Some Haiku design questions

Hi all,
I’m pretty new to Haiku and I’d like to ask some questions about design aspects:

  1. As far as I understand everything is run as root. Is there any reason for this design decision?
  2. Again, as far I understand there is no multi-user support in Haiku, is this wanted or is this just a WIP feature?
  3. Webpositive is OK, so is Otter. I’d like to know if there is any techinal reason to not port other opensource browser like Firefox and Chromium, or isn’t really any interest in having these browsers on Haiku?

Thanks for your time,

Haiku is a single user system.
Technical design is root. In Haiku it is called “boot”

Yes. You can see process tree on screenshot. It is not design decision, kernel support multi-user and permissions, but some GUI parts are still not support multi-user. It will be eventually implemented. I have managed to run GUI environment with separate non-root user and its own home directory, but with test GUI server that run in a window on top of system GUI server. GUI user session switching is not yet implemented.

Interest exists, but porting modern Firefox or Chromium is difficult. There are old Firefox 2 port called “Bezilla” and some attempts to port Chromium engine Blink (can render some pages but crash often) exists.


@X512 Thank you very much for this detailed answer!

I think the major reason is the the group who asks for this and the group who can do this have really small section.

1 Like

root in this context refers to the root user (superuser) not to the root filesystem

1 Like

There is no interest from most of the people developing Haiku in making it just another OS to run Linux applications. Our goal is to develop the OS to run native Haiku applications. Including the web browser. This is why we decided to put our effort in WebPositive and base it on WebKit, which is one web engine that can be integrated very tightly with our native APIs.

This should result in a faster browser and less memory use. I think the amount of work required is similar to porting other engines (a lot of work in all cases).

I think it makes sense to not spread efforts in porting 3 different engines when we could focus on just one. But I’m sure eventually someone will port the other two anyway.