“A production version of the free GNU operating system is likely to be available by the end of this year, according to the president of the Boston-based Free Software Foundation (FSF). “We actually have the GNU kernel working, and we can now produce the GNU system, as opposed to the GNU/Linux system that people have been using so far,” said Richard Stallman, who is in India this week to attend a GNU/Linux Day in Pune.” Read the rest of the report about GNU/Hurd at IDG.net.
GNU/Hurd to be Released This Year
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Do we really need GNU/Hurd , thanks to Linux
and the many BSDes out there ???
CHF
Yeah Right. With no sound card support and only 1 gig partitions support.
is it needed? no. I does, however, provide yet another choice, which is never a bad thing.
And, to Mr. HAHA man, this is a project not supported by nearly as many developers as there are hackers to the linux kernel, which also has industry support. Plus, getting hardware support under the linux kernel as not always an easy task either.
Maybe we won’t have sound card support when the next release of hurd comes, but I think it’s safe to say they’ll increase the filesystem limit and fix MANY other issues… It’s supposed to be production ready when it is released, while it’s unclear what that means exactly, I’m sure it means thing such as raising that 1 GB limit.
Do we really need GNU/Hurd? Well, what about all the other people out there who bitch that other OS’s don’t have this feature and/or how all OS’s lack some other feature that would be great to have, or how the BeOS zealots complain that other os’s don’t their features? etc etc. Hurd should have a lot neat new features (before you complain about X… remember that X is getting better, and that several new GUI systems are in the works and perhaps we’ll be able to use those with the hurd).
Well, BeOS doesn’t run on a microkernel exactly, but I believe there are microkernelistic properties about the beos kernel. Also, hurd, is at some point supposed to have some cool buzzwords like “pervasive multithreading”, everyone, especially the beos users should like that. I also think that SMP at some point will take off… I have heard often that microkernels are made with SMP in mind…
Once it’s ready, it’ll run basically most of the linux software (well bsd/linux/etc software) out there and eventually I’m sure we’ll have pretty solid Linux emulation.
Which will be nice for the occasional closed source linux specific app like quake3 or wolfenstein.
Having another OS, ESPECIALLY a free one is a good thing! Even if it wasn’t, how could it be a bad thing? If you’re one of those people who say “… yet ANOTHER os…” well, then perhaps osnews isn’t the best site for you? I come here to learn about new os’s and os’s in general, I think that’s why most people come here…
its nice to see a free microkernel based system (not just a microkernel, this one has a system, too) and i will be playing with it.
its unreasonable to expect a level of maturity similar to linux’s, since the hurd has very few developers and even fewer full-time ones. besides, due to the nature of the license, hurd can cannibalise linux’s drivers and “server-ise” them. if they can automate this (maybe by writing wrappers) then hurd can enjoy the linux drivers from now until whenever
eventually I’m sure we’ll have pretty solid Linux emulation.
look no further than L4/Linux, which includes linux emulation by running a complete linux kernel as an L4 task. not just source compatible – this is even binary compatible
http://os.inf.tu-dresden.de/L4/LinuxOnL4/overview.html“> ” rel=”nofollow”>http://os.inf.tu-dresden.de/L4/LinuxOnL4/overview.html
MkLinux does a similar thing… Mach as the actual kernel, running a Linux microkernel server to handle the emulation. Interersting stuff!
“Do we really need GNU/Hurd , thanks to Linux
and the many BSDes out there ???”
Well, that’s kind of like saying, “do I really need more socks? I already have my right foot sock and my left foot sock.
I don’t know much about Hurd, and to be honest until today I haven’t really cared.
However, I read a very detailed article about Hurd’s architecture today and I think it looks very interesting. I probably won’t use Hurd as my main OS, at least not for a while, but I will definitely be running it on at least one of my systems.
I wish the Hurd team the best.
This is exactly the reason why free sw and open src will
never suceed: All developers are constantly reinventing
the wheel. “More chocie is never a bad thing”. Yes, it is.
It is not strategic if you want to take some market share.
We don’t need another kernel with some fancy design. We
need a desktop system worth the name. We need intelligent
and aiding systems that don’t get in the way of confuse the
developers and users. That is what we should concentrate on.
Soon it is everyone’s hobby to write an OS, and therefore
I will – against my will – end up running Windows.
*Sigh*, double-*Sigh*.
If Hurd were a linux-wannabe, i would most probably agree with you. But this it is actually quite a bit different than Linux and all the others. Someone who wishes to make an OS exactly like another shouldnt, but I doubt their aim is to copy another kernel. So I say, give them a chance at least, they deserve it.
The best of luck to Hurd.
Who ever said developers wanted to take market share? Some may, some may not. There is no way you could ever force a developer to work on a project they don’t want to, unless you pay them. If I feel like working on project X, that is my right.
I understand how you feel, but do you really believe that, say, the linux kernel is “intelligent and aiding”. Have you ever tried to build the kernel? What a pain.
Anyway, for a desktop OS, you need better apps. I have not seen many truly great free apps. Sure, there are some, but the majority come comercial entities. And, if that wasn’t true, why would I need VMware on my linux desktop at work? There are plenty of apps that I love to use, that there just is not a viable “free” alternative to.
teejay wrote:
>It is not strategic if you want to take some market share. We don’t need another kernel with some fancy design.
I think that you are misunderstanding the general audience that Hurd is going after. They are not looking for “market share” or anything like that. Not in the money-making sense anyway.
The early adopters will mainly be hobbyists and free-software believers. If the system is good, then it will be stealthily integrated into back offices like Linux was in many places.
I certainly do no think that their main thrust has anything to do with a desktop system.
I certainly look forward to trying it.
I like it as project and I know for sure that is progressing, even if in stealth mode
I for one, prefer microkernel architectures over Linux’s and FreeBSD’s monolithic designs. It’s not another kernel with a fancy design. It’s a completely different design, and a design that I think it better. Microkernels are easier to modify and debug than either BSD’s or Linux’s kernels are. Putting most of the services in user space servers also make it easier to maintain and debug those components as well. It’s also possible to totally redesign any one of the components without effecting any of the others since they don’t know anything of eachothers underlying structure (which I must admit is the beauty in NT, but I still don’t like it for other reasons). I like FreeBSD and Linux, but I would like GNU/Hurd more.
> This is exactly the reason why free sw and open src will
> never suceed: All developers are constantly reinventing
> the wheel.
Quite the opposite… this is the precisely the reason that free and open source software *has* and will *continue* to succeed! A huge and very accesible body of knowledge and usable code. The knowledge is there to take it where you want. If there are no current projects that are working toward your goals, create one yourself! Or fork a system that’s in the right ballpark and give it a new direction.
Why the hell should anyone create a Windows replacement for you when that’s not what they want? If you can’t code, it seems the only resource you can exchange is money. And if that’s the case, find someone you can pay to create your Windows-killer. The beautiful thing about free software is that it’s still relatively untainted by artificial stupidity like “marketshare” and other greed-driven frontiers.
> “More chocie is never a bad thing”. Yes, it is. It is
> not strategic if you want to take some market share.
Who said HURD, or anyone other project is out to take over the world’s market share? Success doesn’t equal marketshare any more than happiness equals the amount on your bank statement.
> We don’t need another kernel with some fancy design.
Why not? Who are you to decide what the developers and users of HURD should want? There are people who want this kernel, so let them have it.
> We need a desktop system worth the name. We need
> intelligent and aiding systems that don’t get in the
> way of confuse the developers and users.
*You* need a desktop system worth the name that fits in with your definition of an intelligent and aiding system. The HURD people need a Free kernel for the GNU operating system.
> That is what we should concentrate on.
Who is we? You seem to be an end user, not anyone with the desire or skill to write the code to create what you think should be concentrated on. If you have the ability, then do it! Why should HURD or other developers work on what *you* want, if it’s not of any use to them? If you want to give them a reason to work on something they don’t find interesting or useful personally, pay them.
Implement something on top of HURD, Linux, Darwin or NetBSD. No one is going to stop you, ye, they can’t stop you, provided you’re working within the license terms. That’s the beauty of it!
> Soon it is everyone’s hobby to write an OS, and therefore
> I will – against my will – end up running Windows.
Against your will? Pffft. You’re the one who is choosing to run Windows, you’re not being forced. Your will could choose to *learn* how to be productive on a non-Windows operating system, or your will could choose to *learn* how to program, and create the perfect system for what you think a desktop should be.
That’s exactly what I’m doing, but in the world of PDAs. PalmOS and PocketPC blow. The Linux PDA options blow even more. That is, they are not the kind of systems that I think would be perfect for me. They work for some people- great- but not for me. So what am I doing? Complaining that someone’s not giving me exactly what I want for free? No, I’m creating my dream, fulfilling my vision myself.
I don’t mean to rip on you personally, so don’t take it that way. I’m just sick of people who think that what they want should be what every developer wants so they can get something for free, when they should go Get Stuff Done themselves, or learn to deal with the options.
Now they need to leave behind the Xfree license and get a GNU GPL X disaster. And to just name it GNU OS, to make it even fancier.
<blockquote>This is exactly the reason why free sw and open src will never suceed: All developers are constantly reinventing
the wheel.</blockquote>Yeah, unlike all those closed-source companies.
>>Soon it is everyone’s hobby to write an OS, and therefore I will – against my will – end up running Windows.
It’s not “soon”. Plenty of people have written interesting OSes – FreeBSD, NetBSD, OpenBSD, OpenBeOS(NewBeos), SkyOS, etc. The more the better. Do you want to eat cheese burgers 24×7 for the rest of your life? I’d like Chinese food one day, Italian the next, and French surely looks tempting too.
*Sigh*, double-*Sigh*.
Keep sighing.
now was that real Linus or someone else??
>> Soon it is everyone’s hobby to write an OS, and therefore I will – against my will – end up running Windows.
Here’s another one for you: http://www.rtmk.org/nova/
What? A waste of time! There are no operating systems worth working on other than those which replace Windows for the desktop! Resistance again the One True Way is futile!
IP address would indicate that this is a “real” leenas. Notice that this ip address is also used for Bill Doors.
[quote]What? A waste of time! There are no operating systems worth working on other than those which replace Windows for the desktop! Resistance again the One True Way is futile![/quote]
Nova looks impressive to me considering that this is one person effort in such a short time. The primitive GUI Crust looks good on my 15″ LCD. Keep up the good work, Johan.
Johan,
I didn’t do much reading on Crust, so curse if you like. Am I correct that Crust isn’t based on top of the turtle X? I like the features of the kernel, but you need to pick a better name for it. rtmk sounds plain sucky, man. Let me know if you need people to play with your os.
>> This is exactly the reason why free sw and open src will
>> never suceed: All developers are constantly reinventing
>> the wheel.
>Quite the opposite… this is the precisely the reason that >free and open source software *has* and will *continue* to >succeed! A huge and very accesible body of knowledge and >usable code. The knowledge is there to take it where you >want. If there are no current projects that are working >toward your goals, create one yourself! Or fork a system >that’s in the right ballpark and give it a new direction.
This is to typical open source advocates. I don’t have
the time to wrestle with a massive code-body to create
an environment I want. College students and people with
huge amounts of spare time or unemployed people may do that.
But I have other things in life, and can only spend a
limited amount of time. If ‘hacking it to my liking’ is not
an option. A fate that I share with the 99.999% of Earth’s
population.
>Why the hell should anyone create a Windows replacement for >you when that’s not what they want? If you can’t code, it >seems the only resource you can exchange is money. And if >that’s the case, find someone you can pay to create your >Windows-killer. The beautiful thing about free software is >that it’s still relatively untainted by artificial >stupidity like “marketshare” and other greed-driven >frontiers.
I won’t comment on greed or Windows-killer and the debate
on capitalism and what it brings us. But I do have a couple
of truckloads of money on my backyards that I now till
pay Linus to create me a Windows killer. 🙂
>> “More chocie is never a bad thing”. Yes, it is. It is
>> not strategic if you want to take some market share.
>Who said HURD, or anyone other project is out to take over >the world’s market share? Success doesn’t equal marketshare >any more than happiness equals the amount on your bank >statement.
I would say “Success euqal marketshare in the same way happiness equal positive bank statements.”
Anyway, the more diluted the open source/fsf movement
will be the less likeley they will produce a usable
desktop. Divide and conquer, simple – only that it is
the developers themselves the to all the dividing and
M$ all the conquering. There’s a big difference in Focus.
>> We don’t need another kernel with some fancy design.
>Why not? Who are you to decide what the developers and
>users of HURD should want? There are people who want this >kernel, so let them have it.
Anyone is free to do what they want. But it is not
prodctive.
>> We need a desktop system worth the name. We need
>> intelligent and aiding systems that don’t get in the
>> way of confuse the developers and users.
>*You* need a desktop system worth the name that fits in
>with your definition of an intelligent and aiding system.
So the rest of you are satisfied with KDE or Windows, then?
I must have exceptionally high demands on my desktop. 🙂
>The HURD people need a Free kernel for the GNU operating >system.
No, RMS needs some limelight. He’s clearly frustrated that
Linux has stolen all the attention and really wants it to
be called GNU/Linux. Free software? Sure, as long as RMS
gets the attention he wants.
>> That is what we should concentrate on.
>Who is we? You seem to be an end user, not anyone with the >desire or skill to write the code to create what you think >should be concentrated on. If you have the ability, then do >it! Why should HURD or other developers work on what *you* >want, if it’s not of any use to them? If you want to give >them a reason to work on something they don’t find >interesting or useful personally, pay them.
We all are end users. I have yet to see a person interact
with create_process() or vfork(). 🙂 (I’ve got a PhD,
and am designing hard real time schedulers for a living).
I’ve been using and coding under FreeBSD, NetBSD, BSD/OS,
Linux, OSE and BeOS so I know at least something about
OS:es and kernels. I haven’t seen any desktop as nice as the
BeOS’ one. I cannot spend time on writing code on my spare
time because I have none. Nor do I have the money to pay
people to write code for me, I’m not that rich.
I end up in a dead-end: I don’t like Windows, there are no
other commercial OS’s with a decent desktop, there are free
OS’s that people work on and (other) use but I cannot
contribute in my current life-situation, nor am I allowed
to have comments on the work (despite I in the past
have contributed to the open source projects). So what
should I do? Sit here and shut up, I guess. 🙂 And
and a cup of free tea.
I’m merely stating that open src/fsf will never produce
an viable alternative to Windows and that I will end
up using Windows.
>Implement something on top of HURD, Linux, Darwin or >NetBSD. No one is going to stop you, ye, they can’t stop >you, provided you’re working within the license terms. >That’s the beauty of it!
NetBSD lacks any decent SMP capabilities and has horrible
latencies. It is not suited for a responsive desktop.
2.5.x and 2.6 of Linux has SMP and finally accepted
the kernel preemptive patches and SHOULD be better suited
for a desktop system. The problem still remains in the
user-space: X and its applications are not threaded to
any larger extent, so the gain will be limited. I’ve read
about HURD and Darwin. Darwin has borrowed a lot from
Net/FreeBSD and has its limited capabilities too (I guess,
I have’t actually looked at the code). HURD: It’s Mach,
isn’t it?
Then there’s the GPL: don’t get me started. 🙂
Even if they were good enough, technically, it would still
take an enormous effort (in the order of 100 man-years)
to create a system like BeOS. I’m 32, now… maybe
if I’m lucky enough to be 132. Then my children can use
it (only aged 101 by then… :-).
>> Soon it is everyone’s hobby to write an OS, and
>>thereforeI will – against my will – end up running >>Windows.
>Against your will? Pffft. You’re the one who is choosing to >run Windows, you’re not being forced. Your will could
More or less. I’m not working alone: colleagues are using
windows and their office suite, the official support policy,
the network.
In theory: yes. In practive: no. In any event, if you
choose a non-MS system, you have to be prepare to put
in a lot of extra work to interact with you MS environment.
>choose to *learn* how to be productive on a non-Windows >operating system, or your will could choose to *learn* how
As I could choose to *learn* how to be productive
on a Windows platform: I can conform, but I don’t like
it. As many of you.
>to program, and create the perfect system for what you >think a desktop should be.
Learn how to program? Yes, maybe I should… 🙂
>That’s exactly what I’m doing, but in the world of PDAs. >PalmOS and PocketPC blow. The Linux PDA options blow even >more. That is, they are not the kind of systems that I >think would be perfect for me. They work for some people- >great- but not for me. So what am I doing? Complaining that >someone’s not giving me exactly what I want for free? No, >I’m creating my dream, fulfilling my vision myself.
I don’t need anything for free: I will happily pay for
a decent desktop system. If I have copious amount of
spare time, no wife, no children, no work or better things
to do: sure I could code my dream. But I’d rather buy it.
There aren’t many secret ideas in software dev…. use of Ext3 or ReiserFS eventually leads to better filesystems for closed or semi-closed OS’ like OSX and XP-Longhorn..
KDE and Quartz interfaces for different variants of *niX beget Luna on WinXP…
Standards-complient browsers from Opera and Mozilla encourage better compliance in IE6
why bitch about it.. if someone does a project, on any OS, and they get it right, in a year or so similar ideas will become available on other platforms. In the long run, Windows users are seeing benefits brought about by better Linux/GNU/GPL software.
I freely admit to no knowledge of programming…. responsibility for any errors in this is my own.. its just how I see it.
OK, OK.. the above wasn’t all that well thought out cos I was thinking more about the Windows Longhorn FS database….. than about Hurd. I clearly need more coffee….
> This is to typical open source advocates. I don’t have
> the time to wrestle with a massive code-body to create
> an environment I want…
This argument is typical because it is rock solid. Most people haven’t the time or the abilities to create software systems that are exactly to their liking. But the option is there, and very available. No one is keeping you from creating such a system but you and your choices regarding what to do with your own spare time. People write code in their free time because they are creating what they want. Just like you don’t have the time to program the system you want, we don’t have time to work on the system that you want either, because we’re busy with more important things- one of which is creating the system that *we* want. Which is to say that each individual wants as far as their own project is concerned, it’s not that there are two perfect environments, one for hackers and one for the rest of you. But unless you give someone incentive, why would they put the goals and wants of someone they do not know over their own?
> I would say “Success euqal marketshare in the same way
> happiness equal positive bank statements.”
Your happiness is based solely on how much money you have? It sounds like a rather empty and unfulfilling life.
> Anyway, the more diluted the open source/fsf movement
> will be the less likeley they will produce a usable
> desktop.
Of course it is less likely to produce something that is a Windows replacement for you- that’s not the goal!
> Anyone is free to do what they want. But it is not
> prodctive.
Why is it not? It is productive for the person implementing, using it, or studying it, for their own reasons. How do you spend your free time? Surely, you’re spending time in discussions here on OSNews, time surely spent in an unproductive way. Why aren’t you off doing whatever worked you’re told will profit someone else?
> So the rest of you are satisfied with KDE or Windows,
> then? I must have exceptionally high demands on my
> desktop. 🙂
Of course they aren’t. But the people that are creating KDE seem to be pleased with it. It’s not surprising, as they are creating a system for themselves to use. They are trying to make it usable for others along the way, but the immediate intent is to create a system they’d want to use. Those of you who run Windows do so at your own discretion. If you don’t know how to program, why not apply your skills to generate the funds to pay for a team to create your own Windows replacement?
> No, RMS needs some limelight. He’s clearly frustrated that
> Linux has stolen all the attention and really wants it to
> be called GNU/Linux. Free software? Sure, as long as RMS
> gets the attention he wants.
RMS has quite the desire for attention, no doubt. But surely you cannot believe that’s the only reason that HURD exists! Looking at the ChangeLog, it appears that RMS himself doesn’t contribute code to this project, so there is only so much attention that he can gain from it. But there are people out there using it, so obviously others want it. There are people who prefer the architecture of HURD to that of Linux. Or do you think that all users developers of HURD are robots programmed (in e-lisp, of course) by RMS himself to pretend that some people have a genuine interest in the project? Perhaps that’s why you use Windows! You’re just a robotic pawn of Bill Gates!
> I’ve been using and coding under FreeBSD, NetBSD, BSD/OS,
> Linux, OSE and BeOS so I know at least something about
> OS:es and kernels. I haven’t seen any desktop as nice as
> the BeOS’ one.
Then why don’t you continue to use BeOS? If you’re so disgusted by Windows and the Linux desktops, why not get a Mac running Mac OS X?
> I cannot spend time on writing code on my spare time
> because I have none. Nor do I have the money to
> pay people to write code for me, I’m not that rich.
That’s too bad. But that’s your choice, and your problem. While I’d like to have a 1 GHz XScale laptop running LispOS that has 12 hours of battery life, I don’t, and I can’t. Lacking the ability and time to design and implement the hardware and software, I deal with it and use the next best thing. It would be unfair to expect others to do this work for me, just because I demand it.
> I’m merely stating that open src/fsf will never produce
> an viable alternative to Windows and that I will end
> up using Windows.
Do you think that OpenBeOS will bear any fruit? Or is it, by being open source, doomed to never producing anything of value? That seems to be a project where your desire to have a good desktop system seem to be in alignment with the goals of the developers. But you are using Windows by your own choice, not because you’re being forced to by a cabal of dirty and unemployed open source programmers who do not bend to your every wish.
[comments about specific problems with HURD, Linux, *BSD]
Then fix them, or wait until someone else does. Or choose to continue to use Windows.
> In theory: yes. In practive: no. In any event, if you
> choose a non-MS system, you have to be prepare to put
> in a lot of extra work to interact with you MS
> environment.
*You* are the one choosing to run Windows. You could have a different job. You could work around the fact that all of your colleagues run Windows. For instance, you could have run Mac OS X, and have Virtual PC for specific instances where you need a Windows system. You could choose to try to get your institution to switch to methods of computation and communication that didn’t require you all used Windows. Don’t have the time? You’re the one who chooses to manage your time like you do.
> But I’d rather buy it.
Unfortunately, the way the market works, you’re at the will of companies who produce the software you’d potentially want. So why isn’t it their fault that you don’t have a usable desktop system? If it weren’t for Microsoft, you wouldn’t be using Windows.
They’re kdding….right?
I asked about hurd back in 1996,
‘when will this be done?’, I asked.
Now we’re deep in the next millenium
and RS wants to lead us all back to 1971.
Now there’s a ‘mutually recursive anacronism’.
For all the money that pours into the FSF, you’d
think that this would have been
*just a bit more marvelous Dick*
> RS wants to lead us all back to 1971.
What’s wrong with 1971? UNIX was born around then, and Linux is basically a UNIX clone. Both work very well in most instances. GNU HURD (which stands for “Hurd of UNIX-Replacing Daemons”) is designed to go beyond the UNIX concept. If anything, it is a forward-thinking architecture, not a backward-looking one.
> For all the money that pours into the FSF, you’d think that
> this would have been *just a bit more marvelous Dick*
They don’t really get much money, and there are nowhere near as many HURD developers as there are for Linux.
> HURD (which stands for “Hurd of UNIX-Replacing Daemons”)
From the HURD page:
<em>
According to Thomas Bushnell, BSG, the primary architect of the Hurd:
`Hurd’ stands for `Hird of Unix-Replacing Daemons’. And, then, `Hird’ stands for `Hurd of Interfaces Representing Depth’.
</em>
Also note, the HURD is just the collection of servers that run (in user-mode) on a microkernel. It’s *supposed* to be microkernel-agnostic. I think that, at the moment, it’s running on the Mach u-kernel but it turns out that the HURD relies quite a bit on Mach-specific stuff. There seems to be a movement to port the HURD to run on an L4-based u-kernel (from what I’ve read).
http://os.inf.tu-dresden.de/L4/
My guess is that GNU-HURD + L4 would bring home the bacon and fry it up in the pan.
> Johan,
> I didn’t do much reading on Crust, so curse if you like.
> Am I correct that Crust isn’t based on top of the
> turtle X?
Correct!
> I like the features of the kernel, but you need to pick a better name for it.
> rtmk sounds plain sucky, man. Let me know if you need people to play with your os.
rtmk is the name of the microkernel, and stands for real-time microkernel.
The OS kernel is called Nova.