Will systemd make FreeBSD more popular?

Status
Not open for further replies.
I have seen lots of posts from Linux users who say they are moving to FreeBSD, or have already moved.

I have been using Linux myself for over ten years, systemd is the reason I am trying out FreeBSD.

Poettering called FreeBSD an irrelevant toy. It would be awesome if there was a surge in popularity for FreeBSD.
 
I have seen lots of posts from Linux users who say they are moving to FreeBSD, or have already moved.
Well, they are welcome. I just hope they do not expect to come here asking why FreeBSD is not like Linux, like many Windows users do when they switch to Linux and then asks why the desktop is not as good as in Windows.

Poettering called FreeBSD an irrelevant toy
The feeling is reciprocal. Moreover, considering the fact that every time he create something new, users starts looking around in search of an OS without that new feature, I personally couldn't care less of his opinion.

I would be awesome if there was a surge in popularity for FreeBSD.
I disagree. I've been a Debian user for about 10-12 years before definitely switching to FreeBSD. In the beginning (I'm talking about the Debian 2.0 era) using Linux was a pleasure. Users constantly asked for more popularity, so they can have more drivers and more software. Then came Ubuntu, and also popularity. And chaos. And the OS became an ever-changing entity.
 
And the OS became an ever-changing entity.
Although I am by far no expert of operating systems, let alone the Unix like ones, I guess that there can be good and bad changes. FreeBSD grows and changes also, gets lots of new features, very very interesting ones, like f.e. bhyve.
When it comes to systemd, one of the founders of FreeBSD, Jordan Hubbard, gave a talk where he explained why they needed launchd when creating OSX 10.4 (I think it was 10.4). If you have to worry about power consumption, when you want to make things more 'lego like', when your machine gets switched on and off lots of times and lots of devices can be plugged in and dragged out and what ever, in short when you have f.e. a mobile device or a laptop, you need another way to handle things, as I understood Hubbard.
 
Maybe among hobbyists. I would be seriously surprised to see any business moving to FreeBSD just because they dislike systemd.

I work for pure UNIX shop and we use OpenBSD, FreeBSD, and Red Hat (I have played a lot with DragonFlyBSD as well so I can say something about that). We have to run Red Hat because MATLAB and bunch of our proprietary software doesn't run on any BSDs (I know everything about MATLAB alternatives but that is not how real world works). We tested Red Hat 7.0. We dislike it a lot and already made decision to stick with branch 6 till the end of its life which is 2020. I think this is more or less what is going to happen with most labs and businesses.

I will tell you another example from my Lab where we had even more compelling technical reason to chose FreeBSD over Linux and it has not happened. We currently run bunch of our proprietary software out of KVM instances running Red Hat on host and guests. Guests must have access to common storage pool so I set up soft RAID 6 on the guest OS (XFS) and tried to mount to guest using pass-through 9p distributed file system. 9p comes from Plan 9 and works like charm on its native platform. It turns out that 9p like many things on Linux is just a vaporware so I was forced to use NFS. We are hitting some serious bottle necks with that NFS now. Being BSD guy I suggested a radical but very simple solution. We would use FreeBSD as host OS, common storage pool will run ZFS, guest OSs will run in Jails
and share common data storage via nullfs(5). All technologies I suggested are very mature and work like charm. However some effort was required to clean mostly our proprietary code to run on FreeBSD (should not be a big deal). My idea was more or less immediately rejected by developers (read Linux users) and management. Developers are Linux users and are afraid of FreeBSD while management doesn't want to be stuck with obscure platform on our code when everyone else is using Linux. Instead I started experimenting with ZFS data storage pool on Linux and Dockers. I got things to work as advertised and then just demonstrated how a person with the access to a Docker instance which has common data pool mounted can use it to gain the root access to host machine. I also show that Dockers don't work well off ZFS. The crew is reconsidering FreeBSD. Stay tuned.
 
Last edited:
More popular? Maybe. That's a big "maybe." It is of course worth keeping in mind that the kind of person who spends time and energy bitching and moaning about systemd is not the kind of person who's going to contribute something to the community, or even express gratitude. They'll remain silent and contribute nothing, until the people doing the work decide to change something in a way they don't like. They'll then either hop ship to the closest thing resembling what they're used to (perhaps FreeBSD), or they'll suck it up and get used to the changes, eventually forgetting what they were ever complaining about. In the meantime, new users and developers enter the community at large totally unaware any controversy exists, and their only experience with the Linux ecosystem is with the stuff everyone previously thought controversial. They grow accustomed to it and then, one day, someone actually maintaining the system they love so much decides to make another change, and the bitch cycle continues. This pattern has been repeating itself in the Linux world for decades now.

I switched to FreeBSD as my primary OS after seven years of using Linux, and did so for several reasons. systemd was not one of them.
 
Most of my clients are now running FreeBSD on their servers rather than Linux and part of that (though certainly not all) was due to changes like systemd. These are mostly small businesses, not big enterprises, but it still shows some people are migrating toward FreeBSD. Another big drawing point was ZFS and boot environments. While Linux has Btrfs, it is not fully supported in most distributions yet and ZFS is stable on FreeBSD these days.

I guess what I'm saying is there might not be a huge surge toward FreeBSD, but there is a trickle of people/businesses migrating to FreeBSD.

And despite what ANOKNUSA suggested, some of the people complaining about systemd are contributors, in one way or another. Some are looking at helping with port maintaining, most of them are planning to donate to the FreeBSD Foundation. More users means more funding and more bug reports.
 
I know the other thread was brought up (and somewhat beaten to death), but I didn't throw my hat in the ring there as someone who is relatively inexperienced with the BSDs (which was the original focus of that thread, though it veered off topic). Certainly I'm not qualified to criticize the technical direction of the project, I would say that. I have some thoughts as a recent convert to FreeBSD on all of my devices (laptop/desktop) if anyone is interested, though.

I'd been using Slackware (as well as some other distros, on and off) since the late 90's. Systemd didn't have anything (at all) to do with why I switched, considering there is guaranteed to be a supported Slackware without systemd until - at least, knowing Pat maybe longer - 2020. Let me break down my reasons for switching:

1. Linux is becoming more and more bloated as time goes by, in spite of dropping support for old hardware that I still use (there are distros that use older kernels to counteract this, but I wouldn't connect one to the Internet). A real use case example of bloat: I can run several emulators (emulators/dolphin-emu, emulators/pcsxr) with no slowdown on my older Core 2 Quad PC with a GTX 550 Ti and x11/nvidia-driver on FreeBSD. I can't do this on Linux any longer. I have tried minimalist Linux distros, but can't achieve that with a modern one. FreeBSD is much more efficient these days, just in my experience. Granted, that's not my main desktop, but it isn't obsolete hardware either. Also, FreeBSD still runs on my obsolete hardware (PII/PIII era) while a newer Linux kernel doesn't.
2. Btrfs is just not ready as a next generation file system, but ZFS on Linux lags far behind ZFS on FreeBSD. I swore by LVM with XFS for many years, but the advantages of next gen file systems have made it less than ideal for me (especially in preventing bit rot) these days.
3. FreeBSD hardware support is very good now. I get better performance with my desktop NVIDIA cards, plus my sound cards and onboard Intel all work with my recent (say 2010 or newer) motherboard and sound card purchases. Sound support used to be one of the things that stopped me from switching, this isn't the case anymore, and it hasn't been for awhile.
4. FreeBSD still uses OSS, which I was very annoyed to lose in Slackware.
5. FreeBSD still has (in my opinion) a more complete and properly supported ports system than any modern Linux distro (Gentoo, Void, etc.)
6. FreeBSD has clear leadership and direction, you don't see statements like this - http://ostatic.com/blog/linus-systemd-indifference-pclos-review-and-rebecca#buzz
I am personally hesitant to donate to and support (via forums, bug reports, etc.) projects when I don't see a clear road map for the future. As much as I think Pat Volkerding knows what should happen with Linux, he is just the BDFL of Slackware and can't control decisions made by Linus and Red Hat.

Also, one other thought I would put out there: I don't see why people are so apt to assume FreeBSD will make the same mistakes that were made with systemd when/if a new init system is implemented. To give an example, DEVD has been introduced in place of HAL and is excellent (rather than being another HAL); after watching Jordan's presentation on YouTube, I feel like he is being misrepresented a bit. If he had said "We're bringing the greatness of systemd to FreeBSD!", then I'd be worried, but he didn't say anything like that at all. As I understood it, the hope is to please the embedded device makers (and supporters of FreeBSD) taking advantage of lessons learned from launchd and still adhering to the UNIX philosophy (with the Lego analogy being my source for that).

In short, I think FreeBSD is becoming more popular, but systemd has little to do with it.
 
And despite what ANOKNUSA suggested, some of the people complaining about systemd are contributors, in one way or another. Some are looking at helping with port maintaining, most of them are planning to donate to the FreeBSD Foundation. More users means more funding and more bug reports.

I neither doubt nor deny this. There's just no way the number of people actively contributing to any projects (be they alternatives to systemd or their distro/OS of choice) in the Linux world has even approached the number of people content to spend all their time and energy simply complaining about systemd. There's just no other conclusion to draw from the thousands upon thousands of rants that have been made about it, relative to the tiny number of people actively working to create alternatives in the couple years since the push toward systemd began. Most people who "hate" systemd will do nothing more than hop from one distro/OS to another until they're left with no other options, at which point they'll likely suck it up and acclimate to it. They want things the easy way. They want someone else to fulfill their expectations. That's unlikely to change for most people who wind up choosing FreeBSD for no reason other than that it doesn't have systemd.
 
Maybe among hobbyists. I would be seriously surprised to see any business moving to FreeBSD just because they dislike systemd.
For what it's worth, someone on reddit, a few days ago, posted that their company runs 60 servers and were going to switch them all to FreeBSD cause systemd was the final blow among other reasons.
 
For what it's worth, someone on reddit, a few days ago, posted that their company runs 60 servers and were going to switch them all to FreeBSD cause systemd was the final blow among other reasons.
The key ward here is final. In my book that means that they had many other reasons and systemd was just the last drop. I hope your read the rest of my original post. I am just doubtful that systemd alone will force anybody in the next couple of years to move away from Linux. 4-5 years from now is anybody's guess.

IMHO here in U.S. the main reason for lack of adoption of OS like FreeBSD is the lack of large legal entity which can be sued if thing and which provides support for OS. I am afraid that iXsystems is just not big enough and that would take a player like IBM for more people to adopt FreeBSD. On another hand individual technologies are different story. For example, in some sense every car in U.S. runs OpenBSD by way of QNX. OpenSSH is also in your car for the same reason. And those two things are just a tip of the iceberg when it comes to technologies which came our of OpenBSD. On another hand you have even people on this forum presumable FreeBSD users who think OpenBSD is irrelevant.
 
Maybe among hobbyists. I would be seriously surprised to see any business moving to FreeBSD just because they dislike systemd.

I think it might depend on which business. I have been corresponding with a few business owners who have switched from Linux, to FreeBSD. One has a small web-hosting business, the other a small software development business.

In both of those cases, the reasons for moving where not entirely related to systemd. They just thought FreeBSD was a better fit.
 
I would be awesome if there was a surge in popularity for FreeBSD.

No, it would not. FreeBSD (or any BSD for that matter) is like this beautiful, quiet beach you've discovered. You don't want this beach to become a fscking tourist attraction.

This is exactly what happened to Linux (as pointed out by someone more than 10 years ago):
In the beginning, Linux was pretty cool. It was free (always a plus), had a rapid development cycle, a moderately knowledgeable user base, and a funny mascot. Then the pinheads arrived.
 
No, it would not. FreeBSD (or any BSD for that matter) is like this beautiful, quiet beach you've discovered. You don't want this beach to become a fscking tourist attraction.

This is exactly what happened to Linux (as pointed out by someone more than 10 years ago):

Truer words have never been said. I really hope this beach does stay the idyllic place it has been for the last 4-5 years(for me), or I will be moving to a new one. That would be a sad day.
 
This conversation reminds me of this scene.
While I get what you mean, there really is a problem of perception in the tech media. There are so many clickbait articles about systemd that attribute people switching to FreeBSD and BSDs in general (or considering doing so) due to an init system that a lot of them don't even understand in the first place (i.e. painting people who switch in a buffoonish light), when there are many valid reasons that users are switching because of. The story always seems to be about what Linux is doing wrong (the equivalent of saying "Well gee, I guess you're stuck with a BSD then.") when there are so many positive things about FreeBSD that should get mentioned. I just wanted to rant a bit, I realize negative articles get page views and writing an article like "The 10 features that will make you want to use FreeBSD" won't generate the clicks of "SYSTEMD ARMAGEDDON APPROACHING! WHERE WILL THE SURVIVORS TURN?".
 
The key ward here is final. In my book that means that they had many other reasons and systemd was just the last drop. I hope your read the rest of my original post. I am just doubtful that systemd alone will force anybody in the next couple of years to move away from Linux. 4-5 years from now is anybody's guess.

IMHO here in U.S. the main reason for lack of adoption of OS like FreeBSD is the lack of large legal entity which can be sued if thing and which provides support for OS. I am afraid that iXsystems is just not big enough and that would take a player like IBM for more people to adopt FreeBSD. On another hand individual technologies are different story. For example, in some sense every car in U.S. runs OpenBSD by way of QNX. OpenSSH is also in your car for the same reason. And those two things are just a tip of iceberg when it comes to technologies which came our of OpenBSD. On another hand you have even people on this forum presumable FreeBSD users who think OpenBSD is irrelevant.

This is something I would like to see, a more or less commercialized version FreeBSD with commercial support. As solid as FreeBSD is as a platform is, "who are we going to call if something breaks" is a common concern of IT managers, and should definitely be addressed should companies decide to use FreeBSD. I'm not sure what iXsystems is doing with TrueOS but they don't seem to marketing it much as a viable alternative platform to Linux/Windows. It doesn't take much to advertise FreeBSDs' administrative and/or development features to steer people this way.

I yearn for a commercialized version of RHEL from iXsystems, and yes, with a GUI also. It'll attract a wider audience. Even though I think GUIs on a server are kind of a sin. :)

The key really is in advertising from a commercial entity, whether big or small. Just look at guys from Joyent with SmartOS.
 
IMHO here in U.S. the main reason for lack of adoption of OS like FreeBSD is the lack of large legal entity which can be sued if thing and which provides support for OS.
I hear that reason used often but never in practice. At least the "suing" part. No one choose Windows so they can sue Microsoft should it ever fail. Support for IIS? Sure but 75% of the web runs on non-Microsoft products so the evidence isn't there for the majority. For Fortune 500 companies? Maybe that is true.
 
I hear that reason used often but never in practice. At least the "suing" part. No one choose Windows so they can sue Microsoft should it ever fail. Support for IIS? Sure but 75% of the web runs on non-Microsoft products so the evidence isn't there for the majority. For Fortune 500 companies? Maybe that is true.

I have done a lot of work in IT for government, government contractors, and large corporations.

In some cases, I think there may be a "you scratch my back and I will scratch yours" decision making process. Ever notice how often IT projects for governments go way over-budget, and are badly botched? Ever notice how a decently working system is replaced, at great expense, with a worse system? I think there have been cases of a top exec leaving an institution, then starting his/her own company, and doing business with his/her old buddies back at that institution.

In other cases, I think there is a huge avoidance to try anything new - because it's your ass on the if things go wrong. They used to say "nobody ever gets fired for buying IBM" then it was Microsoft, then Linux.

I also think that sometimes corporations tend to identify with, and therefore trust, other corporations. If you are not making money with your product, how do I know it will still be around six months from now? There is something fishy about software being given away for free.

But, I think most of that stuff is more relevant for huge institutions, not smaller businesses.

All JMHO, of course.
 
  • Thanks
Reactions: Oko
FreeBSD . . . is like this beautiful, quiet beach you've discovered. You don't want this beach to become a fscking tourist attraction.

I am not sure how far you can take that analogy.

Problem is, can you count on a lot of support for an OS that practically nobody uses?
 
Maybe among hobbyists. I would be seriously surprised to see any business moving to FreeBSD just because they dislike systemd.

"Just because of systemd?"

I dunno. I think systemd is a very big deal. Systemd is not just a replacement for the init system, systemd completely changes everything.

No more: init, login, PAM, getty, syslong, udev, mount, cryptsetup, cron, at, dbus, acpi, cgroups, gnome-session, autofs, tcpwrapper, audit. No more run levels, no more text logging, no more POSIX, no more UNIX philosophy.

Great gif describing this:
http://giphy.com/gifs/hungry-systemd-5xtDarAgrjoOrBxSVYk/fullscreen
 
No more: init, login, PAM, getty, syslong, udev, mount, cryptsetup, cron, at, dbus, acpi, cgroups, gnome-session, autofs, tcpwrapper, audit. No more run levels, no more text logging, no more POSIX, no more UNIX philosophy.
Well, consider this thing from another point of view: we are moving towards an OS made of a single monolithic file. No config files, no libraries, and only one thing to keep up-to-date. Not to mention the install procedure... a dream for every sysadmin! :D
 
"Just because of systemd?"

I dunno. I think systemd is a very big deal. Systemd is not just a replacement for the init system, systemd completely changes everything.
I agree but if you read the rest of my post you will see that it is not like you have to chose between systemd and FreeBSD. Many of my colleagues have decided just like me to stick with Red Hat (or clone of it) 6.XXX which will be available until 2020 and postpone the decision until dust settles. Now I don't know many people who run Ubuntu and Debian in production and I am guessing they might have to decide earlier.
 
Status
Not open for further replies.
Back
Top