What is the future of the BSD?

[0xBADCODE]

So has been Hammer2 working good recently?

Overall design seems to be not bad, but looking on their manpower I can estimate it would take forever and half to complete it (unless they'll try to do some blatant cheats like trying to mainline it to Linux kernel).

Or ZFS on BSD not crashing every moment, unlike on opensolaris?

Looking on some mail lists it still does, though not as bad as before. Also it's not a BSD devs achievement at all. It's some piece of Sun code (ironically, under copyleft license, but since "it's not GPL!!!111" , it seems to be okay, haha ). For these reasons it haves load of strange quirks like very own memory cache management, not really integrated with kernel memory management. It's okay if machine is a pure fileserver which lacks any memory hungry programs. But in all other cases you can see some obscure memory allocation failures since cache memory can't be reliably revoked by kernel. So it's crappy choice for desktop or app server, etc. Sure, you can reduce cache size ... and enjoy by slow ancient design performance while half of RAM does nothing. Instead of serving as disk cache, dammit. It's also ancient and slow design which does not even uses extents. So each and every benchmark shows us it performs quite poorly unless backed by multiple gigz of RAM to hide crappy speed of underlying structures. Sure, RAM caches are fast. But that's not ZFS disk layout achievement. Any RAM based data storage is fast. Geez. The only thing Sun did good was MARKETING. Everything else is mediocre.

Or the softupdate journaling mess?

And thats the best FBSD devs can do themselves. Some ancient and crappy design which internally can challenge something like EXT2 in levels of disk technologies used. These pathetic losers do NOT use extents as of 2014! Sure, it requires to remake ancient disk structures. Ext4 did it and it's performance predictably skyrocketed. But UFS... oh, lol. Another example of shitty project management.

[0xBADCODE]

The battle with AT&T just proves that BSD was leading the way to open source; they did not have the advantage of having a reference as to what problems they could get into.

And now we can see, Linux on other hand got a load of powerful corporations sharing their efforts to make a decent operating system. While they compete here and there, there seems to be room for collaboration between them all and community around: operating system is some kind of common base. Virtually everyone needs some OS.

Next, are you forgetting about jails (way before Linux)?

And now what we can see? VPS hosters are all about OpenVZ, also some clouds/VMs/etc based on Xen or KVM, those are also Linux things most of time. And where're BSD jails? BSDs lost on their own battlefield. How pathetic.

ZFS? DTrace? Capsicum? What are the modern equivalents in the Linux world? Btrfs? Don't make me laugh.

Btrfs internals seems to be seriously superior to ZFS. And it mostly starts to stabilise. Granted that ZFS in FBSD has been released with unimplemented/bugged sendfile(), being capable of deadlocking whole OS, which still haven't prevented BSD nuts from claiming it "stable", I would not trust much to such judgement. There was epic outcry in mail lists when BSD zealots sweared ZFS is rock solid. Haha. Nice that Linux devs do not want to cheat their users about properties of released code. When they call something stable, usually it is. Also, there is Linux port of ZFS if someone needs it. Dtrace? Again, it's here in Oracle's linux. Though Linux haves other tracing/debugging tools as well. Capsicum? Hmm, some strange hack. Maybe not really bad, but not widely used as it requires to rework program code IIRC. Also IIRC Google ported it on Linux. Though it seems to be in low demand and nobody really cares about that.

Could you explain how an infrastructure such as Netflix's Open Connect depend on a "third rate" system?

I guess these DRM-inclined fuxx do not want to share any code. What you expect from DRM-inclined fuxx? And if BSDs are not third rate crap, good luck to boot into BSD desktop and then watch video from Netflix. Ah. what? I'm supposed to use proprietary systems to do that? Then, Netflix itself considers BSDs as third rate crap. isn't it?

FreeBSD is world class, and that hurts you.

Word-class footpads who serve to most uncooperative and troublesome DRM/proprietary fuxx. It does not really hurts. But makes some disgust.

So, what is Linux's market share in the desktop?

Over 9000 times of FBSD, lol. Even most pessimistic assumption of 1% is waaaaaaay more than 0.01% FBSD can show. This is literally some few thousands of requests per whole continent per day. Rough estimation gives some several thousands of users for whole planet.

Bad news to you: BSD will never die.

It's not like if I care too much, to be honest. I would not see difference, no matter if it alive or not.

Long live the GNU Communist Party!

Long live BSDM party. And I admit corporations tend to prefer "master" role somehow.

[Sergio]

And now we can see, Linux on other hand got a load of powerful corporations sharing their efforts to make a decent operating system. While they compete here and there, there seems to be room for collaboration between them all and community around: operating system is some kind of common base. Virtually everyone needs some OS.


And now what we can see? VPS hosters are all about OpenVZ, also some clouds/VMs/etc based on Xen or KVM, those are also Linux things most of time. And where're BSD jails? BSDs lost on their own battlefield. How pathetic.


Btrfs internals seems to be seriously superior to ZFS. And it mostly starts to stabilise. Granted that ZFS in FBSD has been released with unimplemented/bugged sendfile(), being capable of deadlocking whole OS, which still haven't prevented BSD nuts from claiming it "stable", I would not trust much to such judgement. There was epic outcry in mail lists when BSD zealots sweared ZFS is rock solid. Haha. Nice that Linux devs do not want to cheat their users about properties of released code. When they call something stable, usually it is. Also, there is Linux port of ZFS if someone needs it. Dtrace? Again, it's here in Oracle's linux. Though Linux haves other tracing/debugging tools as well. Capsicum? Hmm, some strange hack. Maybe not really bad, but not widely used as it requires to rework program code IIRC. Also IIRC Google ported it on Linux. Though it seems to be in low demand and nobody really cares about that.

You seem to have problems debating... You claimed that BSD is "failing to recognize modern requirements"; I'm providing evidence that this is not the case. If that stuff is being used or not, or if Linux's stuff is the only being used is irrelevant to the point.
Now, "Btrfs internals seems to be seriously superior to ZFS"... Really? Could you provide some proof about this and the other things you claim?

And if BSDs are not third rate crap, good luck to boot into BSD desktop and then watch video from Netflix. Ah. what? I'm supposed to use proprietary systems to do that? Then, Netflix itself considers BSDs as third rate crap. isn't it?

You seem to have some concentration issues... I asked you to show me how it was possible for a "third-rate crap" to handle "30% of nighly home Internet traffic". Please, focus.

Word-class footpads who serve to most uncooperative and troublesome DRM/proprietary fuxx. It does not really hurts. But makes some disgust.

So, you admit it is world-class...

Over 9000 times of FBSD, lol. Even most pessimistic assumption of 1% is waaaaaaay more than 0.01% FBSD can show. This is literally some few thousands of requests per whole continent per day. Rough estimation gives some several thousands of users for whole planet.

Man, you really need to focus... You argue that BSD's irrelevant position in desktop is some sign it dying (or at least not worth investing resources). Yet, Linux is also irrelevant on the desktop...

[profoundWHALE]

BSD is more relevant when it comes to things like servers and development. If Linux and FreeBSD were the two largest OS out there how Windows and OS X is, we still wouldn't go wrong and there would be a lot of good cross platform developer relations.

So don't be an idiot and freak out at one or the other when we are both on the same team.

[brosis]

The fact is that I haven't found any real definition of Libre Software beyond "it's another term for Free Software". It's a term that you are applying arbitrary standards to. Even Stallman agrees that the BSD/MIT licenses are Free Software; he prefers the GPL for obvious reasons, of course.

I cut my response, because unlike you I don't care how things break in proprietary world, or that you solve bugs by cutting features.
Your mileage, no problem.

I prefer Libre software, due to "Free" issue in English language of multiple definitions.
Libre exactly refers to freedom of speech, not price.

But Stallman insists on Free even if he probably recognizes its a bug, but he 100% refers to libre definition, not free price (ie GPL vs Freeware).
Renaming FSF to LSF would only do good, but keeping aliases to trademarks, domains etc is to persist regardless.

[JX8p]

BSD is more relevant when it comes to things like servers and development. If Linux and FreeBSD were the two largest OS out there how Windows and OS X is, we still wouldn't go wrong and there would be a lot of good cross platform developer relations.

So don't be an idiot and freak out at one or the other when we are both on the same team.

Unfortunately some people want to drive a wedge between two open-source communities that would get along fine (remember, Linux also is not going GPL v3) otherwise. Helps you see who the real proprietary stooges are: it's quite clear that they are the people who are trying to divide and conquer free software.

[JS987]

BSD is more relevant when it comes to things like servers and development. If Linux and FreeBSD were the two largest OS out there how Windows and OS X is, we still wouldn't go wrong and there would be a lot of good cross platform developer relations.
So don't be an idiot and freak out at one or the other when we are both on the same team.

OS X is based on FreeBSD. If OS X is competition for Linux, then FreeBSD is also competition for Linux.

[brosis]

Unfortunately some people want to drive a wedge between two open-source communities that would get along fine (remember, Linux also is not going GPL v3) otherwise.

Linux is not going towards GPL3+, because Linus has made a mistake of using GPL2 only license. People contributed under this license and its near impossible to find them to ask for GPL3+. I am more than sure 95% would approve the change though, but Linus thinks its not worth it. Time will show..

Helps you see who the real proprietary stooges are: it's quite clear that they are the people who are trying to divide and conquer free software.

The real proprietary stooges are those who use or support proprietary license.
Nothing else.

I am also not sure of your "divide and conquer" claim, because:
1) the licenses are different
2) the original goals are different
3) as time unfolds, each license with its iteration still follows original goals.

So the "get along" thing is nothing more, but a ethereal, outside effect. Following it will only damage original goals of each approach.
"Can we all get along?" - says who?


Lets look at your inner guts (license) first.

[yogi_berra]

Linux is not going towards GPL3+, because Linus has made a mistake of using GPL2 only license.

Except that's not how copyright works. Nice troll though, you get a gold star for effort.

[profoundWHALE]

I love you yogi bear!

@JS987
That's what I meant. I want there to be competition in the Open Source arena, but at this point, it's like we all split into different teams, all trying to beat each other these proprietary guys who simply stayed in one piece and stayed relatively consistent.

Like, to me it's stupid when we all basically have the same packages and the same core, but then start splitting and forking when there could be more collaboration.

A lot of people (myself included) will probably think of Wayland/Mir, but it doesn't apply in this situation the same way Unity was made from forking Gnome (I think). They have been forking in order to create a unified experience across their OS and make major changes to existing packages. Now, they could have tried to push their changes to the existing package, but there is never a guarantee that it will even be accepted. More often, it's guaranteed that it will never be accepted.

Still, the result is the same: Fragmentation. Now, that's not saying that I don't like alternatives, what I don't like are clones with minor changes. Like, imagine if each different Android phone didn't just have a different theme, but also had different packages that were running everything. D: