15 ms·

This is interesting. I wonder what goal they are aiming towards. There isn't much information about the capabilities of workstation. Is it a GUI OS? Can one ru

by Osiris 5y ago

This is interesting. I wonder what goal they are aiming towards. There isn't much information about the capabilities of workstation. Is it a GUI OS? Can one run Flutter apps on it?

owaislone 5y ago
Yes it has GUI. Flutter is the main/official way to build apps on it. The OS shell is also written in Flutter if I remember correctly. So far, the OS has been seen on or confirmed to release in near future on one Google Home device. Technically though, it seems to be designed for all kind of consumer devices ranging from phones to desktops. IMO this is more or less an experiment which if successful, will end up merging/replacing Android and Chrome OS into a single OS. Android and Chrome runtimes can be ported to Fuschia and the underlying OS can be swapped on many devices. New devices can ship with Fuschia + an android compatibility layer. Developers would be able to ship native Fuschia or android apps. Would work on phones, IoT, Chromebooks and could be install-able on desktops. This is all just speculation though and a lot of things need to go right for this to happen but I'd imagine this would be the ideal/desired outcome for its creators.
- pjmlp 5y ago
  > ... Android and Chrome runtimes can be ported to Fuschia ... Android is already being ported to Fuchsia for quite some time now.
  - cryptos 5y ago
    I thought Google would no longer follow the idea to replace Android with Fuchsia. Do you have a source for this statement?
    - pjmlp 5y ago
      Yes, check the AOSP Gerrit commits related to Fuchsia, quite a few of them there. You can start with this one, https://android-review.googlesource.com/c/platform/manifest/+/847437 https://android-review.googlesource.com/c/platform/manifest/...
    - mhoad 5y ago
      There is no compelling reason to keep Android running as some legacy option when they are making Android and more generally Linux compatibility as a goal. It would just be a slower, less secure option with years of cruft and a huge dependency base they don’t control.
      - pjmlp 5y ago
        Project Treble and Project Mainline where Linux drivers are considered "legacy", and already Rust's adoption even with nightly features, don't really show that. They are only upstreaming to decrease the development efforts in Linux kernel itself, tomorrow that kernel can be Zirkon instead.
        mhoad 5y ago
        I don’t know I disagree with anything you said other than the end goal. But yes, I agree, I think getting the Linux kernel out of Android is an obvious next step. Does Android just become another Fuschia “product” at that point the same way the workstation and a stripped down IoT one are? I’d argue it goes further than that but my reasoning kind of gets pretty deep into how Fuschia for example handles things like app delivery etc… I think Android will continue to exist as a supported interop style solution for years but ultimately you will end up seeing a total transition to 100% Fuschia across all Google platforms (server, IoT, Workstations, mobile devices etc).
        pjmlp 5y ago
        I would say that the facts speak for me, it is the Project Treble HAL documentation call refers to Linux drivers as "legacy". https://source.android.com/devices/architecture/hal-types https://source.android.com/devices/architecture/hal-types > "HALs expressed in HAL interface definition language (HIDL) or Android interface definition language (AIDL). These HALs replace both conventional and legacy HALs used in earlier versions of Android. In a Binderized HAL, the Android framework and HALs communicate with each other using binder inter-process communication (IPC) calls. All devices launching with Android 8.0 or later must support binderized HALs only." While Linux folks are still arguing if Rust makes sense to be adopted in the kernel and what features still need to be stabilized, https://source.android.com/setup/build/rust/building-rust-modules/android-rust-modules https://source.android.com/setup/build/rust/building-rust-mo... And as I pointed out in another comment, here are the ART commits supporting Fuchsia, https://android-review.googlesource.com/q/fuchsia https://android-review.googlesource.com/q/fuchsia I should also note that while people pat themselves on the back, because Android uses the Linux kernel, from Google's point of view that is an implementation detail, not supported as public API for NDK code. https://developer.android.com/ndk/guides/stable_apis https://developer.android.com/ndk/guides/stable_apis From NDK point of view for app developers, the underlying OS is a generic OS, with ISO C, ISO C++ standard libraries plus a couple of additional stuff. Something that Termux guys have a hard time to swallow.
- rjzzleep 5y ago
  Maybe it's just my laptop, but most Flutter apps I've tried were clunky and draining CPU like crazy. On top of that other people here have mentioned problems with the framework concept itself which causes certain bugs to just linger for years, because they can't be fixed. I don't really understand the excitement around flutter.
  - aero-glide2 5y ago
    I have experienced the same performance performance on laptop and phone (Google pay).
  - defer 5y ago
    I'm very conflicted with flutter. The developer story is tempting, cross platform consistency but extendable to each system's specifics. On the flip side there's the odd programming language, the bundle size, the game-engine like rendering (seemingly wasteful, but may improve as hardware evolves).
    - asiachick 5y ago
      except their web story is horrific. they render the entire thing in a canvas which means a giant F.U. to non English-FIGS and a big F.U. to accessibility as well as extensions
      - defer 5y ago
        Definitely, I remember from reading the documentation that they could also target HTML5/css (https://docs.flutter.dev/development/tools/web-renderer https://docs.flutter.dev/development/tools/web-renderer) instead of canvas but I'm not sure how complete / coherent the output is.
      - mijoharas 5y ago
        what does FIGS stand for?
        wffurr 5y ago
- eklitzke 5y ago
  Fuchsia does a lot of interesting/new things, but the primary motivation for Fuchsia was to make it possible to update the OS independently of drivers. Everyone complains when their Android devices stop receiving software updates, and nearly always the reason that new Android versions can't be backported is that the devices have custom hardware drivers (perhaps closed source) that would need to be be updated to work on newer versions of the Linux kernel. Since Fuchsia was started, Google has committed to maintaining a stable kernel ABI for AOSP, what they call the GKI. This is a huge amount of work for kernel developers at Google but it will allow (in theory) updates to newer kernels without the need to update these old drivers. I believe ChromeOS is also going to use the GKI (not 100% sure about this though). I think the big question for Fuchsia is how successful the GKI initiative is, since it takes a lot of wind out of the sails of Fuchsia. Linux already works, has a lot more features than Fuchsia, and has much better performance.

by Osiris 5y ago

This is interesting. I wonder what goal they are aiming towards. There isn't much information about the capabilities of workstation. Is it a GUI OS? Can one run Flutter apps on it?

owaislone 5y ago

Yes it has GUI. Flutter is the main/official way to build apps on it. The OS shell is also written in Flutter if I remember correctly. So far, the OS has been seen on or confirmed to release in near future on one Google Home device. Technically though, it seems to be designed for all kind of consumer devices ranging from phones to desktops. IMO this is more or less an experiment which if successful, will end up merging/replacing Android and Chrome OS into a single OS. Android and Chrome runtimes can be ported to Fuschia and the underlying OS can be swapped on many devices. New devices can ship with Fuschia + an android compatibility layer. Developers would be able to ship native Fuschia or android apps. Would work on phones, IoT, Chromebooks and could be install-able on desktops. This is all just speculation though and a lot of things need to go right for this to happen but I'd imagine this would be the ideal/desired outcome for its creators.

pjmlp 5y ago
> ... Android and Chrome runtimes can be ported to Fuschia ... Android is already being ported to Fuchsia for quite some time now.
- cryptos 5y ago
  I thought Google would no longer follow the idea to replace Android with Fuchsia. Do you have a source for this statement?
  - pjmlp 5y ago
    Yes, check the AOSP Gerrit commits related to Fuchsia, quite a few of them there. You can start with this one, https://android-review.googlesource.com/c/platform/manifest/+/847437 https://android-review.googlesource.com/c/platform/manifest/...
  - mhoad 5y ago
    There is no compelling reason to keep Android running as some legacy option when they are making Android and more generally Linux compatibility as a goal. It would just be a slower, less secure option with years of cruft and a huge dependency base they don’t control.
    - pjmlp 5y ago
      Project Treble and Project Mainline where Linux drivers are considered "legacy", and already Rust's adoption even with nightly features, don't really show that. They are only upstreaming to decrease the development efforts in Linux kernel itself, tomorrow that kernel can be Zirkon instead.
      - mhoad 5y ago
        I don’t know I disagree with anything you said other than the end goal. But yes, I agree, I think getting the Linux kernel out of Android is an obvious next step. Does Android just become another Fuschia “product” at that point the same way the workstation and a stripped down IoT one are? I’d argue it goes further than that but my reasoning kind of gets pretty deep into how Fuschia for example handles things like app delivery etc… I think Android will continue to exist as a supported interop style solution for years but ultimately you will end up seeing a total transition to 100% Fuschia across all Google platforms (server, IoT, Workstations, mobile devices etc).
        pjmlp 5y ago
        I would say that the facts speak for me, it is the Project Treble HAL documentation call refers to Linux drivers as "legacy". https://source.android.com/devices/architecture/hal-types https://source.android.com/devices/architecture/hal-types > "HALs expressed in HAL interface definition language (HIDL) or Android interface definition language (AIDL). These HALs replace both conventional and legacy HALs used in earlier versions of Android. In a Binderized HAL, the Android framework and HALs communicate with each other using binder inter-process communication (IPC) calls. All devices launching with Android 8.0 or later must support binderized HALs only." While Linux folks are still arguing if Rust makes sense to be adopted in the kernel and what features still need to be stabilized, https://source.android.com/setup/build/rust/building-rust-modules/android-rust-modules https://source.android.com/setup/build/rust/building-rust-mo... And as I pointed out in another comment, here are the ART commits supporting Fuchsia, https://android-review.googlesource.com/q/fuchsia https://android-review.googlesource.com/q/fuchsia I should also note that while people pat themselves on the back, because Android uses the Linux kernel, from Google's point of view that is an implementation detail, not supported as public API for NDK code. https://developer.android.com/ndk/guides/stable_apis https://developer.android.com/ndk/guides/stable_apis From NDK point of view for app developers, the underlying OS is a generic OS, with ISO C, ISO C++ standard libraries plus a couple of additional stuff. Something that Termux guys have a hard time to swallow.
rjzzleep 5y ago
Maybe it's just my laptop, but most Flutter apps I've tried were clunky and draining CPU like crazy. On top of that other people here have mentioned problems with the framework concept itself which causes certain bugs to just linger for years, because they can't be fixed. I don't really understand the excitement around flutter.
- aero-glide2 5y ago
  I have experienced the same performance performance on laptop and phone (Google pay).
- defer 5y ago
  I'm very conflicted with flutter. The developer story is tempting, cross platform consistency but extendable to each system's specifics. On the flip side there's the odd programming language, the bundle size, the game-engine like rendering (seemingly wasteful, but may improve as hardware evolves).
  - asiachick 5y ago
    except their web story is horrific. they render the entire thing in a canvas which means a giant F.U. to non English-FIGS and a big F.U. to accessibility as well as extensions
    - defer 5y ago
      Definitely, I remember from reading the documentation that they could also target HTML5/css (https://docs.flutter.dev/development/tools/web-renderer https://docs.flutter.dev/development/tools/web-renderer) instead of canvas but I'm not sure how complete / coherent the output is.
    - mijoharas 5y ago
      what does FIGS stand for?
      - wffurr 5y ago
eklitzke 5y ago
Fuchsia does a lot of interesting/new things, but the primary motivation for Fuchsia was to make it possible to update the OS independently of drivers. Everyone complains when their Android devices stop receiving software updates, and nearly always the reason that new Android versions can't be backported is that the devices have custom hardware drivers (perhaps closed source) that would need to be be updated to work on newer versions of the Linux kernel. Since Fuchsia was started, Google has committed to maintaining a stable kernel ABI for AOSP, what they call the GKI. This is a huge amount of work for kernel developers at Google but it will allow (in theory) updates to newer kernels without the need to update these old drivers. I believe ChromeOS is also going to use the GKI (not 100% sure about this though). I think the big question for Fuchsia is how successful the GKI initiative is, since it takes a lot of wind out of the sails of Fuchsia. Linux already works, has a lot more features than Fuchsia, and has much better performance.