FreeBSD Plans For The Next Ten Years

[wikinevick]

I am not so worried about "systemd" on FreeBSD because I am quite sure the developers will do a great job (and they are looking at xml / plist files rather than binary shite)

And I am not worried because they will take their time on it and it won't have the ugly land-mines the linux crowd has already stepped on, like binary logs.

[kpedersen]

mobile devices like tablets, smartphones, and embedded devices

Nope

PC desktop

Nope not really

FreeBSD is still mainly about the servers! Desktop is perhaps fun but not the main aim of the project. Gnome 3 is pretty damn low on their priority list that the base is very unlikely to bend over for any gnome port.

They are perhaps focusing on getting the thing running on ARM servers. Something that OpenBSD / NetBSD has maintained continuously so might be a little ahead. Not sure though.

I remember hearing that the FreeBSD project (or was it OpenBSD?) cannot be arsed with consumer ARM devices like tablets or smartphones until they get a standardized boot system (like a bios, so you don't need a separate image per damn device).

[Xaero_Vincent]

Nope



Nope not really

FreeBSD is still mainly about the servers! Desktop is perhaps fun but not the main aim of the project. Gnome 3 is pretty damn low on their priority list that the base is very unlikely to bend over for any gnome port.

They are perhaps focusing on getting the thing running on ARM servers. Something that OpenBSD / NetBSD has maintained continuously so might be a little ahead. Not sure though.

I remember hearing that the FreeBSD project (or was it OpenBSD?) cannot be arsed with consumer ARM devices like tablets or smartphones until they get a standardized boot system (like a bios, so you don't need a separate image per damn device).

While watching the presentation, developers kept talking about how FreeBSD is being used as a base for other products, such as FreeNAS and Playstation 4's Orbis OS. They want FreeBSD to become more like "lego blocks", essentially a bunch of components that can be used to produce OS solutions for all sorts of devices, such as embedded in your car or smart cables, game consoles, routers & firewalls, mobile devices, etc. Gnome 3 was probably a bad example as FreeBSD already has a full Gnome 3.14 port but meant that future programs may begin to depend on systemd, which may become troubling.

Actually, judging from the presentation, FreeBSD devs long term vision seems to be to move FreeBSD away from being a server centric OS or at least branch it into new directions. According to marketshare reports, FreeBSD's use on the server has been declining sharply in favor of Linux and Windows Server, so it doesn't make much sense for the devs to limit FreeBSD's future as a server-only optimized OS in the long term. I understand the FreeBSD desktop is a bit of a stretch, but hey I'm on Phoronix, where I think most of the readers here are interesting in desktop *nix OSes and seeing how their graphics stacks and gaming capabilities are improving. It's only natural for us to ponder the idea of *BSD improving in this space as Linux has been.

[kpedersen]

FreeBSD is being used as a base for other products, such as FreeNAS and Playstation 4's Orbis OS..

True but FreeNAS is effectively 100% a server product and the FreeBSD developers had no input into the PS4 OS, that was all Sony.
PC-BSD is perhaps the official desktop distribution of FreeBSD but that does not seem to be swaying the core OS at all.

Gnome 3 was probably a bad example as FreeBSD already has a full Gnome 3.14 port but meant that future programs may begin to depend on systemd, which may become troubling.

This is not quite a full port, it is missing a lot of functionality like WiFi, automount (perhaps currently due to systemd but before then it was hald, procfs etc... etc...). The fact of the matter is that Gnome 3 took a long time to get to FreeBSD because it was such low priority, even OpenBSD (a much smaller project) got an official Gnome 3 in ports months before FreeBSD.

Future software (especially for the server) will never depend on systemd because Windows and Mac OS X do not provide it. Open-source Unix software has always been portable to those platforms in the past so why suddenly would people stop writing portable (correct) software just because they *really* want to use some Linux-centric backbone API haha.

[profoundWHALE]

While watching the presentation, developers kept talking about how FreeBSD is being used as a base for other products, such as FreeNAS and Playstation 4's Orbis OS. They want FreeBSD to become more like "lego blocks", essentially a bunch of components that can be used to produce OS solutions for all sorts of devices, such as embedded in your car or smart cables, game consoles, routers & firewalls, mobile devices, etc. Gnome 3 was probably a bad example as FreeBSD already has a full Gnome 3.14 port but meant that future programs may begin to depend on systemd, which may become troubling.

Actually, judging from the presentation, FreeBSD devs long term vision seems to be to move FreeBSD away from being a server centric OS or at least branch it into new directions. According to marketshare reports, FreeBSD's use on the server has been declining sharply in favor of Linux and Windows Server, so it doesn't make much sense for the devs to limit FreeBSD's future as a server-only optimized OS in the long term. I understand the FreeBSD desktop is a bit of a stretch, but hey I'm on Phoronix, where I think most of the readers here are interesting in desktop *nix OSes and seeing how their graphics stacks and gaming capabilities are improving. It's only natural for us to ponder the idea of *BSD improving in this space as Linux has been.

Something I keep thinking about is the NT kernel. It was originally designed for servers (Windows Server 2003) and enterprise, but then decided to replace their DOS based kernel with NT on their desktop products. Server kernels seem to focus on performance and stability, which I would say that FreeBSD is great at, but the big difference (if I understand this right) between FreeBSD and Linux has to do with FreeBSD believing that there should be essentially one right way to do something. The great part about this is you get the stability (things written for it are guaranteed to work) but you lose flexibility. Linux has a lot of flexibility, and actually due to the nature of the multiple ways of doing things, it seems like programs are written so that there are multiple ways of having it work (and break).

Two sides of a coin, and I like it that way.

If they want more market share, all they'd have to do is emphasize that their product is free, here is paid support if you need that, these very successful businesses and platforms are doing this. Less than 1/10 of people I talk to have maybe heard of Linux. They've heard of Android or Ubuntu, but not Linux. Less than 1/100th have maybe heard of FreeBSD.

Ubuntu has done one thing VERY well, marketing. What would happen if someone came out with a BSDbuntu that had FreeBSD preconfigured with Linux compatibility layers and everything that Ubuntu comes with?

[nasyt]

And none of them contributed back. They should get rid of those leeches.

They contribute back. The Darwin OS is under the Apple Public Source License. Grand Central Dispatch was pushed back to FreeBSD and Launchd is about to be incorporated (maybe). Both are Released under the Apache License 2.0.

http://en.wikipedia.org/wiki/Darwin_...ting_system%29 says

In 2000, the core operating system components of Mac OS X were released as open-source software under the Apple Public Source License (APSL) as Darwin; the higher-level components, such as the Cocoa and Carbon frameworks, remained closed-source.

[Guest]

They contribute back. The Darwin OS is under the Apple Public Source License. Grand Central Dispatch was pushed back to FreeBSD and Launchd is about to be incorporated (maybe). Both are Released under the Apache License 2.0.

http://en.wikipedia.org/wiki/Darwin_...ting_system%29 says

In 2000, the core operating system components of Mac OS X were released as open-source software under the Apple Public Source License (APSL) as Darwin; the higher-level components, such as the Cocoa and Carbon frameworks, remained closed-source.

I know, but I was talking about things FreeBSD plans to do in upcoming years. Somebody said most of them are already done at apple. If that's true apple doesn't want to help FreeBSD by contributing those changes back. Btw. parts of Darwin that were open sourced by apple are hardly interesting for BSD in my opinion.