“Right now, someone, somewhere is developing the killer operating system feature of the future – a feature that will change computing and make us wonder how we lived without it. However, the person responsible probably isn’t grafting away in the labs of Microsoft, Apple or Red Hat – he or she is more likely to be working in a bedroom or loft. […] We’ll look at the best alternative operating systems, with the potential to change the computing landscape over the next decade.”
Am I the only one who thought that the author of that article didn’t seem to know a lot about what he was writing? Glad to see Haiku come out number one though.
Edited 2011-03-14 09:12 UTC
You’re not the only one.
“DexOS
BeOS clone hobby project ”
Says it all.
Yeh, it was hard to overlook that one! lol
No, you are not the only one.
Still a nice, quick overiew.
Edited 2011-03-14 11:41 UTC
I was gladdened to see mention of the GNU/HURD. Totally new to me was Inferno, not something I would have a personal need for but interesting none the less.
Yup, we may see The Hurd released one day. I mean, Duke Nukem Forever is coming out soon, so anything is possible
You may have recognized the screenshot, though: Plan 9, screen mostly taken up by an acme window. Apparently Inferno can run as a daemon on p9, which makes absolutely no sense since it can do all those things already. o_O I guess it’s for sharing resources with other host platforms.
Inferno isn’t really well known, but it’s basically Plan9.
More like the successor to plan9. Still good to see its name in lights, as it were.
Nice little overview, but no real surprises for regular readers of OSAlert.
The thing that annoys me is this statement:
However, the person responsible probably isn’t grafting away in the labs of Microsoft, Apple or Red Hat – he or she is more likely to be working in a bedroom or loft.
Insinuating that innovation only happens outside big companies seems a little absurd to me, since the big push we are currently seeing on for instance mobile platforms and the web, is primarily being driven big corps. I actually think it likely that this person will soon be an employee of one of these companies, working in their labs.
Edited 2011-03-14 09:33 UTC
I actually think it likely that this person will soon be an employee of one of these companies, working in their labs.
Which validates the point of the author. Innovation starts outside of the big corporations and when deemed feasible is assimilated into professional products.
The thing is though, that many of the alternatives shown were not developed in a bedroom, but by a colaboration of many people, out in the open, sometimes for years and years.
Nice to see an article about alternative OSes, that’s the real reason i come to this site and it’s been very scares for a long time.
its missing key C# OSes – although jNode is nice it doesn’t seems to include kernel-level typing for asm and kernel level driver sandboxing and the like – all which Singularity has, and probably Mosa (ex SharpOS)
While all hobby OSes are fun I think those are the ones with a future, as they are the ones actually _solving_ real issues. Security for one, is a joke in current OSes. Singularity brings a much higher confidence on that point of view.
Finally, HURD while separating drivers is still built upon the ancien model.
The creators of Unix saw some flaws and recreated it, and called it Unix v2.0 – the name? “Plan 9”. Inferno is a variant of Plan 9. I think Inferno is proprietary and Plan 9 is open?
Inferno is OSS – from wikipedia:
The Inferno Business Unit closed after three years, and was sold to Vitanuova. Once Vitanuova owned the rights, it continued development and offered commercial licenses to the complete system, and free downloads and licenses (not GPL compatible) for all the system except the kernel and VM. They ported the software to new hardware and focused on distributed applications. Eventually, Vitanuova released the source under the GPL license and the Inferno operating system is now an Free/Libre/Open Source Software project.
[edit]
I am a little disappointed that Genode did not make the list. Is this because Genode is an operating system framework as opposed to just being an operating system?
I feel the same way about VMS that BeOS fans feel about it — does that make me weird?
yes
but ithink its more to the fact that some of these Oses were so far ahead of their time, we’ll always feel given a chance they would have done so much more
Allow me a comment regarding number 8, FreeVMS.
The article states: Back in the ’70s and ’80s, the main competitor to Unix on big-iron hardware was VMS.
The term “big iron” traditionally refers to mainframes (often in relation to IBM) and supercomputers. The systems that ran VMS usually were called “small computers” (to illustrate that they are not mainframes, and to draw a line to midrange systems. Those “big iron” either ran Unix or, as one can easily conclude, one or more of IBM’s mainframe operating systems.
Famed for its stability, running on chunky fridge-like boxes called VAXes, it included advanced clustering and security features for its time, along with an automatic versioning file system.
The plural of VAX – also traditionally grown – is VAXen. Or VAXera, if you come from a scientific background.
And as I’m already in the mood, let me continue with number 4, OpenBSD.
Security before style
Regarding style of code and documentation – the quality developers are interested in, OpenBSD is first class, it doesn’t have to hide anything “behind security”.
I’ve got nothing more to add at the moment.
The list basically ignores research OSs, while focusing mainly on projects trying to reverse-engineer decades-old operating systems (Unix variants, Windows, BeOS and the Amiga OS). As such, it should have mentioned also Open RISC OS, which is in a similar category. Inferno seems to be the only example looking forwards rather than backwards.
But there are lots of research operating systems that are more future-oriented. These, for example, use various forms of code verification to formally guarantee safety (rather than just rewriting a few library routines while still using unverified code in an inherently unsafe programming language), use functional programming from kernel-level and up, use message passing without shared memory to ensure both safety and distribution or seamlessly integrate GPGPUs and CPUs to share a workload.
Such are more likely to form the basis of tomorrows OS than those in the list.
I think jnode qualifies as future-looking and not just a clone of an older system.
So of the top 10 of Operating Systems that will bring “the killer operating system feature of the future”, there are lots of recreations of old (and sometimes not even succesful) Os’s and a couple of concepts that have been “showing great promise” for decades now.
and at first the article is about the killer operating system FEATURE, but it actually seems to be more about the entire OS than one “feature that will change computing and make us wonder how we lived without it”.
A nice summary of alternative Operating Systems but nowhere does it mention that killer feature or that one whizkid that is coding it from his bedroom
Meanwhile I am seeing big companies develop and innovate at a crazy pace in the mobile world and I see “autorecovery / selfhealing” being added to the next version of Windows. THAT is what I call I killerfeature (assuming it will work at least half as good as advertised)
Another killerfeature is the linking between the HP devices we have recently seen appear with WebOS
And a phone that you can turn into your desktop computer by docking it might just be a variation of a notebook+dockingstation but it IS a killerfeature in my book
(I personally consider plug-and-play the best killerfeature ever and the big Operating Systems had that first as well)
ooh.. I remember that being advertised as well… For windows 98 SE. Obviously “it will be better this time”, etc, etc. Colour me jaded by repeated failed promises. I wish the guy that installed every version of windows from 1 to win 7 had listed the installation promises that scroll by as it installs. They’re pretty much all the same: “Faster Performance”, “better hardware support”, “More secure”, “More stable”, “More multimedia capabilities”, “Integrated into the web”.
I mean they have improved all of those areas ( jury is still out about the continuous performance increase, security was joke for a long time. ), but its just funny seeing them pop up over and over.
Couldn’t they just have a stripped down browser and let me surf the web in that space instead of forcing me to read their own marketing BS? Inn fact all Operating systems should do this, I don’t even care if its just lynx. Just let me do something with the computer while I wait for the install.
Great idea
Maybe I’ll add an enhancement ticket to let Haiku’s installer launch one or two of the included games while the user is waiting for installation to completely.
Technically, these days most Linux installers can run from a LiveCD environment – so you should be able to do what you desire anyway, no?
Nice! I’d love to see it. Games work too.
Some Linux distros have supported this. I remember SuSE, for example, let you bring up game of solitaire.
These days a web browser is a better idea, if the network has been brought up.
Edited 2011-03-15 10:20 UTC
Not as far as I recall. The Acorn Archimedes had plu-and-play already in 1987 when it came out. And it was a more real plug-and-play than you see on, say, Windows.
Windows implements plug-and-play by including drivers for all known devices in the OS (so you don’t get p&p for new devices, unless these can reuse older drivers). Acorn implemented p&p by requiring devices to carry their own drivers, which they would upload the the computer when they were plugged in. Hence, the OS did not need to include any drivers and would be prepared for any future products.
On the downside, this made devices more expensive, as a device would need to contain a small ROM with its driver and logic for allowing the computer to read it. For a mass-market OS this wouldn’t matter much, as device manufacturers would readily do this, but since RISC OS was a niche OS, it meant that the selection of devices was small. Some third-party companies made such devices by adding the extra parts to standard PC devices, but that could easily double the price.
Still, I think this was the right way to do p&p before the Internet became ubiquitous. Now a better way would probably be for the device to contain an IP-address from where the driver can be automatically downloaded. This will allow the driver to be more easily updated.
If you were active in the PC world in the late 1990s, you may remember BeOS. Designed for the PowerPC, it was ported to the x86 PC architecture, offering a unique experience that was designed from the ground-up for desktop computing.
I think BeOS was originally developed for the Hobbit CPU, then ported to PPC and later to x86.
Just to answer some questions and correct a few misconceptions I saw earlier: both Plan 9 and Inferno are FOSS and both have some Unix-y characteristics. Inferno essentially started out as a port of an early version of Plan 9 to a Java-like VM system, but they soon diverged, with Inferno putting more focus into potential use in embedded and hosted systems with things like a proper GUI stack for creating user-oriented interfaces with and a robust portable host-based VM to allow for use on top of other systems, while Plan 9 continued as a sort of “ultimate expression of Unix ideals”. Both were open sourced in the early 2000s when it became clear that they had little commercial potential and now serve a useful function as a place for nerds to play around with cool OS ideas that will one day rock the world but whose time has not yet come.
This might be the first and last article that lumped HURD and OpenBSD in the same category. Catfight!
I was half hoping to hear that SkyOS was back in development and listed as number 5 or something… But then I visited http://www.skyOS.org and found the website has been parked. I will cry now, as I never got the opportunity to try it.
What the hell is doing OpenBSD listed as an alternative operating system?
That’s understandable as they said they were interested in anything other than WinMacLinux.
Lol, I wanted to listen to my old track “infinity” in sawteeth, from the time I used/tried BeOS, but Haiku just gave me KDL.
Well, everyone’s a critic it seems, even the Haiku OS ;D
It was a major smash among the 4 BeOS users in its day!
http://haikuware.com/directory/view-details/multimedia/audio/synthe… (demosong).
The future may be a kernel that recognise all hardware. Once that’s achieved, users will choose a desktop and applications to build on top of it.
This is more probable to happen than the Apple solution of closed hardware and related software. Unless Apple is going to take the world on its pocket, which I don’t believe.
Simple fix : revert the current one-driver-per-hardware trend and make all hardware use standard interfaces, things like VESA VBE or PS/2 keyboard/mices but designed a bit better and in a more evolutive fashion. OS hardware compatibility problems and various reliability issues associated to multiple drivers are definitely solved.
Edited 2011-03-15 16:58 UTC
“OpenBSD was the first non-research OS to integrate many features we now take for granted, including address space layout randomisation, which puts libraries and memory sections in random locations in RAM, so crackers can’t assume their location.”
OpenBSD is an “alternative” OS? Isn’t that statement really making OpenBSD null as a selection for this list? OpenBSD shouldn’t be on this list.
DexOS looked interesting at first, until I thought “Why would I buy a seperate computer when I already have this console?”
This list is fun for nerds looking for something to play with, but I don’t consider it to be a glimpse into the future. More like hanging on to the past.
I agree OpenBSD should not be on this list … what is alternative about it? By that same token you could put Solaris on here. I also wonder about the choice of HURD for other reasons, but hey, at least it was #10.
Windows Azure is a service that allows developers to run applications and store data on servers owned and operated by Microsoft. Its simplicity is that it utilizes your existing skills in familiar languages such as .NET, Java and PHP to create and manage web applications and services.
http://www.whatisguide.net/0314-what-is-azure.html
How the fuck is this relevant to the news item?