HaikuOS Hits Kernel Milestone

Professor Cool Linux writes "HaikuOS kernel team has a working implementation of fork(), exec(), and waitpid(), as well as a TTY layer. 'Despite some minor issues (you still don't see what you type), we can now have a fully functional bash running on our kernel. A lot of other shell commands, even those using our own libbe.so are working, as well.'Things are really moving."

Comments

Kudos to the Team

[RAMMS EIN]

BeOS still has some interesting features that other systems lack, and an open-source implementation of it will allow the whole world to improve it and benefit from it.

Kudos to Michael and the other hard-working people.

Re:Kudos to the Team

[Anonymous Coward]

For example?

Re:Kudos to the Team

[MarkusQ]

IIRC, some of its nifty features were:

• Threads as a way of life, rather than as an afterthought.
• A semantic file system.
• An interesting object model/UI fusion.
• The support of famous SF authors.

I've no idea if these will all be true of the port though.

-- MarkusQ

Cosmoe

[RAMMS EIN]

Also check out Cosmoe, which is a port of the AtheOS API (similar to BeOS's) to Linux. I think it's a really cool idea that is not getting the attention it deserves. It shares some code with HaikuOS, so you can help both by helping one.

Nearing Completion

[RAMMS EIN]

Noth that HaikuOS is actually much further along than the kernel, broken keyboard driver, and shell. They are basically approaching a complete system from all directions, and have application server, media kit, etc. already in various degrees of completion. Before you know it will all fall into place and there will be a fully open-source BeOS clone, binary compatible with the original.

I don't want to rain on anyone's parade but...

[deminisma]

"Things are really moving."

That may be true, but they're still a long, long way away from their destination. They still don't have an alpha-quality kernel or interface kit, 3 years after work began. There's a heck of a lot of work to be done, by a pretty small team. Also, the developers are trying to clone BeOS 5, an OS released in 2000 - so when they finally get something usable and stable, at a guess sometime in 2006, they'll have a copy of an OS that is over six years old. Sure, there will be some differences, but that is the gist of it. Tiger will be out by then, Longhorn will possibly be out in that timeframe, and GNOME and KDE will be more advanced than ever. Meanwhile, Haiku will still be single-user!

I hate to be such a pessimist, BeOS was a great OS for its time, but it seems circumstance will relegate Haiku to the fringe, used only by hobbyists. It's a shame really that they were so ambitious (maintaining binary compatibility, pretty much doing it all from scratch), and that the efforts to implement the Be API over a Linux kernel stalled.

Re:I don't want to rain on anyone's parade but...

[mnmn]

BeOS was a great OS for its time. Its design is still that of a great OS even now. BeOS is lean and geared towards media, with opengl built in. Its threading system, filesystem were pretty advanced. Think OS X.. only simpler and faster.

Few OSes now are efficient for the desktop, linux and bsd are generalists, and windows is not efficient at all OS X itself comes close, but still not so well designed for being a desktop OS inside out (OS X doesnt have to be. it runs on awesome hardware, and has an awesome kernel+libraries, the most efficient for being a general OS, but not internally designed to be a desktop OS).

Instead of Haiku, I was hoping for changes to the Linux/BSD kernel to make it more desktop-geared like BeOS was. That, and establishing API and package standards should bring Linux/BSD out of the dark ages and onto grandma's and grandkids' computers.