ASK HN: Has anyone adopted Chrome OS as their primary OS? I'm in the market for a new device (and also an OS), having used exclusively Mac OS for the most of last decade. Recently I used Chrome OS, in break time, and I like the simplicity. I'm wondering and interested to know if Cloud9, EC2 and such platforms kind of enable remote development well enough? Specifically, mostly interested in how it would work out for web-development and Data Science. |
Later I switched to working on a tablet instead.
So i enable dev mode then installed Ubuntu in a chroot.
These days I use an iPad Pro as my laptop and it's not as good, mostly because even when iOS is connected to a 4k monitor, Safari tells the remote website "hey, send me your shittiest mobile site please". Apple should really make the iPad Pro browser tell things it's a desktop.
I’ve used my iPad Pro as a portal to Google Cloud and AWS for almost 2 years in personal projects. It works fine via SSH and the apps, but Cloud Shell isn’t great via the native app, so I end up running an extra instance. AWS, I just ended up doing an EC2 for command/control/dev.
Observed a few people at work on ChromeOS and they seem to be productive, so most likely you need something like a high end pixelbook to be safe on local tasks for data science.
did these two desktops also use Chrome OS? This sounds pretty nice
On other Chrome OS: It syncs your Chrome browser + the OS settings.
2) I do a lot of work on airplanes, so not everything can run on a cloud instance. Apple cost too much (and has a remarkable lack of native apps for things like netflix). I also need it to fit on a airplane without jamming my elbows into my sternum due to size. The Chromebook has the form factor, but not the flexibility to work when the net is spotty(again, Crostini probably helps this out).
What I ended up doing ironically was a Surface Go. Great keyboard, nice form factor, and WSL support. 8GB of RAM (which is really needed by notebooks) and it works on plane.
If the quality of Apple’s own Stocks/News/Home conversions is anything to go by it’s not going to make for good software. But I’d take it for offline Netflix viewing.
Main reason is good driver support out of the box. Also not perfect(trouble with sd-cards, that can cause freezes), but my ASUS C213 runs very smoothly and long compared to my other laptops and it is sturdy and not expensive. Perfect for working while traveling ...
But ChromeOS itself ... is quite horrible. You feel cripled anytime you want to do more than Surfing and basic writing.
With the developer switch and some pain, I could set up my work enviroment with node and chrome dev tools and now that works pretty good, but every time I am forced to use that joke of a file browser I get annoyed.
But once linux apps will be fully supported (stable and gpu accelerated), things might get interesting. But I still hate beeing limited and restricted in many ways. I like to fully tweak my system to my taste. It is Linux underneath but different. And not very well documented.
You want to change key-binding or just want to invoke a shutdown after a long script finished? Or ... dare to create new logins(without google-connection)?
Good luck, I gave up on that.
If you need to run any specific stuff locally, get a Chromebook that has an Intel CPU, not an ARM. Some stuff (especially proprietary binaries) are not available for ARM. If you get a device with a touchscreen, Android apps sometime will be good enough for things that do not work well on Linux (e.g. Skype).
A noted limitation is that there's no good X11 story. The new Linux subsystem looks very interesting, but it only works on some hardware, and you'll have to do heavy research before buying to figure out if your new Chromebook works.
All that said, I'm thinking of moving away from it. Using Chromebooks for everything requires trusting Google a lot, and because of their various recent behaviors, I'm willing to do that less and less these days.
Your Google account (containing your personal details such as your location and date of birth) is now tied to all your activity in the OS. Even if that data is 'anonymised' (a pretty meaningless term) or aggregated with other user data, this is a colossus sweep of user data on an industrial scale.
I presume all the people who are happy to use Chrome OS would also be happy to sign-in to Windows/Ubuntu/Mac OS with a personally-identifying account and have everything they do recorded and tracked by those operating systems?
Chrome OS for me is perfect, especially since they introduced linux support and android apps. I can run anything I need, I do not feel like I took a step back, in fact I feel like I leaped forward by pushing towards a cloud first experience. I was already using Google for all my apps, spend most of my time in the browser and I personally have not felt any pain beyond adjusting to my new keyboard layout.
I highly recommend everyone ditch their macbooks and buy a Pixelbook.
Good keyboard, decent screen, long battery life.
The only thing I’d have done differently would be to have gotten the i7 version.
We live in weird times when there is (almost) only one browser left... Chrome (chrome-clones). The argument may soon be moot regardless of market share or other variables.
(Of course, I'm stretching this a bit for effect but it's scary to think about).
I need MS to install IE, so I can use it to go and get IceWeasel, or whatever. Leave them alone!
Software was distributed offline as well. It seems like a long time ago now, but there was no shortage of AOL installation discs at the time.
(In fact, I believe both Internet Explorer and Netscape Navigator were distributed on floppy disks at one point, although that would have been long before the antitrust case.)
If your workflow is just remote access with SSH and an occasional Web site, a smartphone or tablet can already do the job. Just deck it out with peripherals, a stand, etc. You can get very, very far with remote access. But for a full local development environment, especially if it involves graphics, there's a tendency to depend on a "desktop x86" stack that tends to creep in no matter what you do.
With crostini, I can do local dev (python, js, golang) 90% as well as I can on a full fledged OS, without having to rely on cloud IDEs, etc.
I would like to be able to run Emacs on it. It doesn't have to be full blown dev machine, but I'd like some of my simple tooling, and backup to cloud storage to share with my main work machines.
I currently using an iPad with a Brydge keyboard attached (I paid $350 for the setup, ipad on sale for $250 + $99 keyboard), which I love, but I don't have native Emacs and some other stuff, so I'm actually looking at a Chromebook. I think I could do most stuff with native Android apps and the web browser. I'd still like to figure out the Emacs/lightweight development type of thing. I've tried some of the online tooling like repl.it, but have a hard time with the portable keyboard bindings...
The biggest surprise for me was that I bought a $450 14" HP x360 chromebook for a booth at a tradeshow, and I've been using that chromebook as my main laptop for the last few months, instead of my 16GB top-end Pixelbook. Why? The larger 16:9 screen is way better for coding (for me), and the keyboard is amazingly for me (I consider it much better than the pixelbook keyboard). This HP chromebook weighs a pound more than the pixelbook, but I'm not carrying it around that much so it doesn't matter. Also, it's nice when I do carry it around knowing that it cost $450 instead of $1599, so if it is stolen or damaged, it is much less of a loss.
Conclusion: I've been using ChromeOS as my primary OS off and on since 2013, and I'm impressed with how ChromeOS (and the Chromebook ecosystem) have improved over the years. And I'm very surprised, because today I choose a $450 Chromebook over a $2000 Apple laptop, whereas from 2007 to 20017, I would choose the Apple laptop. Also, for me, everything is a company (or university purchase), so even though "money is no object", I still prefer the relatively cheap Chromebook.
Another, thing -- my old Chromebook Pixel had LTE data, since it always had internet. My iPad pro also did. However, with modern tethering (I have a Pixel 2 XL), a phone provides perfectly good tethering --there's no need to have it built into your laptop.
For example: To use 'adb' to connect to my phone I know of three options. 1 Use Chrome but that is extremely limited to web. 2 Activate developer mode and upgrade adb to current. Or 3 Connect to another device like rpi.
I chose to activate developer mode for convenience. I run OpenBSD via qemu for a decent CLI. I keep a loopback ethernet cable and proxy settings to fool ChromeOS into letting me use alternate network set ups.
Also the nag screen on reboot is alarmingly bright and, i think, just gives directions to disable developer mode.
Always backup your files if you use developer mode.
Always backup your files if you use the linux system.
Always backup your files.
I got my wife a Samsung Chromebook 3 and she uses it for her job as preschool director. Being able to install Android apps is great. I feel like I could use that machine for most of my personal projects. I would NOT use a remote development tool; try to do it all locally when you can.
Originally I expected to write code directly on the Chromebook using Crostini (edit: Crouton), but it required a lot of hand-holding. I eventually gave up. But that was years ago, so maybe they've smoothed out the edges.
In any case, Crostini has gotten much better since last year -- but the chromebook itself is too underpowered compared to a desktop to make a good experience. This is even true for the Pixelbook.
Crostini also seems to have problems with storage; the filesystem constantly soft-locks, for seconds at a time, under any appreciable load. This might be because they use Btrfs.
So chromeos is more my mobile/meeting computer.
When I code, I generally do it with Emacs in crouton, not with a remote IDE like Cloud9.
New Chromebooks are even better for development because they support docker, but I can't test it with my "too old" Chromebook.
Before I gave her the Chromebook, I set it up for myself to try it out. After updating it (an Acer Chromebook 14), I was able to install Android apps which all ran pretty well. I also installed Sublime Text and VS Code in developer mode. Sublime ran much faster, but they both had issues compared to my nice fat PC. However, Chrome Remote Desktop works well enough that I could have just used the Chromebook as a remote screen for my PC. I don't think I could ever bring myself to use Chrome OS as my primary OS though...ever. It's too limited.
So, I switched to Manjaro Linux with XFCE for all of my web development and other work needs last year. I highly recommend it! (Before that I was mostly on Windows and sometimes on Mac.) I installed it on an Acer E5-575G and I got everything working including precision touchpad gestures which perform just as well as on any Macbook (I own a Mac Pro and also a Macbook Pro for comparison). It took some work to get the touchpad configured correctly, but not much. Then I got a 2nd PC workstation (a refurbished i7 desktop from Amazon for like $400) and I put a new 500GB ssd with Manjaro on that... Honestly, it's great. All the development tools that I need just work and there is plenty of software for Arch/Manjaro. I'll probably never go back to Mac or Windows for work, despite the fact that Manjaro updates have twice caused me a minor setback which were solved easily with help from threads like this - https://forum.manjaro.org/t/19-february-2019-system-update-i... (No big deal - I've had similar scares from Windows and Mac updates over the years.)
A Linux desktop is the only desktop OS that makes me feel like a kid again. The reasons I failed at moving to a Linux desktop before were because things broke all the time or required hella work just to get the proper drivers...or becuase I didn't like UI (on Ubuntu for instance), etc - but Manjaro with XFCE was a breeze and changed all that for me. I'm so glad I decided to try it out. I only ever tried Gnome or KDE before that.
In fact they hardly took over outside the US school system.
http://gs.statcounter.com/os-market-share/desktop/worldwide/
The chrome crostini project [0] basically gives you a little container in which you can run a terminal (probably more, but a terminal is good enough for me). Docker is supported so I can run additional containers within the crostini.
I've pulled all of my dotfiles into the crostini terminal, and have encountered very few issues experimenting with docker, python/flask, ruby, and C/C++ locally. I can't speak to emacs, but vim/tmux run as expected.
The C630 in particular has worked out remarkably well: the hardware is much better than I'd expected, the battery is very good, the touchpad is second only to my macbook (this is what piqued my curiosity initially), and the display, while not matte, is big and bright enough to split a browser and terminal side-by-side.
The C630 would probably fit the bill for a nice travel box. It's not as pricey as a macbook, but at $600-700 it's a little more than most chromebooks.
The only real concern I have is the reliance on Google and the G-ecosystem.
[0] https://chromium.googlesource.com/chromiumos/docs/+/master/c...
I can't rely on available internet connection (with this portable setup).
But, I'll check out the Inspiron 11.
The world of 2001 was very rapidly becoming the connected world we're in now, and imposing a restriction on anyone like that now would just seem ludicrous and in fact user hostile, as it was. To me and I think anyone else paying attention back then, that was very obvious.