Submitted by Does anyone remember PC-BSD, the FreeBSD-based distribution aimed at desktop users? After being acquired by iXsystems and renamed to TrueOS, the graphical installer was removed in 2018 because TrueOS served more as a base for iXsystems’ other offerings, such as FreeNAS, And now, in April of this year – we missed it – development has been halted entirely.
TrueOS source code will remain available on GitHub for others that may want to continue the work that we started so many years ago. I can’t explain just how much we appreciate you all being loyal fans of TrueOS and PC-BSD in the past. We’re confident that even though this is a hard decision, it’s also the correct decision because of the exciting new projects that we’re all becoming more involved in like TrueNAS CORE.
End of an era, but PC-BSD forks such as GhostBSD have taken up the mantle.
We understand why, yet it will be sad for some and a relief for others, that’s the irony of OS development! My heart goes out to those who poured love into that project.
It makes me wonder, is niché ever sustainable, is it ever anything more than transient?
Essentially all viable alternative OSes are Unix type kernel + Linux userland. No matter which one someone picks, they’re all essentially the same user experience and pain points that turn the vast majority of people that try them back to the commercial OSes they came from. Those smaller OSes simply don’t have the budget and resources to do what Apple and Google have done by writing their own GUI and userland while scrapping the desktop Linux stuff.
dark2,
The state of indy OS dev is kind of unfortunate. While I don’t think it’s impossible for someone to create something new on their own, it is virtually impossible to create the much needed ecosystem that users will need & expect short of turning the project into a kind of clone taking APIs and technology from other more popular operating systems, but you may have to compromise on your own originality in the process. Hardware support has gotten more difficult than in the past and on top of this I get the feeling modern consumers are less savvy and experimental than in the 80s and 90s. The market is so hard to break into outside of niche areas. Even with a working product in hand, companies may have to spend hundreds of millions just to convince people to look at it. There’s no guarantee of reaching critical mass, the incumbents are regarded as good enough.
It might still make for a satisfying project though, depending on one’s reasons for doing them. You can learn alot by trying your own hand at things!
I think there is also another factor, the PC clones was an accident that IBM failed to prevent, everyone else was selling a vertical integrated experience.
With the exception of a few die hard gamers with their rigs, everyone else has moved into laptops, or is using a tablet as their first computer and in some world regions their only computer is a smartphone and the few hours they might spent at some computer coffee shop.
So that singularity of PC clones is gone, the large part of consumers are back into that vertical experience and alternative OSes like GNU/Linux only managed to come this far, because the likes of IBM and such decided to cut development costs on their own UNIX clones.
moondevil,
That’s possible, we’re really fortunate that they didn’t employ DRM the likes of apple’s on IOS. It would have crippled a lot of the innovation.
When I went to university laptops were still relatively rare (even in computer science). You’d go into the dorms and everyone had a desktop. I didn’t have a laptop until sometime in the 2000s. However I’d wager that you are right about today.
Yeah, alot of those things that came out during the 90s would have far more questionable viability if they were starting out today. We don’t have the same level of opportunities and the status quo is the path of least resistance.
I won’t even start to explain what’s wrong with your first sentence, but let’s gloss over that and let someone else moan about it.
The reason that the alternative OS world is dominated by UNIX-likes is simply because UNIX-likes have had a massive head start in development, and also, they’re quite simple to reverse engineer.
BSD, which was a rewrite and reimplementation of Bell Labs UNIX, was first released in the late 70’s. At this point, “UNIX Mania” had kicked in with full force around many computing circles, with companies either rewriting original Bell Labs UNIX, modifying the open-source BSD code, or reverse engineering from scratch. Every computer company wanted a slice of the UNIX pie, especially since it was the OS of choice by the US Government. Many of the modern proprietary UNIX flavours came out of this UNIX mania, including SunOS/Solaris, SCO UNIX, Microsoft Xenix, NextSTEP etc etc, and even non-UNIX systems, like BeOS at least boasted some POSIX compatibility.
Because of it’s simplicity, and prolificness, it was a no-brainer that most open source OSes would be based on the simple, but elegant, UNIX base, since a majority of proprietary systems followed the same logic. With so many developers, and so many projects (GNU, BSD, Mach, MINIX etc etc…) all working towards building UNIX-like systems, it was quite easy to combine several open source projects into one complete OS, just as Linux has done.
Non UNIX OSes, however, do indeed struggle. Most of the time, they are cloning something much more integrated and complicated9 than UNIX, with development teams much smaller than it’s UNIX brethren. Things like BeOS and Windows Server 2003 (which both have “cloning” efforts) are so undocumented and complicated, it can take decades to make progress, to even approach some form of compatibility. Much of the reason these projects lag behind isn’t because of lack of money, or lack of manpower (although manpower can be a big factor), it’s actually lack of documentation. When you have a black box in front of you, you have to probe it and poke it and try and work out how it works inside, without ever getting a chance to open the box and look. Obviously, more complicated OSes like Windows have poor documentation on their internals, and many of the components are highly intertwined, making reverse engineering complicated.
In short, open source UNIX-likes got a 20 year head start over other non-UNIX systems, and non-UNIX systems are often much more complicated, so they take longer to develop, and as such take much longer to accrue a userbase.
I’ll stop you there. BeOS is perhaps the best documented closed source OS in existence. The BeOS Bible, Be Advanced Topics (I own a physical copy of both), and tons of archived information from its developers have helped to make Haiku into a nearly complete OS, from scratch. Yes, it took much longer than adapting an open source operating system would, but the level of accomplishment the team has achieved over the years is nothing short of phenomenal. Compare it to ReactOS, which is understandably having a difficult time making something usable out of Microsoft’s closed source and developer-hostile Windows 2003 base.
You’re focusing on your rehearsed technical argument you can win, not the most important part: “Linux userland.” If every BSD, Linux distro, etc. just port over the same software and package managers from Linux as the fast way of getting a working system, it’s all the same to the end user. As for smaller things like BeOS/Haiku, they’ll never be popular enough to get decent hardware support while Linux takes the top spot of alternate OSes, and development will also dwindle as nostalgic programmers of the 80’s die out. Things are 100% stuck in place with the current status quo.
“Linux userland”
No, It’s GNU userland.
BSD uses their own userlands.
I once had PC-BSD running on my laptop. Linux was not supporting my wifi-card … BSD did not either, but the kernel-drivers on BDS are much more straight forward than on Linux, so I managed to modify an existing driver, to get my card working.
Overall I enjoyed the experience of PC-BSD and kept it on for a couple of years until my old laptop broke down…. and was replaced by a MacBook. At least the kernel is also based on BSD….
cybergorf,
Yeah, sometimes I think we would have been better off if BSDs had taken over instead of linux, not that there’s much I can do about it. Things might have been quite a bit different if the ATT lawsuit didn’t happen…
http://www.softpanorama.org/People/Torvalds/Finland_period/att_lawsuit_as_a_launcher_for_linux.shtml
Ah, the butterfly effect. Trivial events can have such major consequences. Like this world-wide pandemic now that a couple trivial events are responsible for.
“sometimes I think we would have been better off if BSDs had taken over instead of linux”
Maybe yes, maybe not.
JimRaynor,
That’s the most non-committal answer ever, haha
Also, note the qualifier I used: “sometimes”.
Obviously linux has done a lot for open source, but arguably other projects have more disciplined engineers. And a separate criticism is that in its quest to become a unix clone, linux missed opportunities to make some things better (like plan 9 did). It does make me wonder if linux could have succeeded if linus tried to make it more than a clone since being a clone was linux’s main claim to fame.
my favorite feature was the usage of openrc as init system, that I wish that freebsd adopts someday