Re:NetBSD Concurrency Model
on
NetBSD 1.5ZB
·
· Score: 2, Informative
Er, you seem to be confusing userland and kernel threads. The two are not connected. The way you implement pthreads in users space and whether you have a preemptable kernel and such aren't connected at all.
In terms of pthreads, NetBSD has chosen the approach used in Solaris, Digital Unix (now Tru64) etc. of using Scheduler Activations. On an MP system, two threads will indeed run on different processors if you like.
It is true that NetBSD does not yet have a preemptable kernel and that our SMP support is not incredibly stunning yet. We're working hard to fix that.
When 1.6 will branch....
on
NetBSD 1.5ZB
·
· Score: 2, Informative
As someone involved in NetBSD release engineering, I'm guessing 1.6 is going to branch "soon", likely within weeks (though no promises.)
Our hope is to pick up the pace of releases now that we have a lot more infrastructure for doing fast release engineering. A lot of that was developed only in the last six months.
Re:NetBSD stopped being useful once I forked OpenB
on
NetBSD 1.5ZB
·
· Score: 2, Insightful
I presume that someone is just forging Theo's name as I doubt Theo would be foolish enough to put up a post like that.
However, just to set the record straight, I will point out that:
0) NetBSD is very much alive and vibrant. If you look at the sheer number of commits per day to the NetBSD tree, one will see that pretty quickly. There are a lot of NetBSD developers and users, and the developers are very active. 1) Multi-platform portability is pretty damn useful in the embedded systems world. Maybe running on a StrongARM or a low power MIPS design isn't interesting to you, but it is very interesting to people building things like routers and set top boxes. We pay our bills at Wasabi thanks to this. How many platforms will we ultimately port to? Well, people keep paying Wasabi to port to new things, and there are people outside of Wasabi doing NetBSD ports, too. As long as people keep designing new computers, I don't think NetBSD will stop adding ports. 2) NetBSD is successful enough in terms of design wins to support our company fairly nicely. It is also successful enough in terms of developer resources that I'm proud to say we've got a damn good operating system and it keeps getting better all time. There are a couple hundred very good engineers who commit to the NetBSD tree, and a cast of thousands submitting patches and updates. 3) Generally speaking, the OpenBSD guys are smart and nice people -- I get along with a lot of them very well. Guys like Todd, Niels and Angelos (to name a few) are fine engineers and I have plenty of respect for them.
Re:NetBSD Concurrency Model
on
NetBSD 1.5ZB
·
· Score: 3, Informative
Not entirely clear what you're asking. We implement POSIX pthreads (the really good version of that, which uses scheduler activations, is on a branch pending the cut of 1.6 -- it will be integrated into -current shortly.)
Anyway, pthreads, which is pretty much the Unix standard, has a set of mechanisms available for synchronizing the multiple threads.
Inter-process communication between threads, processes, etc. happens pretty much the way you want it to -- message passing via sockets, shared memory, whatever you like. It is pretty much the way any POSIX style system works.
The advantages of NetBSD are not primarily in the API, which it shares with most POSIX systems, but in the license and the quality of the implementation.
What's the point of all this? You can just run ethernet over fiber, and a lot longer distance than 1km. You can also get ethernet in a whole lot better than 10Mbps these days.
The equipment is perfectly standard, and if you count your time, a whole lot cheaper.
Replying to trolls is always a deadly business, but what the hell. I'm bored.
I don't know how many NetBSD users there are, but somehow, I suspect you don't either. NetBSD is in use every day in far more places than you think. Your DSL router or a cache box at your ISP may very well run NetBSD without your even knowing it. Ditto for large numbers of users of NCs -- NC/OS is NetBSD 1.3. There are users of hundreds of thousands of things like airline reservation terminals and such that use NetBSD every day and don't even know it. NetBSD is a very portable, very clean OS with a BSD license attached, so it gets put into LOTS of embedded hardware.
There are also a pretty large number of people who use NetBSD day to day and are very well aware of the fact. Just look at the NetBSD mailing lists if you don't believe me.
Why do people like NetBSD? Well, that would be a very long discussion. Suffice it to say, though, we've done a lot of careful architectural work on NetBSD over the years and it has paid off handsomely. There is a reason we port to new hardware quickly for instance, and it isn't that we have more fanatics than the Linux crowd. We've also got a BSD license on the code, and without trying to start a fight some people prefer the BSD license to the GPL. There are also other people who like having their whole OS build out of a single coherent source tree -- Linux code integration is a big pain which is why most people don't build full Linux systems from sources on their own.
Why is it news that there's now an OS that's gone multiuser on the x86-64? Well, that's probably more to do with the x86-64 being a neat new design than to do with us. We do a lot of ports and most of them don't get slashdotted. However, I'd say that given the fight brewing between Intel and AMD, this wasn't the silliest story for slashdot to cover.
Why do lots of people seem to think there are no BSD users in the world? I don't think anyone but a troll would claim that with a straight face...
Well, I manage to earn enough to meet a pretty large payroll off of what real companies pay us to make the "toy" work on their embedded platforms, and our customers (of which there are a lot) seem pretty happy.
However, what would I know? I just run a company that exclusively focuses on NetBSD.
Yes. It can run both 32 bit NetBSD and 32 bit Linux binaries, though of course, given that the simulator is slow, this hasn't been very thorougly tested.
The Z50 has pcmcia, and runs the NetBSD drivers for all supported ethernet, serial and wireless cards, so there is no trouble running you Z50 with any sort of net connectivity (including mobile) that you like!
It is fairly straightforward. The NetBSD hpcmips crowd has been running on a wide variety of WinCE handhelds for a while. If you want to learn more, the web site and (more importantly) the mailing lists can help you out.
Open source systems have been taking over the embedded systems market for quite a while. There are millions of devices out there that run NetBSD, for instance. One issue here for the embedded people is, of coures, GPL vs. non-GPL licensing. One reason BSD derived systems are already widely deployed in embedded markets is for this reason.
This doesn't change anything for Microsoft. The NCs already were not running Microsoft OSes. They were running NetBSD. This is just a shift towards the Linux "brand" of open source.
Microsoft HAS put pressure on various NC makers in the past, of course -- see the famous pressure they brought to bear to destroy the DEC DNARD -- but in general, one should keep in mind that NC OS is NetBSD, and that it isn't likely that NC is going to start shipping Win CE or some crap like that instead.
Most IBM NCs already run an open source OS, specifically NetBSD. NCOS is just a hacked version of NetBSD 1.3.2
Rumor has it that some people have "real" NetBSD running on IBM NCs but I haven't checked.
It isn't much stated, but NetBSD is located inside of literally hundreds of thousands of deployed network computers, and millions of embedded device applications.
Just wanted to point out that NetBSD originated the Linux syscall emulator FreeBSD uses. By definition, we run anything they do in this stuff -- we created the code, after all.:)
We are, pretty much, destined to become gods. It is obvious that with the tools we'll have in less than a century, we will gain the ability to do anything that isn't inconsistent with the laws of physics. Between nanotechnology and artificially enhanced intelligence, very little should be beyond our capacities.
Perhaps you haven't noticed the growing technological changes around you. Perhaps your belief that we are too "stupid" impairs your ability to notice that we're rapidly learning how to fix that.
In any case, however, neither quasi-religious optimism nor quasi-religious pessimism stemming from political beliefs about how horrible the human race is are likely to have a long term impact on what actually happens.
Why is it that when people get Linux running on new Mac hardware, the/. crew immediately accepts the news story, but when NetBSD does so (we run on virtually every Mac around) the announcement is never considered newsworthy?
NetBSD has the Kame code integrated in -current and it works just fine -- you can telnet to a v6 host if you like.
I suspect your Linux userland utilities just haven't been updated to be v6 aware. Either complain to the maintainers of those programs or send them patches. There are lots of patches on several FTP sites for making utilities v6 aware, by the way.
By the way, NetBSD runs great on iMac and G3 -- presumably it will run on the G4 (possibly with a little work -- no one that I know of has gotten their hands on one yet.)
I interview people for jobs on nearly a daily basis. A solid undergrad engineering degree is probably a really good way to learn your trade, but as an employer, for the most part, I totally ignore credentials. Why? Because I've found little correlation between them and the ability to do one's job, that's why. I've met far too many people who can't do the job who have professional degrees and certifications. Since they don't help me figure out who to hire, they're not something I pay attention to. Passing my technical interview usually indicates that a candidate will succeed, so that's most of what I care about.
They aren't OPEN sourcing it, they are giving it the boneheaded license they've used for their other "no where to go but into the ground" projects. Who wants to work for free to increase Sun's profit margins? The use of the term "community" in the license is laughable.
I'm especially worried because the New York Times story about the purchase indicates that Sun intends to use this as a way of pushing their silly thin client plans. Word processing over the web! Gack.
If they really want to kill Microsoft, they should just open source StarOffice FOR REAL and support development. The benefits would be astounding. Sun, of course, will never do this.
This is the same company that buldozed System V into the offices of users who wanted to stick to BSD, who unbundled their compilers from their OS, etc. They don't ever back down when they make a mistake.
The API issue is a bit funny. Unfortunately, lots of the v4 API made assumptions about address length and about there only being one address family in use at a time. The v6 API fixes this. There is an RFC that describes the suggested API in detail -- so far as I'm aware, most implementations to date have tried to follow it fairly closely.
Slashdot Mirror
Er, you seem to be confusing userland and kernel threads. The two are not connected. The way you implement pthreads in users space and whether you have a preemptable kernel and such aren't connected at all.
In terms of pthreads, NetBSD has chosen the approach used in Solaris, Digital Unix (now Tru64) etc. of using Scheduler Activations. On an MP system, two threads will indeed run on different processors if you like.
It is true that NetBSD does not yet have a preemptable kernel and that our SMP support is not incredibly stunning yet. We're working hard to fix that.
As someone involved in NetBSD release engineering, I'm guessing 1.6 is going to branch "soon", likely within weeks (though no promises.)
Our hope is to pick up the pace of releases now that we have a lot more infrastructure for doing fast release engineering. A lot of that was developed only in the last six months.
I presume that someone is just forging Theo's name as I doubt Theo would be foolish enough to put up a post like that.
However, just to set the record straight, I will point out that:
0) NetBSD is very much alive and vibrant. If you look at the sheer number of commits per day to the NetBSD tree, one will see that pretty quickly. There are a lot of NetBSD developers and users, and the developers are very active.
1) Multi-platform portability is pretty damn useful in the embedded systems world. Maybe running on a StrongARM or a low power MIPS design isn't interesting to you, but it is very interesting to people building things like routers and set top boxes. We pay our bills at Wasabi thanks to this. How many platforms will we ultimately port to? Well, people keep paying Wasabi to port to new things, and there are people outside of Wasabi doing NetBSD ports, too. As long as people keep designing new computers, I don't think NetBSD will stop adding ports.
2) NetBSD is successful enough in terms of design wins to support our company fairly nicely. It is also successful enough in terms of developer resources that I'm proud to say we've got a damn good operating system and it keeps getting better all time. There are a couple hundred very good engineers who commit to the NetBSD tree, and a cast of thousands submitting patches and updates.
3) Generally speaking, the OpenBSD guys are smart and nice people -- I get along with a lot of them very well. Guys like Todd, Niels and Angelos (to name a few) are fine engineers and I have plenty of respect for them.
Not entirely clear what you're asking. We implement POSIX pthreads (the really good version of that, which uses scheduler activations, is on a branch pending the cut of 1.6 -- it will be integrated into -current shortly.)
Anyway, pthreads, which is pretty much the Unix standard, has a set of mechanisms available for synchronizing the multiple threads.
Inter-process communication between threads, processes, etc. happens pretty much the way you want it to -- message passing via sockets, shared memory, whatever you like. It is pretty much the way any POSIX style system works.
The advantages of NetBSD are not primarily in the API, which it shares with most POSIX systems, but in the license and the quality of the implementation.
What's the point of all this? You can just run ethernet over fiber, and a lot longer distance than 1km. You can also get ethernet in a whole lot better than 10Mbps these days.
The equipment is perfectly standard, and if you count your time, a whole lot cheaper.
Replying to trolls is always a deadly business, but what the hell. I'm bored.
I don't know how many NetBSD users there are, but somehow, I suspect you don't either. NetBSD is in use every day in far more places than you think. Your DSL router or a cache box at your ISP may very well run NetBSD without your even knowing it. Ditto for large numbers of users of NCs -- NC/OS is NetBSD 1.3. There are users of hundreds of thousands of things like airline reservation terminals and such that use NetBSD every day and don't even know it. NetBSD is a very portable, very clean OS with a BSD license attached, so it gets put into LOTS of embedded hardware.
There are also a pretty large number of people who use NetBSD day to day and are very well aware of the fact. Just look at the NetBSD mailing lists if you don't believe me.
Why do people like NetBSD? Well, that would be a very long discussion. Suffice it to say, though, we've done a lot of careful architectural work on NetBSD over the years and it has paid off handsomely. There is a reason we port to new hardware quickly for instance, and it isn't that we have more fanatics than the Linux crowd. We've also got a BSD license on the code, and without trying to start a fight some people prefer the BSD license to the GPL. There are also other people who like having their whole OS build out of a single coherent source tree -- Linux code integration is a big pain which is why most people don't build full Linux systems from sources on their own.
Why is it news that there's now an OS that's gone multiuser on the x86-64? Well, that's probably more to do with the x86-64 being a neat new design than to do with us. We do a lot of ports and most of them don't get slashdotted. However, I'd say that given the fight brewing between Intel and AMD, this wasn't the silliest story for slashdot to cover.
Why do lots of people seem to think there are no BSD users in the world? I don't think anyone but a troll would claim that with a straight face...
Perry
Well, I manage to earn enough to meet a pretty large payroll off of what real companies pay us to make the "toy" work on their embedded platforms, and our customers (of which there are a lot) seem pretty happy.
However, what would I know? I just run a company that exclusively focuses on NetBSD.
Perry Metzger
CEO
Wasabi Systems
Yes. It can run both 32 bit NetBSD and 32 bit Linux binaries, though of course, given that the simulator is slow, this hasn't been very thorougly tested.
The Z50 has pcmcia, and runs the NetBSD drivers for all supported ethernet, serial and wireless cards, so there is no trouble running you Z50
with any sort of net connectivity (including mobile) that you like!
It is fairly straightforward. The NetBSD hpcmips crowd has been running on a wide variety of WinCE handhelds for a while. If you want to learn more, the web site and (more importantly) the mailing lists can help you out.
Open source systems have been taking over the embedded systems market for quite a while. There are millions of devices out there that run NetBSD, for instance. One issue here for the embedded people is, of coures, GPL vs. non-GPL licensing. One reason BSD derived systems are already widely deployed in embedded markets is for this reason.
This doesn't change anything for Microsoft. The NCs already were not running Microsoft OSes. They were running NetBSD. This is just a shift towards the Linux "brand" of open source.
Microsoft HAS put pressure on various NC makers in the past, of course -- see the famous pressure they brought to bear to destroy the DEC DNARD -- but in general, one should keep in mind that NC OS is NetBSD, and that it isn't likely that NC is going to start shipping Win CE or some crap like that instead.
Most IBM NCs already run an open source OS, specifically NetBSD. NCOS is just a hacked version of NetBSD 1.3.2
Rumor has it that some people have "real" NetBSD running on IBM NCs but I haven't checked.
It isn't much stated, but NetBSD is located inside of literally hundreds of thousands of deployed network computers, and millions of embedded device applications.
Just wanted to point out that NetBSD originated the Linux syscall emulator FreeBSD uses. By definition, we run anything they do in this stuff -- we created the code, after all. :)
We are, pretty much, destined to become gods. It is obvious that with the tools we'll have in less than a century, we will gain the ability to do anything that isn't inconsistent with the laws of physics. Between nanotechnology and artificially enhanced intelligence, very little should be beyond our capacities.
Perhaps you haven't noticed the growing technological changes around you. Perhaps your belief that we are too "stupid" impairs your ability to notice that we're rapidly learning how to fix that.
In any case, however, neither quasi-religious optimism nor quasi-religious pessimism stemming from political beliefs about how horrible the human race is are likely to have a long term impact on what actually happens.
Why is it that when people get Linux running on new Mac hardware, the /. crew immediately accepts the news story, but when NetBSD does so (we run on virtually every Mac around) the announcement is never considered newsworthy?
NetBSD has the Kame code integrated in -current and it works just fine -- you can telnet to a v6 host if you like.
I suspect your Linux userland utilities just haven't been updated to be v6 aware. Either complain to the maintainers of those programs or send them patches. There are lots of patches on several FTP sites for making utilities v6 aware, by the way.
By the way, NetBSD runs great on iMac and G3 -- presumably it will run on the G4 (possibly with a little work -- no one that I know of has gotten their hands on one yet.)
I interview people for jobs on nearly a daily basis. A solid undergrad engineering degree is probably a really good way to learn your trade, but as an employer, for the most part, I totally ignore credentials. Why? Because I've found little correlation between them and the ability to do one's job, that's why. I've met far too many people who can't do the job who have professional degrees and certifications. Since they don't help me figure out who to hire, they're not something I pay attention to. Passing my technical interview usually indicates that a candidate will succeed, so that's most of what I care about.
They aren't OPEN sourcing it, they are giving it the boneheaded license they've used for their other "no where to go but into the ground" projects. Who wants to work for free to increase Sun's profit margins? The use of the term "community" in the license is laughable.
I'm especially worried because the New York Times story about the purchase indicates that Sun intends to use this as a way of pushing their silly thin client plans. Word processing over the web! Gack.
If they really want to kill Microsoft, they should just open source StarOffice FOR REAL and support development. The benefits would be astounding. Sun, of course, will never do this.
This is the same company that buldozed System V into the offices of users who wanted to stick to BSD, who unbundled their compilers from their OS, etc. They don't ever back down when they make a mistake.
The KAME people have ported a lot of open source apps already. Most of the better open source browsers have patch kits available.
The protocol identifier was used by an experimental protocol that was never widely deployed.
Well, you can help! Get onto the 6bone, port open source applications, and pressure your ISP to offer native v6 service!
The API issue is a bit funny. Unfortunately, lots of the v4 API made assumptions about address length and about there only being one address family in use at a time. The v6 API fixes this. There is an RFC that describes the suggested API in detail -- so far as I'm aware, most implementations to date have tried to follow it fairly closely.
There is also data on applications and implementations on www.ipv6.org