Other than moving it to github, is there actually anything new here? They're still talking about the xmas release (2008), and of targeting Darwin9 (2007) and Darwin10(2009).
Exactly -- the latest major release is Darwin 15 (corresponding to OS X El Capitan / 10.11). I'd love to see this (or a similar) project succeed and remain relevant, but it just seems to be falling further and further behind the official releases.
(To be clear, I know there are good reasons for this, and I'm not trying to imply that a lot of work hasn't gone into what has been released ... but I can't help but wonder if there's no longer any hope of it catching up with (semi-closed) Darwin.)
My interest in having a running Darwin system is the ability to build OSX things on an openly licensed platform. I know that it is permissible to virtualize OSX on top of OSX, but even that is a PITA and more so in a TeamCity/jenkins/etc setup
Giving what the mingw project has contributed to being able to build Windows binaries on Linux, I'm pretty sure it would be possible to do something similar for OSX, but since so much of it is already open source, it seems a waste to discard the exact code that OSX is running
This is something I've been waiting for. After years of standby I've been working on Arch OS X again lately (fully operational but currently waiting for trademark approval from the Arch Linux team) and Arch Darwin is something I was contemplating to do for fun at some unspecified timeframe in the future.
If you don't mind me asking: Other than intellectual curiosity, why would you want to run Darwin, and not just one of the BSDs? Aren't a large number of the drivers still closed source?
The kext mechanism is pragmatic and interesting: as opposed to Linux kernel modules, binary kexts are compatible across whole kernel versions.
Also, having alternatives is always interesting. I was also interested in making the core of OS X† have some form of FOSS liveliness, but I've been quite disheartened about OS X and openness (however limited) since this[0] happened.
† I do believe that FOSS is very important, but I also recognise the critical effect of closed source innovations, and believe that a delicate balance and synergy between both aspects can produce fantastic results.
Strange The OSS kext I use for SMART USB HDDs [0] is also signed, but obviously not the source. You can download the compiled and signed driver here [1].
Darwin was cool until Apple quit releasing ISO files and just the source code that needed a lot of work to debug it and put it back together.
I'm glad to see someone took it over and has a goal of making Darwin ISO files. I hope it also gets an OSX themed skin for whatever Desktop GUI they decide to use with Cario Dock or something to look like OSX.
Wow, it isn't built on Objective-C? That's completely the reverse of my expectations. I'd picture an "Open Darwin" as a ground-up FOSS rewrite of Mach + XPC, followed by XNU + IOKit, followed by libobjc and CoreFoundation (and libdispatch), followed by most of Cocoa, then all the stuff like LaunchServices to rotate the userland around to an app-bundle-based perspective... and and then maybe getting to the Aqua stuff, sometime around when HURD is usable. You know, basically like the order Mono occurred in.
More to the point, I'd expect the first-milestone goal of an "OpenDarwin" project would be to be able to install and run OSX's Server.app and act as an OpenDirectory master, NetBoot Restore manager, etc. That proves compatibility, and is a useful thing all on its own (because, among other things, it would mean being able to manage OSX from non-Apple hardware†.) Doing a GUI first? Crazy talk!
† ...without spending weeks tearing your hair out trying to conform a Linux LDAP+Kerberos+whatever server to OSX's idea of what an OpenDirectory controller is supposed to look like. I tried, I tried so hard...
I remember there was a crowdfunding campaign or so, about reimplementing this runtime (including graphics libraries) using opensource ObjC implementations.
The fit and finish of linux desktops, especially KDE, are nowhere near as good as the Mac, and adding a theme which kind of sort of makes it look like OS X only serves to highlight where it doesn't match up.
As a long time Mac user and now Linux user the only desktop which is vaguely tolerable is Elementary OS as at least someone there with a good eye for detail went through and made sure it all fits together.
> The fit and finish of linux desktops, especially KDE, are nowhere near as good as the Mac
I beg to differ, but only with the "especially KDE" part. KDE is a different beast than most other modern desktop environments, in that it does not try to mimic OS X style in any way.
KDE was started way back when with a goal of replicating a Windows-like UI, with a Start-menu like launcher and a Start-bar like task bar with applets.
It may not be to your tastes, but the fit and finish of KDE3, 4 and now 5 is excellent. Integrated application sets (K[Anything]) with integration into a centralized control/config panel for shortcuts, MIME type handling, ...
The only problem with this is that other popular applications do not bind this tightly to KDE (understandable from application developers - KDE is not the only game in town).
Previously, GTK applications in particular acted and looked like they were from '93 when run outside of GTK-centric desktop environments (XFCE, Gnome). However, Qt has developed a GTK2 and 3 engine that uses Qt and Qt themes under the hood, largely solving that problem when properly configured (try any stable OpenSuse release, for instance).
Macs are nowhere near as usable as Linux systems in any respect, especially at the GUI level. The Mac GUI is just, fundamentally, a mistake Apple made in the 1980s and never moved on from. Even Jobs recognized the Mac's error when he made NeXT, which now lives on in WindowMaker.
Why? Every time KDE (or even GNOME) undergoes major changes, so many tweaks are also broken. One of the reasons the OS X UI/Ux is so consistent and reliable and mature is because its developers aren't saying "f it, start over" every few years.
And there are bugs and annoying aspects to Aqua, it's not perfect, but it is stable.
you're being downvoted, but I've converted many a macbook to linux, and KDE was what switched them all. 4.x was far nicer than Snow Leopard-Lion, and 5.x is now vastly superior to Mavericks-current.
KDE is the interface that both Apple and Microsoft steal from/reinvent nowadays.
