Upcoming 'Supervised User' Feature of Google Chrome | Hacker News Reader

Upcoming 'Supervised User' Feature of Google Chrome

browserfame.com

19 pointstwapi35 comments

I've never understood the concept of multiple users in a browser. Isn't that what accounts on my computer are for?

Isn't this already a long solved problem?

I have a personal and a corporate Gmail account, and I want to keep them separate, with their own bookmarks and extensions.

I also don't want to shutdown all of my running applications just to switch.

I can understand the use case of having a personal & corporate Gmail. However, thats still a solved problem; Google has a multi-user account signin option.

This incessant integration of google specific features into the browser is essentially the same thing as vendor-lockin.

Multi-login doesn't solve the problem of separate bookmarks, extensions, saved passwords, etc.

What exactly is "google specific" about it? Firefox supports multiple profiles, too. I'm sure they can add a supervised account feature if they think their users want it.

It requires using special command-line flags (default behavior of Firefox prevents running multiple sessions concurrently, on purpose). It is also, performance-wise, was not so good.

I am a diehard Firefox user, and I only use this crap feature on Chromium anyway. Firefox is what I use for my personal lif, as it is in my opinion a better browser, but a more trustworthy org. Chromium/Chrome is for work, and I do not entrust Google for anything more than that these days

Multi-user is hilariously broken on the products where it is implemented, and not implemented on most Google products.

You can just use multiple browsers

You can do this with chrome using the --user-data-dir option, which works fantastically well. For my work gmail I have the following shortcut:

chromium-browser --user-data-dir=~/.profiles/work-mail-profile -app=https://mail.google.com

But why would you want to do that instead of just using the built-in multiple profiles feature? The only thing I've encountered that required a separate browser instance was using non-default command-line flags for testing something with different proxy options.

Why would I do that when I want to use Chrome only?

I can have both my Github accounts logged in, both of my gmails, etc, and get my favorite interface. I'm glad I don't have to launch Firefox or Safari anymore.

Same reason I use it. Our office insisted on using Google Apps, and this is the only way I can manage multiple email accounts effectively (using Chromium though; I do not use Chrome, because I do not trust their "only so much source" approach, and thus use the free bits only).

you might want to take off that tinfoil at some point, let your brain breathe

way to put someone down for being not only security aware, but principled as well.

Thanks for trolling. It is not tinfoil hat. I just prefer to use open source and see the WIP of Chrome, which I get by downloading the nightly alphaware. I find security in open source not because I distrust Google, but I trust the open source development model to be more transparent. You can disagree, but I thought that was a common viewpoint on HN. But it is ok to make fun of me for it.

But I guess my brain does not breath as much as yours.

Obviously, but in my experience, at least, normal non geek home users want a computer that just turns on, they don't want to be logging on to different accounts, let alone managing them. Even I gave that lark up. With 6 kids it almost became an unpaid support role.

Yes, of course there would be some of that for the browser, but managing browser logins, and what not, is less overwhelming for mum and dad.

Also, given all the integration we now have, I guess that a chrome login account will transfer from device to device. So, perhaps the setting on the home PC's chrome, will still apply on say the tablet's chrome.

Non tech people don't use multiple accounts on their home (windows ?) computer. Yes they should but they don't and from what I see they don't even fully understand the concept (why are my programs not there ? I want to start skype !), so here we are solving the same problem at another more limited level.

I use it to manage multiple gmail accounts (personal/business). They put an icon in the top right corner of the browser which makes it really easy to switch profiles. Apart from this though I can't think of any other reasons to need multiple profiles.

Testing web apps? That's the first thing I thought of.

Google is hoping the browser will replace your operating system.

This appears to be for a classroom full of Chromebooks. You could restrict students to one website during a test, for example.

Modern computer use never integrated multiple accounts well. For example, you can't share an itunes library between multiple users. I don't know of any modern OS that even has a standard "public shared folder for all users" location. If I turn on itunes sharing on one account, I need to leave itunes open and "switch users" rather than logging out. That is, I must log in as that user every time I turn on the computer.

The problem isn't separation of powers. It's maintaining modern usage patterns with separation of powers.

Really? I have been an on-and-off sysadmin since college. The "public folder" for sharing data on the Windows operating system is accessible via the %ALLUSERSPROFILE% environment variable [0], and has been so since Windows XP. That means at least 15+ years now. On OS X, there is /Users/Shared, but I am not sure how old this convention is old enough for me to presume it started with the earliest versions of OS X. Unix has been way ahead of the game, but in different variations, and if you include/exclude Linux as standardized, you can read the FHS standard regarding where files ought to go.

Having said that, this does not mean these folder arrangements detract from your argument. Windows was not originally designed to be a multi-user/multi-tenant system, and evolved from a one-user turd to a multi-user turd (albeit a shinier, polished, more refined one). OS X roots in Unix dictate it should be better but it is not. Outside of Windows, it is largely software developers obvious intentions or programmatic error. In the case of Windows, there later turn around was not honored by developers, some long gone (out of business or not interested), who refused or continue to refuse convention regarding standard, like the All Users Profile. Windows has a lot of indirection APIs that layer over the OS so people can write to the same registry and file locations they should have not been using for 10+ years, that are in system-file only resource areas or require admin privileges to run and access system features they most certainly do not need. iTunes, as a different kind of example, regardless of OS, is a trend of a-hole software companies dictating how and where you use data you own. You can move a iTunes library, but it is annoying and DRM means the benefits are useless anyway. It is not like you need an exclusive lockfile to play a MP3, you need to buy into their exclusive ecosystem of garbage and put a lockfile on all of your computing choices. But that is another topic.

Point being, only Unix was designed for multi-user tenancy. I am not saying well, but it was made for by large corportations and research groups to share data. The others were modified later to do so, and at a cost, as developers out of the game or too disinterested to pivot ignored/ignore these changes. The end user does not care, so why follow the rules? I get your point, but I think it is wise to assume this issue is far more socio-political then technical.

We are humans, we no likey sharey.

[0] http://ss64.com/nt/syntax-variables.html

You can share an iTunes library between accounts by pointing a preference at the same directory for each account. /Users/Shared is a nice one because it alrwady has friendly permissions.

Many places where Chrome runs don't have user accounts (phones/tablets/GoogleTV). In many more places users don't know how to setup an OS account.

Additionally there's a lot of cross-over between features being implemented in Chrome (the browser) and ChromeOS.

Often the browser is used to surface features which are implemented differently in ChromeOS to how they are implemented in other ChromeBrowser implementations.

It's hard to install two versions of the same browser separately, and changing accounts is just overkill if you only want to separate bookmarks and sign ins or whatever.

It's also really useful if you want to open pages without extensions or without being signed into the website. You can do that with incognito, but this is more flexible. You could have an account that is configured to a proxy and one that isn't, if I understand correctly, and other settings as well.

Also browsers are pretty much becoming their own little operating systems.

You can't use multiple user accounts at the same time. Easily anyway, that I know of.

This feature has a terrible name.

"Superuser" is an existing term for a user with more privileges than a non-superuser.

"Supervised user" is Chrome's new term for a user with fewer privileges than a non-supervised user.

These two terms sound nearly identical but mean opposite things. I'm surprised Google apparently didn't think about this when they named the feature.

This feature is aimed at the general public, who by and large, probably won't have such a confusion.

"Supervised Accounts" is apparently going to be some kind of parental control, but the article lacks any details on what it actually will do. My assumption is it that it will include some kind of content filter, possible based on keywords?

If it's a parental control thing, then one of its main uses will be increasing the number of computers on which someone has covertly installed Firefox.

Okay, Google, now how about making Chrome work correctly with OSX Parental Controls?

bloat.

Exactly.

Maybe this makes sense on Chrome OS. But for other main stream OSes like Windows, User accounts should be best left on OS. maybe a quick way to change profiles but that should be limit. Any further than this would be just reinventing the wheel & duplicating the features(parental control for example) that are already present in Modern Operating System.

on a related note, has anyone noticed memory usage of Google Chrome? Its crazy bonkers?

It makes sense on Windows and OS X too. I have a couple sites where I have multiple accounts, and it's nice to just fire up another instance of Chrome where different credentials are saved. Different extensions per account and so on. I shouldn't have to switch users at the OS level (which is slow) just to manage multiple online identities for me.

This is very much needed on my iPad.

Reply on news.ycombinator.com