Feat: Add custom/global Sandboxes

[SpecialAro]

Pre-flight Checklist

Please ensure you've completed all of the following.

• I have read the Contributing Guidelines for this project.
• I agree to follow the Code of Conduct that this project adheres to.

Description of Change

Enables custom sandboxes or global sandbox

Motivation and Context

This feature fixes #1084 and enables the grouping services in the same custom sandbox (by default, services still have their own sandbox). It was also introduced a feature to disable sandboxes completely, which puts all services in the same "global" sandbox.

⚠️ KNOWN ISSUES:
Services in the same sandbox share the same webviewContents - that is why I introduced a workaround in the Service.ts. For this reason, some issues may arise, e.g. zooming in/out/reset in a service within a sandbox affects other services in the same sandbox.

Screenshots

In the screenshot, services on the right are services on its own sandbox (without a custom sandbox), and services on the left are services on a custom sandbox (named "GOOGLE" in this example). Should we add a title to each column so it is clearer?

Checklist

• My pull request is properly named
• The changes respect the code style of the project (pnpm prepare-code)
• pnpm test passes
• I tested/previewed my changes locally

Release Notes

[vraravam]

@SpecialAro - how is this "stored" for folks using the ferdium server?

[SpecialAro]

I haven't updated the logic for using Ferdium server, the file is only stored locally. Users can replicate or have different sandboxes in different machines. But we can also implement it if you think it is worth it!

[vraravam]

But we can also implement it if you think it is worth it!

im actually against this feature itself - since it violates the security aspects of sandboxing. but, since you have implemented it, i would assume that you also have found a need for it.

[SpecialAro]

I understand you and I'm really up to make all the warnings possible that turning this feature on is a security risk - as we loose the sandbox itself.

The only use I found for it (and I think it is what the users in the issue were requesting as well) is not having to login independently on each service (for instance, I can login once in GMAIL and it will allow me to use that same account without having to login in other google services inside the custom sandbox. I can also add services that use Google auth to login more easily without the need of inputing the password) - as a regular web browser does.

Nevertheless, I totally agree this is somewhat of a security risk... So I won't be sad if this isn't merged at the end or if we should work on it a bit more 😁

I just think that users should be the ones to know if they want to risk it or not - for instance, that's the same with the webRTC settings for exposing IPs - at the end, the default option should always be the safest one, but users may be able to change it if they need it.

Let me know what you think!

[maddin79]

Hey, thank you for this. I really need this feature. I have 2FA auth and SSO in my company and have several applications running in Ferdium. I have to login several times a day and that is really annoying.
I must admit that I do not really see the benefit of sandboxes. Yes, applications can not talk to each other and that is more "secure". But nowadays, especially in a company environments, everything is build up that applications know from each other or share a context, especially when using SSO.
I also think that developers should not take control over the decisions of users. The sandbox can be the standard, but if the user decides to turn it off, then this is fine (of course with warnings). That is actually the reason why I hate Windows and love Linux ...

So, please merge it. I'm really looking forward to check this out. @SpecialAro Thanks again for your work and thanks to all the Ferdium people... great tool ... love it.

Best
Martin

[vraravam]

@maddin79 - the principle of being secure has been there for quite a long time - even pre-Ferdium.
There are other followup issues that would come with each implementation/feature. I dont disagree that this can be useful for some swathe of users, but the default should be what the pre-existing users are expecting.

Also, as called out multiple times, unless the code-contributor-base increases, any support for new functionality will need effort which might not happen. (Remember that this is a volunteer-run project.) So, with each new added functionality, the breadth of such coverage will also need involvement. Are you or others signing up for this? (I am definitely not willing to do so due to other commitments).

[maddin79]

@vraravam As I said, the current state of Ferdium should be the standard setting. But a user should be free to choose if sandboxes are enabled or not.

You referring to security, but what security do you mean? Are browsers without sandboxes not secure to use and nobody should use them?
Security comes from knowing what I'm doing and being careful of what I'm sharing, clicking or opening in the web. If users are forced to certain things they don't want, they go around it and mostly with insecure solutions.

I totally see your point of the lack of contributors and I'm sorry that I can not provide more help. But is it really a new feature to turn of sandboxes? And @SpecialAro did the work for this.

[vraravam]

But is it really a new feature to turn of sandboxes? And @SpecialAro did the work for this.

im equally concerned with the support for this feature and who will do that support. If @SpecialAro is signing up for this, that concern can be mitigated.