38 comments

[ 3.9 ms ] story [ 56.8 ms ] thread
> “A beta release of the Play store is available to users now on the Acer R11 and Asus Flip (and coming soon to Pixel 2015) and can be enabled from the Settings page,”

Not really ...

Care to elaborate?

I'm kind of considering getting an Asus Flip (or maybe a Pixel) as it's clear that a tablet without a keyboard is pretty limited in its functionality for people who like to create rather than simply consume content. Having some apps available would greatly add to the appeal of those devices.

> Google Play starts rolling out to Chrome OS stable

Made it sound like it was coming to all Chrome OS devices (or, at least the ones Google said would be getting it on dev sometime in "late 2016"~). It's just out for two Chromebooks. Got me hook, line, and sinker Venturebeat!

(Maybe I was dumb to think that was happening. I saw this when I first woke up. Thought it was a huge surprise designed to beef up the hype before the Oct. 4 event.)

From talking with the chromebook team I got the impression that this feature is meant for newer devices. Maybe someone from Google can chime in?
This, and I'm posting from my R11 right now, I see the button to enable the Play store but it's Greyed out.

I'm also on the developer channel, apparently you need to be on the stable channel to work. Which requires a powerwash (equivalent of a factory reset for the uninitiated) to revert back to. Damned if I do, damned if I don't because I rather like having Arch Linux a few keystrokes away, but I also want to play with a few Android apps on the toilet :P

The software is still beta and is limited to just a few devices. The only thing that has changed is that it is now in the stable channel.

So basically, nothing has changed, this is not a real roll out

After Samsung's awful rollout of the Note 7 with battery problems, and the costs of bringing a product to market without enough testing, I have some sympathy for this slow Play on Chrome OS rollout.

If there are end user problems running Android apps on a Chromebook, it is better to have a smaller and more manageable test set of users.

Um, rolling out software has no comparison to pulling out millions of phones and replace them all around the planet. This is just how Google ships stuff - from the announcement to the day when 0 people care.
I should have explained myself better. The Android on Chrome OS is I think a big deal because one thing that Chrome OS has going for it is its security model. I could be wrong, but I consider my Android phone less secure than my Chromebook. If the security gets botched up on Chromebooks, that would be bad.
In principle i agree about security, but for the common non-tech user, Android's security is probably good enough.
I agree about being a big deal, one might even think if the Android update problem could be surmounted by eventually bringing Chrome OS as a host running Android apps to mobile devices.
> After Samsung's awful rollout of the Note 7 with battery problems

You should give Samsung some credit here. When the issue became known they wasted absolutely no time in recalling 2-3 million devices.

I like Samsung. Every phone I have bought is from Samsung. They have my sympathy over the battery issue but I think they will go through tougher testing procedures in the future.
I've been pleasantly surprised by the breadth of apps that work on my Chromebook Pixel. Pretty much everything I use on my nexus phone works, albeit some with a small window size.

For me, the biggest advantage is the performance boost of android media players, such as Netflix, Hulu, and You Tube.

What do you mean by the performance boost of Android media players? Do they perform better on Chrome OS than watching video in the browser? I'd imagine that in a lot of cases, the max bitrate served to mobile clients may be lower than the web, although I'm not sure this is true for all the services you named.
Compared to running in a browser window, the android apps for Neflix, Hulu and YouTube exhibit far less jitter, dropped fames, delayed frames, etc. This is true even on a Pixel with an i7 processor and 16G of ram, so it's not a question of CPU or memory resources, and also true on a symetrical 1GB internet connection.

For me, this eliminates need need to carry a tablet for consuming video or audio media.

How is the cast functionality with the apps? Let's say you use the Netflix app, will it cast like an android phone? This way you don't need to cast your chrome tab.
fwiw i already cast netflix from chrome without casting the tab
Unfortunately there are some media companies that only implement casting in their mobile apps, not on the web.

So you either cast from the mobile app or tab cast.

I know that Netflix was probably just an example, but for the record, Netflix is not one of these media companies.

You can directly cast Netflix when using it from a browser, you don't need to cast the tab.

Or if you're talking about Amazon Prime Video, you just cast the tab because you can't cast the stream directly from the browser (like Netflix) or from any app on your phone because that only works if you're sending to a Fire Stick.

Somewhat tangential, if you have a TV running Android, you can install HBOGo but if you're on Comcast, they won't authorize the app so you can't use it. You have to install it on your phone and cast the mobile app to the built-in Cast receiver in AndroidTV (or do the same with the website on your other computer).

I'm legitimately curious: Did you actually buy the Chromebook Pixel at the normal RRP ($500-600)?

I've met people who have had one but most of them got them for free (either via Google employment or Google I/O). I'm yet to meet anyone in real life who purchased one retail.

PS - My question is about the Pixel range in particular, other Chromebooks sell very well.

I've bought two: a first generation Pixel used, and after two years, a Pixel 2015 directly from Google Play the month they came out.

The browser is so much faster than my desktop, that I tend to use it even when at my desk.

The best part, to me is that, unlike windows, I never need to troubleshoot it. I logged into the 2nd Pixel with my gamil account, and 2 minutes later it was synced, configured just like the previous. That's one of the reasons I bought a new one the second time -- consistency and trouble free use.

The next step is the play store for Chrome on Linux. Then Mac. Then windows. That'll be awesome
Ugh, I don't doubt that will happen and I'm afraid it's going to bring back the browser monopoly situation.
It is unrelated technology. First version of Android runtime for chrome was ARC welder , if I remember correctly. But that didn't go anywhere. Right now they are using something similar to docker technology which runs Android apps in sandbox in chrome.

You can watch the talk in Google io 2016.

So in this scenario there not feasible way to port this runtime to other platforms because it is not using chrome runtime.unless they redesign it.

My understanding of how this works, mostly from reading commit messages and poking around the repo, is to add a wayland compositor/server to the aura component of chromium. Aura is the window manager, compositor, and render of ui items such as the task bar and status icons of ChromeOS. It looks like the Android portion is largely based on the same fork as ARC, with the rendering performed through Wayland. It also looks like a large number of Android APIs had to be reimplemented which may excacerbate the fragmentation problem.
Cool , did they used Wayland? I didn't know that. It is quite fascinating, Wayland finds its actual place in industry.

About API reimplantation, I do remember a little bit, you are correct, the guy in Google talk mentioned it is simultaneous development to Android and they are not using Android underlying technology on top of namespace,etc (Linux syscalls).

I would love to read a blog post about this, and since your poked around in source code you are right person. So please, if you have any spare time, write a blog post.

A Linux port is actually feasible:

> the Android container itself has no plans to be released (which is basically just the rootfs of an Android system). everything else is already released, and as noted, is using standard Linux technology to isolate things (namespaces, etc...).

> with that in mind, there's nothing stopping people from implementing this approach elsewhere.

https://groups.google.com/a/chromium.org/forum/#!topic/chrom...

Yeah of course , but we are talking about chrome as browser. I would love to see semiofficial Android runtime for Linux.
I was part of the ARC program as a publisher and it was financially successful for my team. I can't praise Google enough for it. The tech worked well although I don't oppose their new approach. The seem to have learned a lot from ARC and decided to move forward.
Does anyone know how this works (under the hood)? Is it just a very good Android emulator?
There is no need to emulate. They have the same runtime on chrome as they have on android. There is very light weight wrapper which enables the runtime to work on chrome.
It's not an emulator, but rather integrated directly into the chrome OS, like the previous "android runtime for chrome" (aka ARC) was.
Perhaps the new 'Pixel' phone is ChromeOS based - yet another platform but like BB10 and Sailfish including Android apps.
Probably too soon for that, considering they basically have Play in beta for Chrome OS. But I could bet that this might be a future someday... But even if it makes updates easier, it will involve carriers and Android update situation might happen again.
I think it's definitely a way to start combining the Pixel branding with the ChromeOS branding under "Pixel" to make any future crossover easier. Other than that, though, why would carriers have to be involved? Sure, if you buy a carrier-branded device they get their hands in it but it could just as easily work the way Nexus did on any non-carrier-branded device. If anything, I see the "Pixel" move as a way to start moving away from that side of Android and all it entails.

Idle speculation: I wonder if the end goal is for "Android" to stick around as the "anyone can port it to their device" OS as it is currently, with "Pixel" being Google's more in-house label that they can treat like Apple does iOS or Google currently does with standard, unbranded Nexus.

just more adware. What is really needed is access to the F-Droid repository.