https://chromium.googlesource.com/chromiumos/docs/+/HEAD/con...
ChromeOS got good and is wildly underrated for software development. ¯\_(ツ)_/¯
App sandboxing on standard (non-ChromeOS) Linux distributions is painful and finicky, while it "just works" on ChromeOS.
I wish there was a non-hacky way to use Chromebooks without a Google account.
The hacky options are:
- "Switch to dev mode". But I don't want to be prompted to factory reset each boot.
- "Create a dummy Google account and use that". But I don't want file syncing and tracking to reach Google at all, not even on a dummy account.
- "Create a dummy Google account and use guest mode". But I want persistent storage.
(1) Is there a de-googled version of Chromium OS?
(2) Is there a non-hacky way to install it to a Chromebook?
Developer mode simply puts up a splash at boot warning you that the OS is custom, you just press enter to boot. The requirement to do a drive wipe is a one-time thing when you enable it the first time (for obvious reasons, to prevent exfiltration of data stored by a secured OS).
A few years back (when I last looked in depth), it was very easy at every boot (not only first boot) to accidentally wipe everything - e.g., https://www.reddit.com/r/Crouton/comments/3be2su/reducing_ri...
I also remember that it made a loud beep on every boot!
(CS major, coding happened mostly over SSH to the school servers, but I did run RStudio locally)
I do remember that running a certain R program consistently caused the Chromebook to turn off, which was quite an issue for one particular office hours session!
If I close the lid, when I get back _I_ need a clean slate. I'll never remember what I was doing anyway.
Modern machines boot and launch apps so fast I'm not really saving any time sleeping an empty desktop either.
Other devices, maybe; it's at least as much firmware as OS. To this day the fastest wake from sleep I've seen was a random midrange Lenovo laptop that I got used, which is slightly maddening.
For me ChromeOS’s top UX fault is the password. You either do it with pin, or with a Bluetooth device to unlock. Otherwise it’s pain to enter the password.
Both pin and phone unlock are supported, as of today.
There were days that you need to enter developer mode to turn on pin unlock, and Bluetooth devices came later on. I remember that pain of needing to enter my very long password in each boot. Horrible UX.
I interacted with a recent (a couple of months) Chrome OS Flex device (a regular old PC laptop of a senior that I know and helped to establish a basic computer for him), and it didn’t work on the first boot. It was a huge problem for him to enter the password, especially a strong one. I set up a pin with his birthday, which never worked. It worked only if you logged off the account (but didn’t turned off the computer). But if you start fresh, then you won’t be able to use the pin. I ended up setting his Google password to his birthday with dots and some letters, e.g. Qw25.12.1935, which Google allowed and which worked for him. He enters his password letter by letter and for him it’s a worse UX than before, when he had Windows XP that just boots and has Chrome (outdated with no option to upgrade) installed. But I convinced him this new way of things is better. At least it loads momentarily, which he likes.
And on top of that, built in Bluetooth adapter doesn’t work on that very laptop to connect his Android smartphone. The Bluetooth module works, but it doesn’t with the software for some reason. Brief googling showed me it’s easier to buy usb Bluetooth module and try with it. Which I did, but haven’t checked that yet, as he lives quite far away from me. As of now he uses the laptop somehow.
When I put my machine e away I am usually 'done', and when I open it, it isn't to see where I left off.
Linux just works without all the WSL issues, I can run Visual Studio Code, Docker and anything that I tried so far.
It's actually a delight to work on. I can only imagine how much better yet it will be with a top tier Chromebook
I agree though, this claim seems unlikely to me.
* Slower IO -- not sure if this can ever be fixed, but quite a bummer * cpu utilization at 100% due to an issue with WSLg #6982 * no access to Internet when host is connected to VPN #5068 * very slow network in some situations #4901
These are not some edge cases -- they affect things that I use every day. Some of them have workarounds but overall this is still not very usable. I decided that it was easier to just get a separate machine and install Linux on it.
Sort of like how a Pentium 3 feels snappy on Windows XP and yet, this 13th Gen Intel Core i5 on Windows 11 feels sluggish. There is WAY more processing power on my newer machine but it doesn't translate to a snappy interface.
For Gui. Too hard. Done it once when win 11 coming up.
It is now a dreadful wait for win4 to come. Wsl2 …
For those suggest dual boot I have a whole thinkpad X now only in windows and only if it’s do a bit action in the boot loader.
May be I should try Chromebook.
Depends on programming language. Not every compiler benefits linearly from 16 cores.
I have been interested in buying a Chromebook for dev purposes.
Then install VSCode and Docker as you would on any Linux system.
You can then launch them from the ChromeOS desktop environment as if they were native ChromeOS apps. It's really quite seamless, much more so than WSL (and better performance IME).
Next laptop will be a framework, so I guess I'll see how that goes.
For old models as well
I have a kevin (Samsung Plus v1) and flashed it with libreboot for Linux at that point.
You can even install ChromeOS Flex on a non-ChromeOS PC.
[1] https://chromeenterprise.google/intl/en_ca/os/chromeosflex/
I recently got a Chromebook and I'm amazed at how snappy and fast it is. And its built in Linux is amazing.
I've had some pretty severe issues with a chromebook running out of memory because the devs decided swap wasn't allowed (and I could never get the kernel to build and install properly). The builtin chrome definitely doesn't stop single websites from using most of a gigabyte of RAM.
300 euro from 2009, upgraded with 8GB and 1TB SSD, and with native graphics that beat WebGL 2.0, running latest Xubuntu.
- no need to worry if upgrading some Xorg will kill Xserver. Yes, it does happen if one has autoupdates
- Will my zoom/camera/mic work after I make pulseaudio->pipewire-><whatever latest>? No idea but chromeos - just works
- There are software engineers that have no installation/hardware/sysadmin experience (or they dont want to learn).
- Just close and open device. Chromeos is instant. (yes, one can fiddle with pm-utils to get modules (un)loaded - no life is too short to fix these.
- Many programs like netflix need some form of terrible DRM and dont work with linux. Some people still need to live with DRM.
> ChromeOS got good and is wildly underrated for software development.
There are laptops sold with Linux support in mind, those Asus netbooks were such laptops.
As for DRM, just install Chrome, the Web nowadays is ChromeOS anyway.
https://cloud.google.com/blog/topics/developers-practitioner...
To actually dogfood Chromebooks internally like Chromebooks for Education and for Families, Google would need to deny its devs containers behind a greyed out "Turn on Linux" button and deny them access to Colab notebooks and AI platform etc (while only allowing Play Store apps); nothing to SSH to, no notebooks, no devpod, no local git repos (!), no local terminal to run e.g. pytest tests with, no devtools JS console, etc.
It's a different market with different product segmentation and feature sets, it just happens to run on the same hardware and OS. And it's true that schools and parents often demand that their kids not hack stuff that the "admins" don't understand. And maybe that's bad or misguided or whatever. But it's not a statement about using a Chromebook as a developer client, which IMHO works very well.
Googlers only code remotely on Chromebooks.
The kids MUST be able to run `pytest arithmetic.ipynb` on whichever platform.
(It was not appropriate for Google 2016-2020 (?) to decide that we should all accept WASM runtime vulnerabilities in sandboxed browser processes running as the same user instead of per-app SELinux contexts like Android requires since 4.4+, or instead of containers and VMs like almost all of Google's hosted apps and internal systems)
Cannot believe there's not even a JS console on these for the kids (because "Inspect Element" and "Turn on Linux" are blocked for them all)
But you're responding to a thread saying "Chromebooks are good for Developers", where's it's just not responsive. There's absolutely no reason schools or parents can't hand kids unmanaged/unrestricted Chromebooks; they just choose not to (for some good and some bad reasons). Take that up with schools and parents, I guess?
Holy cow. Are you serious? Kids break stuff for fun.
This is basically a “You’re holding it wrong” argument. Schools are gonna have heavily managed chromebooks, because it’s the reality on the ground that they can hardly afford anything, much less insane IT support for every student. Like, kids are going to actively sabotage the software on their school equipment if they can as a way to avoid doing schoolwork. This was common when I was in high school over a decade ago, they solved it with Deep Freeze software and extremely restrictive Microsoft policies.
Google is the one with a wealth of IT and developer talent, not school systems. They are making the product, not the schools. Devs have to come from somewhere, and going out a limb here, I don’t think smartphones (the other computer that [low income] students are likely to have access to) provide an on ramp for developers, so it would be good if these chromebooks did.
And then an instructor could specify policy that would affect which apps and containers run on the machines attached to the management domain (e.g. at test-taking time and also all of the times)
Comparing etckeeper and rpm-ostree's diffs of /etc lately
(where e.g. LVM or btrfs CoW snapshots would need eventual compaction)
There should be a specification of things that we need the computers we buy for the students to support; a rubric to consider in acquisitions and discussions with vendors attempting to solve for the needs of education and learning.
Seriously, compare the results of canned flash games (with metrics) vs locally coding math to do a scientific experiment or solve immediately-graded exercises with code.
I've reviewed the curated app catalogs here and TBH the tragic gap is perhaps at "how to computer math" [in notebooks in a version controlled (e.g. git) repo] [in Python [with JupyterLite or vscode.dev w/o devpod/codespaces [due to Chromebooks]]]; just a video of Arithmetic in notebooks instead of a calculator.
GeoGebra and Desmos are neat. Geogebra has a CAS (Computer Algebra System) built-in, but it's not Python with test assertions. And when the canned math app e.g. fails with weird complex exponents of e, it's a good idea to reach for Python (or Julia, or R, or) instead of only the apps in the Play Store.
Exercise: Install Git and Python (maybe in Termux from Fdroid) on a Chromebook, then run repo2podman {with a FamilyLink account,}
Exercise 2: Create a Jupyter Notebook on a Chromebook and save it to work on from home {with Gapps Edu and Family Link} when Colab isn't allowable and JupyterLite doesn't have a gdrive plugin yet.
Containers in a local devpod (~codespaces) VM for the students might solve.
This should also work on computers to support real-world STEM workflows:
<alt-tab> make test
make clean
<alt-tab> <up arrow> <enter>But nonetheless schools don't (and never have, FWIW), and it's not the OS's fault that they use important and desirable manageability features to do it.
Internally, Google has a huge git monorepo and gn and IIRC gerrit for code review (?) and externally many repos in various orgs on GitHub (some now contributed to e.g. the Linux Foundation's CNCF).
A Google Colab JupyterLite-edition could optionally integrate with Google Classroom.
There's an issue in a repo re: a JupyterLite gdrive extension and fs abstraction so that it would work with other non-git cloud storage providers.
vscode.dev supports git commit and git push from the browser tab to GitHub, which can then run the code with GitHub Actions. vscode.dev can also connect to GitHub Codespaces (or devpod). GitHub Codespaces spawns a container in GitHub's k8s cloud for interactive use with {VScode, Jupyter, and IIRC SSH} instead of batch headless use with logs like GitHub Actions.
You can write a simple git post-receive hook script in .git/config that runs a script on every push to the repo, and then later realize why each execution of a post-receive hook needs to run the job in an isolated container to prevent state leakage between invocations of the script; there should be process and data storage isolation for safe and secure GitOps.
Research institutions like e.g. CERN and ORNL have e.g. GitLab with k8s; there, pushing to a Pull Requests branch can cause a build to run the tests in a container auto-provisioned somewhere in the private cloud and report the results back to the Pull Request thread. UCBerkeley developed ottergrader (to succeed nbgrader and okpy) to grade Jupyter notebooks in containers (optionally with k8s, or locally) and post the scores to e.g. Canvas LMS.
Can the students grade their own work with containers in Crostini/Crouton//ARC without homomorpohic encryption (like Confidential Computing) running potentially signed but unknown code on their local workstations?
JupyterLite and vscode.dev+jupyter+pyodide work in Chrome on Chromebooks today, but it's really suboptimal from a mainframe-era perspective.
Gerrit was deprecated a while ago.
You say "important and desirable manageability features" in your other comment and I strongly disagree. Disabling the web inspector is not important and it is the exact opposite of desirable.
For most Google developer developing on a Chromebook basically means running ssh and chrome remote desktop.
For Chromebook sysadmins, Instructors, and Students learning about how {Linux*, ChromiumOS, Android, Git, Bash, ZSH, Python, and e.g. PyData Tools supported by NumFOCUS} are developed, for example;
When you git commit to a git branch, and then `git push` that branch to GitHub, and create a Pull Request, GitHub Actions runs the (container,command) tasks defined in the YAML files in the .github/workflows/ directory of the repo; so `git push` to a PR branch runs the CI job and the results are written back as cards in the Pull Request thread on the GitHub Project; saving to the server runs the (container,command) Actions with that revision of the git repo.
Somewhat-equivalent GitOps CI Continuous Integration workflows (without Bazel or Blaze or gtest or gn, or GitHub Enterprise or GitHub Free due to the kids' intererests) that might be supported at least in analogue by Education and Chromebooks: k8s with podman-desktop in a VM, Gitea Actions (nektos/act; like Github Actions), devpod
devpod: https://github.com/loft-sh/devpod :
> Codespaces but open-source, client-only and unopinionated: Works with any IDE and lets you use any cloud, kubernetes or just localhost docker. (with devcontainer.json, like Github Codespaces)
devcontainer.json is supported by a number of tools; e.g. VScode, IntelliJ,: https://containers.dev/supporting
repo2docker has buildpacks (like Heroku and Google AppEngine).
repo2docker buildpacks should probably work with devcontainer.json too?
repo2docker docs > Usage > "REES: Reproducible Execution Environment" describes what all repo2docker will build a container from: https://repo2docker.readthedocs.io/en/latest/specification.h...
jupyterhub/repo2docker builds a Dockerfile (Containerfile) from git repo (or a Figshare/Zenodo DOI) that minimally has at least an /environment.yml and /example.py (and probably also at least a /README.md to start with), and installs a current, updated version of jupyter notebook along with whatever's in e.g. /environment.yml per the REES spec. [1][2][3]
[1] repo2docker/buildpacks/base.py: https://github.com/jupyterhub/repo2docker/blob/main/repo2doc...
[2] "Make base_image configurable" https://github.com/jupyterhub/repo2docker/commit/20b08152578...
[3] repo2docker/buildpacks/conda/environment.py-3.11.yml: https://github.com/jupyterhub/repo2docker/blob/main/repo2doc...
[4] "When to use [TLJH or Z2JH]" https://tljh.jupyter.org/en/latest/topic/whentouse.html
SLSA; Supply-chain Levels for Software Artifacts:
https://security.googleblog.com/2022/04/how-to-slsa-part-2-d...
https://slsa.dev/blog/2023/06/slsa-github-worfklows-containe...
How do ContainerSec, DevOpsSec, and UI/UX apply to getting a perhaps necessarily at least sometimes restricted set of containers provisions to support STEM lab module workflows for learning?
1. Instructor: Specify e.g. docker.io/busybox:latest, and docker.io/buildpack-deps:container_tag as the necessary containers for the course/module/day
2. Student: Login to that context (as allowed by domain policy if necessary), have the container images pulled according to idk like a syllabus.yml before the schema.org/CourseInstance :eventSchedule :Event begins or at the start of class per a distributed URL/QRcode.
3. Student: Review and manage which containers are running with something like the podman-desktop indication bar applet.
4. Student: git push a repo of notebooks to be graded with ottergrader, okpy, nbgrader (and get feedback on a Pull Request)
Gaps/Challenges/Opportunities:
Instructors don't have CI workflows implemented in curricula, but easily could create a git repo that works with repo2docker and/or devcontainers.json.
Education Organizations don't have control over which containers are running on students' managed workstations, do have such monitoring for which apps are so errantly provisioned; and so students don't have containers but do have apps.
Android Apps are installed as APKs, which are a ZIP file with a manifest. Binaries on Android devices will not run without SELinux labels. SELinux labels for executables on Android can only be granted by (root or) a the process that installs the APKs.
This means that other installers like pip and apt (in Termux or any app off the Android Google Play Store with the new SDK) cannot work in an Android App in a VM on a Chromebook. This means that e.g. Termux could work from the Play Store only if they repack every installable package as an APK on the Play Store. Python is installable with Termux on Android with Fdroid; but the Fdroid Android App repository ("Allow third party sources" (with a different cert trust root)) doesn't work on Chromebook Android, so FWICS you can't install Python with Termux apt on a managed Chromebook, and you can't keep sideloaded APKs from Fdroid updated with an unmanaged Chromebook: a [Google] dev would just use containers in a VM (or CI) instead.
JupyterLite and VScode.dev work in a browser tab because everything is compiled to WASM (WebAssembly), but the browser traps shortcuts like Ctrl-P.
So then you might say, "wrap JupyterLite in Electron or similar so it works offline as an app with all the keyboard shortcuts that you need". (Electron apps run a web app locally in a browser without the chrome (the address bar and back buttons); but if 5 apps each depend upon unique copies of Electron that they must repack and that users must keep updated, app sizes are suboptimal due to lack of a proper package dependency model for Android APKs.) Flatpack is neat, supports package dependency edges, and supports having multiple versions of a package installed, but most flatpaks have similar boundary violations where the guest container is insufficiently isolated from the host machine. VScode flatpak, for example, can call commands on the host with `flatpak-spawn` or `host-spawn` or the additional package for the vscode flatpak copy of the `podman-remote` go binary. distrobox and fedora/toolbox make it easy to mount your entire $HOME into the container with the correct UID and file permissions: `distrobox create arithmetic; distrobox enter arithmetic` provisions a container and opens an interactive shell within the container. ChromiumOS's zygote messages also cross container/vm boundaries IIRC.
gvisor is considered good enough to contain containerized workflows for shared multitenant workloads at Google; and so it should also probably be useful for getting devcontainers.json and at least vscode working on Chromebooks for the kids.
chromiumos/docs/+/HEAD/containers_and_vms.md > Can I run Wayland programs? with Sommelier: https://chromium.googlesource.com/chromiumos/docs/+/HEAD/con...
But that doesn't solve because containers and vms aren't on and aren't supported for their accounts.
A school chromebook's access to containers could be controlled by setting the containers.conf repo URL to a Container Image Repository controlled by the school. GitHub, Gitea, and GitLab all support storing (OCI) container images.
An instructor would import a container image with an associated Pull Request that causes an Action to run to (1) scan the container and its SBOM Software Bill of Materials; before (2) hosting the container image for the students and (3) regularly (along with e.g. Dependabot, which for security regularly checks for references to outdated versions of software in GitHub repos) .
It looks like GitHub supports 3rd party code scanning tools, too; so Instructors and Students could auto-scan for security vulnerabilities and get reports back in the Pull Request https://docs.github.com/en/code-security/code-scanning/intro...
GitHub Project Templates are designed to be forked; e.g. like an assignment handout to be filled out (that already has an /.github/workflows/actions.yml and README.md headings). Cookiecutter is another way to create a project/assignment/handout/directory/git_repo skeleton; with jinja2 templates to generate file names like `/{{name}}/README.md` and file contents like {% if name %}<h1>Hello World, {{name}}{% endif %} . jinja2 is a useful skill also for ansible [collections of roles of] playbooks of tasks.
chromebook-ansible installs a number of apps by default (including docker and vscode (instead of podman and vscodium or similar)), but because there are variables in the playbook, you can change which parts of the playbook runs by specifying different parameters with ansible inventory: https://github.com/seangreathouse/chromebook-ansible#include... https://github.com/seangreathouse/chromebook-ansible/blob/c8...
It would be helpful to be able to provision [Android and Chromebook] devices with [Ansible] like it is possible with Mac, Windows, and Linux devices (without a domain controller; decentralizedly and for bootstrapping). It appears that there happens to be no way to `adb install play-store://url` with Ansible, but there is news about Ansible support for Enterprise Chromebooks.
There are [vscode] IDE mentions in the chromebook git repos IIRC. The [vscode] [docker/podman extension] could work with aforementioned functionality to limit which containers can be pulled or are running at a given time.
USE CASE (for a "STEM workstations for learning" spec): Create a minimal git repo project from a project template with cookiecutter-pypackage or similar.
A minimal project [template] would have at least:
/README.md # h1, badges, {{schema:description}}, #install, #usage, #license, #citation
/LICENSE || /COPYING # software/content license terms
/devcontainers.json # for vscode and other IDEs
/environment.yml # which packages to install (conda, pip, repo2podman)
[/src]/example/__init__.py
[/src]/example/main.py
/tests/
/tests/__init__.py
/tests/test_main.py
/example/tests/ # this dir would ship with the package, so that that tests Could run in production
/example/data/ # will be installed in site-packages
/data/ # will not be installed in site-packages without a special setup.py
/docs/
/docs/conf.py # sphinx
/docs/index.rst # .. include('../README.md') .. toctree:: # sphinx
/docs/_toc.yml # jupyter-book
/docs/_quarto.yml # quarto
/docs/_metadata.yml # quarto
# When we need to vendor _other projects in
# we then need src/ (and/or lib/) for _those (and maybe ours_ 2)
[/src][/example]/tests/
[/src][/example]/tests/__init__.py
[/src][/example]/tests/test_main.py
[/src][/libexample2]/COPYING
I used that laptop to skip over the lousy overheating butterfly MacBook generation. Now I have an M1 Pro MacBook like everyone else.
They also offer Windows on the cloud VMs.
1. chrome-os device
2. install a debian VM
3. do all their development in a VM where they can easily install packages they need etc?
4. so they are doing dev in a VM on a core-i3 intel machine I guess?
(Disclaimer: I'm the current maintainer of the Eupnea project.)
[1]: https://github.com/eupnea-linux-backup/depthboot-builder
For many legacy command line tasks, you SSH to a cloud workstation. More than ever though, you can get away with a lot of programming tasks without ever touching a terminal.
For those who really want a traditional Linux Desktop environment, you use Chrome Remote Desktop to connect to the same cloud VM that you SSH to.
Genuine concern with anything google. Current and future risks.
* window management was a hit and miss. I was not able to do split screen for a Chrome window and a Linux window (vscode). You can do that for two Chrome windows. A big disappointment * many GUI applications are only available through... Android Play Store, e.g. Cisco Anyconnect. Fortunately this works, but it's not an enjoyable experience. * Chromebook built-in keyboards usually don't have F11 or F12 keys, but you need them as a software developer.
And there are a number of other smaller issues that made me feel it would be easier to stick with a "standard" laptop, although I do think there is a lot of potential. I sold that Chromebook later.