Will systemd make FreeBSD more popular?

[scottro]

With respect to both parties, I think that there might be some misunderstanding here. We'd have to define behind the puck, but I took it as meaning less well-known and less supported. drhowarddrfine is seeing the issues that can cause. Were it more popular, there would be less resistance.

Somewhere, there must be a happy medium (not that I know it, nor have any idea how to get there) where it's popular enough to be well supported but not so popular that it becomes the target of those who would add all the things that would stop you from doing dumb things but also get in the way of stopping the smart things.

 

[BSD-Kitsune]

It's attitudes like this that will keep FreeBSD way behind the puck as markets fundamentally change. If you want more people to use our code or to even stay relevant, we to have to learn to adapt to current trends. I'm all for adopting something like launchd as long as it doesn't change FreeBSDs overall structure and design ethos. Think mobile, for example. Something like launchd could greatly improve projects like PC-BSD where targeting the mobile space (tablets/laptops) are more of a challenge. If you've watched Jordans talk, he states more Unix machines are running on batteries than any other form of power, which is true.

Just my two cents.

I against OS X and its usage. I'd rather use Linux or Windows over OS X - the OS is laid out very poorly, and I do not appreciate its engineering or design. An init system has nothing to do with power management - you should leave that up to a separate subsystem.

I'm not saying that rc should or should not stay, but if you're going to change it out for something different, then it should ideally be something like Runit, or OpenRC. They're under more BSD compatible licences vs launchd, they're simpler and they already run on BSD with little to no modification. As I said, I'll fight tooth and nail if Jordan Hubbard continues to push this agenda, and if he does, I'll fork the last non-launchd branch of FreeBSD and start working on making changes that I've already wanted to see in FreeBSD for a long time.

 

[Beastie7]

Meh, I enjoy the user experience and ecosystem, and it's still Unix. It's well designed enough for it's purpose and place in the market as a solid Unix Workstation.

 

[Oko]

I against OS X and its usage. I'd rather use Linux or Windows over OS X - the OS is laid out very poorly, and I do not appreciate its engineering or design. An init system has nothing to do with power management - you should leave that up to a separate subsystem.

In generally I like your posts and even better I like your enthusiasm but I respectfully disagree with your assessment of OS X. For the record my desktops run OpenBSD so I have no horse in that race. Due to the nature of my job I had to play with OS X and these are observation of a UNIX geek.

GOOD:

1. OS X in its heart is UNIX and opening shell reveals that.
2. With a little effort it can be configured to feel and have similar functionality as for example OpenBSD (of course not nearly as secure and simple as OpenBSD)
   
3. Includes more modern version of PF than FreeBSD for example
4. DTrace in the base
   
5. MAC ports are solid and having all that proprietary software is nice

BAD :

1. Lack of centralized rc.conf file is pissing me off
2. PList should be banned by law. The only format which kernel should be able to parse is plain text files of course
3. I am not a big fun of Apple proprietary file system. Too bad Oracle bought SUN Microsystem because OS X would include ZFS (third party OpenZFS doesn't count sorry people). However I would not be surprised to see HAMMER2 (if/when finished) ported to MAC before FreeBSD
4. I kind prefer traditional FreeBSD userland over Linux-ism (bash is default shell of OS X) but in whole honestly FreeBSD was moving in that direction (userland development was neglected for couple of years and FreeBSD was full of ugly GNU userland software) itself for a while (thinks are better in last year or two) so I understand why Apple moved to that direction

UGLY:

1. One has to fight GUI all the time
   
2. Network drivers are very flaky comparing to OpenBSD
3. One has to go through Apple store/itunes (whatever it is called) and register to download Xcode Tools (you must be fucking kidding me)
4. Turn on that PF first thing and write the most restrictive pf.conf file you can because otherwise Apple controls your machine.
5. Most OS X users are click clack moo GUI users and have the same or lower computer literacy as Windows users. However they have an additional thing working against them. They are snobs and buying into a life style. Most of them are plain clueless when it comes to technology.

As I said, I'll fight tooth and nail if Jordan Hubbard continues to push this agenda, and if he does, I'll fork the last non-launchd branch of FreeBSD and start working on making changes that I've already wanted to see in FreeBSD for a long time.

I don't think that Jordan Hubbard has that kind of influence over FreeBSD. People like Hiroki Sato (who is current core team member) have done and are currently doing far more damage to the project. Jordan Hubbard has a corporate mindset and he/ixSystems have done great job with Free/TrueNAS as well as PCBSD/TrueOS. People like you and me might prefer vanilla FreeBSD for our own reasons but those are good products for typical American consumers.

 

[BSD-Kitsune]

Meh, I enjoy the user experience and ecosystem, and it's still Unix. It's well designed enough for it's purpose and place in the market as a solid Unix Workstation.

Not by my definition, but okay. I'm not going to argue with you on that.

In generally I like your posts and even better I like your enthusiasm but I respectfully disagree with your assessment of OS X. For the record my desktops run OpenBSD so I have no horse in that race. Due to the nature of my job I had to play with OS X and these are observation of a UNIX geek.

GOOD:

1. OS X in its heart is UNIX and opening shell reveals that.
2. With a little effort it can be configured to feel and have similar functionality as for example OpenBSD (of course not nearly as secure and simple as OpenBSD)
   
3. Includes more modern version of PF than FreeBSD for example
4. DTrace in the base
   
5. MAC ports are solid and having all that proprietary software is nice

BAD :

1. Lack of centralized rc.conf file is pissing me off
2. PList should be banned by law. The only format which kernel should be able to parse is plain text files of course
3. I am not a big fun of Apple proprietary file system. Too bad Oracle bought SUN Microsystem because OS X would include ZFS (third party OpenZFS doesn't count sorry people). However I would not be surprised to see HAMMER2 (if/when finished) ported to MAC before FreeBSD
4. I kind prefer traditional FreeBSD userland over Linux-ism (bash is default shell of OS X) but in whole honestly FreeBSD was moving in that direction (userland development was neglected for couple of years and FreeBSD was full of ugly GNU userland software) itself for a while (thinks are better in last year or two) so I understand why Apple moved to that direction

UGLY:

1. One has to fight GUI all the time
   
2. Network drivers are very flaky comparing to OpenBSD
3. One has to go through Apple store/itunes (whatever it is called) and register to download Xcode Tools (you must be fucking kidding me)
4. Turn on that PF first thing and write the most restrictive pf.conf file you can because otherwise Apple controls your machine.
5. Most OS X users are click clack moo GUI users and have the same or lower computer literacy as Windows users. However they have an additional thing working against them. They are snobs and buying into a life style. Most of them are plain clueless when it comes to technology.

I was an OS X user for 7 years, I left shortly after Yosemite made its debut, because I had a love-hate relationship with Linux, and because I at the time despised Windows heavily. I left for a variety of reasons, one of which was the progressively declining build quality of their systems. In the PowerPC days, the price made sense and OS X was an okay OS, also I was significantly less computer literate. The nail was that my Powerbook G4 of 3 years (used) met its end, and I had to have new computer for college, so I blew $2500 on a shiny new rMBP. and I had so many issues with it:

The SSD had to to be replaced no less than 4 times, twice by Apple, once by an Apple store, and once by myself. When I cracked that case open, I was mortified because they glued the damn battery to the case - what were they thinking!? I was also increasingly fed up with the instability of newer versions, the brittleness of the plastic keys, the piss poor performance of code I wrote, as well as so much rigmarole with the libraries, plist files, App-Store bullshit, outdated libraries and hell knows what else. I replaced the SSD a final time, and sold it for $1000 in February of 2014, after roughly 16 months of use. I've been Apple-free for nearly a year, and before that my servers running Fedora and desktops running Arch were forced into that straight-jacket called systemd, I've had enough with these "relevant", "walled garden" or other bullshit consumer OSes. About the only OS that I have any respect for is Android, and that's because its very well designed compared to conventional Linux, and I love that everything but the kernel is permissively licensed, if Apache License ( I agree with DeRaadt here, the AL 2.0 is bullshit )

I don't think that Jordan Hubbard has that kind of influence over FreeBSD. People like Hiroki Sato (who is current core team member) have done and are currently doing far more damage to the project. Jordan Hubbard has a corporate mindset and he/ixSystems have done great job with Free/TrueNAS as well as PCBSD/TrueOS. People like you and me might prefer vanilla FreeBSD for our own reasons but those are good products for typical American consumers.

I use PC-BSD, and their *NAS products for various things. My main system is FreeBSD 10.1, and relatively minimal and simplistic. I love Lumina, but just like I don't want Windows in my Linux, I don't want Apple in my BSD.

As I said, if the day comes where launchd or anymore poorly designed Apple bullshit comes into FreeBSD, I'm going to fork it, like Matt Dillon did. Looks like I have at least another year, since the change will have to come at release 12 or later, which is good because my prohibition from contributing open source expires May 19th of this year. Note, I'm not ragging on clang, because clang is good, but most all other change.

I'll invite any disgruntled users of FreeBSD, any developers and such to the project as well.

 

[Deleted member 13721]

Not by my definition, but okay. I'm not going to argue with you on that.

I was an OS X user for 7 years, I left shortly after Yosemite made its debut, because I had a love-hate relationship with Linux, and because I at the time despised Windows heavily. I left for a variety of reasons, one of which was the progressively declining build quality of their systems. In the PowerPC days, the price made sense and OS X was an okay OS, also I was significantly less computer literate. The nail was that my Powerbook G4 of 3 years (used) met its end, and I had to have new computer for college, so I blew $2500 on a shiny new rMBP. and I had so many issues with it:

The SSD had to to be replaced no less than 4 times, twice by Apple, once by an Apple store, and once by myself. When I cracked that case open, I was mortified because they glued the damn battery to the case - what were they thinking!? I was also increasingly fed up with the instability of newer versions, the brittleness of the plastic keys, the piss poor performance of code I wrote, as well as so much rigmarole with the libraries, plist files, App-Store bullshit, outdated libraries and hell knows what else. I replaced the SSD a final time, and sold it for $1000 in February of 2014, after roughly 16 months of use. I've been Apple-free for nearly a year, and before that my servers running Fedora and desktops running Arch were forced into that straight-jacket called systemd, I've had enough with these "relevant", "walled garden" or other bullshit consumer OSes. About the only OS that I have any respect for is Android, and that's because its very well designed compared to conventional Linux, and I love that everything but the kernel is permissively licensed, if Apache License ( I agree with DeRaadt here, the AL 2.0 is bullshit )



I use PC-BSD, and their *NAS products for various things. My main system is FreeBSD 10.1, and relatively minimal and simplistic. I love Lumina, but just like I don't want Windows in my Linux, I don't want Apple in my BSD.

As I said, if the day comes where launchd or anymore poorly designed Apple bullshit comes into FreeBSD, I'm going to fork it, like Matt Dillon did. Looks like I have at least another year, since the change will have to come at release 12 or later, which is good because my prohibition from contributing open source expires May 19th of this year. Note, I'm not ragging on clang, because clang is good, but most all other change.

I'll invite any disgruntled users of FreeBSD, any developers and such to the project as well.

I will definitely climb aboard as a user and a potential donor if things come to that, I really hope it doesn't come to that though.

 

[BSD-Kitsune]

I appreciate the support. Let's hope it doesn't. I'm a capable developer, but I'm nowhere near capable enough to tackle a project like that by myself. Communities are weakened by fragmentation - for instance the NetBSD-OpenBSD schism has left NetBSD without much steam, especially after the collapse of Wasabi Systems.

 

[Oko]

Not by my definition, but okay. I'm not going to argue with you on that.

I was an OS X user for 7 years, I left shortly after Yosemite made its debut, because I had a love-hate relationship with Linux, and because I at the time despised Windows heavily. I left for a variety of reasons, one of which was the progressively declining build quality of their systems. In the PowerPC days, the price made sense and OS X was an okay OS, also I was significantly less computer literate. The nail was that my Powerbook G4 of 3 years (used) met its end, and I had to have new computer for college, so I blew $2500 on a shiny new rMBP. and I had so many issues with it:

The SSD had to to be replaced no less than 4 times, twice by Apple, once by an Apple store, and once by myself. When I cracked that case open, I was mortified because they glued the damn battery to the case - what were they thinking!? I was also increasingly fed up with the instability of newer versions, the brittleness of the plastic keys, the piss poor performance of code I wrote, as well as so much rigmarole with the libraries, plist files, App-Store bullshit, outdated libraries and hell knows what else. I replaced the SSD a final time, and sold it for $1000 in February of 2014, after roughly 16 months of use. I've been Apple-free for nearly a year, and before that my servers running Fedora and desktops running Arch were forced into that straight-jacket called systemd, I've had enough with these "relevant", "walled garden" or other bullshit consumer OSes. About the only OS that I have any respect for is Android, and that's because its very well designed compared to conventional Linux, and I love that everything but the kernel is permissively licensed, if Apache License ( I agree with DeRaadt here, the AL 2.0 is bullshit )

It looks like most of your distaste with Apple comes from previous problems with hardware not with OS X. I thought we were discussing OS X not hardware. Yeah I agree with you. Since Apple switched from PPC to Intel their hardware is peace of shit although other products are getting significantly lower quality (my beloved ThinkPads for example). For the record I always preferred Sparc64, Mips, and in particularly Alpha to PPC. I never had money for the latter two. However I still run Sparc64 eagerly awaiting for ARM to mature.

I use PC-BSD, and their *NAS products for various things. My main system is FreeBSD 10.1, and relatively minimal and simplistic. I love Lumina, but just like I don't want Windows in my Linux, I don't want Apple in my BSD.

As I said, if the day comes where launchd or anymore poorly designed Apple bullshit comes into FreeBSD, I'm going to fork it, like Matt Dillon did. Looks like I have at least another year, since the change will have to come at release 12 or later, which is good because my prohibition from contributing open source expires May 19th of this year. Note, I'm not ragging on clang, because clang is good, but most all other change.

I'll invite any disgruntled users of FreeBSD, any developers and such to the project as well.

Why not join DragonFly BSD? The project is super cool and they are craving for more developers. 2 dozen or more capable developers to DragonFly and DF to FreeBSD will look more like OpenBSD to NetBSD.

 

[BSD-Kitsune]

It looks like most of your distaste with Apple comes from previous problems with hardware not with OS X. I thought we were discussing OS X not hardware. Yeah I agree with you. Since Apple switched to Intel their hardware is peace of shit although other products are getting significantly lower quality (my beloved ThinkPads for example).

My thinkpad T510 is a hella better than my rMBP ever was, even if it was a step down. Also, since OS X is locked to Apple hardware, talking about OS X without talking about the hardware would be like trying to talk about Buddhism with no mention of Siddhartha Gautama. I have plenty of major issues with OS X in my post, but those things I don't care to elaborate on because I've forgotten most of the details in any case.

Why not join DragonFly BSD? The project is super cool and they are craving for more developers. 2 dozen or more capable developers to DragonFly and DF to FreeBSD will look more like OpenBSD to NetBSD.

I've considered it, and its still on my mind, but again, I can't join any FOSS project officially till after my legal agreement sunsets. And the only way I would join DFBSD is if Matt and the team allowed me to work on porting it to MIPS and ARM mostly, although I would consider taking on Nouveau as I'm a dedicated Nvidia user. I'm also not keen on some design choices by the DFBSD team, namely, it seems historically Matt is opposed to any work on other architectures, he's willing to making DFBSD more Linux-like, which I'm worried may pollute it with GPL code ( It's bad enough that he's not using clang ) and I'm also likely to piss other people on the team off - I do best when I'm in charge of a project. At the very least though, if I lead the charge in forking FreeBSD, I will probably ditch ZFS for HAMMER2 and improve UFS2.

 

[Crivens]

Okay, some strong emotions are voiced here. Could we, maybe, get back to finding a solution for the original question? I quite appreciate the strong feelings, but this is one of the points that get such threads closed. And then the next one gets closed sooner, and no solution can be worked out at all. I would not like that, because I would like a solution for the problem at hand. Lets discuss all the options, we can discuss if we want OpenRC or launchd, but that will not stop all of upstream to force systemd down our throat.

As I once explained to a project lead, the result of multiple changes in a design without syncronizing the effords is like shooting a rabbit, while you are kneeling in a boat on the lake. Hasty attempts to fix that mess then turn out to change the boat into a big rubber duck, giving huge buckets of coffee to the rabbit, and dancing the weather dance to get some hurricane, also. In the end someone makes a hole into the duck.

On a side note, said project lead would not listen and said project is now filed in Davy Jones Locker. I would like to spare FreeBSD from that.

Changing FreeBSD to another init system will not change a bit about the problem we have, we need to solve that at the root. And fragmenting FreeBSD more seems, I would say, counterproductive to the situation. I'm more interested in getting Matt and Theo into the same boat.

 

[BSD-Kitsune]

Lets discuss all the options, we can discuss if we want OpenRC or launchd, but that will not stop all of upstream to force systemd down our throat.

Changing FreeBSD to another init system will not change a bit about the problem we have, we need to solve that at the root. And fragmenting FreeBSD more seems, I would say, counterproductive to the situation. I'm more interested in getting Matt and Theo into the same boat.

I am in agreement with this post. I've rested my case with regards to init.

On the subject of systemd dependencies, I think its good the systembsd project and the Devuan project are working on offering some alternatives to systemd services. But that's not the best answer to the issue. We should instead pull together and work on becoming independent. Now's the best time to become more independent - we're getting a lot of new Linux refugees, we are reasonably up to date on a lot of things, and with the GPLv3 becoming evermore prevalent, we have a lot of incentive for leveraging away from Linux.

However, I am for forgoing porting efforts on GNOME, KDE and XFCE, as well any other desktop environment or component reliant on systemd-logind. I know that's likely to be an unpopular opinion, but hear me out. We now have Lumina, and CDE has just been open sourced in the last few years. In both cases, no dependency on systemd-logind or Consolekit. Now while neither of them offer the range of options that GNOME, KDE and XFCE, if we can somehow convince the people who maintain those projects to instead either roll another new, BSD-compatible DE or fork the current options and excise the Linux specific code from them. If we can get some of the people focusing on the desktop environments from Linux on BSD specific ones, we can remove a lot of the systemd dependency pressure.

As far as other software goes, our best options are to either maintain the systemd shims, and hold out for a systemd replacement on Linux's end, or we should start contributing patches to the projects we care about so they can work on this.

I understand this is unlikely to happen, but I had to throw it out there. The shims only can only carry us so far, and it looks unlikely that systemd will be replaced, unless RedHat somehow collapses. It has too much power and control of the Linux ecosystem.

 

[drhowarddrfine]

The first question I would ask, as to how involved FreeBSD needs to get involved in systemd, is, "Why?". Unless there is a compelling need to do so, there is no need to follow Linux or be led around by the nose.

Does FreeBSD no longer wish to be a Unix-like system? (Linux is no longer a Unix-like system and is Linux only unto itself.)

 

[wblock@]

Some interesting trains of thought in this thread. As far as the subject, yes, systemd has clearly generated interest in FreeBSD. I would suggest that concern over launchd or other rework of the FreeBSD init system is probably a bit misplaced. For one thing, it's a bit early. Work on porting launchd (as an alternate, not a replacement) has been going on intermittently for years, and other working init systems, usually offering concurrent startup, have been available in the meantime. Secondly, the idea that FreeBSD developers would work in the same way or produce the same results as systemd is pretty skeptical. The FreeBSD ecosystem tends to reject under-engineered stuff. Have some faith that if or when new systems are offered, they will be up to the level expected by FreeBSD users.

 

[BSDBernd]

When it comes to init systems like launchd, I think that I do no harm to the conversation to mention the puredarwin project here:
http://www.puredarwin.org/
This is the open source core of Mac OSX. It is an independent project. These people have nothing, no hardware donations, no donations in general, I once talked to them in their IRC channel. They need support . It would be cool if their OS would grow also, this would then presumably a more BSD-like OS for mobile devices and Laptops etc.
Edit: Although, when I think about it, launchd came with Mac OSX 10.4. Hopefully they have launchd in their OS, they are not so up to date ...

 

[walterbyrd]

This is not a popularity contest

I cannot help but notice that more popular OSes have better hardware, and software, support. It is just a fact.

 

[walterbyrd]

The first question I would ask, as to how involved FreeBSD needs to get involved in systemd, is, "Why?". Unless there is a compelling need to do so, there is no need to follow Linux or be led around by the nose.

Does FreeBSD no longer wish to be a Unix-like system? (Linux is no longer a Unix-like system and is Linux only unto itself.)

Linux software is being made to be dependent on systemd. Could that affect porting to FreeBSD? If so, is that a compelling reason for FreeBSD to be concerned with systemd?

 

[BSD-Kitsune]

The first question I would ask, as to how involved FreeBSD needs to get involved in systemd, is, "Why?". Unless there is a compelling need to do so, there is no need to follow Linux or be led around by the nose.

Does FreeBSD no longer wish to be a Unix-like system? (Linux is no longer a Unix-like system and is Linux only unto itself.)

That's what I meant by the above post, moreorless. Most of Linux went off the cliff like lemmings, but we're not lemmings - we don't have to follow them.

Some interesting trains of thought in this thread. As far as the subject, yes, systemd has clearly generated interest in FreeBSD. I would suggest that concern over launchd or other rework of the FreeBSD init system is probably a bit misplaced. For one thing, it's a bit early. Work on porting launchd (as an alternate, not a replacement) has been going on intermittently for years, and other working init systems, usually offering concurrent startup, have been available in the meantime. Secondly, the idea that FreeBSD developers would work in the same way or produce the same results as systemd is pretty skeptical. The FreeBSD ecosystem tends to reject under-engineered stuff. Have some faith that if or when new systems are offered, they will be up to the level expected by FreeBSD users.

I will admit, I'm less faithful in the FreeBSD project then when I was when I first came here. I'm a little dismayed that many users do not use it as a desktop OS and instead use other OSes, they reinforce the stereotype that FreeBSD is no good on the desktop.

Secondly, while I don't doubt any init system replacement will solve any limitations of rc, I believe in the UNIX philosophy. Launchd violates this principle, in my opinion.

I refuse to associate with Apple, and thus I will not contribute to or support Darwin. I see them as worse than Microsoft these days, tying the OS to subpar, overpriced hardware, and with a sellout philosophy making the OS useless for power users.

I'm fine with Android as a mobile platform, so I'm not interested in FreeBSD going mobile, or becoming more popular. From what I know FreeBSD has a higher userbase of Japanese speakers than does Linux, at least this was 5 or so years ago when I read this.

 

[scottro]

I would doubt that, but my evidence is purely anecdotal. Most Japanese speakers I know use Windows or Mac. There is a Tokyo Linux Users group (which also has many FreeBSD users), but actually the majority of its members are gaijin. I haven't been to Japan since the early double oughts, either.

However, I repeat my evidence is anecdotal, and most of the Japanese folks I know aren't involved with tech. (Debates a take my wife....please joke and decides against it.)

 

[BSD-Kitsune]

The reason I say that is that, according to my contact in Yokohama, the BSD community was the first to really get decent Japanese support going in the OS, and much of Japan lives in the stone age of computers, so many corporations into the 2000s used the original BSD on a lot of VAXen and such, so its only natural that FreeBSD would have a large Japanese community. Most of the Linux users in Japan I've met are indeed foreigners. Hell, I know a guy who has a PC-9821 and runs an old FreeBSD version on one HDD, and the NEC DOS on another.

 

[scottro]

Japanese support started getting easier around the early 2000's, in both Linux and FreeBSD. I think the first system where I got Japanese input working was RedHat 5 or 6 or so, and around that time, FreeBSD had kinput2 and canna as well. Anyway, we're digressing, and apologies to others on the thread. I'm sure your friend is right--by the time I got to FreeBSD, it was already pretty easy to get Japanese working, and that would have been the early 2000's.

 

[Oko]

Linux software is being made to be dependent on systemd. Could that affect porting to FreeBSD? If so, is that a compelling reason for FreeBSD to concerned with systemd?

Yes check Google Summer Code project on OpenBSD website (I didn't make a mistake it is OpenBSD). They are working on the simple code which would solve that problem.

 

[roddierod]

(Debates a take my wife....please joke and decides against it.)

I'm guessing less than a handful of people on this board get that one...

 

[BSDBernd]

One point by Hubbard is that these days you have these cheap ARM-Chip devices, many many of them. It is like with digital photos, they are cheap, lots of Teens do hundreds and hundreds of them. Because digital images are so important these days, a whole branch of computer science and mathematics grew immense because of that fact: image processing. This is a remarkable subject. An image is nothing else than a function, and with functions you can do a lot to make your image look smoother, reduce the noise, make colors more intense, make borders look more sharp, etc.. I would like to see FreeBSD (or something that uses FreeBSD at its base) also getting good at running on these devices because this is where a big interest is. Why not have several init versions in the ports tree, where creative people can then set up FreeBSD for their needs?

 

[BSD-Kitsune]

One point by Hubbard is that these days you have these cheap ARM-Chip devices, many many of them. It is like with digital photos, they are cheap, lots of Teens do hundreds and hundreds of them. Because digital images are so important these days, a whole branch of computer science and mathematics grew immense because of that fact: image processing. This is a remarkable subject. An image is nothing else than a function, and with functions you can do a lot to make your image look smoother, reduce the noise, make colors more intense, make borders look more sharp, etc.. I would like to see FreeBSD (or something that uses FreeBSD at its base) also getting good at running on these devices because this is where a big interest is. Why not have several init versions in the ports tree, where creative people can then set up FreeBSD for their needs?

I don't have a problem with launchd being an optional init system, but I do not want a single port to rely on it. Not one. We don't need our own systemd. Why is launchd necessary for ARM devices? You don't - that's the key. Power management doesn't need to be handled by init. My battery on my laptop lasts longer than Linux or Windows 7 ever did, 3.5 hours to be exact. Again you're not answering the question: Why launchd, or why Darwin?

 

[scottro]

I'm guessing less than a handful of people on this board get that one...

But the ones who do will laugh.