Live data from Hacker News

DirectX is coming to the Windows Subsystem for Linux

devblogs.microsoft.com

251–260 of 548 posts

Re: DirectX is coming to the Windows Subsystem for Linux

#251
post #230

Earlier quoted context omitted.

Yup, from one of the MS staff replies further in the thread[1] > There is a single usecase for this: WSL2 developer who wants to run machine learning on his GPU. The developer is working on his laptop, which is running Windows and that laptop has a single GPU that Windows is using. Can't say I can get behind MS trying to shift maintenance for a Windows only "feature" onto the Linux devs here. 1. https://lkml.org/lkml…

subtle gendering there.

I don't see it, but if it reads that way it wasn't my intention.

Re: DirectX is coming to the Windows Subsystem for Linux

#252

It appears that Microsoft has now initiated the "Extend" phase of their classic Embrace, Extend, Extinguish playbook. The key is that the proprietary Microsoft specific API added by this patch is only usable in a WSL environment as it relies on many pieces of proprietary closed source software that Microsoft is unwilling to open source. This patch does nothing but fragment the Linux ecosystem and encourage people to…

This feels more like a reverse takeover to me. MS actually has a legacy problem in the form of the windows ecosystem. Under Ballmer this was a problem that required defensive moves and rhetoric to not lose people to Linux and OS X and it didn't work. Under Nadella, a bait and switch was initiated where the problem (windows legacy) is slowly switched out for Linux. This is not the embrace and extend policy of last century. They don't care about software licenses; they care about SAAS revenue these days.

In the process, they are rapidly regaining developer trust by building productive relations with the same developers they were alienating under Ballmer.

It's smart. They own Github now. Most stuff that happens there is not windows centric. But increasingly the MS developer ecosystem is being untangled from windows in any case. That's necessary to future proof it.

While the windows kernel is nice and the driver ecosystem around it serves MS well, it's actually been a problem for them as well. They've failed in the phone market (repeatedly) because linux was a better fit for OEMs. Also they've had Google and Apple compete with them effectively with the ipad and chrome os in a market where MS was peddling crippled laptops. These are all examples where windows legacy was part of the problem for MS. They weren't able to compete there. Their crippled laptops were too expensive and uncrippling them would kill their high end market. So people bought ipads and chrome os laptops instead.

Increasingly desktop software that is not web based is getting more and more of a niche. Even office at this point runs well in a broswer and MS actively supports native applications on platforms of all their competitors (android, IOS, OSX, etc.). At this point that's not optional from a revenue point of view. A windows only office would be a problem at this stage. They support it and they actually do a decent job too. This too is something that happened very quickly under Nadella. A lot of the Azure revenue is in fact office revenue.

IMHO this move to support GPU virtualization and ultimately running linux desktop apps, gets them access to a lot of niche OSS applications for Linux, the entirety of the machine learning ecosystem, and the community of professionals using that. Not all of them will switch to a windows laptop of course. But some will and they tend to be the type that spends money on their tools.

Additionally, they are doing a clever play with APIs, github, cloud native stuff in Azure, development tooling etc. where they are not blocking using non MS things but merely make the choice to buy into their premium SAAS subscriptions more attractive. All of this stuff is usable without doing that but if you are using VS Code already, have your code on Github, and are doing some AI stuff in the cloud, etc; it's a small step to buy into a well integrated ecosystem provided by MS where you are using a windows laptop with their oss tools, while deploying to Azure, and maybe doing your office stuff using office 365. The value is no longer in selling the OS but up-selling SAAS and making sure the choice to buy into that is logical, easy, and natural no matter what you use.

Github codespaces is a great example of that. I bet it will be really easy to setup and come with some nice SAAS subscription. It's all OSS and you are welcome to run it on AWS or your own cloud. But I bet it will be easier to run in Azure. MS tools, MS cloud, MS as the easiest place to run linux developer tools (!!!), etc. They won't force anyone to switch. They don't care about individual developers but they do care about what their bosses sign up for in terms of SAAS services. That's where the money is.

Re: DirectX is coming to the Windows Subsystem for Linux

#253
post #236

Earlier quoted context omitted.

To be honest I don't quite understand what stopped those developers from running machine learning on their GPUs under Windows itself. Most frameworks work just fine. I've been doing quite a lot of TensorFlow with both Python and .NET. The only time I faced the need for Linux box is trying a demo project from OpenAI, which did not use the features it required Linux for on a single machine anyway.

But Windows is a painful OS to use for anything other than gaming. Ideally, I'd like to see the exact opposite of this: run Linux with a Windows subsystem just for gaming.

Apart from photoshop video production audio and many other professional uses.

Like the previous response what does this buy me as compared to developing and running natively in windows - as there are native compliers that support cuda etal on windows.

Re: DirectX is coming to the Windows Subsystem for Linux

#254

Earlier quoted context omitted.

I disagree, i do not think that anyone would want to start running windows just to run the same software they already run while losing a chunk of control and support. However for the real argument, ask yourself: am I switching to using windows because of this?

I think it's the other way round: Do I still need to dual-boot or switch to Linux if this works? Or can I stick with Windows and have the best of both worlds?

I do not agree with it being "The best of both worlds".

You do not control windows, it controls you. You must adapt to it being in your life and the choices made by its designers, this is both not secure and insanity for a software engineer to allow as an ongoing situation.

Even dual booting windows will mean its updater will overwrite the partitioning table of your drive and hide any other OS you have installed, and as I understand this people only do it out of need and not out of want.

Re: DirectX is coming to the Windows Subsystem for Linux

#255
post #248
post #118

Earlier quoted context omitted.

> the year of Linux on the desktop will never come also thanks to this. I think you're giving MS too much credit here. Apple has one UI, and Windows has one UI (2 if you count windows in tablet mode). Linux has, what? 30? Quoth the Torvalds: "I still wish we were better at having a standardize desktop that goes across all the distributions… It’s not a kernel issue. It’s more of a personal annoyance how the fragmentat…

I think that both Ubuntu and Red Hat family being based on Gnome is a good thing for Linux on the desktop, as there is a clear target for Linux GUI applications. I like the fact, that you still have the freedom to run any kind of desktop environment of Linux, but a stable default can only help Linux.

Its better if the competing ecosystem offers multiple stable defaults IMHO. Have you looked at KDE5? its amazing.

Re: DirectX is coming to the Windows Subsystem for Linux

#257
post #245

Earlier quoted context omitted.

That's exactly what Wine is.

It's exactly what Wine wants to be. Sadly that's not quite the same thing.

What do you mean? Wine has a different architecture because distributing microsoft binaries is not legal, so technically it's not the same thing, but it still does an amazing job and a lot of apps/games works flawlessly.

Re: DirectX is coming to the Windows Subsystem for Linux

#259

I really wish Microsoft would just tell NVidia (and AMD) to support SR-IOV on all their GPUs. Then we wouldn't need any major software changes to enable CUDA acceleration within VMs, just a configuration change to pass through a virtual function of the GPU to make it available to existing drivers.

AFAIK that's up to motherboard manufacturers. SR-IOV often doesn't work on desktops, I doubt there mass produced notbooks support it at all. MS can work with manufacturers, but they have their own incentives and it is unlikely to expect that all of them suddenly start doing a good job on this front. Meanwhile MS wants to cast widest possible net, so it needs to enable more and more workflows on WSL2 today, not tomorr…

Other than enabling the IOMMU (which any system with Thunderbolt should have on by default), what does the motherboard firmware actually need to do for SR-IOV? I can't see any reason why it would need to be concerned with enumerating Virtual Functions or anything like that, but I don't actually have the SR-IOV spec on hand to dig through.

EDIT: Looking through an old Intel whitepaper, it looks like the system firmware at most has to reserve some extra config space for SR-IOV devices when enumerating the PFs, so that the OS can enumerate VFs after creating them. But Linux includes an option to re-allocate this stuff if the BIOS doesn't reserve space, so this apparently isn't a hard requirement.

Post reply on HN