store working (Boox eReaders for example), but those are increasingly rare now, as years of sanctions has led to Huawei de-Googling itself fast, with others following.

Kindle Fire? Your smart TV? Your car infotainment? Android is everywhere even without Google

Android TV is not AOSP, it is an Android variant actively developed by Google and requires Google Play Services.

Kindle Fire is based on AOSP, that was my mistake. I will say Amazon is a rare OEM in this regard; they're tablet-only, and happy with AOSP because they have their own app store that don't need Play Services integration.

They are shutting their app store down, at least for non-first-party devices: https://www.androidpolice.com/amazon-android-app-store-shut-...

Your point is actually a good one because Amazon has abandoned its app store infrastructure. And it likely will just integrate Play Services in future.

In the tablet / phone world where end users expect to install applications from the playstore yes but there are many other device types that do use AOSP.

For instance some parking meters / train ticketing machines / payment terminals run Android. source: I work for company that makes some of them though we are moving away from Android to more classical embedded Linux.

Google play services doesn't matter in these contexts.

Interesting? So Android replaced Windows CE in some cases?

The Horizon OS for Oculus headsets too.

All of the chinese and indian phones?

> Google Play Services, which is required for many popular apps to work

And here's where we need to apply antitrust pressure.

So one of the giants will build their version of microg and it will be like with wine, proton and steam os.

Except that attestation (Play Integrity) exists, and its usage by the apps on the Play Store seems to be increasing exponentially (if I'm not mistaken it only takes enabling a switch now to have it in your app).

Play Integrity and Attestation are distinctly different things. Attestation is a far better solution as it's not literally controlled by Google like the Play Integrity API is.

GrapheneOS actually explicitly pushes for services to switch away from Play Integrity API or SafetyNet API towards the Android Attestation API for that reason.

https://grapheneos.org/articles/attestation-compatibility-gu...

And notably, long term if Google becomes an unreliable maintainer for Android, the other major Android providers are likely to coordinate a list of "approved signing keys" so that apps can use those with the standard Android HW Attestation API.

Ok, yes, I meant Play Integrity indeed (which is an attestation system).

That's what Google has been strongly pushing to developers.

DJI/Hasselblad use it for their X1D/X2D line of cameras, based on the (accidentally) non-encrypted firmware for the X1D Mk II which has android stuff in it.

We're all going to have to live in ze pod and eat ze boogs when it comes to software platforms and ecosystems because aside from perhaps cloud deployments at scale, there just isn't an incentive to support open source from an ecosystem standpoint.

Give you an example: Recently the Open Source Initiative held board elections. Three of the candidates were disqualified. Two were disqualified for refusing to use proprietary software with Stallman-like stubbornness after explicitly being told that use of proprietary software was non-negotiable for board participation. Lunduke tried to weave it into his neo-Nazi narrative of cultural bolsheviks infesting open source, but the reality is the proprietary software in question is DocuSign -- and there is no alternative in the open source world that does what DocuSign does.

So if the steward organization of open source cannot function without proprietary software, what hope do the rest of us have? Especially with online services using remote attestation and refusing to function unless you're using a known, approved stack from boot to UI layer. May as well buy a Mac and an iPhone and be done with it. Save you lots of hassle and you'll look less like a dweeb.

Android is going to go closed source. There's no benefit to Google when a Chinese company takes Android, guts the Google parts, and shoves it on a cheap ass phone for a market with a few billion people.

Maybe longer still... I could see Google fully jettisoning the "other" phone makers and going all in on its own hardware to boost profits further. The iPhone is highly profitable, wouldn't you as a Google share holder want them to maximize the value there?

Personally I gave up on Android years ago. It was never great, but the endless poor support of software updates (mostly stemming from Qualcomm blobs as I understand it) was enough to drive me away. Apple really does support iPhone's for a very long time, and their sole goal isn't to sell more eyeballs. Good enough reason for me.

> There's no benefit to Google when a Chinese company takes Android, guts the Google parts, and shoves it on a cheap ass phone for a market with a few billion people.

There's nobody at Google who knows what to do with that kind of market leadership, so they're just folding up shop?

> There's no benefit to Google when a Chinese company takes Android, guts the Google parts, and shoves it on a cheap ass phone for a market with a few billion people.

Imitation is the sincerest form of flattery! This might give them more credibility.

Anyway, I think we're a half decade past where android source has been worth stealing: China now has strong domestic competition.

But at least Windows isn't the number one OS anymore.

Yeah, yet my 2009 laptop still gets updates.

Well, I hope that if they do so, the FSF and all copyright holders of the Open Source software used to make Android Great In-the-First-Place will sue their a$$e$ off.

This is an entirely predictable next step - the only surprise is that it took so long to happen.

They progressively replaced all the default apps with the Google alternatives.

A Chrome-Chromium type split was only a matter of time.

I mean nothing changes, it was already open source, and few people outside Google know how to dev things around android, it's now a very complex os

They are giving reasons that claim it will be better for everyone in the end, but this is obviously not true for devs of LineageOS and Graphene.

Why is it OK for corporations to make self-interested sociopathic decisions that harm consumers then slap a post-hoc rationalization that is basically a lie to appease the public? It's so common that almost every corporation does it, and most people don't even take note.

Can we please just go back to having some sort of Debian-based GNU/Linux on a smartphone. Just create a really great mobile interface on top of that.

The more I learn about Android, the more I think the whole thing is a waste of time.

You say “go back” but that was never more than a niche thing. For better or worse, Google’s Android has pretty much always been the chief representative of mobile Linux. And far from being a waste of time, Android is the most widely distributed mobile operating system on the planet.

> You say “go back” but that was never more than a niche thing.

but you could walk to a phone shop and buy a Linux phone, for example FirefoxOS.

Firefox OS lived a short and obscure life years after Android entered the market and certainly wasn’t a GNU or Debian-like distribution of Linux.

Yeah, and there was also Maemo/Meego that was also sold on stores. Neither had a very high market share.

> just go back to having some sort of Debian-based GNU/Linux on a smartphone

Two years of full time engineering man hours

> Just create a really great mobile interface on top of that.

Probably another 3 years of full time engineering man hours

It was never a thing since OpenMoko, with exception of Nokia's N9.

There's still Sailfish OS... which is not open-source.

Here you go: https://mobian-project.org/

postmarketOS is based on Alpine Linux: https://postmarketos.org/

Still there is yet to be a device that is completely supported, see the following section which is currently empty:

https://wiki.postmarketos.org/wiki/Devices#Main

Interesting. I know the Librem 5 is usable as a day to day phone (I have been using it as such for 3.5 years), and a few phones used to be in that catagory. I guess they changed how they consider main vs community.

It will be interesting when the world move to HarmonyOS

Reply on news.ycombinator.com