201–210 of 245 posts
Earlier quoted context omitted.
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/+/...
Earlier quoted context omitted.
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…
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…
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.
The obligation is specifically to provide the source code (without certain usage restrictions) for binary releases when requested, and no more.
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.
>The source will still be released Lol nice pipe dream. The source isn't even fully released today.
Note that the demand for secrecy is primarily driven by device manufacturers, not Google. Manufacturers want to keep their "secret sauce" from their competitors.
[flagged]
Not really clear how moving this project to a private repo is evil. If you're going to argue it is a "bait and switch", that implies intent to deceive. Do you think that was the plan? Release the first version in 2007 and then wait 18 years and pull the ol switcheroo?
it effectively kills off all hope of virtually all open source projects and most corporate forks keeping up with upstream changes.
The Linux kernel is bad enough but at least you know what you'll get and can keep up with the development. But this here? Everyone writing ROMs of their own will get a truckload of changes that have to be integrated and local forks rebased.
Earlier quoted context omitted.
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.
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 masterc…
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).
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?
> not much will change overall. It's a major step in the wrong direction but without a large direct impact on us
Earlier quoted context omitted.
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.
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?
Earlier quoted context omitted.
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/i…
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.
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.