Ubuntu on Windows [LWN.net]

Ubuntu on Windows

Posted Mar 30, 2016 17:46 UTC (Wed) by rsidd (subscriber, #2582) [Link] (25 responses)

This is insanely neat and exactly what MS needed to attract the geeks who moved over to OS X over the past 15 years.

Now, what has kept me on Linux (Ubuntu) and away from OS X, apart from the proprietariness of the latter, is my desire for a tweakable interface and, these days, a tiling wm in particular. If Windows can let me run i3 and the Ubuntu userland, I'd be tempted to jump in. (But the other thing I'm unwilling to abandon on Ubuntu is zfs. What chance that will appear on Windows too?)

Ubuntu on Windows

Posted Mar 30, 2016 17:50 UTC (Wed) by kokada (guest, #92849) [Link] (1 responses)

Windows already have a next-gen filesystem, ReFS (https://en.wikipedia.org/wiki/ReFS). However, you can't use it as a boot disk yet, at least in desktop versions of Windows.

Ubuntu on Windows

Posted Mar 30, 2016 17:57 UTC (Wed) by rsidd (subscriber, #2582) [Link]

Interesting, but sounds like it's not production ready. Nor is Ubuntu-on-Windows, of course.

Ubuntu on Windows

Posted Mar 30, 2016 18:42 UTC (Wed) by einstein (subscriber, #2052) [Link] (20 responses)

This doesn't appear to be, ultimately, anything more than a re-hash of the old "MKS utilities for windows", the microsoft "unix services" or the "posix" layer in nt - in other words, not any sort of real, usable Linux environment. To get an idea of how much spin there is here, the same article repeats the marketing hype about linux containers running natively on windows - which is, of course, nonsense, as lxd is a function of the Linux kernel. The whole "lxd on windows" mirage really only amounts to running lxd on Linux, inside a VM on a windows pc.

So take everything here with a grain of salt. Technical details, not marketing speak, will tell the tale.

Ubuntu on Windows

Posted Mar 30, 2016 18:50 UTC (Wed) by rsidd (subscriber, #2582) [Link] (19 responses)

I'm not sure what article you read but the one linked here makes explicit that it's not a container on Windows. Nor is it a recompilation of Linux binaries. It is the unmodified Ubuntu binaries running on Windows, via an interface in the Windows kernel that maps linux syscalls to windows syscalls and with access to the unmodified Ubuntu userland (libraries etc). The closest analogue I can think of is the linuxulator on FreeBSD. But FreeBSD is much more like Linux than Windows is, so this is much more impressive. The most impressive part being that it's largely done in-house at Microsoft.

Ubuntu on Windows

Posted Mar 30, 2016 19:11 UTC (Wed) by ssmith32 (subscriber, #72404) [Link] (6 responses)

So the inverse of wine? (But in kernel space)

Trying to think of a catchy acronym...

wile ?

(Windows is not a linux emulator?)

I used my windows wiles to run wily werewolf on windows?

Ubuntu on Windows

Posted Mar 31, 2016 0:00 UTC (Thu) by asaz989 (guest, #67798) [Link] (4 responses)

Sounds like it.

And if I may digress, the fact that it's possible to write Wine is user space is a credit to the Windows ecosystem. They clearly thought long and hard about backwards compatibility before putting together the win32 API, and made userspace programs go through a shared library rather than directly making syscalls. The intention was to make it easier for Microsoft itself to swap out the kernel (which they have done), but the ability for third parties to swap in *nix has been a happy side effect.

Ubuntu on Windows

Posted Mar 31, 2016 5:20 UTC (Thu) by roc (subscriber, #30627) [Link] (3 responses)

WINE would have been just as easy to write, or easier, if Windows offered stable syscall APIs. Linux already supports a variety of syscall personalities and WINE would have implemented another one.

Not having a documented, stable syscall interface means that many tools (e.g. strace, rr) are difficult or impossible to support on Windows.

Ubuntu on Windows

Posted Mar 31, 2016 5:27 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link] (2 responses)

> WINE would have been just as easy to write, or easier, if Windows offered stable syscall APIs.
It does. It's called "native API" and is fairly compact and is extremely stable.

There were projects to add native Windows "syscalls" to Linux: https://en.wikipedia.org/wiki/Longene - not many people are interested in it, though.

> Not having a documented, stable syscall interface means that many tools (e.g. strace, rr) are difficult or impossible to support on Windows.
Way back before Docker, brave people from Parallels wrote container support for Windows: https://virtuozzo.com/support/pcw/ - without getting access to Windows source code, btw.

They did it by providing a thin layer above the native API and running multiple userspaces on top of it.

Windows kernel is pretty neat, although it's showing its age.

Ubuntu on Windows

Posted Mar 31, 2016 18:40 UTC (Thu) by xilun (guest, #50638) [Link] (1 responses)

I don't think the NT API is particularly stable. At least not at binary level http://j00ru.vexillium.org/ntapi_64/ and probably also not at source level.

Ubuntu on Windows

Posted Mar 31, 2016 19:26 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]

It is stable. Windows still supports several implementations of the USB stack, just because some drivers depend on a 17-year-old first stab at USB support.

Things do break from time to time, but MS actually cares about old drivers and system software.

It's much easier in Windows, as it's built on top of a "message passing". Almost all operations involve sending a uniformly formatted "message" (IRP) which can pass through multiple layers that can filter and/or modify it. So as long as the message format is preserved, keeping the compatibility is doable (not simple, but doable).

Linux kernel API is completely ad-hoc so it simply can't achieve the same stability level. The flip side is much better performance and maintainability.

Ubuntu on Windows

Posted Mar 31, 2016 21:03 UTC (Thu) by Seegras (guest, #20463) [Link]

Yes, this is totally Wile. E. Coyote.

Ubuntu on Windows

Posted Mar 30, 2016 19:30 UTC (Wed) by adler187 (guest, #80400) [Link]

We have a thing like this on IBM i, where the IBM i kernel (SLIC) provides (a subset of) AIX system calls for applications to use. We call it PASE (for Portable Application Solutions Environment, previously Personal Address Space Enablement). Many times I wish we'd have implemented the Linux syscall interface instead, since it would make porting OSS to IBM i much easier. I'm sure if we were doing it now instead of 15+ years ago it would be based on Linux.

https://www.ibm.com/support/knowledgecenter/ssw_ibm_i_72/...

Ubuntu on Windows

Posted Mar 30, 2016 19:41 UTC (Wed) by SEJeff (guest, #51588) [Link]

Or perhaps binfmt_misc on Linux?

https://en.wikipedia.org/wiki/Binfmt_misc

https://www.kernel.org/doc/Documentation/binfmt_misc.txt

ie: PE :D

Ubuntu on Windows

Posted Mar 30, 2016 20:21 UTC (Wed) by einstein (subscriber, #2052) [Link] (9 responses)

> I'm not sure what article you read but the one linked here makes explicit that it's not a container on Windows.

Right, we know this is not linux containers on windows, which doesn't exist anyway. Pardon the reference, I had been looking at SJVN's article on the announcement, and in that article he also mentioned the lxd thing. Let's be clear, nobody has said this has anything to do with containers, only that the same level of breathless hype appears in the two claims.

So, to recap, it's not really any different from what has been trotted out several times before. It might have a temporary niche, as a curiosity, but beyond that, it seems merely to be a solution searching for a problem. Linux users sometimes have a need to run some legacy windows app - but do windows users have some need to run some Linux app that's not available on windows? Just not feeling it.

Ubuntu on Windows

Posted Mar 30, 2016 20:35 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link]

> Right, we know this is not linux containers on windows, which doesn't exist anyway.
You're saying?...

http://www.colinux.org/

Ubuntu on Windows

Posted Mar 30, 2016 21:01 UTC (Wed) by rahulsundaram (subscriber, #21946) [Link] (5 responses)

> but do windows users have some need to run some Linux app that's not available on windows? Just not feeling it.

Yes they do but you are probably not "feeling it" because you are not a Windows user.

Ubuntu on Windows

Posted Mar 30, 2016 22:29 UTC (Wed) by donbarry (guest, #10485) [Link] (4 responses)

WINE was developed to help bridge users from the proprietary Windows OS as they converted to free software.

This is intended exactly the reverse way, to leverage the free software community to promote a proprietary OS.

That Ubuntu would sponsor this is simply more revelation that their agenda is not friendly towards free software.

And those who see this in purely opportunistic terms: that's a consequence of the "open source" movement and its "pragmatism" -- which is really another way of saying its adaptation to corporate interests.

Ubuntu on Windows

Posted Mar 30, 2016 22:51 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link] (3 responses)

After such posts, I'm starting to think that _anything_ that is "bad for free software" in the end is actually a really great stuff for everybody (except Stallman).

Ubuntu on Windows

Posted Mar 31, 2016 4:48 UTC (Thu) by floriansnow (guest, #107824) [Link] (2 responses)

Yes, great for everybody who does not care about computing freedom. I think the technical achievement here is neat, but all it does is let me run more Free Software on a non-free OS. And that additional Free Software is probably not all that useful to people who aren't familiar with it already, so people who could also run a completely free system. So if this is more than just a tech-demo, then it is a way to lure people back to non-free software.

Ubuntu on Windows

Posted Mar 31, 2016 5:44 UTC (Thu) by jspaleta (subscriber, #50639) [Link]

Uhm...
There's a large collection of workers in enterprise environments who are using computing resources dictated by corporate it policy.

Any technology advancement which makes it easier for them to use free software to do their job, even if they are forced by corporate policy to run a mix of free and non-free software, is a win...

Its called progress.

Hats off to MS and Canonical for getting things this far. It's an interesting technical achievement, with a lot of potential.

The work is far from complete from the sounds of it, even for getting a usable strictly command-line linux developer's environment working. But there is obvious potential here to make it easier for people to pick up traditional linux user-space and make use of it even if they are forced to use windows (or choose to use windows) compared to having to run a VM or container.

I skeptical of the claim that Kirkland made that "most" of the ubuntu repository packages are going to work... but i give him the benefit of the doubt... its the first public demo of this..and if MS continue to put engineer resources into this then the claim might become reality. The only way to really know is to test packages and see what works and what doesn't.

My biggest concern is if MS takes the effort just far enough to serve their own purposes..but not far enough to serve actual use cases very well. I'd hate to see users asking Ubuntu and Debian packagers to service bugs specific to the windows 10 specific implementation of the linux ABI. At the end of the day, a lot of the magic here comes down to MS willing to spend the engineering resources to service the linux compatibility layer because they are the only ones who can do the work. The lack of open development, with a clear contribution friendly development model might keep this from being able to jump from interesting potential to real solution for real users. If people can't reasonably expect to be able to use a random ubuntu package successfully.. then developers, the stated target group, are just going to use a VM as a more reliable alternative, even if its a performance hit.

Ubuntu on Windows

Posted Mar 31, 2016 6:51 UTC (Thu) by ncm (guest, #165) [Link]

This is Microsoft's response to the decreasing fraction of competent developers willing to shackle their careers to the Windows ecosystem. Not long ago it was considered perfectly sensible to lock yourself into MS-only languages, libraries, and tools, but it's finally recognized as extremely short-sighted. Offering the ability to work like a normal developer helps to keep them closer to competitive.

Part of this is that developers who still are locked into MS are competing with each other in a shrinking market at cut-rate prices. So, there are still plenty of them, but they are the ones who work for less because they must.

Ubuntu on Windows

Posted Mar 31, 2016 0:19 UTC (Thu) by aryonoco (guest, #55563) [Link]

If you think the ability to run Linux binaries unmodified on Windows, for Windows to provide Linux API compatibility, is "the same as has been trotted before" and "only a niche", you're not living in the real world.

This is a colossal improvement over Cygwin and everything else that's been available for Windows. It's even an improvement over OS X, I don't have to do gymnastics to replace all the ancient BSD utilities that Apple ships with OS X with their GNU counterparts.

And those of us who teach/train people in Python/Ruby/Go don't have to fret anymore when someone shows up with a Windows laptop.

Ubuntu on Windows

Posted Mar 31, 2016 12:03 UTC (Thu) by nye (subscriber, #51576) [Link]

>Linux users sometimes have a need to run some legacy windows app - but do windows users have some need to run some Linux app that's not available on windows?

Fuck yes. If this is even half as good as it sounds, this is the best product announcement I've heard from *anybody* in *years*.

Ubuntu on Windows

Posted Mar 31, 2016 10:29 UTC (Thu) by khim (subscriber, #9252) [Link] (1 responses)

I still fail to understand why people never try to apply logic to announces. Microsoft have failed - and now it's time to change the strategy. By now it's obvious that Linux will not die any time soon. It's future on the servers it's secure (heck, HPC is almost entirely Linux nowadays) and it's winning on mobile.

These successes mean that Linux desktop could survive on coattails of server/mobile development. What Microsoft needs to do to protect itself? Well, the answer is obvious: make sure developers could develop all these mobile and HPC applications on Window. And it's also entirely obvious that development of desktop Linux applications should be as hard as possible. That's not hard to achieve, really: just make sue there's no GUI.

Ubuntu on Windows

Posted Mar 31, 2016 12:10 UTC (Thu) by nye (subscriber, #51576) [Link]

>And it's also entirely obvious that development of desktop Linux applications should be as hard as possible. That's not hard to achieve, really: just make sue there's no GUI.

Just use Cygwin's X server via TCP. Performance won't be amazing, but outside of games it's unlikely to matter.

GNU/Windows

Posted Mar 30, 2016 18:02 UTC (Wed) by atai (subscriber, #10977) [Link] (1 responses)

surreal indeed

GNU/Windows

Posted Apr 1, 2016 18:34 UTC (Fri) by smitty_one_each (subscriber, #28989) [Link]

Got to set it to play a measure or two of "Running with the Devil" for the initial bash instance.

Ubuntu on Windows: good to prepare user to migrate to Ubuntu?

Posted Mar 30, 2016 18:17 UTC (Wed) by sgourichon (guest, #88605) [Link] (1 responses)

Okay, that's funny and has some really interesting real-world use cases.

Not for us geeks. Any time I have the choice of OS, I see no point of running Windows over Debian/Ubuntu, not least because running Windows means losing the little bit of control about what the machine does, that we have when running a free OS.

The use case is the reverse: this announces a world where you can develop for Linux, and reach *Windows users that have not switched yet* more easily than with previous solutions (cygwin, mingw, etc.). Windows users get familiar with free software ecosystem and can switch more easily.

Ubuntu on Windows: good to prepare user to migrate to Ubuntu?

Posted Mar 30, 2016 18:22 UTC (Wed) by sgourichon (guest, #88605) [Link]

Seen at the end of the article: "And what a wonderful vector into learning and using more Ubuntu and Linux in public clouds like Azure."

Okay, so what Microsoft wants is reduce the number of Linux instances on the cloud to favor Windows. That's the real motivation for Microsoft.

So, perhaps Microsoft will make that available on their cloud only, not on desktop Windows... That reminds me of the old SFU which was seldom used. https://en.wikipedia.org/wiki/Windows_Services_for_UNIX

Ubuntu on Windows

Posted Mar 30, 2016 18:39 UTC (Wed) by dfsmith (guest, #20302) [Link]

I've been running the Cygwin bash (and their excellent terminal window) as my Win7 command line for a while. The most serious deficiency with Cygwin seems to be device performance. Dumping data to tape drive maxes out at about 10MB/s, where it should be 150+MB/s. Hopefully this new version will perform properly in /dev!

Ubuntu on Windows

Posted Mar 30, 2016 19:42 UTC (Wed) by Zizzle (guest, #67739) [Link] (9 responses)

I seem to remember some stories saying the reason OS/2 failed was that it did Windows emulation.

No sane developer would want to write 2 different versions of their application, so write to Win32 and run on either windows or OS/2. Windows captured all the developer mind share.

I wonder if the same argument can be made now with Linux.

Ubuntu on Windows

Posted Mar 31, 2016 1:36 UTC (Thu) by jeleinweber (subscriber, #8326) [Link]

I went to one of IBM's dog-and-pony shows when OS/2 warp was released, and walked out knowing it was doomed. TCP/IP was still an extra cost add-on at just the time that the Internet was taking off commercially, developer tools were expensive and they had no outreach program to attract new ones, and they didn't have compatibility with windows 95 apps. So clearly there was going to be no app ecosystem and no pool of consumer users to attract developers, meaning it was going to wither in decreasingly small commercial verticals. Microsoft made much smarter management and marketing decisions even if you neglected their predatory monopolist tactics.

Ubuntu on Windows

Posted Mar 31, 2016 10:40 UTC (Thu) by khim (subscriber, #9252) [Link] (7 responses)

Microsoft understands that problem very well, I think. That's why there's no support for GUI (and I'm pretty sure there are no such plans, either). Just how many important popular Windows non-GUI tools you really know?

Ubuntu on Windows

Posted Mar 31, 2016 11:38 UTC (Thu) by dufkaf (guest, #10358) [Link] (6 responses)

umm, you can install Xming http://www.straightrunning.com/xmingnotes/ (or Cygwin/X http://x.cygwin.com) and X apps should work

Ubuntu on Windows

Posted Mar 31, 2016 13:01 UTC (Thu) by jengelh (guest, #33263) [Link]

And then run a x11-backend weston, providing Wayland on Windows...... oooh the circle closes! :-D

Ubuntu on Windows

Posted Mar 31, 2016 13:26 UTC (Thu) by khim (subscriber, #9252) [Link] (4 responses)

Sure, but would it work good enough to make development of "native" Windows app superfluous? I seriously doubt it.

Microsoft's goal is not to make it impossible to run Linux apps on Windows (heck, you can do that today with many virtual environments), but to make it unfeasible.

Ubuntu on Windows

Posted Mar 31, 2016 21:37 UTC (Thu) by clump (subscriber, #27801) [Link] (3 responses)

Microsoft's goal is not to make it impossible to run Linux apps on Windows (heck, you can do that today with many virtual environments), but to make it unfeasible.

Absolutely not. Microsoft's strategy is Azure, where revenue is based on consumption, not on preferred operating systems. Given that so much of "the cloud" is Linux/open source, Microsoft has a bit of an ecosystem issue.

Ubuntu on Windows

Posted Apr 1, 2016 17:56 UTC (Fri) by khim (subscriber, #9252) [Link] (2 responses)

Microsoft still makes lots of money on desktop and it does not look like there are any threat of them losing the desktop (except they would do something totally stupid).

I'm not really sure if Azure even supports GUI Linux applications, but even if so these are so tiny that revenue from them does not matter - but they could threaten desktop.

Thus they need to make sure server apps are well-supported and run well, while desktop apps don't work well. I doubt they would spend too much time crippling them, though: this would happen automatically as a consequence of doing nothing.

Ubuntu on Windows

Posted Apr 1, 2016 19:51 UTC (Fri) by clump (subscriber, #27801) [Link] (1 responses)

Microsoft still makes lots of money on desktop and it does not look like there are any threat of them losing the desktop (except they would do something totally stupid).

No disagreement.

Thus they need to make sure server apps are well-supported and run well, while desktop apps don't work well. I doubt they would spend too much time crippling them, though: this would happen automatically as a consequence of doing nothing.

This doesn't make sense. As you've said, Linux isn't a threat to Microsoft on the desktop. "Cloud" is a different market, a growing market, unlike the desktop market. Microsoft doesn't care at all whether you're using Linux or not in Azure because their revenue is based on consumption.

The desktop is in retreat and Microsoft is betting heavily on the cloud, where they don't care what you're running so long as you're there.

Ubuntu on Windows

Posted Apr 1, 2016 22:25 UTC (Fri) by khim (subscriber, #9252) [Link]

As you've said, Linux isn't a threat to Microsoft on the desktop.

Except they would do something totally stupid: supporting Linux GUI would be "totally stupid". MacOS was also "not a threat". Then Microsoft saved Apple - and Apple ditched Microsoft what it become possible to do. Microsoft wouldn't repeat the same mistake twice, would it?

"Cloud" is a different market, a growing market, unlike the desktop market.

It may be different market, but it still uses the same OSes: Windows and GNU/Linux.

The desktop is in retreat and Microsoft is betting heavily on the cloud, where they don't care what you're running so long as you're there.

Right - as long as are in the cloud Microsoft does not care. But who in their right mind would use Mir or Wayland in the cloud? Yes, some people may try to use it - but why? To bring their apps to desktop, of course!

In the end there are very small upside from supporting of Linux GUI and desktop environments (if any) - and very large downside. Just why would Microsoft do that?

Ubuntu on Windows

Posted Mar 30, 2016 21:07 UTC (Wed) by Karellen (subscriber, #67644) [Link] (3 responses)

As mentioned elsewhere, are we sure this wasn't accidentally posted a couple of days early?

Still, I wonder - the article mentions that Windows has effectively re-implemented the Linux syscall interface. Does anyone know if that means that they're also running the stock Ubuntu glibc? Actually, thinking about it, the syscall ABI is probably easier to reimplement than the outward-facing libc ABI, or than messing around with some hybrid of the two.

Second, the Ubuntu filesystem appears to be installed per-user?!? Does that seem odd to anyone? Although... the screenshots make it look like the Ubuntu "system" is running as root. Does anyone know if the Linux subsystem supports users? Or does everything think it's running as Linux UID 0, even when running as a non-privileged Windows user? (Are the Ubuntu packages OK with that? I thought some packages would probably have postinst errors if they couldn't set permissions to some non-root UID) I guess some kind of PAM module would be the best way of creating a better mapping between Linux subsystem users and real Windows accounts? Could you even get sudo working if that were the case?

Ubuntu on Windows

Posted Mar 31, 2016 3:12 UTC (Thu) by jdub (guest, #27) [Link] (2 responses)

Yes, it's the stock Ubuntu user land, including glibc. Exactly the same bits you'd expect to find after installing Ubuntu, or booting a VM, minus the kernel and related utilities.

Yes, it's an implementation of the Linux kernel ABI (syscalls, special file systems, etc).

Yes, the rootfs is installed per-user, under AppData. The combination of all of the above means it's not just restricted to Ubuntu, and different users can run different flavours. Seems there can only be one flavour per user, though.

Yes, in the first public beta release, everything runs as root, but that will be fixed soon. Ultimately "root" on WinLS has the same permissions as the executing NT user, so it doesn't mean very much (but the safety and semantics of uid awareness will be good).

Based on what I've seen so far (particularly the per-user rootfs), I doubt there'll be any mapping between different Windows users within the Linux subsystem. It's theoretically doable, but way too much to bite off to start with.

Ubuntu on Windows

Posted Mar 31, 2016 4:57 UTC (Thu) by joey (guest, #328) [Link] (1 responses)

Hey jdub, do you know if this allows a Linux process to exec() a .EXE?

And does it support inotify?

These would be my likely pain points for dropping git-annex's windows port in favor of this. Fingers crossed.

Ubuntu on Windows

Posted Mar 31, 2016 7:15 UTC (Thu) by jdub (guest, #27) [Link]

I don't have any inside info, but I suspect "no" on the former (but not outside the realms of possibility), and "not yet" on the latter (but I reckon *notify will land because lots of dev tools use it to watch for code changes).

Ubuntu on Windows

Posted Mar 30, 2016 21:54 UTC (Wed) by callegar (guest, #16148) [Link] (11 responses)

If Qt and GTK applications from Ubuntu run fine and the few non-Qt, non-GTK applications can work with a little Xserver in Windows, then eventually there is a complete linux userspace with working (though proprietary) drivers for all possible hardware. With a minimum Windows 10 core stripped of all the bloat, a working package manager and all the linux software available for free, Windows may become a more desirable Linux than Linux in many environments. This may be particularly true if the emulation layer ends up working also with other userspaces than Ubuntu's (which is likely) and someone starts shipping Linux userspace distros specifically tuned to integrate well on the Windows desktop. If this is successful, Linux applications may soon start incorporating features specifically designed for the Windows host.

Ubuntu on Windows

Posted Mar 30, 2016 22:53 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link] (2 responses)

Countdown to Debian/kWindows:
3...
2...
1...

Ok, joking.

Ubuntu on Windows

Posted Mar 31, 2016 2:30 UTC (Thu) by TRS-80 (guest, #1804) [Link] (1 responses)

People did try in the past.

Ubuntu on Windows

Posted Apr 18, 2016 5:48 UTC (Mon) by TRS-80 (guest, #1804) [Link]

The link I was actually looking for: http://www.debian-interix.net/

Ubuntu on Windows

Posted Mar 30, 2016 22:56 UTC (Wed) by larryr (guest, #4030) [Link] (7 responses)

Same. When I can use this on a laptop and not be able to easily tell it's running Windows underneath except for some keyboard shortcuts that take me out of the fullscreen X server (using mesa or equivalent) window, that will be great.

Basically like running in VirtualBox without running VirtualBox.

Ubuntu on Windows

Posted Mar 31, 2016 8:14 UTC (Thu) by callegar (guest, #16148) [Link] (6 responses)

I guess that for most things you might not even need the X server. Modern Linux applications tend to be Qt or GTK and both have a modularized structure with backends ready available for Windows. It may end up being the same as supporting them in Wayland or Mir. If graphics works and performance is decent, I really see this as replacing standard deployments of Linux on laptops and particularly on convertible/hybrid laptops that are now a hot thing and are a pain to use in linux proper due to lack of drivers. If this is the case, even those niches where the linux kernel and linux desktop environments are on the desktop/laptop are going away. In fact this may eventually end up affecting linux more on the desktop space than on servers where the Linux kernel maintains some advantage in terms of performance and flexibility. Ironically, the first things that may struggle because of this are Mir, the Unity desktop and the Ubuntu tablet effort.

Ubuntu on Windows

Posted Mar 31, 2016 10:48 UTC (Thu) by khim (subscriber, #9252) [Link] (5 responses)

The whole thing is ridiculous. Just WHY would Microsoft make something like this? Give them SOME credit, will you? Of course they thought about that - and taken counter-measures. As I wrote before.

Plan A have failed spectacularly. Time for plan B: separate desktop Linux from server Linux and mobile Linux - and watch it weather on the vine. Support for GTK/QT apps would interfere with that plan and thus, obviously, should be made as hard and glitchy as possible.

Ubuntu on Windows

Posted Mar 31, 2016 10:53 UTC (Thu) by khim (subscriber, #9252) [Link] (4 responses)

Oops. Commented the wrong message. So you essentially agree with me, just don't go far enough. We are thinking along the same lines only you think that Linux desktop may be an accidental casualty while I think that it's prime target of that move. Not a big difference, really: to hurt Linux desktop with that plan Microsoft does not need to do anything obviously nefarious. It just need to NOT do some suicidal moves. REALLY hard to argue about that in any kind of court. Well played, Satya!

Ubuntu on Windows

Posted Mar 31, 2016 21:32 UTC (Thu) by raven667 (subscriber, #5198) [Link] (3 responses)

I doubt they are doing this because they are terrified of the Linux desktop and are trying to harm it, they are doing it because MS is trying to re-join the rest of the tech industry so as to not be left behind in a little Win32 backwater. Clearly based on the number and scope of work that's been done over the last year or two, there is a HUGE amount of pent up demand within MS and with their customers for transparency and compatibility, look at the MS Github for C# or MSSQL for Linux, that Ballmer must have been suppressing while Nadella is allowing the MS developers to run with.

Ubuntu on Windows

Posted Apr 1, 2016 18:34 UTC (Fri) by callegar (guest, #16148) [Link] (2 responses)

Obviously, you are right from the point of view of MS. I was looking at it from the point of view of Linux distros. I think that at the same time "MS is trying to re-join the rest of the tech industry", they may end destroying even those niche spaces where Linux on the desktop has succeeded so far, by providing a good *nix experience on top of Win10. To MS it may change almost nothing (+1 % of desktop installations). To Linux distros it may change a lot (e.g., -80% of desktop installations, including a large amount of laptops, and all convertible tablets and standard tablets - i.e., all machines where Linux requires some effort due to hardware drivers). If you consider that the largest cost for distros is to have the desktop stuff working and nicely integrated, this may end up affecting distros quite significantly.

Ubuntu on Windows

Posted Apr 1, 2016 19:09 UTC (Fri) by raven667 (subscriber, #5198) [Link] (1 responses)

I dunno, the critical mass needed for sustainable development of the Linux desktop is pretty small, and something already achieved. I don't think the XFCE or GNOME developers are just going to quit because of anything MS does, otherwise they would have quit long ago. There are plenty of tools within the Linux ecosystem that continue to be maintained and used because there is a critical mass of developers who keep them going, like EMACS. Heck, FVWM still seems to exist and make releases, they have a Github page now.

I think there will still be enough people who want to build desktops that desktop distros and applications will still exist, even if a number of Ubuntu users head off to Windows land.

Ubuntu on Windows

Posted Apr 1, 2016 22:47 UTC (Fri) by khim (subscriber, #9252) [Link]

I dunno, the critical mass needed for sustainable development of the Linux desktop is pretty small, and something already achieved.

Really? Is that why it's constantly in half-broken state?

I don't think the XFCE or GNOME developers are just going to quit because of anything MS does, otherwise they would have quit long ago.

Probably not core developers. But as users will go away so will do contributors, too. And if nobody would care about desktop then sooner of later support for the hardware will disappear, too.

Today Linux desktop rides on coattails of it's Linux-on-server adoption. If people will stop caring about desktop completely... not even for development of stuff for servers... look on FreeBSD (or is it DragonFly BSD? who cares, really). How well desktop apps work there?

This wouldn't happen in one day, of course. Such things take time. But in next 10 or 20 years... who knows?

Ubuntu on Windows

Posted Mar 30, 2016 23:08 UTC (Wed) by dowdle (subscriber, #659) [Link] (8 responses)

Microsoft says this is really just for developers... and not really end users (not even sure if graphical stuff works or not)... nor is it for running Linux server applications.

So, I don't think this is really much more than an novelty especially when compared to Cygwin or MobaXterm.

If you think Microsoft is going to spend a lot of time making it perform well for the Linux desktop and server applications, I'm pretty skeptical.

So, it is going to be released as an App in the Microsoft Store. Hopefully they'll charge money for it and Canonical can *FINALLY* have a profitable revenue stream. :) Of course, Microsoft would get a 30% cut. Maybe Microsoft can sue themselves for enabling Linux (which they claim violates Microsoft's patents) to run on Windows. :)

Ubuntu on Windows

Posted Mar 30, 2016 23:27 UTC (Wed) by juliank (guest, #45896) [Link] (1 responses)

cygwin requires you to recompile apps. this does not. that's a huge advantage. If I guess correctly, it's also implemented in kernel, so there's no performance overhead compared to native applications - cygwin on the other hand is hilariously slow.

Ubuntu on Windows

Posted Mar 31, 2016 0:14 UTC (Thu) by dowdle (subscriber, #659) [Link]

If I use Cygwin's package manager, I don't compile anything. It seems to have a fairly reasonable set of packages... at least for my limited tastes. I realize everything has been recompiled, but so what? So far as performance goes, I haven't really noticed a significant degradation... but then again, I haven't done much other than run some simple tools. Supposedly I/O is a lot slower... but then again, how is Ubuntu's userspace installed for Windows 10? In a directory tree on the host's filesystem? How many issues will that cause? Or are they using a disk image with native ext4 or something?

Again, Microsoft says it is for developers... and not really for users nor running server applications... and in that use case, they are probably similar. Except at this time, Cygwin's terminal emulation might be a tad better.

Ubuntu on Windows

Posted Mar 31, 2016 3:22 UTC (Thu) by jdub (guest, #27) [Link] (3 responses)

This is absolutely 1000% better than Cygwin or MobaXterm. It's an NT subsystem that implements the Linux ABI. It's the Right Way to do it… and we had to wait for Microsoft, because only they can do it this way. (Only one other company has ever built an NT subsystem, and that was because Microsoft goofed up. They didn't mean for it to be doable by anyone else.)

The NT subsystem itself won't be released on the Windows Store. It will ship as a Windows "Feature", which (at least so far) you'll only be able to turn on in Developer Mode.

It's the Ubuntu root filesystem that will be distributed on the Windows Store, and the way all of this is built means other distros ought to be able to release their own flavours. Technically. Whether Microsoft has an exclusive agreement with Canonical is currently unknown.

Ubuntu on Windows

Posted Mar 31, 2016 10:30 UTC (Thu) by tzafrir (subscriber, #11501) [Link]

Any limitations with that store on content under some specific licenses?

Ubuntu on Windows

Posted Mar 31, 2016 15:02 UTC (Thu) by dowdle (subscriber, #659) [Link] (1 responses)

Why is it 1000% better? I think you greatly discount all of the work the Cygwin folks have put into not only porting software but in building packages and maintaining them. I also have to wonder just how much porting effort is/was required. I would imagine that Cygwin provides its own compatibility layer and to assume it is terrible and the new thing is so much superior... is a big assumption. Time will tell.

Ubuntu on Windows

Posted Mar 31, 2016 23:04 UTC (Thu) by jdub (guest, #27) [Link]

I'm not discounting the work of the Cygwin folks when I say that building a Linux ABI compatible NT subsystem (which only Microsoft could do) is a better way to run Linux software on Windows than creating a Win32-based POSIX/GNU compatible build target.

Ubuntu on Windows

Posted Mar 31, 2016 10:02 UTC (Thu) by callegar (guest, #16148) [Link] (1 responses)

Or it may, as a finally efficient Cygwin-lookalike working with standard distributions, end up with Canonical loosing most Ubuntu installations on desktops and laptops as soon as distributions finely tuned to it come out. Could be marketed as "Linux" without the burden of the missing hardware drivers and the buggy desktop environments. For some reasons, it looks to me as a much more credible approach to get a distro for being productive on a tablet or a convertible laptop than the Ubuntu tablet or the various DEs tablet efforts, which in many senses is not good news. For sure, it has already put all my interest for the Ubuntu tablet on hold.

Ubuntu on Windows

Posted Mar 31, 2016 21:28 UTC (Thu) by kmccarty (subscriber, #12085) [Link]

Surely marketing it as being "Linux" would be a trademark infringement? Because the one thing that is actually *missing* from the Ubuntu user-land to be distributed within Windows is the Linux kernel itself!

Ubuntu on Windows

Posted Mar 30, 2016 23:22 UTC (Wed) by kjp (guest, #39639) [Link] (26 responses)

fascinating. would love to see some details on how they emulate for posix compat 1) fork 2) atomic rename (overwriting) of files.
(those are the main windows pain points I've seen in the past).

Ubuntu on Windows

Posted Mar 30, 2016 23:25 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link] (25 responses)

Windows kernel supports fork() just fine. Lots of people re-implemented it many times for end-user apps.

The more interesting problem is locking - Windows does not allow deletion of open files (it DOES support atomic renames).

Ubuntu on Windows

Posted Mar 31, 2016 0:49 UTC (Thu) by aryonoco (guest, #55563) [Link] (2 responses)

For me, the interesting question is how devices will be treated. Can I do dd if=/dev/zero of=/mnt/d for example?

Ubuntu on Windows

Posted Mar 31, 2016 3:23 UTC (Thu) by jdub (guest, #27) [Link]

No support for arbitrary devices thus far, only the ones needed to be ABI compatible. Doesn't sound like they have any desire to wade into that mess, either.

Ubuntu on Windows

Posted Mar 31, 2016 3:57 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]

That's not impossible. Windows actually has a hidden device namespace, they just need to create symlinks into it.

Ubuntu on Windows

Posted Mar 31, 2016 5:24 UTC (Thu) by roc (subscriber, #30627) [Link] (1 responses)

All the fork() reimplementations are bad hacks.

Real COW fork() would be really great and it needs kernel support.

Ubuntu on Windows

Posted Mar 31, 2016 5:31 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]

And Windows has it. You can specify CoW behavior when creating a new process using Zw|NtCreateProcess.

It's used by: http://midipix.org/ for true POSIX compat and by many other projects. E.g.: https://github.com/jonclayden/multicore/blob/master/src/f...

Ubuntu on Windows

Posted Mar 31, 2016 8:08 UTC (Thu) by sourcejedi (guest, #45153) [Link] (5 responses)

Actually even Win32 allows deleting open files, if they were opened with the flag FILE_SHARE_DELETE.

Ubuntu on Windows

Posted Mar 31, 2016 8:13 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link] (4 responses)

Unfortunately, it's not as easy.

FILE_SHARE_DELETE means that the file can be opened with DELETE_ON_CLOSE mode, meaning it'll be deleted once all open handles for this file are closed. I don't think it's possible to do it otherwise.

Ubuntu on Windows

Posted Apr 4, 2016 20:38 UTC (Mon) by Wol (subscriber, #4433) [Link] (3 responses)

Shame they can't do what Pr1mos did ... they had the same problem with "how do we replace open files when upgrading the Operating System". (Or when recompiling, or whatever.)

What it did was rename open files to some weird (but standard) name and put the new file in its place. So every now and then you ran a clean-up on all files with this standard filename layout and deleted them.

The thing about Windows that really gets me, though, is when damn background processes I don't want open files I want to access, and lock the file/directory. I typically copy directories off my camera, and the first thing I do is rename the directory - that is - if I can - if Windows hasn't damn well locked it first ...

Cheers,
Wol

Ubuntu on Windows

Posted Apr 6, 2016 20:52 UTC (Wed) by nix (subscriber, #2304) [Link] (2 responses)

This is the same kludge NFS servers usually use. Not for nothing is it called 'silly-rename' in the Linux implementation.

Ubuntu on Windows

Posted Apr 6, 2016 23:42 UTC (Wed) by anselm (subscriber, #2796) [Link] (1 responses)

The silly renaming in NFS is necessary in order to maintain Unix file semantics on a (supposedly) stateless NFS server. In Unix, you can open a file and unlink it immediately afterwards, but as long as the process keeps the file open it can still use it even though the file doesn't have a name in the file system anymore. The NFS server, being stateless, doesn't keep track of which files a client has opened. Hence the NFS server must rename the file instead of deleting it outright, in order to be able to find it again when a client which might still have it open (from its POV) accesses it later, even though the file's original name may already be in use for a completely different file. It is silly but there is really not a lot to be done about it.

Ubuntu on Windows

Posted Apr 12, 2016 16:18 UTC (Tue) by nix (subscriber, #2304) [Link]

I didn't mean to imply it wasn't essential, and indeed I use it all the time. But that doesn't mean that it's any less silly :)

Ubuntu on Windows

Posted Mar 31, 2016 19:45 UTC (Thu) by kjp (guest, #39639) [Link] (13 responses)

> (it DOES support atomic renames).

Not if the target file is open (even with all sharing modes). I just tried - it's just like the "can't delete if open" problem you mentioned.

What garbage. I'm just trying to have a GUI QT program write out an updated conf file, at the same time a inetd-like daemon could be concurrently reading it. If the daemon has it opened for read access, even with full sharing (read/write/delete), I can't MoveFileEx on top of it. Looks like windows explorer and del.exe do the 'delete after all handles closed' crap. (And explorer attempts to hide the file even though if you open another window or cmd.exe it's still there). *Bangs head on desk.*

It's too bad MSFT can't fix these bone headed, frustrating problems in win32, instead of coming up with a whole new BS subsystem that will never interoperate with the rest of the system right.

Ubuntu on Windows

Posted Mar 31, 2016 19:49 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link] (2 responses)

> Not if the target file is open (even with all sharing modes). I just tried - it's just like the "can't delete if open" problem you mentioned.
Yes, that's the way Windows work.

But on the bright side, it does support directory hardlinks :)

Ubuntu on Windows

Posted Mar 31, 2016 20:31 UTC (Thu) by viro (subscriber, #7872) [Link]

> But on the bright side, it does support directory hardlinks :)

All together: Always look on the bright side of life!

Ubuntu on Windows

Posted Apr 1, 2016 12:37 UTC (Fri) by micka (subscriber, #38720) [Link]

Always using delete while open (like, everyday, every hour, multiple times).
Never using directory hardlinks.

Ubuntu on Windows

Posted Mar 31, 2016 23:50 UTC (Thu) by zlynx (guest, #2285) [Link] (4 responses)

You cannot move a new file over top of an old file. That would delete the old file. What you can do is move the old file out of the way, then move the new file to the old filename.

I have a function that I named "remove_dance" which deletes the backup filename, moves the current file to the backup filename, then attempts to delete the backup filename. I use it a lot in Windows code.

That way if something prevents deleting the file it still gets shoved out of the way.

Ubuntu on Windows

Posted Apr 1, 2016 0:48 UTC (Fri) by viro (subscriber, #7872) [Link] (3 responses)

... at which point any atomicity warranties are, of course, gone.

Ubuntu on Windows

Posted Apr 1, 2016 6:53 UTC (Fri) by zlynx (guest, #2285) [Link] (2 responses)

There's always Transactional NTFS. Much better guarantees than anything POSIX can do.

Ubuntu on Windows

Posted Apr 1, 2016 13:23 UTC (Fri) by madscientist (subscriber, #16861) [Link] (1 responses)

You mean, the API that Microsoft has deprecated and is actively discouraging people from using?

> Microsoft strongly recommends developers investigate utilizing the discussed alternatives (or in some cases, investigate other alternatives) rather than adopting an API platform which may not be available in future versions of Windows.

Ubuntu on Windows

Posted Apr 4, 2016 11:05 UTC (Mon) by nye (subscriber, #51576) [Link]

>You mean, the API that Microsoft has deprecated and is actively discouraging people from using?

It's generally not a great idea to rely on a specific feature of one filesystem anyway, even if that filesystem *seems* to be married to the platform. I think NTFS itself will be slowly deprecated over the coming years, in favour of ReFS, which doesn't support the same transactional API and may never do so.

I'm pretty impressed by ReFS so far. It seems (based on admittedly limited testing on my part, where I didn't look at performance) a lot like what you'd expect to see as a result of investigating ZFS, paying particular attention to its weaknesses, and then throwing endless piles of cash at coming up with a competitive response.

Ubuntu on Windows

Posted Apr 1, 2016 7:44 UTC (Fri) by Seegras (guest, #20463) [Link] (4 responses)

Most of it is still there: http://seegras.discordia.ch/Blog/whats-wrong-with-windows/ Including mandatory locking.

Ubuntu on Windows

Posted Apr 4, 2016 11:13 UTC (Mon) by nye (subscriber, #51576) [Link] (3 responses)

>Most of it is still there: http://seegras.discordia.ch/Blog/whats-wrong-with-windows/ Including mandatory locking.

The bits of that post that I do know something about are astonishingly ignorant. Is there any reason to believe the rest is different?

(I particularly love the claim that 'no Unix-Filesystem does fragment'. That one got a wry chuckle.)

Ubuntu on Windows

Posted Apr 4, 2016 21:01 UTC (Mon) by Wol (subscriber, #4433) [Link] (2 responses)

Well, it's almost true :-) Seriously, which unix filesystems DO do defrag? Which ones NEED to do defrag? Which is what the author was getting at - unix file systems don't have a defrag because they don't need a defrag.

It was quite an interesting read, if a little out of date :-)

Cheers,
Wol

Ubuntu on Windows

Posted Apr 4, 2016 21:24 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link]

> Well, it's almost true :-) Seriously, which unix filesystems DO do defrag? Which ones NEED to do defrag?
All of them, but perhaps a tad slower.

> Which is what the author was getting at - unix file systems don't have a defrag because they don't need a defrag.
http://manpages.ubuntu.com/manpages/precise/man8/e4defrag...

Ubuntu on Windows

Posted Apr 5, 2016 10:56 UTC (Tue) by itvirta (guest, #49997) [Link]

Interesting? More like filled with opinion-based ranting. :)

Ubuntu on Windows

Posted Mar 31, 2016 1:52 UTC (Thu) by anguslees (subscriber, #7131) [Link]

That's pronounced "GNU/Windows"

Finally, a successor to coLinux then?

Posted Mar 31, 2016 6:08 UTC (Thu) by AndreE (guest, #60148) [Link]

Sometimes, needing to work on Windows (at work or for a specific application at home), it's just plain useful to use the Linux suite of tools to get things done. I'm just hugely familiar with bash and the various standard Linux utilities and use them productively for a whole bunch of different tasks. In lieu of learning powershell at letting my *nix skills atrophy, the only real alternative on Windows has been to run a virtual machine, a clearly heavy-handed approach.

coLinux allowed me to avoid this altogether, but being 32-bit only hasn't been practically usable for years. This seems almost as good, although I haven't read anything anywhere about device driver support. With coLinux, since it was an implementation of a Linux kernel as windows process, I could mount and ext3 drive in windows and have it treated natively by the linux kernel running under windows. I'm not sure that a syscall translation layer can actually do the same thing

Ubuntu on Windows

Posted Mar 31, 2016 8:56 UTC (Thu) by rvfh (guest, #31018) [Link] (1 responses)

A bit early for an April fool...

Ubuntu on Windows

Posted Mar 31, 2016 12:52 UTC (Thu) by jengelh (guest, #33263) [Link]

Considering things like UTC+14, I'd say "right on time".

Is there a clear statement on how it's licensed?

Posted Mar 31, 2016 10:20 UTC (Thu) by k3ninho (subscriber, #50375) [Link] (2 responses)

I want to know where the interaction of the VMWare GPL enforcement trial, Canonical's view of the GPL in light of their ZFS modules, and the Google-Oracle Java API retrial put this from the perspective of respecting the moral rights of authors to share their work as they see fit. This might be a terrible confluence of 'nope', or might cause MS to fight in court to uphold the GPL -- I'm waiting to see what SFC, mjg59, Christoph Hellwig and Harald Welte might say.

K3n.

Is there a clear statement on how it's licensed?

Posted Mar 31, 2016 10:58 UTC (Thu) by khim (subscriber, #9252) [Link]

Microsoft have separated Ubuntu userspace from the kernel at the EXACT line kernel developers specified as GPL-friendly. What COULD they say, really? "We don't like it but are powerless to do anything about it", probably.

VMWare have picked completely different line, thus the very different reaction.

Is there a clear statement on how it's licensed?

Posted Apr 1, 2016 17:29 UTC (Fri) by lsl (subscriber, #86508) [Link]

This appears to be an independent reimplementation of the Linux system call interface. So really, it's stuff that ought to be possible without Linux authors having a say in it. OTOH, there's Oracle and Java, of course.

Ubuntu on Windows

Posted Mar 31, 2016 17:53 UTC (Thu) by flussence (guest, #85566) [Link] (27 responses)

How well does Mir run on it? Windows 10 with a good UI would be an interesting combination.

Ubuntu on Windows

Posted Apr 1, 2016 18:01 UTC (Fri) by khim (subscriber, #9252) [Link] (26 responses)

There's no GUI - and it would really surprising to see it added. This wouldn't help Microsoft with Azure and could hurt them on desktop - why would they do that?

bash in windows shellon bahWindows

Posted Apr 1, 2016 22:23 UTC (Fri) by h2 (guest, #27965) [Link] (25 responses)

If I understand the linked articles right, this is simply about running a shell inside of cmd.exe, that is, replacing the primitive windows cmd window with one running ubuntu bash with access to most key tools on the cli.

I view this far more as a tacit admission on microsoft's part that they actually cannot be as good as bash or rsync or ssh.

If you couple this with microsoft's recent decision to use openbsd's openssh, coupled with a very large donation to the openbsd project to help fund this work and support the source of openssh, this could actually be microsoft throwing in the towel on shell tools. their powershell commands are so awkward and clunky compared to the mature shell tools of *nix that you really can't even compare the two.

I also was wondering at first about how they could possibly run any x driven tool, and the answer is, of course they can't, but they can run bash in a windows window.

The notion above in this thread that this would lose any linux desktop users is odd to put it mildly, this is just a shell, of great use to any dev environment with a lot of people running windows but interfacing with non windows systems via ssh, a good friend of mine works at one such large scale place, and I know they can use this directly as is, and probably will, because running all these emulators and putty's and other stuff just isn't as good as running native tools that hook right into the NT kernel.

So what this will simply do is let development environments where users already run windows for whatever reason now use better shell tools directly.

I don't believe the support will be anywhere near complete, for example, I doubt this will have the linux kernel core interfaces like /sys and maybe or maybe not /proc (though that's been emulated before on other platforms), so I just see this as getting the basics running so you can work in a real shell without indirect logic. bummer for stuff like cygwin and their project, but as with wine, I'm personally not a fan of emulators, I generally will just skip it, or run a vm if I need the feature. Apt on windows, lol, that's really funny though. Risky, because apt is so radically superior to any other package management system, particularly in the non existent version in the windows world, if anything, exposure to apt is very likely to push people to gnu/linux systems running apt.

Every time I run windows, I always wonder where apt is, then remember I have to do the download package, install, pray, thing, make sure to avoid 3rd party sources, all types of stuff, and I think if you show windows developers/admins these tools, after a while, you just get used to apt-get install, apt-cache search, etc, to install something you need, a little risky on Microsoft's part actually. I'm surprised the commenters in this thread haven't caught how significant apt on windows bash is, huge package pool by default.

Interesting development, for sure. I wonder if they have decided to forego internal openssh in powershell, or if they are going to go both ways, hard to say, powershell has a real userbase too, so maybe they are going for both, it's smart on their end.

bash in winlinksdows shellon bahWindows

Posted Apr 1, 2016 22:48 UTC (Fri) by h2 (guest, #27965) [Link] (1 responses)

http://www.hanselman.com/blog/DevelopersCanRunBashShellAn...

That I think clarifies it well.

and this

Posted Apr 1, 2016 23:11 UTC (Fri) by h2 (guest, #27965) [Link]

http://www.zdnet.com/article/head-of-oracle-linux-moves-t...

bash in windows shellon bahWindows

Posted Apr 1, 2016 23:09 UTC (Fri) by khim (subscriber, #9252) [Link] (22 responses)

their powershell commands are so awkward and clunky compared to the mature shell tools of *nix that you really can't even compare the two.

It's not that. PowerShell is pretty good - but you couldn't use it to drive Linux servers.

You still would need PowerShell BTW: bash would have no access to Windows services. At all.

Microsoft admits that it have lost the battle for server thus now it's time to make sure developers would at least stay with Windows. They have killed Xamarin Studio, for example - because otherwise people would try to use it and apt instead of Visual Studio and NuGet.

As long as developers stay on Windows Microsoft could stage a comeback and/or counter-attack.

And I'm not sure how well it'll work in the end but this approach has potential: today a lot of developers are forced to stay with Linux to develop stuff for Linux - on servers. Microsoft just made sure that they would be able to go back to Visual Studio - and many would do that, I'm sure.

Sure, I don't think people who have chosen Linux voluntarily would drop it. But bet you would be surprised to find out how many have chosen Mac because it has real bash and now they could even develop stuff for Linux without leaving cosy Windows world!

Remember that for each user who have picked Linux there are 99 who haven't. It's for them, not for you.

good points

Posted Apr 1, 2016 23:23 UTC (Fri) by h2 (guest, #27965) [Link] (20 responses)

thanks I was actually wondering what types of system hooks bash shell would have access to.

And it's an excellent point re OSX and bash and devs. Not to be pedantic, but I believe the real current number is about 1.8 gnu/linux users out of 100 desktop users, heh. Never gotten much above 2%, and then only for a short while, for very good reason (upgrades, reinstall, no stable api/abi's). It's much more likely that MS is targetting os x here, not gnu/linux, which is so fragmented on its own that MS doesn't even have to contemplate engaging in divide and conquer strategies since we do that so well ourselves already. But OSX, that's a real target, and worth aiming for.

I can see why this though, because my friend was just biting the bullet to try to learn enough powershell so he could write some small utilities for putty connections, but this type of thing would cover that completely and let you run all native stuff.

As long as Linux/Gnu insists on never maintaining stable abi's / apis on the kernel/desktop toolkit front, there's no danger at all of Linux desktop marketshare ever breaking much beyond the slightly tech oriented types who tend to use it now, ie, 1-2% of total desktop user base.

By the way, I was really happy to see windows get the credit they deserve for their api stability here in this thread, it's something they've always had as a very high priority, primarily I think for their corporate desktop users, which is a huge chunk of their market. Every OS does something better than the others, it's fair to give credit where credit is due. I just wish linux kernel devs would learn what stable means, and that toolkits wouldn't fundamentally break everything at every release, but the days of my thinking this is actually going to ever change are now behind me so it's nice to see realistic moves made wherever you may find them.

good points

Posted Apr 2, 2016 16:01 UTC (Sat) by khim (subscriber, #9252) [Link] (8 responses)

I just wish linux kernel devs would learn what stable means

What do you mean? Kernel devs know that very well and follow the mantra religiously. Linus just reasserted that if it does break anything, it needs to be turned off by default. That's a hard rule.

Pehaps you mean that thing? This part is not a problem - and, surprisingly enough, Microsoft and Windows help to highlight that very well, indeed. Old applications work quite well on Windows. Old drivers? Not so much. It was common to lose support for WinModems and WinPrinters, scanners and other random hardware with Windows upgrades. I have quite a collection of old hardware - and I'm forced to keep Windows XP and Windows 98 (sic! Winows98) around to use it. Stable API just does not work for the hardware.

The same is true for MacOS, Solaris and all other OSes (including Linux). Linux developers explicitly don't care, others do care, but the end result is the same: obsolete, unsupported driver are lottery with all of them. It's just simple as that. Sometimes they work. Sometimes they don't work. Sometimes they could be easily upgraded. Sometimes it's impossible to do.

We may discuss the question “why?” for a long time (I have some ideas but have no way to verify them), but in the end it does not matter: stable API works in userspace, but it does not work for drivers. It's just a fact. Kernel developers are not to blame. This particular sin is not a problem for Linux.

good points

Posted Apr 4, 2016 21:15 UTC (Mon) by Wol (subscriber, #4433) [Link] (7 responses)

> Pehaps you mean that thing? This part is not a problem - and, surprisingly enough, Microsoft and Windows help to highlight that very well, indeed. Old applications work quite well on Windows. Old drivers? Not so much. It was common to lose support for WinModems and WinPrinters, scanners and other random hardware with Windows upgrades.

Actually, linux RARELY loses support for old hardware. If Win98 WinModems and WinPrinters were supported by Linux of that era, then they are still supported by the latest linux 4 unless nobody bothered to complain when they were accidentally broken (which is quite likely :-). Linux support stays around for as long as users complain (and help debug) when it gets accidentally broken.

Cheers,
Wol

good points

Posted Apr 5, 2016 8:48 UTC (Tue) by NAR (subscriber, #1313) [Link] (6 responses)

Well, when pulseaudio was introduced, Linux definitely lost support for lots of hardware. Maybe got back later, I didn't wait to find out.

On the other hand I do remember that I had to use a 3rd party bttv driver on WinXP in order to use my AverMedia98 TV capture card. So this is definitely a problem, only mitigated by the relative infreqency of releases (5 releases in 15 years, compared to 23 Ubuntu versions in shorter time).

good points

Posted Apr 6, 2016 7:33 UTC (Wed) by johannbg (guest, #65743) [Link] (5 responses)

"Well, when pulseaudio was introduced, Linux definitely lost support for lots of hardware."

By all means explain to the wider audience how the linux kernel lost support for a lots of hardware when pulseaudio got introduced.

I'm looking forward to hear that explanation.

good points

Posted Apr 6, 2016 7:46 UTC (Wed) by NAR (subscriber, #1313) [Link] (4 responses)

It's simple. Before pulseaudio was introduced, I could listen to music, record from my TV card, etc., could do all the stuff I expect from an audio card. After pulseaudio was introduced, none of it worked. The audio card practically became junk (under Linux). Maybe the audio support recovered since, but I'm afraid Lennart Poettering's reputation never.

good points

Posted Apr 6, 2016 8:18 UTC (Wed) by seyman (subscriber, #1172) [Link] (3 responses)

> After pulseaudio was introduced, none of it worked.

Pulseaudio's introduction didn't make any of the existing audio systems disappear or stop working so this is obviously false.

good points

Posted Apr 6, 2016 8:54 UTC (Wed) by NAR (subscriber, #1313) [Link] (2 responses)

So you know better than me what worked on my system and what didn't. And I do remember that many people had the same experience, see e.g. https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug...

good points

Posted Apr 6, 2016 11:16 UTC (Wed) by nye (subscriber, #51576) [Link] (1 responses)

Sure, the defaults were broken, but broken defaults is a fact of life in Linux distros, especially around that time period which was particularly turbulent. PA wasn't required though, nor even hard to remove; you could always just uninstall it and get sound back - I know that's what I did after every update for a couple of years.

I do feel your pain though - the nightmare year of 2008 was when I started my transition to using Windows on the desktop. PA still wasn't production-ready by the time I'd switched full-time.

good points

Posted Apr 6, 2016 12:07 UTC (Wed) by mathstuf (subscriber, #69389) [Link]

I used Fedora 9 from its Alpha stage (KDE4 was at 4.0.3) and it was (one of, if not the) the first to include PulseAudio. Things were…rough, but it went well enough (was helping the KDE SIG at the time). By Fedora 10 and almost certainly 11, I had stopped having PA issues (well except flat-volumes, but that is an easy fix). However, Ubuntu continued having systemic problems for up to 18 months afterwards. Whether it was due to older kernels not shipping with the driver fixes, older ALSA libraries not getting the bridging right, problems with the default configurations, whatever, I don't know for sure, but as far as I can tell, most of people's problems were the result of Ubuntu botching it up on their end; Fedora showed that the right pieces were available, maybe just not according to Ubuntu's update policies (but no, can't fix things because the new release contains an unacceptable version bump nevermind it might fix 100 bugs in the process).

These days, I keep an instance of mpv streaming music whenever I'm at work and if I need to start using the webcam, start to watch another video, listen to other audio files, I just mute the stream and start the other program. No audio device locking problems, I can reroute audio while programs are running, and none of the problems people complain about today from their experiences 5+ years ago.

good points

Posted Apr 2, 2016 17:04 UTC (Sat) by anselm (subscriber, #2796) [Link] (1 responses)

It's much more likely that MS is targetting os x here, not gnu/linux

It seems that Microsoft's offering is very much a server-side thing (given that, for example, it doesn't come with a GUI). How that is really supposed to be an attack on OS X, which is approximately as popular on servers as Linux is on desktops, in other words not all that much, is something you would need to explain in more detail.

good points

Posted Apr 2, 2016 18:58 UTC (Sat) by khim (subscriber, #9252) [Link]

It's not an attack on MacOS. It's "attack on" developers which are stuck on MacOS. I know a lot of guys who are using MacOS instead of Windows because MacOS is "real Unix" (albeit a poor one) and Windows is not.

They don't comprise a huge percentage of population, thus this "attack" wouldn't directly bring number of MacOS users down. But they represent guys who are influencing others thus effect could be significant down the road.

But as someone on the other forum (don't remember which one) have pointed out that all these ideas fail Hanlon's razor (the infamous: never attribute to malice that which is adequately explained by stupidity).

It looks that Microsoft finally realized that it's ambitious project to save Windows Phone could instead bring down the whole house of card (OS/2 style) and cancelled it then was faced with a question: "could we salvage anything usable from the wreckage?" - and this is the result.

No deep plans, no well-thought attacks, just sheer bestial fear of an Android and the desire to at least use result of hundred man-years of work somehow...

good points

Posted Apr 3, 2016 7:41 UTC (Sun) by jem (subscriber, #24231) [Link] (2 responses)

> Not to be pedantic, but I believe the real current number is about 1.8 gnu/linux users out of 100 desktop users, heh. Never gotten much above 2%, and then only for a short while, for very good reason (upgrades, reinstall, no stable api/abi's).

ABI stability on Windows is no better. Only recently (with Windows 10) the C and C++ standard libraries (what Microsoft calls "C Runtime library") got promoted to a system component called the Universal C Runtime. Before that, the libraries were distributed with Visual Studio and were not compatible with each other.

This is why third party libraries for Windows are offered in a large number of variants depending on which vesion of Visual Studio they are intended to be linked with: VS 2005, 2008, 2010, 2012, 2013, etc...

good points

Posted Apr 3, 2016 19:12 UTC (Sun) by khim (subscriber, #9252) [Link] (1 responses)

Before that, the libraries were distributed with Visual Studio and were not compatible with each other.

They were compatible where that counted: all versions were included and maintained as part of the OS.

This is why third party libraries for Windows are offered in a large number of variants depending on which vesion of Visual Studio they are intended to be linked with: VS 2005, 2008, 2010, 2012, 2013, etc...

Sure, but that only affects developers. For a long time developers have faced a dilemma: develop for Linux (easy peasy lemon squeezy) then find out that you have no way to distribute your stuff to users (no way to produce binaries and deliver them to users and minuscule distributions are trying to impose insane demands on you) or develop for Windows (really hard, you need to pay $$ to get good instruments and many things are just plain out painful to do) then distribute result easily.

Developers have endured great pains because stability of target platform is more important than stability and usability of development platform.

Now with MacOS/iOS and Android situation is changing. No longer Windows is the only game in town! Development tools for these new platforms are not as great yet but they are going there - and in both cases billions of users are there too. What's not to like? Microsoft feels the heat, apparently.

good points

Posted Apr 4, 2016 14:20 UTC (Mon) by rghetta (subscriber, #39444) [Link]

> They were compatible where that counted: all versions were included and maintained as part of the OS.

This is not correct.
If your application is linked to a Visual C++ runtime DLL, you need to bundle the dll with your executable (by including the MSVC Redistributable Runtime installer), the SO itself doesn't include it.

good points

Posted Apr 4, 2016 21:10 UTC (Mon) by Wol (subscriber, #4433) [Link] (5 responses)

> By the way, I was really happy to see windows get the credit they deserve for their api stability here in this thread, it's something they've always had as a very high priority, primarily I think for their corporate desktop users, which is a huge chunk of their market. Every OS does something better than the others, it's fair to give credit where credit is due

Sorry, I call bullshit here ... "Dos ain't done til Lotus won't run". Windows stability matters (a) for Microsoft Office et al, which uses (or at least, used) undocumented APIs, and (b) for all those silly little apps that were important to users but meant nothing to Microsoft.

You could pretty much GUARANTEE that EVERY rev of Windows would contain API breaks that were very damaging to apps that MS considered competitors. As a WordPerfect fan, the list of API breaks from WFWG onwards that caused *serious* problems is long long long ... (Oh - and that's pretty much every version of Windows from then right through to XP - maybe beyond.)

Cheers,
Wol

good points

Posted Apr 4, 2016 21:25 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link] (4 responses)

> Sorry, I call bullshit here ... "Dos ain't done til Lotus won't run".
This is bullshit.

> You could pretty much GUARANTEE that EVERY rev of Windows would contain API breaks that were very damaging to apps that MS considered competitors.
Nope. MS went to great pains to keep applications running. ALL of them, up to including custom workarounds for some apps.

good points

Posted Apr 10, 2016 14:25 UTC (Sun) by Wol (subscriber, #4433) [Link] (3 responses)

Then why did EVERY upgrade to Windows break my "working fine" copies of WordPerfect?

WFWG. Office95. Win98. XP.

I've left out NT4/NT2000 - I don't remember problems with them. But the number of upgrades and/or emergency bug-fixes you needed to keep WordPerfect going as Windows changed underneath was awful.

Cheers,
Wol

good points

Posted Apr 10, 2016 14:50 UTC (Sun) by reedstrm (guest, #8467) [Link] (2 responses)

ISTR one of those windows upgrades involving Microsoft giving WordPerfect early access to some APIs, then yanking/ not releasing parts that wp depended on, but that word did not, just before the release. Result: wp broke on upgrade. Pretty sure there was a lawsuit.

good points

Posted Apr 10, 2016 15:14 UTC (Sun) by reedstrm (guest, #8467) [Link] (1 responses)

Or that it came out in discovery or testimony during the drdos lawsuit.

good points

Posted Apr 10, 2016 16:11 UTC (Sun) by anselm (subscriber, #2796) [Link]

Also ISTR that the Microsoft applications used to use special undocumented APIs that could do convenient and powerful things and were (a) unavailable to third-party applications, (b) not part of any stability or support guarantees because they weren't part of the documented interface, so even if a third-party developer figured out one of them for Windows version N they had no guarantee that their code would keep working on Windows version N+1.

bash in windows shellon bahWindows

Posted Apr 4, 2016 11:29 UTC (Mon) by nye (subscriber, #51576) [Link]

>today a lot of developers are forced to stay with Linux to develop stuff for Linux - on servers

There's more than that though - it's not limited to building stuff *for Linux*. A lot of cross platform FOSS is mostly developed under the assumption that you're cross-compiling[0]. For example, Clementine is already hard enough to build on Linux, but building it on Windows would be like traversing several circles of hell - if it's even *possible*. It's actually *very* common that cross-compiling the Windows version from Linux is enormously easier than building natively.

[0] Hell, even when I was using Linux full time on my desktop I would sometime cross-compile for Windows when working on personal projects, because Wine is a more reliable platform than native libs and I knew that if the Windows build would work in Wine then it would continue working if I launched it in a year, unlike native binaries.

Ubuntu on Windows specific presentation on Channel 9

Posted Apr 1, 2016 23:36 UTC (Fri) by deucalion (guest, #12904) [Link]

Please do check out the presentation (aprox. 29min) on the subject:
https://channel9.msdn.com/Events/Build/2016/C906

It appears to be a rehash of SFU/MKS with a Linux focus, but with a lot less ambition (no full LSASS support at first like SFU, no full NT/kernel API access), but a few nice features like native elf binary execution support.

Instead of Ubuntu on Windows, why not Android apps in Windows Phone

Posted Apr 4, 2016 8:13 UTC (Mon) by NAR (subscriber, #1313) [Link]

I'm bit skeptical how useful Ubuntu on Windows really is (e.g. for Java developers Eclipse on Windows works fine), but I think the real big achievement would be to enable Android apps on Windows Phone. I happen to use one and is much nicer than the Android phone I used to own, but dearly lacks apps.

Ubuntu on Windows

Posted Apr 7, 2016 8:44 UTC (Thu) by gwg (guest, #20811) [Link] (3 responses)

Sounds an awful lot like an updated version of Microsoft's POSIX subsystem/OpenNT/Internix/Windows Services for UNIX.

Ubuntu on Windows

Posted Apr 8, 2016 14:55 UTC (Fri) by mathstuf (subscriber, #69389) [Link] (2 responses)

That was Yet Another POSIX Implementation™ that you had to deal with incompatibilities in (see the porting required to work on glibc vs. musl, Linux vs. BSD, etc.; smaller than *nix -> Windows porting, but still non-zero). Now that they're implementing the Linux syscall ABI, the standard is "what Linux does" and if there's an incompatibility, I'd be more optimistic about the behavior being fixed.

Ubuntu on Windows

Posted Apr 9, 2016 14:18 UTC (Sat) by gioele (subscriber, #61675) [Link] (1 responses)

> Now that they're implementing the Linux syscall ABI, the standard is "what Linux does" and if there's an incompatibility, I'd be more optimistic about the behavior being fixed.

Some incompatibilities have already surfaced:

https://github.com/kernelslacker/trinity/pull/9

> LXCore always returns EINVAL when opening perf events, throwing trinity in an endless loop. Patch fixes the endless loop.

> Why don't they just -ENOSYS ? […] My hope is we can do this without pessimizing the common !lxcore case

Ubuntu on Windows

Posted Apr 10, 2016 4:16 UTC (Sun) by flussence (guest, #85566) [Link]

I've already heard from someone who tried to use it that GNU tar doesn't work properly either (busybox's does).

I'm wondering if this initial release just implements the bare minimum necessary to debootstrap an Ubuntu userspace and put out a press release about it.

Windows sysbystem for Linux and /mnt/c/

Posted Apr 8, 2016 14:58 UTC (Fri) by domo (guest, #14031) [Link]

How does using /mnt/c/ match with http://refspecs.linuxfoundation.org/FHS_3.0/fhs/ch03s12.html
and does anyone care ?