Microsoft Joins the Blender Development Fund | Hacker News Reader

Microsoft Joins the Blender Development Fund

blender.org

377 pointsantoineMoPa166 comments

There is a lot of implied ill-will on HN, I'm guessing the phrase "embrace extend extinguish" in Microsoft related threads gets upvoted? From the post I see that many Microsoft employees (at least in research) use Blender. Microsoft decided to pay for the software their employees are using, which happens to be open source, so the best way to do that is to donate to their foundation. Epic Games, Google, Nvidia, Valve, AMD, Intel, Ubisoft and others have done this (looking at the list of corporate sponsors is easy). This isn't a plot to take over Blender and destroy it from within...

Exactly. According to the membership page[1], Microsoft will be paying between €30k/year and €120k/year[2] to the blender foundation, along Intel and Ubisoft. On the other hand, Nvidia, AMD and Epic games[1] are contributing €120k+/year each.[2]

All of these companies get "direct access to the Blender team for strategical discussions. Roadmaps and priorities will be aligned with your requirements as good as possible."[2]

I don't see how they're going to "extinguish" blender. With this kind of reasoning, OpenBSD was also "extinguished" in 2019.[3]

_____________________________________

[1] https://fund.blender.org/#credits

[2] https://fund.blender.org/corporate-memberships/

[3] https://www.openbsdfoundation.org/campaign2019.html

The extinguish phase is supposed to come after the embrace phase. Even if Microsoft wanted to extinguish OpenBSD, it's no wonder they are still fine if MS just donated last year. That is not "this kind of reasoning" at all, that is just a strawman.

>"direct access to the Blender team for strategical discussions. Roadmaps and priorities will be aligned with your requirements as good as possible."

Call me a language snob but this writing is so poor.

"Strategical"

"As good as possible"

Can I submit pull requests to that page?

I don't think so, this seems to be managed by a wagtail based django project: https://git.blender.org/gitweb/gitweb.cgi/blender-dev-fund.g...

So this text copy is most likely in their CMS database.

If I'd have to venture, it was written by a non-native English speaker.

Yup, Blender are based in Amsterdam.

The whole point of this kind of thing is to be able to influence the software that your business relies on.

It's certainly possible that Microsoft (or any other company with the $$$) can influence the strategic direction of Blender in ways the harm competitors and/or other Blender users. I'm not sure why that's such a controversial concern to have, that kind of thing happens all the time. There are likely mechanisms in place to protect against it, but what are they and are they good enough? I don't think "just trust me" is a good approach to take.

Whenever I complain about any piece of OSS, I get told to submit a pull request or fork it on my own. If people are that concerned about MS having influence over Blender, why don’t they just fork it and maintain their own branch? That’s the point of OSS right?

Theoretically sure you can "just fork it". In practice, for complex software, it carries enormous cost. In that case "just fork it" means "organize dozens of engineers willing to volunteer their time (or pay them)". For software complex enough, forking just isn't a feasible option.

And now we see why big companies are willing to pay the Blender team to continue development.

That was never a mystery to begin with.

I’m having a very difficult time reconciling your initial post that sounded like you were unhappy that Microsoft was paying Blender developers with your statement that it can’t be forked because it requires too much money to develop followed by your conclusion that there’s no mystery to why big companies pay open source contributors.

Either Blender is so big that it can’t be maintained without the financial support of big businesses, or the financial support of big businesses are a negative influence on OSS projects. Which one is it?

> Either Blender is so big that it can’t be maintained without the financial support of big businesses, or the financial support of big businesses are a negative influence on OSS projects. Which one is it?

It can be both. That’s what my original point was. I’m not criticizing Microsoft or disagreeing with this arrangement. Other commenters are assuming this false dichotomy as well, which I think is invalid. A company can certainly support an OSS project financially, while influencing it in ways that harm competitors and users.

I have no opinion on Microsoft’s motivation, but users and competitors would be wise to consider what mechanisms are in place to protect their interests as well.

In theory, yes. In practice, many people hate certain privacy-related aspects of Chrome, so someone created an ungoogled version. You'd think it would become an instant hit, right? Wrong: most people on HN commented they would never install a browser built by a random person. It's paradoxical but this is how things are. You'd need to have a very particular position in order to be successful as a fork (e.g. MariaDB, LibreOffice). Most forks just die out as not even their creator uses them.

Not to mention that Chromium is massive and complex and unwanted parts of the code have been found and removed long after they were added https://archive.is/4VijY.

The mechanism is that the software is developed and released under GPLv2+ terms and that Microsoft doesn't completely control ownership over the software. Ergo, there's no "extinguish" step. They can't make their changes a proprietary fork nobody else can use, they can't force you to use changes you don't want to get access to changes you do, and any other person making changes to Blender can use theirs.

Yeah, Blender make it super clear in the link you provided ( https://fund.blender.org/corporate-memberships/ ). Pay 30K euros per year, and a Blender dev will work on the features you want for 6 months per year. This is way cheaper than Microsoft forking Blender, and paying their own fulltime devs ~$200K USD/year to work on the fork. Plus the Blender devs know the software really well, and are going to be far more productive, don’t need a Microsoft dev manager, etc. No conspiracy theory needed, this is just a cost effective way to get the features you need added to Blender.

Damn that's really cheap. I wonder what happens if deliverables are for whatever reason not delivered? What visibility and cadence does a corporate sponsor have with their assigned dev?

I assume it's not full-time for 6 months. More like this feature is concretely assigned to a dev.

Also it's a donation not a contract.

They probably can't extinguish Blender any more than Oracle could extinguish MySQL (Larry must have been really mad for that), but I remember what happened when they acquired and ported Softimage to NT. It worked for a couple releases, then it was abandoned, after killing killing SGI by validating NT as a 3D animation OS. The Softimage acquisition, as a business, made little sense to Microsoft - it was a low volume high margin business, completely opposite to what MS does.

SGI made plenty of suicidal moves by itself, but I remember the Softimage thing was a big blow.

I always thought Microsoft did this in order to make inroads into the high-end graphics market, which they did.

That was why they bought Softimage (the then leading 3D animation tool) back in the 90's. Today Microsoft has ample presence in high end graphics.

Microsoft acquiring Softimage made perfect sense at the time, Microsoft was wanting to push NT into the CG/VFX market, but because of the risk around companies producing for a platform that noone uses, someone had to throw the capital at it, and so Microsoft did.

Softimage|3D survived for a while after Microsoft sold it off, its eventual obsolescence seemed to came about from the drive to more extensible platforms, hence the new platform Softimage|XSI put out to compete with Maya.

Its also somewhat perfect irony that the company known for its Windows NT based 3D animation software (3DS MAX), Autodesk, would eventually end up owning the products which brought the big players over to that platform.

SGI also helped contribute to IRIX's obsolescence by producing NT based systems for this new Softimage|3D on NT move.

> Microsoft was wanting to push NT into the CG/VFX market

That's what I meant. In isolation, the business meant no sense. It was a PR move to validate NT as a graphics platform. With that goal accomplished, I got the impression the product languished without significant development.

> SGI also helped contribute to IRIX's obsolescence by producing NT based systems for this new Softimage|3D on NT move.

Indeed. Microsoft assisted SGI in its suicide.

MSFT isnt trying to do that, just read the blender article, the link provided says what MSFT uses it for. And I think Blender stands a chance to gain some good AI assisted toolset if that is what they use blender for.

https://www.microsoft.com/en-us/research/uploads/prod/2019/0...

I used blender since the beginning. It makes me so happy to see it evolve over the years, first to open source then to a widely supported project.

Donating my meager highschool pocket money to help buy the source code ended up being the most impactful thing I've done in life so far.

My story is similar. I was near broke, but I was having so much fun with Blender at the time that I made the donation. I had been an Animation Master user on Amiga in the years before Ton bought the source code. Good times and great to see Blender flourish.

The only times I see embrace extend extinguish get brought up on HN is when there is legitimate concern. Obviously, on something like this if someone is bringing it up it would appear inauthentic. However, I frequently see inauthentic articles spewing Google hate that get upvoted and anything that gets brought up suggesting it is manufactured is automatically downvoted.

One day I noticed some wag in a youtube comment section opine that "embrace extend extinguish" should really be called "Control, Alt, Delete".

Normally if management approves this kind of spending, they have to be convinced. You need to properly justify your proposal: what kind of competitive advantage this would give us and so on. Companies don't spend money just to "have some influence". They want to have this influence for their advantage. And unfortunately Microsoft has a very long history of misusing this advantage.

The "embrace extend extinguish" only makes sense when the company doing it has

1) The ability to extinguish 2) Or the business model of the embraced company is seen as a threat.

It's funny when Microsoft launches WSL and "loves Linux" how quick people are to use EEE. But MS has not the capability to kill Linux (1) (that sail shipped long ago) and it's not aligned with it's current business model (2)

About Blender... How is supposed Blender to challenge or threat MS?

They appear to be using blender to generate data for machine learning!

So if you want to be able to recognizes poses or hand gestures from video or image stills, you need data to train your network.

The bottleneck is often high quality 'labelled' data - real data that requires people to laboriously label each frame and region with the 'ground truth'.

However, if you automatically generate the poses using something like Blender, you can also automatically label, and also generate a huge variety of systematic variations in camera angles, lighting, levels of noise etc.

Obviously there might be a worry that the Blender generated images may somehow inject their own bias, but they claim it's working really well.

https://www.microsoft.com/en-us/research/uploads/prod/2019/0...

Reminds me of how the first version of AlphaGo used recorded games played by humans, but later versions were able to surpass that using self-play.

For the final project of the multimedia information retrieval class in my CS Master's I'm doing something similar.

The goal is to predict depth maps from single-view pictures using CNNs. But how to gather input, output pairs to train them?

I'm using a mod for Minecraft I wrote to generate such labeled data, e.g. input: https://i.imgur.com/JRtmCeG.png output: https://i.imgur.com/q4KocXn.png.

I'm not done yet so I can't tell you the results, but it's a fun project.

1) Blender is GPLv2'ed Software,so there is no immediate danger of MS creating proprietary fork of Blender and making money by selling the fork. I beleive that they are contributing to Blender which benefits community as a whole.

2) Offtopic but i wish megacorps collaborate to create FLOSS alternative to Adobe Suit like Photoshop/Other Graphic Design programs. This would super benefit everybody on all platforms.

> create [emphasis added] FLOSS alternative to Adobe Suit [sic] like Photoshop/Other Graphic Design programs

What's wrong with Krita, GIMP, Inkscape, etc.?

Gimp's interface leaves a lot to be desired (to put it mildly), while inkscape has nowhere near the feature set of Illustrator.

What are some Illustrator features that Inkscape is missing? I edited a few SVGs in Inkscape yesterday and didn't feel like anything was missing.

Inkscape is focused on SVG, Illustrator's proprietary format includes a few things it doesn't, like complicated gradients (at least a while ago). Edit: https://commons.wikimedia.org/wiki/Help:Illustrator#Gradient...

“Should the width of a shape be defined by the inside, outside, or center of its border” is (box-model in HTML terms) is a thing I use all the time in illustrator, but is missing in SVG :(

Krita is fine for painting but isn't a full replacement for Photoshop. Inkscape does vectors, which is something else different entirely.

If GIMP prioritized matching Photoshop's UX to make it easy to switch over (shortcuts, UI similarity, etc.), it'd be much more successful at converting Photoshop users. Serious Photoshop users at this point have spent more than a decade (!) committing these things to muscle memory.

GIMP hasn't made replicating Photoshop's UX a priority, so it isn't a replacement for someone more comfortable using the industry standard tool.

If I used Vim for a decade, then you said, "Notepad++ also edits text, why don't you switch over?" Well, I'm not as productive in Notepad++, and I don't think I'll ever be. Same for GIMP.

They said the whole Suite. Inkscape is Illustrator in that sense.

GIMP should do a Blender 2.8 and do a major focused release on improving UX

I suggested this a few years back but the community grilled me for it. I don’t think it’s the kind of ecosystem that wants to change.

There's a project called GIMPShop which was supposed to make the UI of GIMP more like Photoshop; I don't know what the current status of it is though.

Just be aware that the dot com domain for gimpshop was not created by the original author and contains malware in the download. https://news.ycombinator.com/item?id=7481091

They should have an 'industry standard' keymap option on first start like Blender. But it's not nearly as weird as Blender was: there's no right click select, and eg. the cloning tool is basically the same thing as Photoshop's, though the large variety of select tools is a bit intimidating.

I thought that's what gimp 3 is supposed to be?

This is what you're after (or will be): https://glimpse-editor.github.io/

Didn't they received a lot of funding? What features did they implement that are not present in GIMP?

You can see what they've done and what they're working on here: https://github.com/glimpse-editor/Glimpse/projects/1

I've read all the release notes and I couldn't find anything that wasn't upstream sync or redistribution, packaging, irrelevant stuff.

I just want to understand what they delivered since June, 2019. Could you name a couple of examples?

Yeah, they seem to be off to a fairly slow start. This isn't the first Gimp fork to try to work on UX - but biting off a huge old codebase isn't easy and previous efforts have all been overwhelmed or overtaken by upstream and fizzled out, afaik.

Although I agree, one key difference from this case and your analogy between Vim and Notepad++ is you don't save a few hundred/thousand dollars switching from Vim to Notepad++.

That's an extra factor that, at least for me, made it worth switching and investing some hours into adjusting for hotkeys/shortcuts/etc. To be fair to your point though, I don't have a decade of full-time work invested into Photoshop so perhaps my muscle memory was more malleable.

Not only the money is the issue, also your livelihood if you live in a country where the US can decide from one day to the next to sanction it for what ever reason, and cutting your access to any cloud service.

i'm not convinced the issue is GIMP not matching Photoshop so much as it having awkward UX that's hard to adapt to. Also, Photoshop 1.0 was released in 1990 (!)

GIMP is ...not very good. Fixing the UI and UX wouldn't change that. Compared to GIMP, as a project, Krita is in another dimension.

If any software is going to become a successful FLOSS Photoshop replacement it's either Krita (if they decide to go that route instead of just painting) in a few years, a new open source project with an impressive MVP (like Olive the video editor, to Adobe Premiere), or some kind of miracle like one of their commercial competitors deciding to open source it.

Nothing wrong. They are good in it's own sense. But,they are less adopted and there are lot of desirable features that are not present.

1) Megacorps creating and popularizing means greater adoption. Greater adoption is good thing.

2) FLOSS Graphic Design SUITE is good. Ideally, it can be made to run on every platform.

gIMP isn't remotely at the same power as Photoshop. Photoshop's non destructive editing layers, layer effects, smart layers, and 100s of other features are no where to be seen in gIMP. It's almost like comparing Notepad to VSCode or MS Word. gIMP's FAQ says they are trying to add that stuff to gIMP in gIMP 3.2 but it's been part of Photoshop since like 2003

https://www.gimp.org/docs/userfaq.html#when-will-gimp-suppor...

I don't know what other analogy to make but it's a hugely powerful feature and arguably essential feature which basically puts Photoshop in a separate category from gIMP, at least until they add it to gIMP

> It's almost like comparing Notepad to VSCode or MS Word

That's pretty hyperbolic. It's not like GIMP doesn't have basic layers, blend modes, etc. And Krita seems to have better non-destructive editing than GIMP.

I don't think GIMP is going to blow Photoshop out of the water anytime soon, or is even too close to as competitive in its area as Blender is in 3D, but I've used Photoshop and Illustrator for maybe 50 hours and Photoshop wasn't drastically better than GIMP for basic stuff, composing a simple photoreal scene from individual photos, simple graphic design, etc.

To be blunt, if that's your impression then you don't know how to use photoshop. If you did you'd understand the vast and extreme differences in capabilities.

I don't have a good easy analogy to make but the two pieces of software, Photoshop and gIMP are not remotely in the same league.

If that’s your impression you either can’t understand my comment or won’t / are trolling. I took a high school class taught by a professional graphic designer, using PS and Illustrator. I’m not claiming that’s my career now, but not everyone uses Photoshop as a fulltime job, and GIMP's good enough for plenty of use cases (eg. a one-person team indie game with art made entirely with it) (and can do stuff PS can't easily do with better scripting).

Simply put, a FLOSS alternative to professional software isn't an alternative if it's unable to be used professionally. The barrier for freelance and hobbyist work isn't people's concerns with dethroning these proprietary applications.

Can you get hired at a large studio with Blender on your resume? Absolutely. And Krita is slowly heading down a similar path.

GIMP, on the other hand, is dead-set on being relegated to a corner for FLOSS purists. It still doesn't have a complete CMYK workflow. It can't compete with Photoshop, and its authors have stated they don't want to compete. And until something comes along that can, people will continue to learn and use the proprietary option first and foremost.

This seems mostly true, but on “a FLOSS alternative to professional software isn't an alternative if it's unable to be used professionally”: Photoshop is not exclusively professional software. Last I knew, it costed $10 a month. That's a serious barrier to entry to exceptionally poor residents of slums within developing countries, like the disabled boy who lifted himself out of poverty with GIMP and other FLOSS (https://www.gnu.org/education/edu-software-gimp.html#content), but it's basically pocket change to privileged people in the first world like plenty of YouTubers I've seen use PS as part of hobbies like tiny indie gamedev. This seems like an OK niche for GIMP to target while Krita gets more professional. It's tempting to want them to unite and make one raster editor to rule them all, but it doesn't seem necessary. GIMP doesn't seem to need dedicated full time devs and effort comparable to 3D software, which is more inherently complex and in proprietary form is licensed for orders of magnitude more money than PS.

A big workflow in modern Photoshop is stuff like Layer Styles (non-destructive image editing). GIMP has no equivalent to that (most of its filters are still destructive, despite the 10+ year project to switch to GEGL).

There's plenty of other weird awkward UI and mental model decisions (e.g. GIMP makes the user select the width and height of each layer up-front, when Photoshop just assumes an infinite canvas and computes an AABB for each layer, leading).

A good portion of the work I do in Photoshop is vector editing (e.g. any real composition has dozens of shapes and even purely pixel work is going to use a lot of vector masks), GIMP lacks any vector tools and the team has been opposed to adding it, because it thinks Inkscape is better suited to vector art. Which, I mean sure, but it misses the point! Vector masks are incredibly valuable for compositing work, and we shouldn't have to switch between two radically different apps to have a usable workflow.

inkscape and krita are good but gimp can burn in hell. even extremely simple operations like making a section and then moving only that part of the selection on the current layer somewhere else requires triple key combinations you have to google. the interface seems like it's designed to be as frustrating and unintuitive to use as possible

Photopea.com seems to have matched photoshop UI and runs in the browser. It's not open source but from what I understand it's an one man project which means it's possible for a small dev team to create t an open source photoshop alternative from scratch.

Have you tried www.photopea.com? It's a Photoshop clone.

AFAIK, photopea is not FLOSS.

Photopea doesn't look like FLOSS and it is just webapp . It also doesn't have much features.

It's plain to see how Blender benefits the big companies that are supporting it. So how would a FLOSS alternative to Adobe CS benefit anyone except end users who can't/don't want to pay?

Maybe easier extensions like GIMP's Python-fu, Scheme Script-fu, and G'MIC.

Microsoft has dozens of game studios that enjoy using Blender, and no competitor in the space (not recognized as a profitable business). Finding negatives to this move is being too paranoid.

I doubt any of Microsoft Game Studios seriously use Blender, although I'm sure some individuals might use it for specific workflows. AAA games almost universally use Maya (and ZBrush, Substance and Houdini), for example this presentation on Gears 5 workflow specifically mentions Maya and Houdini (https://cdn.gearsofwar.com/thecoalition/publications/The%20V...). This donation specifically mentions digital humans created for Microsoft Research, and MS Research as the user, not Microsoft Game Studios.

It’s more than games. Hololens/mixed-reality has a boundless amount of modeling and more.

Yes, and the article states that. My response is specifically talking about games, because OP addressed Microsoft Game Studios. Microsoft Research and Microsoft Game Studios are different groups. Folks are acting like I said "Blender sux lol". Simply stating facts about usage.

Blender skyrocketed in popularity in last 5 years. I just went through local gamedev job openings site (skillshot.pl) and over half of 3D Artist postings mention Blender.

I never said that Blender is not used in game development, I specifically said Blender is not widely used in AAA (Microsoft Game Studios) game productions, and that Microsoft Game Studios is not the driving force for this donation (which is stated in the origin post!). This should not be a controversial statement.

> I specifically said Blender is not widely used in AAA game productions

This tells me you may not have been keeping as close tabs on industry practices. I can tell you personally, this has been rapidly changing over the past few years, and it's taken a lot of people by surprise. I can't say it's supplanted Maya yet by any means, but that software's quickly being seen as old-hat by each new wave of gamedevs. It's fascinating to see.

If you have examples of AAA game developers that have moved their entire studio (not individuals using Blender, or outsourced contractors making one off assets) to Blender I'd like to hear it. I'm aware that indies, including many successful ones, and potentially some mid-tier or AA studios, widely use Blender. I'm not talking about that. I'm talking about Naughty Dog, Blizzard, Treyarch, DICE, etc. I keep very close tabs on industry practices and work for a major game engine company. You yourself said "I can't say it's supplanted Maya yet by any means", I'm not talking about potential future growth, I'm talking about the current state of the industry in a very specific product category.

You are correct. Perhaps it is more widely used across their research and other divisions. They just went on a game dev studio buying spree though, and I thought there was a correlation.

MS gaming division is also one of the most notorious for lock-in attitude and market manipulation (the whole DirectX push, refusal to support Vulkan and so on). So may be this move is neutral, but they surely don't deserve any trust by default.

Still waiting for those wonderful Khronos SDKs to come out.

Apparently LunarG SDK was the best they could make, which is quite telling.

Love how Vulkan extension count keeps rising every couple of weeks, not long to catch up with OpenGL 4.6 extension jungle.

This isn't a trolling platform for lock-in shill posts.

> Finding negatives to this move is being too paranoid.

and I've been around long enough and been stung hard enough and enough times by MS[0] to say that not expecting negatives in this risks naivety.

[0] I've just wasted 12 hours trying to install visual studio and that's their own bloody product!

I'm surprised how low a Corporate Gold membership is, just €30k a year. They claim that is half of a developer year.

https://fund.blender.org/corporate-memberships/

Is the assumption that beyond that you would just hire your own developer to work on blender?

€60.000/yr seems reasonable. This is the average salary in the Netherlands (home of the blender foundation).

Could you provide a source for this statement because I'm pretty sure it's actually closer to 45k/y for fairly experienced developers

No, ezperienced C++ devs with the math skills you need to do 3d dev make 60k or more, 45k is just above fresh graduate level. I don't have an online source, just from what I know from companies and people I know, and doing hiring myself in the past. I made more than 45k 10+ years ago when I was an employee.

I was on approx 60k eur in Amsterdam a few years ago with 3 years experience

I should have been more clear. My point isn't that the salary comparison was off but that €30k/y is peanuts for a large organization. I would expect to see some higher sponsor levels.

I just started paying attention to Blender again after some years. The tutorials Ian Hubert does really demonstrates how powerful the product has become - https://www.youtube.com/watch?v=lY8Ol2n4o4A

Came here to say this. This video of his really opened my eyes to what Blender can do (https://www.youtube.com/watch?v=lY8Ol2n4o4A) and makes a cool entry point into that world.

Man, this software has come a looooong way since I first touched it back in 1997. That is absolutely amazing.

Microsoft supporting BSD and MIT licensed open source makes sense; this has become commonplace under Satya's leadership. Aside from contributions to Linux to make it work well in Azure or Git to make it work well for Microsoft's rather unique issues, their supporting GPL licensed products is yet another stark reminder that this isn't Ballmer's Microsoft -- and that Microsoft might actually be doing more for F/OSS software than the FAANGs of the world are at this point in time!

I use lots of commercial software, so this is more a kind of philosophical question.

Many that still bash Microsoft for EEE practices, are the first in the line killing GPL based stacks, pushing BSD and MIT licenses, which most of the time end up with selected updates in upstream and gold features only on the commercial products.

Which kind of makes sense, supermarkets don't take PR, but it is a kind of two weight two measures.

Are there any good examples of embrace extend extinguish under nadella's tenure? I keep seeing this meme everywhere (including in this thread) but git is doing fine, Linux is doing fine. The only notable things I've seen extinguished are their own windows phone and phone hardware platforms.

The fax machine was still a core part of business life back when embrace-extend-extinguish was a thing.

It's hard to tell whether those who continue to fixate on this 20-year-old trope are cranky senior citizens who'd otherwise be yelling at people to stay off their lawns or young zealots pining for the days of a long-past battle they were born too late to actually participate in.

The rest of us disposed of our fax machines and rotary dial telephones and moved on as the tech landscape, including 100% of the leadership at Microsoft, continued to turn over as happens with time.

Hell hath no fury like nerds scorned.

Extinguish, I have no examples. But Embrace and Extend, here is a possible one. Not saying I think this is a case of the first two E's necessarily but I know other people have pointed it out.

1. Build WSL2 and incorporate Linux kernel into Windows with great integration. 2. Add DirectX and GPU acceleration support for Linux that works in WSL2 only and not in regular Linux: https://www.phoronix.com/scan.php?page=news_item&px=Linux-GU....

These two steps could certainly be looked at as Embrace and Extend, although I admittedly fail to see how they could take an Extinguish step from here.

Well, the way directx on wsl works is it redirects dxgkrnl escapes/ioctls to host windows system something not possible on native Linux.

Well, that doesn't matter in the end. Part of extend is to take whatever you embraced, and add things you cannot do. So it's a success in that sense. Seems you're trying to put it as "What can Microsoft do if Linux doesn't support it?" when they can, if they want to, do plenty about it.

I think you're wrong there, bringing dxgkrnl to Linux would have been a much more involved process instead of just providing a redirect.

dxgkrnl is by far the heaviest NT module afaik.

> bringing dxgkrnl to Linux would have been a much more involved process instead of just providing a redirect

Of course, that's how collaboration works. Instead of going the harder but "improving the entire ecosystem" way, Microsoft chose the easiest + the one where they can extend stuff in a non-compatible way.

The way you look, you'll find malice in every little thing.

I thought they wanted to bring DX natively to Linux as well but got rejected. That would be an extend that easily leads to extinguish. You get Linux application developers to use the proprietary API to build stuff, then reduce support for everything but "MS Linux". This is why people need to be vigilant about the freedom of the entire stack.

Over time we may see MS playing nice, but there really is no value to FLOSS developers in having a closed source graphics API on Linux.

DirectX on Wine on native Linux seems to work okay already. GPU on Linux is a seriously mixed bag, but this is largely nVidia’s fault, not Microsoft’s. GPU on Windows works okay because Microsoft is willing to play the let-nVidia-do-whatever-it-wants game on Windows. GPU on Mac is a bit dicey because Apple seems to be unwilling to play this game.

nVidia gets away with this garbage on Linux because (as I understand it) they have strong-armed enough vendors into bending over backwards to support their driver model. The result is that nVidia can tell end users to either put up with constant breakage or to use of several not-really-upstream-kernel-based distros to get something stable enough to support the driver properly. Sadly, nVidia has the market power to pull this off.

I would love to see a big enough consortium (e.g. Red Hat, SuSE, Citrix, Oracle AMD Canonical) tell nVidia to buzz off. If all of those vendors dropped support for out-of-tree kernel GPU drivers and instead tracked upstream more closely for improved in-tree support, AMD and Intel would get a boost and nVidia might have to change its tune. And the vendors could stop wasting engineering resources to keep nVidia working. Users get easier-to-work-with systems. Everyone except nVidia wins.

Embracing and extending something, with no possible way to extinguish cannot be seen as steps one and two of an embrace, extend, extinguish strategy. That makes no sense at all

Agreed! Embracing is purely positive. Extending can be negative, but it generally isn't frowned upon. Heck, it's one of the reasons for FOSS existing!

Extinguish is the only purely negative part, people seem to forget that...

The extend stage was always some sort of small convenience that would be only in microsoft's version of a standard so that when people used a standard language or tool on windows, switching platforms would eventually get bogged down with changes and problems, keeping people locked in.

Nadella seems to be largely hands off W.R.T any funding / acquisitions. AFAIK, LinkedIn is still operating independently, Github too.

Since the acquisition I think LinkedIn is just starting to make moves towards even using Azure.

Both companies use Microsoft products instead of their own or from 3rd party companies. On the surface, they are independent companies. Hotmail was also independent company at start, now it is tightly integrated into Microsoft. Even Xbox was independent previously.

Maybe in a decade or so, when they think they could create a profit model from their integration, they would become Microsoft LinkedIn and Microsoft OneHub something.

?> Both companies use Microsoft products instead of their own

This seems like quite a bold statement. From the outside, both LinkedIn and Github use plenty of their own products / 3rd party solutions.

Neither of those could really be counted as "extinguished" though. Hotmail evolved into Outlook, and Xbox is doing quite well. Those seem like pretty solid outcomes.

Not sure if this qualifies but there’s this[0] embarrassing (to say the least) technical decision.

[0] https://daniel.haxx.se/blog/2016/08/19/removing-the-powershe...

Didn't they change their mind in the end?

https://github.com/PowerShell/PowerShell-RFC/issues/16#issue...

@kuldeepdhaka the aliases have been removed from PowerShell Core 6.0 for awhile now

It’s hard to comment on this charitably, but why did anyone think this was ever a good idea, and why resist fixing it? I believe you’re correct though, it was ultimately fixed.

Well, they would have broken compatibility for their existing users, which sucks, too.

Those examples don't count as extinguished. The idea is that they would start by interoperating with an open standard, then they would make their version of the standard that has features they are not sharing, then they would convince companies/users to only use their version of the standard.

This is why we had so many websites designed specifically for IE6, and why many people are currently wary of that trend in Google/Chrome.

Github was already in the middle of their own version of EEE before Microsoft took over, so I don't know if you could blame them for that.

The thing is that quite a few of us were personally inconvenienced by Microsoft not supporting standards properly. When someone comes along and tells us that the stove hasn't burned anyones hand in over a decade, it doesn't mean that we're suddenly going feel comfortable touching it.

Atom, it's slowly dying with now only bug fixes. Language Servers which were open to all ides but newer versions restricted to their own ides and code editors. I think you would see many similar scenarios.

They are trying hard to extinguish Google Chrome on Windows too. That's your embrace, extend and extinguish.

It's only practical from business perspective to extinguish competitors. Imo, when their open source software have monopoly they would update license slowly. Let's face it, even if you fork the code afterwards there wouldn't be users and contributors to maintain it unless it's revolutionary. You could take example from Cloud9 IDE, BitTorrent etc.

I thought maybe there was a risk with C# tooling or Xamarin but so far they seem to be keeping up Linux and Mac support. Windows is still the lead platform but I think I can live comfortably with this level of cross platform effort.

They updated the Linux .net profiling tools and are committed to a cross platform UI solution.

AppGet?

https://www.theverge.com/2020/5/28/21272964/microsoft-winget...

That's called being Sherlocked, I think.

That was discussed here: https://news.ycombinator.com/item?id=23331287

The summary is Microsoft didn't copy anything; The app is trivial and the dev got a lot of free exposure for his startup. [0]

[0] https://news.ycombinator.com/item?id=23341302

Blender is doing so good nowadays it is crazy... Everyone is supporting them.

I hope they keep the team smallish as always!

Tangential but a project I'd like to see supported more financially is FreeCAD, used in many corporations of all sizes, big and small, like Behringer for instance. It's very powerful and capable but incredibly buggy.

Their paint3d app isn't the worst thing for non technical creatives. But it scares me when I see the big MS start "embracing" something I love.

If Microsoft is smart, they will continue to support large productivity applications in major industries that don't fit into an app store model.

The kinds of things that work better on a full Windows 10 machine.

I think the "Commoditize your complement" mantra is used by megacorps to justify their investment in open source. In this case, 3D modeling is probably complementary to their AI investments.

glTF 2.0 breakthrough adoption was triggered by Microsoft contributions to move glTF 1.0 beyond WebGL as target API.

Microsoft actually helped me put Krita in the Windows Store. The revenue from that meant that I not only didn't have to find another day job, but that I could hire two other people to work on Krita. Of course, we're cheapskates, and we don't give ourselves a lot of money, but it's still very nice not to have to work on very security-sensitive C++ code in a place an hour by train from where I live for 45k euros a year without pension plan, health insurance or travel reimbursement.

I wish Sony, Dreamworks, Pixar, Disney and other major CG studios would pitch in. They all probably pay 6 or 7 figures a year for licenses for various pro 3D software. I know Blender is not actually at the same level as much of that software but it could be if enough funds were there to support it. Either fund it directly or fund internal devs to contribute.

Blender wouldn’t really replace any of their existing software though. It’s not really a competitor for Houdini or Katana. Maybe Maya at some point but good luck getting the modellers and animators to switch.

Blender still needs to improve a lot to compete against Renderman, Octane Render, Hyperion Renderer and similar.

Blender isn't competing with those programs, because those are a different type of software. Those are renderers, not 3D production suites. And two of the renderers you listed can be natively used in Blender in place of its own renderer.

Blender is competing against Maya, not Renderman.

Much of Pixar's software is entirely in-house. They already pay internal devs work on programs such as Presto. I'd be surprised to see them give up that sort of control.

They might need to be there to port the software for that new armv8-a windows they have. I don’t think there’re many volunteers implementing support of a platform that hasn’t quite took off.

"Microsoft makes use of Blender to generate synthetic 3D models and images of humans that can be used to train AI models. For researchers, having access to high quality free/opensource 3D software has proven to be of great benefit for scientific projects." Here's what they linked:

https://www.microsoft.com/en-us/research/uploads/prod/2019/0...

Am I in the minority where all I see in this synthetic human stuff is surveillance and propaganda usage?

What other benefits are there?

R&D for prosthetic limb design or physical therapy using augmented spaces perhaps? Maybe Car crash simulations for accident scene investigators?

3D models can be used for testing/training self driving cars.

I would think the primary use case for this would be pose estimation and/or recovering 3d models from 2d images. The primary use cases I'd guess is for AR/VR and building other kinds of interfaces (gestures etc).

Neither of these use cases seem very useful for surveillance. I'm not convinced that facial recognition algorithms would benefit that much from synthetic faces, but I very well may be wrong about that.

Metaverse is what I'm thinking.

Won't be surprised if one day Microsoft pulls off a product that combines Hololens and Minecraft, i.e. a much better version of Minecraft Earth.

“Corporate Memberships: It's like having developers work for you - on Blender! This membership level is for organisations who want the option to monitor in more detail what will get funded with their contributions. They will get direct access to the Blender team for strategical discussions. Roadmaps and priorities will be aligned with your requirements as good as possible.“

Yeah this seems innocent enough. If Autodesk gets on the board, then we need to be worried.

MSFT figures out how to play with Open Source: If you can't defeat them join them

This happened 10 years ago btw.

Reply on news.ycombinator.com