Hacker News new | past | comments | ask | show | jobs | submit login
Embrace, Extend, and Extinguish (wikipedia.org)
170 points by doener on May 20, 2020 | hide | past | favorite | 118 comments

This is brought up every time Microsoft and Github are discussed, and it seems like a very shallow thing to fixate on. Is it impossible that a company, with a new CEO and leadership team, would discard a strategy from 25 years ago (more like 50 years ago by non-tech industry timescales)? Especially now that its market dominance is fundamentally different and reduced (likely permanently) compared to the 90s? What evidence is there that Nadella is a Manchurian candidate for Ballmer forces?

Note that my issue with EXX being brought up is that it is superficial, to the point of being deceptive to the kind of threats Microsoft+Github are most likely to pose if Nadella et al do have their own malicious plans. But I think Microsoft's support for a healthy OSS ecosystem, via Github and things like VS Code, is genuine. Not necessarily out of altruism, but because it's one of the few ways they can have an edge into the future and not fall into stagnation like IBM.

In my opinion, the current mores of the company are pretty irrelevant; power dynamics are what matter.

Tides change. Whether or not Microsoft is benevolent today, they may not be in 5+ years.

Any company large enough to screw me over without feeling the loss of my (and people like me's) business in their pocketbook, is not a company I want in a position of power over me. Adapted to this conversation: we should not tolerate "extend" from any company that could successfully pull off "extinguish".

YES. It's about which realities are in the "adjacent possible", not about the personality of any entity in this "tick" of the timeline. We should avoid realities that make a "dystopic" reality more adjacent to us than we are willing to deal with.

Yeah, but that would honestly be an argument for any large company with leverage getting too much leverage, and this would apply to things like Google, Amazon, Apple, etc. and them using their leverage to control markets in similarly unethical ways.

It feels tiring that any time Microsoft does something good, EEE comes up, but times where Amazon or Google are doing massive damage to our industry or standards for their own benefits it's crickets.

I think you underestimate a few things

- How great it feels to make dire predictions about the future. Only I can see how terrible M$'s latest move is going to be. The other sheeple can't. When I'm right I'll tell everyone I told them so.

- How satisfying it is to educate these younguns with their JS frameworks. They don't have my wisdom and experience so I'm doing my duty to society and HN by sharing with them. I will paste links to wikipedia wherever necessary. Like so -> https://en.wikipedia.org/wiki/Embrace,_extend,_and_extinguis...

- how smart I'll look by talking cynically. Every action that someone takes will be followed by a cynical analysis which assumes the worst intentions. Wide eyed optimists will look naive to any person lacking context.

People were shouted down here 5-10 years ago for predicting that Google might turn evil, now it's almost a mainstream opinion here.

You know, simple predictions are often correct. But ignore it (especially if your salary depends on it).

For any company brought up here, there are people that say it is/will turn evil, so it is not exactly surprising that some of them are sometimes right. A broken clock is right twice a day bla bla.

Salary is no excuse, just avoid working for ill-intention companies. You'll feel better and the world will be as well.

This is the only way to make a difference.

Making conscience-free salary doesn't necessarily mean it'll always be that way. Companies change, people change.

Are you suggesting people change companies every time they uncover their existing employer did something unethical/immoral?

I say this as someone whose employer in the past turned into a patent troll - as a new parent and during a downturn it was difficult to switch jobs for quite some time.

I also think you underestimate how much fun it is for people who see the people seeing and pointing out that you see what their doing.

But we can keep seeing things in what has been seen, and never debate the merits in a circular semantics game.

Do you see? I see what you did, but I guess you saw what someone else did, and boom we achieved faux transparency.

And here is wiki article because I totally see what you did with that wiki thing, like so:


That’s the kind of stuff Microsoft tried to lock people into before.

And we remember a world where the stranglehold by IE had to be ruthlessly shredded with open web standards just to get back to sanity.

Oh I definitely get that appeal. I just would rather see more hot takes like "Github is doomed to be the Skype/LinkedIn of open source".

As with most interesting arguments, the truth is probably somewhere in the middle.

I don't doubt that MS would take advantage of market dominance again if it was achieved somehow, and I am sure they are making these plays with increasing their ecosystem in mind.

One interesting acquisition to look at is LinkedIn which they have really tightened the screws on to the point where everyone looking for employees or trying to find a job is made to pay. This is especially vile for the unemployed IMHO.

On the other hand, the eager monopolist of the day is really Amazon, and MS taking them on is a good thing for the ecosystem.

If there is competition MS seems to be playing a fair game. I hope it stays that way.

> If there is competition MS seems to be playing a fair game. I hope it stays that way.

The structure of US corporations is such that they are fundamentally misaligned with playing a fair game. Corporations have two types of cycles: growth, and profit taking.

Regardless of who is in charge of Microsoft today, when Microsoft reaches a position of dominance, the leadership will be shifted by the board and shareholders to focus on exploiting that dominance, rather than continuing to play fair. We are only seeing fair play because Microsoft's position in the open source world is not yet entrenched.

Yeah, my most optimistic hope is that Microsoft and the other general software/services titans (Google, Apple, Amazon, Facebook) keep each other in equilibrium and in non-dominance, like 1984's Oceania, Eurasia and Eastasia.

And hopefully there's room for smaller titans to provide specialized alternatives, e.g. Spotify, Dropbox, Gitlab, Zoom

> Is it impossible that a company, with a new CEO and leadership team, would discard a strategy from 25 years ago (more like 50 years ago by non-tech industry timescales)?

Is it impossible that a company, with a new CEO and leadership team, would reanimate a strategy from 25 years ago (more like 50 years ago by non-tech industry timescales)? Especially one that the company itself used, and the new CEO witnessed first-hand being used, to great effect to acquire market dominance?

> What evidence is there that Nadella is a Manchurian candidate for Ballmer forces?

What evidence is there that Nadella is not trying to use the same tactics as Ballmer?

Now MS is saying they were wrong about Linux being cancer


They could certainly change their minds again if it suits them

They've always supported their developer ecosystem better than any other company as well. Nobody else has spent anything like the the time, effort and resources they have in helping devs thrive in their ecosystem. Not because "they're good people" but because it's effective.

From my perspective, all of their actions with open source can be classified as embrace or extend. Their open source work has very little use unless paired with their proprietary software.

Classic Microsoft apologist’s rant.

Microsoft still does quite a few detrimental things towards open source movement and community in general.

Microsoft is now a wolf in sheep’s clothing

So, taking into account they are one of the largest publicly traded companies in the world, what could they do to start earning your trust?

If the answer is nothing, then logically they should disregard you and carry on.

I’d say another 15 years of no bad behavior, minimum.

The web in particular is still traumatized from years of IE use/development. You needed a virtual machine to test IE on linux/Mac, think about that for a second. It’s great that they realize ‘woah we didn’t give a flying fuck about developers before, huh?’

We’re still being super nice to Microsoft that we don’t give them a pure cold shoulder and accept and evaluate their contributions.

Need time, can’t just go running back to the ex because they bought you something expensive.

I don't think he's proposing that we politely ask microsoft to quit being exploitative.

This must be one of the strangest pro Microsoft opinions I've ever read here.

Nadella IT cells...

I think the most recent impetus is the DirectX+WSL thing that was posted earlier today https://news.ycombinator.com/item?id=23241040

I like what MS is doing these days but I also think its fine to be incredulous, to keep a large corporation at arm's length, and to not forget what can be done if we stop paying attention.

Emperical evidence says no. They just extinguished Wunderlist.

I knew something like this would be the top comment as soon as I saw the title.

EEE is, among other things, a pattern of incentives.

Embrace: We care about our customers. If someone other than us creates something that would be valuable to them, of course we're going to embrace it. To do otherwise, would be a Not Invented Here failure, would be to let down our customers.

Extend: We care about our customers. We are a major center of innovation, working constantly to improve things for them. We're not going to hold back on helping them just because some standards committee is moving like cold molasses, riven by politics, technically tasteless, or has conflict interests optimizing for things other than the best interests of our customers. And we're certainly not going to later break things for our customers merely because some committee eventually decides it wants things done differently. They're just committees. More people are benefiting from our work than theirs. We're not going to hurt our customers just because some committees have a disproportionate sense of entitlement and importantance.

Extinguish: We care about our customers. And those who are potentially our customers. We want to make it as easy as possible for people to gain the benefits of being our customers. The health of our competitors, and even of people unfortunate to not yet be our customers, are not our motivating responsibility. We're focused on our customers, and what happens to that portion of the industry that isn't, can't be what we optimize for. And we're just a normal commercial company, so of course we're not going to fail to compete with our competitors. If nothing else, we owe it to their customers who may soon be ours. And of course to our stockholders.

Much evil in the world is arguably an emergent property of regrettable incentives.

I mention this only because this post is for whatever reason trending today, and past discussions have repeatedly been of unusually disappointing quality for tech topic on HN. Perhaps this comment will help prune some of the inevitable "but they're good people now, not evil, so there's no problem" comments.

It is different when you go out of your way to extend the product in a way that is incompatible for no gain for your clients (and often in detriment of them). And when you make sure not to support the standard and use your market dominance to push your competitors out.

Nothing on the Halloween Documents resembles good faith. Those cases where companies just want to improve their products are much more benign and fail to have the same extinguishing power.

Extension can include duplicating functionality, or making something barely easier and also proprietary.

I don't see how you could be any more up the ass of bad incentives.

Nobody disputes that Microsoft's incentives are 'regrettable', so what point are you really trying to make?

> so what point are you really trying to make?

Incentives can result in individuals making "good" decisions relative to their environment that result in organizations making "bad" decisions relative to the playing field as a whole.

This is essentially the Tragedy of the Commons on a meta-scale, and requires regulation or sovereign action to overcome.

The general point exists in Finance, right? When you incentivize greed, you get certain systemic outcomes.

In other words, these things tend to play out the same way, hence why the same things are said about it over and over again.

If we could boil it down to a word, we’d have to just call it a pattern.

VS Code. GitHub. NPM. VS Code Online. GitHub CodeSpaces. Integrated Azure.

A ten year old today learning programming will be locked in. As will JS yuppies and the extremely articulate product manager types.

I like it. Gives me an opportunity to become a "thought leader" by promoting the UNIX approach to enterprises in 5-10 years.

Jetbrains. Gitlab. Deno. Aws. We're in a world where there's healthy competition in _all_ of these domains, excepting possibly VS online/GH codespaces which both feel like products too young to have yet found a market.

In what world is a 10-year-old locked in to the MS ecosystem any more than they were in the days of win32 and IE6? If anything, I'd look askance at Apple for Safari's hiding of `view source` by default...

There is a difference. Amazon doesn't have an OS. Google does but it's not for developers, yet. Apple's is but they care only about their own walled garden, the entrance fee is expensive and they don't have the online ecosystem that Microsoft has built and acquired in the last years. Microsoft is in a good position to control the new developers' experience from the OS to the tools to the deployment cloud (Azure).

But there is alternative OS - Linux.

Lately I know only devs that work either on macos or on Linux. I haven't seen Windows anywhere since few years (except when I do cloud gaming).

I know. I've been developing on Ubuntu since January 2009. Before that I was using colinux on Windows XP, a Linux kernel in a Windows process, somewhat similar to WSL.


When did 'locked in' evolve to mean 'using a product'? None of your examples are remotly akin to lock-in.

An example of a lock-in is excel back in the time when the format wasn't public. Wanna use something different? have fun manually re-entering your sheets in your new software.

How can GitHub CodeSpaces even cross your mind when thinking about lock-in? This is like saying the textbox on HN is a form of lock-in.

> An example of a lock-in is excel back in the time when the format wasn't public.

It wasn't public but it was reverse engineered and supported by things like OpenOffice. Anything, especially a file format, that is both proprietary and popular gets reverse engineered at some point.

- VSCode is open source, so I'm not sure how that constitutes lock in. Switching text editors is pretty easy.

- GitHub has some special things like issues and pull requests but in general you can just switch your origin URL to another Git repository that isn't GitHub and boom you are off GitHub.

- NPM you can just point your package.json at any Git repository you want, no need to use the NPM package repo.

- VSCode Online: like I said, just use a different online editor.

Not too much lock in going on here. Integrations and seamless workflow? Sure. But lock in? No I don't think so.

> Switching text editors is pretty easy.

Not only is it easy, VS Code made it easier than ever. By creating protocols for language and debugger support, they've liberated future text editor developers from having to support every language under the sun. By adding support for the protocol, the new text editor automatically gains support for all languages.

It's just a funnel. As a strategy, you 'commodify' the resources that you cannot control, especially those that are profitable to your opponents. MS is losing in languages, they don't own the important languages for mobile phones: java, kotlin, swift and objectivec and they have a hard time on the server against go and all the other contenders. With some luck the web remains open and typescript will remain an important language since javascript will keep on being important.

The moat for the editor is the plugin shop. We will be able to see how committed they are to being open. At least for now, their Live Share plugin doesn't seem to be available for other editors [1]. It's nice to have easy access to debugging but how do you sell your editor if you don't have all the other productivity features?

[1] https://news.ycombinator.com/item?id=23244434

The same way Slack used to have an IRC bridge, but they shut that down once they were the leading market player.

This is literally the Embrace/Extend part of EEE

Oh? And what’s the play after that? How are they going to shut down the protocol or language servers?

Even if they remove support for the protocols, those language and debugging servers still exist. Those servers are controlled by a wide variety of organisations, all of whom are committed to supporting the entire ecosystem of editors.

It doesn’t take much to regurgitate “EEE” on an MS article. If you can’t even imagine how the final E works in a specific case, maybe say nothing at all.

They'll just disable support for it in VSCode. Got a brain up there buddy? Or maybe you're just an MS shill.

If you can live with crappy lagging UI. Since the stupid Windows 10 skin VS code is unusable.

It looks like you replied to the wrong comment.

With regards to VS Code, there is some, although I'd say not much, lock-in into specifically the Visual Studio Code version of VS Code. Namely, many of the Microsoft extensions only working on the proprietary Visual Studio Code and not the open source VS Code. That does function to negate some of the no lock in benefit of open source.

I don't personally think that Microsoft is attempting EEE here but just bringing up the point that it is, to a certain extent, made hard to fork.

- https://github.com/VSCodium/vscodium/blob/master/DOCS.md#pro...

- https://github.com/VSCodium/vscodium/issues/240

> locked in

I've been on the Microsoft train since I was 8. I suffered no problems picking up Rust.

> "thought leader" by promoting the UNIX approach

This is a controversial opinion around here, but the UNIX approach/philosophy isn't particularly good. It is merely beloved. The Microsoft approach isn't much better.

A thought leader would actually solve this problem instead of repeating a tired, old and failed non-solution. UNIX sucks. Windows sucks. This circular appeal to nature (UNIX is good because UNIX is good) is getting really old.

> A ten year old today learning programming will be locked in.

If that's what you call "locked in" things are significantly better these days than they were 20-30 years ago!

As opposed to AWS/GCP and others?

There is almost no lock in to any product in 2020....people just love the developer ecosystem that MS has built these past 5-8 years or so.

to think about it... Microsoft also owns Minecraft, so a ten year old redstone scientist is already locked in, even without Git, NPM and all those things

Locked in ... to a game?

You laugh, but Minecraft Education Edition -- where lots of kids are getting their first exposure to coding -- is locked to Office 365.

yeah, sounds silly

And then there's the Bill & Melinda Gates Foundation that funds the World Health Organization & invests in Bayer/Monsanto among other medical & food entities. The vendor lock in is Cradle to Grave it seems...

Oh no! An "inconvenient" & related fact is mentioned. Better downvote the HN way.

With meaningful competition, the game is elevated for everyone. Those competing must win on merit, and those watching and consuming get the best output.

When Microsoft was the dominant software Goliath, they were winning in business and personal computing. The internet was new and they wanted to win that - not necessarily by being the best, but by using their leverage.

In the current environment, Microsoft is still winning in a lot of areas of business - competitive if not dominant. They have meaningful positions (but also competitors) in cloud hosting, operating systems and productivity software.

In my opinion, we're in a good place for a lot of these vectors. Cloud has Azure, AWS and GCP (though I don't know how well GCP is competing.) Operating systems have the classic gorilla in Windows, the polished closed system macOS and the persistent open source varieties of Linux. They are all meaningfully competitive.

Mobile is probably the weakest area, in my opinion. I wish Microsoft was still competing with Windows Phone here, because I don't love my Android and iOS options. They are good and they do what I need. I'm not confident I could switch between them easily (the way I am mostly confident I can switch between Windows, macOS and Linux).

(Online advertising is clearly the realm of Facebook and Google.)

I'm not sure how to summarize, but I don't think Microsoft is the same 800-pound gorilla that stuck us with Internet Explorer 6 for way too long. They'll (as long as they can) always be ambitiously seeking dominance in profitable software-related markets. They'll never be 100% charitable or philanthropic. And neither will their competitors. Is there a useful rubric or framework to think about all this?

What are personal decisions you can make that aren't too painful, but matter on the grand scale of these massive companies?

+1, and I wish more folks thought about Microsoft/Google/Amazon/etc this way. What we're seeing with Github, LinkedIn, Azure -- all the modernish MS moves -- feels more like the sleeping giant is waking up and _competing_, not like an 800-pound goliath throwing its weight around to extinguish competition. Github vs Gitlab is a _healthy_ competition, to the benefit of everyone (including MS and Gitlab!), same with Azure vs AWS vs GCP.

This is a _good_ state we're in, not a bad one.

Well in that case, Microsoft is really targeting developers.

VSCode, Typescript, Github, WSL, the inevitable tie into Azure...

I wonder how this all ends.

We can short circuit some of the inevitability I guess, still holding out for this: https://panic.com/nova/

Looks like Deno already adopted Typescript, but the world doesn’t have to revolve around Github.

AWS and Azure will keep each other honest.

Edit: I’ll add that I really really don’t think Microsoft wants to proselytize developers into it’s ecosystem just to make apps for Windows. I really think they are going to try mobile again.

Hypothetically, if they're able to provide a seamless developer ecosystem -- and a corresponding target platform for those developers to deploy to -- then they can guide the direction of the next wave of applications, and they may be able to incentivize and pick winners within it.

My fear is that we're in for a tedious decade where we see unironically-announced re-implementations of every existing iOS, Android, and Linux desktop application against this upcoming target platform, with corresponding re-occurrences of previously-seen bugs and security vulnerabilities to match.

I'd prefer to learn that we're going to standardize on open web technologies for delivery of software, and that for each case there'll be at least a few mature open source applications or libraries that fit the bill -- and that all that amazing developer capability and productivity is directed towards maintaining and enhancing those, instead of re-inventing them.

We'll see.

Right, we definitely would prefer the acceptance of open web technologies vs the shepherding into particular domains.

I mentioned in another comment, it’s nowhere near a stretch to see Typescript extended and compiled down to C#, where suddenly we are all on the .NET stack because we all got shepherded via Typescript. It can happen that fast.

Then we’ll forget what basic JavaScript is because we’re so used to scaffolding with .NET UI Kits and server apis. How vulnerable is the web community to this? Pretty vulnerable, a lot of us are tied into React or Angular or Vue, or even in more specific cases tied into a UI library like Material. Apologies for the narrow web centric perspective, I know there is a world outside of web :p

If I had to takeover with my stack, this is exactly how I’d do it.

I'm not following this train of thought.

First of all C# doesn't run in the browser without compiling to WASM which has all sorts of caveats.

Second, I don't see how the fact that one could compile TS to C# would suddenly put everyone "on the .NET stack". Does the same risk exist if we can compile JS to C#? Which would work about just as well (i.e. not). And TS is also open source and produces reasonably readable JS, so it's not like there's even any lock-in.

Wouldn't it make more sense to be worried about compiling C# or other languages to WASM? Which I'm also not worried about.

Your points are all valid, my concern is mostly that it’s a multi phase plan. We will get closer to something like that as we are shepherded.

Did you think Typescript would be the native language in Deno? No, you thought it was something that you might or might not use depending on the situation. Well guess what, that choice is gone for you in Deno’s case. These things compound.

It’s one of the reasons you will see people make a simple blog site with create-react-app, they are so inoculated with the hammer that everything is a nail. Did we think we’d be using React like that and forgo style sheets? Just inline all styles? That’s what we all do now. I had no idea we’d become that nuts, but it happened.

It’s not particularly bad in all cases, but for the web it is because we’ll stop using standard technologies that are supported across all types of scenarios.

Did you think Typescript would be the native language in Deno?

Yes? I thought that was the whole point. Actually I'm confused about the use of "native" here. It supports running both TS and JS files? And the TS files must be compiled to JS before they can be run anyway. So I'm not seeing any choice that was lost.

That’s what we all do now

I mean, some people do, but I don't. Not everyone uses React / Vue / etc, and afaik it's only React that is forgoing stylesheets. And it's not like stylesheets are dead or that styling elements through JS is non-standard. (I do use Vue, with stylesheets)

At this point Deno is really more of a counter-example to our grim dystopian Microsoft-only future: it's built in Rust (Mozilla) with userspace Typescript (Microsoft) running on V8 (Google).

I hold out hope for the nascent Deno community to find some way to build in Swift as a core language, too...

Not to forget LinkedIn, which is now being deeply integrated with O365, which has gained MyAnalytics (or whatever it's called) which is for all intents an automated "productivity" tracking software that aims to capture the time you're working, who you work with and how you spend your time.

I wonder when that data will be exposed on LinkedIn (hey recruiters, this guy is working late hours a lot!).

And then the eventual integration of that dystopian system with MS's developer monopoly. Now we have a truly "boring dystopia".

The worst part is: MS now has all the tools for this and more.

I've been waiting for the integration of active directory and linkedIn. It seems only a slight stretch to think of using corporate data and social data for cross-validation of true Identity.

Microsoft wants nothing more than a world full of .NET developers.

Worst case (MS moves to bad-acting), re: Deno—TypeScript can at least be forked for now https://github.com/Microsoft/TypeScript

I'd love to see Mozilla hop on board with it

What happens when Typescript gets extended and able to compile down to C#? Then it just makes sense to use a .NET stack.

Javascript ecosystem extinguished. Everybody will forget how to write on the open web platform.

_Yawn_. Wake me up on this thread when Blink/Webkit switch to .NET from V8/JavaScriptCore. Javascript (and thus Typescript as a higher-order lang) is _useful_ on mobile (hi, react-native) and the desktop, sure, but we're clearly a _long_ way from replacing JS in the browser.

Witness the delightful, slow-motion failure to launch that was Dart...

I hope you are right, it took no time for Typescript to catch on, and got ordained in Deno.

At this point I'd _rejoice_ if someone could offer a credible alternative to Javascript for web (and other browser-based/cross-platform mobile) dev. Google tried and utterly failed with Dart; Microsoft embraced JS via Typescript (extend and extinguish, sure, whatever; TS is fantastic); Facebook embraced JS via React (and React Native on mobile, though they've since pulled back a bit from that...); and Apple is predictably absent from the room.

There's pretty much no one left in the conversation who could credibly move us past an endless JS-only future of progressively-more-TS-like ES20XXs at this point...

It's sad to me that nobody would consider a new option unless a FAANG, or otherwise huge company, were backing it.

And the Deno team moved from Golang to Rust for their implementation. Maybe that will result in some interesting developments.

I prefer to think of this as Ryan Dahl choosing Mozilla over Google, whatever the relative merits of Golang vs. Rust.

Not that it matters, but I think it's much more likely he chose Rust specifically for the language features over Go. A lot of Rust's type-system and semantics are inspired by ML-family languages which are well-known to be great for writing parsers, etc.

At the initial presentation, Ryan mentioned that a move from Go to Rust was a possibility due to a few things, and they were related to the languages themselves, not to Mozilla vs Google or something.

(And, IMHO, saying that Rust is a Mozilla thing is just plain incorrect.)


microsoft, and apple also, should make desktop native apps easier to develop. and stop with the mobile chase.

only apps from the top 10 in every app store are used on a daily basis. customers are not willing to use anything else, unless required to -- i.e. bank, insurance, etc.

on the other hand, we are always willing to give an improved desktop app a try.

Eclipse is making open-source registry for VS Code extensions https://open-vsx.org

support it if you are afraid that MS could pull the plug on their Marketplace in a few years

I’ll never forget muttering that to myself when Washington state passed a law banning smoking within 25 feet of doorways, but Microsoft made it 30 feet on the signage on campus.

Extinguish only works if it is a monopoly. Back in the day Microsoft had a monopoly on the PC operating system market and microsoft tried to extinguish the web. Microsoft still has lots of monopolistic practices built into the bootloader. Not sure if driver makers are forced to work only with windows these days but thats definitely the case with hololens.

It would be more interesting for microsoft to take a stance on privacy like apple.

I'd love to hear Torvalds' opinion on this DirectX PR. In any case, as someone who hates complexity and loves simple software (in the suckless sense), whenever I see these monstrosities get merged into the kernel, it makes me search for alternatives. There was a time when the kernel codebase was small and easily understandable by 1 person, apart from the drivers. Apparently simplicity and keeping complexity at bay is no longer a goal for Linux.

BSDs do a bit better here, but they are still behind on several metrics. Maybe this will be the main motivation for the next successful kernel? The fact the Linux simply got too big and complex?

As some of the replies on the LKML suggest ( https://lkml.org/lkml/2020/5/19/742 ) having this as a simple accelerator pipe, but simple in the Linux perspective of re-using as many conventions and existing infrastructure as possible, and living _not_ as a display adapter (which will surely become far more complex later) would probably work for everyone; right now.

The question is strongly dependent on future developments. With this being contributed as a GPU driver and the likely hope of this being used to run Linux applications within another OS, the question might be re-framed:

How is this addition to the Linux Kernel API going to benefit _all_ users of Linux, not just users of WSL?

Were Microsoft trying to provide a thin abstraction layer, so that drawing to a vGPU shard from windows on the application facing side of the Linux Kernel API looked like any other standard GPU under Linux, I feel it very likely everyone would feel positively about the changes.

If this were about contributing Microsoft's vGPU driver for Linux, and in so doing also implementing only Freely Implementable Industry Standards (like Vulcan, OpenGL, or maybe others that Microsoft made/makes similarly free), while also not neglecting the expected core standards for other GPU drivers, then that also seems logically acceptable for everyone.

Though my above suppositions are purely my opinion and guesses about the opinions of the maintainers for the software, while trying to think about how they approach their inherent responsibility for the long term maintenance and stability of a project they are clearly passionate about.

Cisco has entered the chat

Saw this while scrolling Reddit, and thought it is relevant here as well.


Was thinking about this earlier, saw some article about how Microsoft is trying to get Windows 10 to run Linux GUI apps.

But also Spotify! There's a strain of thought out there that the podcast ecosystem is resistant to centralization, but Spotify is making a serious go at proving this wrong.

>saw some article about how Microsoft is trying to get Windows 10 to run Linux GUI apps

I thought that exact same thing when I saw that article yesterday.

Interestingly Linux has its own corporate champion: Redhat/Poettering. Currently they're the villain in most Linux circles, but it'll be interesting to see them and MS square off, as they inevitably will.

Windows 10 can already run GUI apps using WSL and an X server. Microsoft is planning to build a native X server so it's frustration free. Personally, I cannot wait.

Anyone have context as to why this is on the front page?

This is about the directx for wsl post yesterday. Some people fear Microsoft is trying to do this to Linux, although others disagreed.

This probably connected to this post.

'DirectX is coming to the Windows Subsystem for Linux' https://news.ycombinator.com/item?id=23241040

People think that EEE is what Microsoft might be doing in this case.


Microsoft is taking over the world. They now own Github, NPM, Citus data, LinkedIn and a ton of other companies. Not to mention they have a trojan horse on almost every new developers machine thanks to TS and VSCode, and all of the data and services that companies are trusting to them on Azure. They are poised to become a data king.

TBH... if I had a bunch of liquid cash to fuck around with I would be buying a ton of MSFT becuase my gut is telling me they are going to really take a huge slice of the open source development pie here in the next few years.

They know that is the market they need to capitalize on which is why all these acquisitions and investments are happening.

There are a lot of young developers that just have no context for what Microsoft did in the past. If you bring up their past, they say "that was so long ago" or "but they've changed!"

But public companies are not people.

Of course they change -- tactics -- and try to gain control again. Everything they do will in the end have to be justified to their shareholders and how it will either gain them profit either now or in the long run, and as soon as the fickle-minded shareholders change their tune, out goes Nadella, and we'll get another Balmer or worse.

I’m on the fence. I resonate with your take on things, but I also have a bit of hope in me as well. At least outwardly their optics are looking much better and brighter, but I’m not sure what the real culture is like at present day.

Your optimism is inherently misplaced. Power is corruptive. Even if the people in power now aren't corrupted, there is no fail-safe in place to pull the plug if they ever hand off that power to someone who can't control it. It will happen, it always does.

There is historically no such thing as a trustworthy super-power unless it subsists overwhelmingly on the support of discerning people.

You may hop off the fence if they do something like, oh, I don't know, buy Canonical to counter IBM buying RedHat. I'm surprised they didn't do it immediately after that.

And this is the saddest part of it all: the industry slowly started to drag itself out of the mire that is Microsoft monoculture, as servers were moving to Linux even in evil-corp style organizations; cloud services offering competition to Office hegemony (especially in education); IE6 dead and its descendants soon-to-be-forgotten; .NET being forced to improve itself and to become only 5 years behind the curve instead of 10 years (but some places still seek developers for the Lovecraftian horror that is Web Forms).

And now all this, all over again? Goddamn this industry.

What are public companies if not people?

It seems to me what you're complaining about is less about Microsoft and more about capitalism and greed in general.

People in groups behave differently from people alone; people considering billions of dollars in profits behave differently from people considering tens of dollars in profits.

I can say that not every new developer loves MS. I use vim, avoid JS, run Linux etc. As things currently stand, I'm considering migrating to BSD due to MS's new... love... for Linux.

It's worth noting this is the 16th time [0] this article has been submitted to HN.

[0] https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Perhaps its from yesterday's discussion on Microsoft Lists which is similar to Airtable:

Microsoft announces Lists, a new Airtable-like app https://news.ycombinator.com/item?id=23241084

- Embrace, Extend, Extinguish is something different. It's about adopting standards in order to break them and gradually lock people into proprietary variants. This guy is talking about just copying existing products.

Embrace (the UI) Extend (integrate with office) Extinguish...

Seems similar enough to me. Just thinking of UI as a feature.

I figured it was a Google thing. Still being worried about Microsoft sounds like that classic Guardian article about the unending Myspace monopoly. But the EEE philosophy is all over Google now.

Assumed it was due to the article yesterday on Facebook "embracing" Shopify.

It is strictly my personal opinion. From the WSL + Direct X announcement of today, I see the same old M$ that I knew 20 years ago. They simply learnt new tricks.

Microsoft does things well. If you don't want EEE to happen, just be better than Microsoft. As a consumer we just don't care whether Microsoft is the axis of evil or something. We don't give a f. We just want things to work and work well.

If you're the "just work" type go buy Apple devices and be done with it. Not all of us want fisher-price-style products.

"fisher-price-style". Some people just want to get to work on a machine they are comfortable with rather than worrying about the looks of their machine.

But I did and am using, along with lots of people.

so apple is ok but microsoft is not? because...?

Microsoft's recent involvement to open source is currently a bigger threat to OSX than it ever was to Linux.

Users who use Linux distros are content with their overall desktop experience in the absence of closed-source libraries such as DirectX, etc. They were never the target audience Microsoft gains from embracing because not many would have felt the incentive to do so. Especially considering their efforts to port and upstream changes to the kernel, lessening the allure. Albeit not that Microsoft is doing this out of the goodness of their hearts, they benefit from doing so.

Meanwhile, the Apple developer ecosystem has not made any significant strides. It has been a while since homebrew that OSX has gotten a useful developer tool, and ever since Lattner departed from both of his projects related to Swift (TF), it has slowly been losing traction.

Many Mac users choose OSX for the neat desktop experience Linux struggled to offer, and the *nix-like interface Microsoft had been missing. WSL and Microsoft's recent contributions to Linux has changed the Windows developer experience significantly, giving many previously hesitant Mac shops the nod.

I wrote this in another comment so I'm just gonna repost it here. I really don't get how is this a problem. Now you can use acceleration for ML under WSL. Great. If you don't want to do that, what's stopping you from using ML on native Linux as you're doing right now? They are not changing Linux. They are adding something to Windows. Why are you all so threatened all the time? Even if Microsoft does... idk what you think it will do, you can still use your favorite flavor of linux with the tools you want, the language you want, the framework you want.

And to quote many people from open source: you still have the source code so if there is something you don't like you can always change it and build a version that better suits your needs.

How is IBM's or Oracle's involvement in open source not a bad things? Is Oracle really someone who's suddenly ok?

From all these comments, all I see is: someone is going to build a different toy and many will want to play with that instead of this.

"Many Mac users choose OSX for the neat desktop experience Linux struggled to offer, and the *nix-like interface Microsoft had been missing. WSL and Microsoft's recent contributions to Linux has changed the Windows developer experience significantly, giving many previously hesitant Mac shops the nod." Okay, so? It's Windows. Feel free to ignore it. How does that threaten your emacs/node/c/python/gnome/tensorflow working environment?

When docker wasn't working on Windows nobody cared. Why should I have to install vmware or dualboot just because someone can't be bothered to make docker work normally on windows?

> Now you can use acceleration for ML under WSL.

Using libdxd12.so, which is closed source and works only under WSL.

> They are not changing Linux.

They are literally asking the Linux kernel to pull in their changes to add a proxy to Windows's device driver model to the Linux kernel. No changes to Windows; all changes to Linux.

> They are adding something to Windows.

Windows already has DirectX, so no, they are not adding something to Windows.

> Why are you all so threatened all the time?

Because MS has a proven track-record of screwing others over. It is only wise to heed a potential threat.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact