Slashdot Mirror
Andrew Tanenbaum Announces MINIXcon (minix3.org)
LichtSpektren writes: Andrew Tanenbaum, author of MINIX, writes: 'MINIX has been around now for about 30 years so it is (finally) time for the MINIXers to have a conference to get together, just as Linuxers and BSDers have been doing for a long time. The idea is to exchange ideas and experiences among MINIX 3 developers and users as well as discussing possible paths forward now that the ERC funding is over. Future developments will now be done like in any other volunteer-based open-source project. Increasing community involvement is a key issue here. Attend or give a presentation.' The con will be held on 1 February 2016 at the Vrije Universiteit in Amsterdam, the Netherlands.
"The idea is to exchange ideas and experiences among MINIX 3 developers and users..."
I wish all seven of them have a good time exchanging ideas and experiences.
Perhaps they could use email.
You know.
Like through a linux server.
Because.
That's how it works.
"If the OSI developers are emailing each other, it's over TCP/IP" -- Steven Belovin, before you were born
"If the Minix developers[sic] are emailing each other, it's on Linux systems" -- me
Ehud
I hope that Minix 3 gets more interest. When I have time I really want to see if I can use it for NAS.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
Would it really have been so hard to link to the announcement in the summary:
http://www.minix3.org/conferen...
I went to MINIX.com, and found I could get an email address - that works right in my browser! - for only 35 dollars a year.
Oh, minix CON.
Nevermind
Dark Reflection
Minix 3 is not the same as the Minix you might still have on floppies or dead trees. Different basis, different goals. Yes, still a microkernel, but meant to be used in production and not (just) as a teaching aid. Check it out at minix3.org before disparaging :).
You think Tannenbaum lost that argument? Linus made a complete ass of himself.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Please don't be offended, but...
Is that a version number or a head count?
MINIX is obsolete.
Even if assuming that's the case: okay, so what? Things that are considered 'obsolete' are used in many places, every day, doing their thing. Often better than if done by a modern 'equivalent'.
From what I've read, MINIX has some unique features that mainstream OS'es don't have. For that reason alone there's a place for it. And it's useful as a way to learn the inner workings of an OS. Not as big and complex as an OS that supports everything under the sun.
Still not good enough hey? How about as a research vehicle? To try some new concepts that haven't been tried elsewhere. Do things that have been done elsewhere just a little different, and see how that works. Or just for the fun of it.
Especially us /. users should applaud and appreciate projects like this. There used to be a time when it seemed as if every company were working on some OS or programming language of their own. When hobbyists where beating bare metal of their PC's in assembly, even up to a GUI or 3D games. These days... not so much. Most software news these days is new releases of existing software. New versions of existing operating systems. Some new way to make existing software X work with existing software Y. Projects like MINIX that are still developed (even if slowly) are few and far between.
Last but not least: if you're not interested: fine, that's OKAY. But no reason to mock an interesting project simply because it's not your cup of tea.
The children on this thread are hilarious. So what if people want to get together and talk about a project that interests them but may not interest you. Why do you care? Keep using Linux if thats what you prefer but at least grow up enough to not be but-hurt when other people enjoy working with something else.
Yes, I say he "lost" in the court of public opinion, because the rough gist of his initial argument was that Linux was going nowhere due to its Kernel design, and Linux and Minix had very different trajectories than this argument would have predicted. How much of that was due to kernel design is debatable, but the argument that Linux was "obsolete" 23 years ago certainly looks silly now.
If one digs into the details of the debate, not only were a lot of Tannebaum's detailed points quite correct, but Linus wasn't even arguing against them at the time. He was just arguing that for what he was trying to accomplish, Linux's design was better. So you could say they were both right on the details, but it isn't the details that made the argument famous.
Tanenbaum: Do you think I should hire a limo?
Mrs Tanenbaum: To shuttle them to the conference, dear?
Tanenbaum: No, to hold it in.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
You can just see how Tannenbaum won the argument by looking at how Minix installations outnumber Linux ones.
Tannenbaum may think he won. You may think Tannenbaum won. The world, apparently, thinks differently.
Most people see that argument being over which kernel design is better.
The world didn't pick Linux because it was a better design, they picked it because it was free and could do more than any other free OS.
Linus's brilliance was in the way he organized the project and guided development. It meant a LOT of ports and a lot of drivers early on, which meant "easy to deploy." The kernel was a mess at the time. Some would say it's still a mess, since it constantly has to be revised to accommodate new features. The MINIX kernel is stable and flexible enough to go unchanged for a few years at at time.
never heard of it. I'm gnu to this stuff.
If the court of popularity gets to decide truth, then artists like Brittney Spears represent the pinnacle of music composition.
After all, just look at all the air time she gets compared to people like John Cage, who in the court of musical popularity are obviously doing it wrong. And don't let me get started on Tom Waits, who despite being a major influence on just about every musician on the west coast is irrelevant when one compares radio air time.
Perhaps popularity isn't the right measure of correctness.
More desktop users are using Windows 10 than use Linux - of any kernel number. By that logic, Windows must be better. I'm not sure I agree. I use Linux but that's because it's handy and has a huge ecosystem. I do have MINIX in a VM but, honestly, I've not found a real use for it, for me. I do think the idea, the microkernel, is sound and that it should make a more secure and stable system. There's a loss in speed but we've pretty speedy hardware now. He may win that argument in numbers still. Doubtful but, then again, do the numbers really tell you which is better? Is a Ford Focus better than a BMW 6-series?
"So long and thanks for all the fish."
Dear Andy, you kept Minix to yourself for decades. What happened to make you change your mind, 30 years late?
no, I don't have a sig
"More desktop users are using Windows 10 than use Linux - of any kernel number. By that logic, Windows must be better."
Apples to oranges. Windows 10 is an OS. Linux is a kernel. If you mean GNU/Linux, then yes desktop Windows, by that logic, is the better desktop OS (can play latest and greatest games, easier to install, better hardware support, etc). You can include Android and various other systems (embedded or otherewise) that use the Linux kernel, then you have an installed base larger than all Windows versions combined. Many of them power the Net of things, but Android is a front-facing Linux-based system (one of many). For many mobile use cases, Android is the better OS (more difficult than IOS but less confusing than Winphone).
Those are not desktops. Which is why I quantified it. If the number in use for a task is a mark of a better product than grains of sand would be better than food. The Ford Focus would be better than a BMW. Android 4.2 would be better than 5.0. PHB would be better than competent people.
There are lots of arguments to be made that say Linux is better - I might even agree with them. But using the number of them in use as a metric is not a very good deterministic approach for quality. There are more ants than there are people. Ants must be better... :/ Better at what?
No, the number of devices in use don't actually mean the OS is any better. That's like saying that Windows is better on the desktop (Windows also has a kernel) simply because more people use it there. That's like saying Ubuntu is better on the server because more people use it than use Red Hat. It might be truly better but the number of installs isn't the metric to determine that. It's not even a useful metric. If that's the only metric they've got (and it shouldn't be - I can come up with a few just sitting here) then they're really gonna need to go back to square one and try it again.
"So long and thanks for all the fish."
I've poked at Minix off and on for the last two decades, I've even taught OS classes using Minix as the source case. It's pretty cool. Its pretty clean code, and it's pretty easy to follow. Now it would be nice to have a portable platform to be able to use. There is more than a few Pi's around, it would make a great place for people to play. For an OS learning environment Minix is great. It's in the realm of Unix V6 or FreeDOS. Linux is a great OS (using it now) but for a lean teaching tool it's too big to manage. I'd love to be able to teach on the Pi. But the port to the ARM platform has been stalled. It would be nice to have a Pi port.
I dunno. Just because he can do a split and punch you in the balls, does that really make him such a great musician?
Their versions from 3 onwards were pretty different from 1 & 2. About as different as Windows 95 and Windows NT. 3 onwards was a microkernel OS. While the versions 1 & 2 were indeed small kernels, Tanenbaum never claimed that they were microkernels, the way he did about Amoeba, or later, Minix 3.x.
I don't have any opinion on the state of Minix today but I'm certainly grateful to Andrew for his books and the knowledge he's passed to us.
50 million elvis fans....
...can't be wrong. (I'd forgotten that adage.)
I like Linux. I use it. I'm using Lubuntu at the moment but that changes kind of often. Hmm... I have been playing with BSD via a virtual machine lately. I don't know a whole lot about the kernel. I wonder if it's more or less monolithic. Of course, well, the BDSs don't really have a single kernel, I don't believe. I think they're all based on the same kernel from back in the mid 90s.
Wikipedia classifies it (the BSD) as a monolithic kernel. When reading about the microkernel there's a wiki page on Tanenbaum's debate with Torvalds. It's interesting reading though I'm still trying to grasp what makes a nanokernel different and what, exactly, makes one qualify for hybrid status. Ah well... It's good to learn new things.
"So long and thanks for all the fish."
Indeed. Thanks! The rump server and the protocol makes it look like it might be trivially easy (comparatively speaking) to speak with devices - not just as a server. Some of the BSDs have great documentation. I don't have a NetBSD VM to spin up but I have the resources to do it. I can just connect to a machine at home and spin it up there. Strange, I know, but that's the situation.
I am definitely going to have to give that a spin. I don't have a lot of time today but I can get a VM spun up, tested, and make sure it's running well enough. I use VMWare so interacting with the USB would be easy - if I were home to actually plug a USB device in. I've got a spare laptop with me that has VMWare so maybe I'll spin it up there.
My only real foray into BSD-land has been with a small test of FreeBSD and then an extended test of GhostBSD. I've been really enjoying GhostBSD in a VM - it's been shut down once and I go hammer on it once in a while to see how stable it's going to be. I want to move it to bare metal but I really dislike Firefox and there's an exceptionally small number of choices in that regards. Sadly, it is only my love of Opera that keeps me away from putting it on bare metal on one machine and just sitting down to learn more about it over a few month period.
I know Opera devs share BSD source files internally and do some work on it. Maybe I can get them to build me one if I ask nicely? Probably not but, I guess, I can try.
"So long and thanks for all the fish."