Multi-Account Containers

berektes

@berektes said in Multi-Account Containers:

@xorel Thanks for informing me that the ~~Edge Workspaces isolates sessions...~~ (EDIT: Unfortunately, IT DOES NOT especially for MS accounts) I had assumed to be same thing that Vivaldi does.... Only thing really holding me on Vivaldi is the Tab Stack... Otherwise Edge has convinced me on a lot of things being an enforced part of Windows and always running in the background, would not want to run second browser.

Just wanted to correct, my MS Edge claim.

Now, loving Vivaldi even more, but it must break-through blockages and face challenges! For example, this issue could potentially be tackled by one dedicated person (ignoring how long it would take, put it in-progress)... (if I may assume) so far there have been too many UI features (with little user usage) but not much to added on top of Chromium overall that is not existing in other Chromium browser, something that would make it standout... like this feature??! Thank you for efforts in any case.

berektes

@TyrionTargaryen said in Multi-Account Containers:

@iaMMast said in Multi-Account Containers:

Really low hopes of this becoming a thing after reading the posts above. Absolute dealbreaker for me if this isn't an available option. Don't see how I can currently maintain a good workflow without juggling multiple windows.

I work self-employed for customers, I also work for a company, I have my own projects (hobby) and of course partake in leasurely browsing. It's not uncommon for this to mix during the day from hour to hour (flexible hours). The issue however is that I have shared tooling (Google, Figma, Github, etc) with different accounts. So currently I have two options (as far as I can see):

Keep switching accounts constantly

Create multiple profiles

The last option is the best option. However some issues:

Juggling windows

Not being able to use features like "workspace rules" for tabs

I have to maintain all settings separately (but they are always the same for me tbh)

I have to maintain all bookmarks separately (again, they are mostly the same and syncing them is annoying)

Syncing tabs becomes a hassle because I have to setup syncing separately.

Also other minor annoyances; but not really important.

Don't think anyone really cares, just wanted to put my +1 out there. Also I have two genuine questions:

Does anyone have a workable solution that could help me out?

I really think my use case is normal and wanting to separate concerns (accounts, privacy, etc) shouldn't be that weird. How come most users don't care? Maybe I'm making thins harder for myself. Love to hear how other people manage these challenges.

I made a suggestion a while ago in this thread, but it got buried.

Basically, since MAC cannot be done by the Vivaldi Team and we don't know if/when Chromium would implement this (I know there's a feature request in the project itself to open Profiles as Tabs, not the same as MAC but close enough), my suggestion was to make a one-click "duplicate Profile" (without Sync), and have a ruleset like Workspaces, where a page can be automatically opened in a specific Profile in a new window.

Not pretty, but it would be something until Chromium implements the feature.

(I'm wondering if the settings are in a JSON file somewhere, in which case it could be copied into a new Profile manually. But I'm not 100% sure about the browser's structure.)

Edit: @yngve I hope you don't mind if I tag you, since you know about these things. How feasible is this? Or even if a manual workaround is possible while we wait for Chromium? Thanks in advance.

Edit 2: Not "seem to", bad wording.

Interesting idea, and it would likely meet the needs of most with this one, including mine, sync is not a requirement for me.

ctb

This seems so easy to me. Perhaps it's my Linux scripting mind that is making it this way. Here is an idea how this could work I just thought of.

User is user, like [email protected] and all default browsing is done under that except what is not.

Then we create user foo1 foo2 foo3 foo4 etc. and each of the foo people are a container. So user foo1 we will call the google container. When user foo1 visits google or YouTube, it stores a cookie separate from user janedoe and google cannot see the cookies from that profile. Now, let's say user foo1 has another google account they want to log into to, the browser would create foo2 for the user and they name this container google2. Now foo2 has it's own cookies and cannot see foo1 or the cookies of janedoe.

This could continue on and on and since chrome uses a database for cookies, this is as simple as creating new databases for each container. This could be made possible by some UI competent that uses some kind of back end script to create the container and direct all cookies to said created container when active. So think tab 1,2,3 and tab one is storing cookies in the chrome default database, tab 2 is storing cookies in foo1, and tab 3 is storing in foo2.

Seems simple to me just knowing scripting. UI is outside my wheelhouse as far as coding, but this does seem easy.

yngve

@ctb That profile design is the easy part.

The tab part is the difficult part, because all tabs are hard associated with the Browser Window they are contained in, and the Browser Window (one or more) is hard associated with the Profile, in which the Cookies are stored. It is not possible to have more than one Profile in one Browser Window (which why Other Profiles, Private Window (which is a new profile), and the Guest Profile all opens as new Browser Windows, associated with the selected Profiles).

Having Multiple Profiles in a single Browser Window will require a MASSIVE redesign of the Chromium Profile management, and it will have to be maintained. We don't have the resources to do that.

ctb

@yngve

That is very sad because Firefox is making questionable choices and we have no real alternative. Sure, there are things like the Zen browser but I do wonder how long it will take for the choices of the Mozilla board to trickle down to those browsers. People are leaving Firefox in troves because of these choices of the board and we Firefox users are sort of stuck. This feature, I use it for privacy and have used it for years and just cannot live without it.

I know switching profiles is a choice but it is not a choice that really works. I have ten containers setup and to replicate that in profiles would be a huge pain. And since the containers work automatically, I would forget to switch profiles and all the privacy of that would fly out the window.

Thanks for the reply.

ctb

What about something like this from Edge:

"Automatic profile switching (available in Microsoft Edge 83 and higher)

We’re also excited to announce a new feature to help you get to your work content more easily while using multiple profiles. We call it Automatic Profile Switching. If you’re a multiple profiles user, you can check it out by trying to navigate to a work site (a site authenticated with your work or school account) while in your personal profile.

When we detect this, we will prompt you to switch to your work profile to access that site without having to authenticate to it. When you choose the work profile you want to switch to, the website will simply open in your work profile.

We hope that this will help you keep your work and personal data separate and help you get to your work content more seamlessly. In case this doesn’t work for your flows, you can choose “Don’t ask me again,” and it will get out of your way."

yngve

@ctb I have no idea how they implemented that. My guess is that they "cheated" (And one "cheat" is that they can't do that without loading all the profiles into memory when launching the browser, and check their cookies etc. for each navigation; which AFAIK requires quite a significant patch on the navigation system inside Chromium).

I'll also note the major difference everybody seems to forget: we have a couple of dozen developers, developing for 5 different operating systems (plus Chromium updates, which take 3+ weeks each time). The Edge team likely has multiples of that, working on a single Operating system target, or project.

becm

Vivaldi is also (according to chrome:inspect/#apps) running as a Chromium App.

@yngve I guess the decision on accessible resources is likely already made by the enclosing (Window) instance and the only way around that would be something like:

ditch large parts of the existing UI code and do something completely dfferent
have the tab → profile decision be made by C++/native-level Vivaldi code
create something like identities API to tie JavaScript UI code to container-like providers?

karlemilnikka

@yngve Last month, the developers of one of the other major Chromium based browsers stated that they are “actively considering the much requested functionalities of Containers and Workspaces” (source). Let’s hope they extend Chromium with containers and that it benefits other Chromium based browsers as well.

berektes

@karlemilnikka great news from the Brave team, hopefully they'll push their solutions back to the open source which will be quite a significant contribution, and the Vivaldi team can use it. Their current ETA is like end of year, which is great.