Keivan Beigi, the developer behind AppGet, a package manager for Windows, claims Microsoft copied his software. He was contacted by Microsoft as a possible hire, and flew in to Microsoft’s headquarters to talk about AppGet, and after suddenly being ghosted, Microsoft announced WinGet – what he claims is pretty much a direct copy.
Realistically, no matter how hard I tried to promote AppGet, it would never grow at the rate a Microsoft solution would. I didn’t create AppGet to get rich or to become famous or get hired by Microsoft. I created AppGet because I thought us Windows users deserved a decent app management experience too.
What bothers me is how the whole thing was handled. The slow and dreadful communication speed. The total radio silence at the end. But the part that hurts the most was the announcement. AppGet, which is objectively where most ideas for WinGet came from, was only mentioned as another package manager that just happened to exist; While other package managers that WinGet shares very little with were mentioned and explained much more deliberately.
This is the kind of stuff big tech does, so it really shouldn’t come as a surprise to anyone.
Microsoft trying to hide their photo copier.
CaptainN-
They possibly exploited him for free ideas, but I don’t see why they’d need to hide it since it doesn’t seem they did anything illegal here.
Illegal? Probably no.
Looks awful? Yes.
Thom Holwerda,
Indeed, this was microsoft’s M-O throughout the 90s. They would screw over partners and do business in bad faith all the time, taking tech from others and stomping out the little guys. I haven’t seen as much of this behavior from microsoft since the antitrust cases, but I wouldn’t trust microsoft’s intentions for a second.
The more things change the more they stay the same.
I came here to say the same thing. While I’m sure he invested a lot of time in it, being an open-source project, I imagine he didn’t have a lot of money invested in it. In that regard, while Microsoft definitely did him dirty, he is losing way less than companies did against the Microsoft of the 90s.
This entire history is very shady. If this is real, this is MS from 90s at its worst, just with a “open source” savoring.
And this is the reason why I see with concern this whole “MS open source” embrace, in particular regarding Linux. To me, WSL is a MS attempt at embrace, expand (to incompatibility) and extinguish strategy that they are so famous for.
You may ask “why MS bothered to copy the concept with a straight face (not fork, and without giving credit) instead to adopt it since it was open source”? From my POV, this is about control. They want total control of the repository, on what is committed and who can commit to it, so they can have total control regarding the project’s direction.
The real problem that big corps have with open source is not fear that their “secrets” get copied. It’s about control: they don’t want that the “community”, and maybe other corporations, has a real power to where the project is heading to.
So, when a corporation embrace open source, they do like that unless they absolutely can’t: they see a concept that they like, and either fork it or copy all it’s concepts to its own new “open source” product, in they own tight controlled software repository.
Our old pal Red Hat used to be the poster child of that mentality.
The source of both projects is open source. If you just download and compare them you will see that the word “copy” or “fork” is completely out of place here. Even “inspired by” is only correct in the way that MS had a good look and incorporated some of the features. I was under the impression that Win-Get was just an evolution of their https://github.com/OneGet. Apparently Microsoft just restarted https://github.com/microsoft/winget-cli/issues/186 and the only reason they could come up with above other installers is
So far I am completely unimpressed with Win-Get. It currently lacks even update functionality which means it is just a way to quickly install a bunch of software and then let it bitrot. Of course this will be added to Win-Get before it reaches 1.0 but for now there is no use in giving Win-Get any serious thought.
Visual Studio has support for npm and nuget, Windows Store has a very nice sandboxed+auto-updating mechanism and PowerShell has https://docs.microsoft.com/en-us/powershell/module/packagemanagement build in and supports multiple PackageProviders and whatever PackageSource you want.
People that act like Windows needed the App-Get developer to provide a magic piece of software that they just ripped off are not looking at any facts but are just pulling out arguments from the 90’s. This guy literally made his software open source with a license that allowed MS to fork it and close it up without any problem. Instead MS told them what they wanted to do, discussed hiring him, opensourced their tool that is clearly not just a fork/clone/copy and they let him know just before publishing this to the world and they mentioned him on their page. Even the guy himself mentioned that he didn’t know if he wanted the job and that his only complaint is that Microsoft didn’t mention him more prominently.
avgalen,
You are right, however I don’t see why you replied this way to CapEnt. What he said was correct. Both CapEnt and the blog post are already in full agreement that they copied the concept and not the source code implementation. I think you missed his point: microsoft could have forked it as open source, the reason they didn’t is because they wanted full ownership, which probably doesn’t surprise anyone here.
I responded to CapEnt because he said “this is MS from 90s at its worst, ” which is a ridiculous statement in this situation when all Microsoft didn’t do nicely is give enought credit to AppGet as an inspiration, which they aren’t required to do in any way but is just the right thing to do of course.
On the other hand, I don’t like that he is describing himself like this: Founder of AppGet (WinGet?), Sonarr.
(source: https://medium.com/@keivan/appget-what-chocolatey-wasnt-61f2b658d95b)
Also, I don’t understand why both of you mention wanting full control/ownership as the reason that they didn’t just fork it. This project had a license that allowed taking the source, forking it and having complete and utter control over that fork just like they have over WinGet now. That is the whole reason you create a fork…to get control. AppGet already had updating working so that would have saved WinGet some effort. A more likely reason they didn’t fork it is because they didn’t like the code that existed for whatever reason. One hint for that would be that AppGet was written in C#, WinGet is written in C++.
avgalen,
Well that was the crux of his point, which that as the owner of a work microsoft sets it’s own terms and isn’t subject to anyone else’s license terms. Ownership naturally has more value to companies than merely being a licensee, as suggested by organgtool. Microsoft owns winget completely whereas it would merely be an appget licensee, which has legal implications to microsoft.
Note that appget uses the apache license, whereas microsoft winget uses the MIT license…
github.com/appget/appget/blob/master/LICENSE
github.com/microsoft/winget-cli/blob/master/LICENSE
One difference between these two licenses is that appget explicitly grants patents to the licensee whereas winget does not, which could provide microsoft with more ammunition in lawsuits like the one between oracle and google, or hypothetically in the future against competing threats like reactos.
Well, the most obvious reason is that owning is legally advantageous to forking. Ownership grants you more control with no terms or conditions over the work. Also, I don’t know if this was a factor or not, but by creating their own incompatible implementation with a different license microsoft may have sought to prevent winget code from being “upstreamed” back into appget, which would have been a much greater risk as a fork.
That’s the dark side of FOSS, particularly GPLed FOSS, that people don’t acknowledge. The entity which contributes the most controls the project. Red Hat is basically GNU/Linux. No one else contributes to the Linux ecosystem at their level, and the Linux world revolves around them.
Also, one of the reasons I don’t feel bad at release my junk code under MIT or BSD licenses. In the very infinitesimal chance someone finds my code useful, I don’t want to have to deal with their changes.
Any yet, they’re still one of the best stewards of FOSS applications in the business. Canonical has tried something similar and failed miserably.
Lots of projects could be accused of the samething as well. Apple, FreeBSD, OpenBSD, LLVM, Redox, all the grep clones, . That’s the nature of the business. People don’t like product X for whatever reason, and they build product Y which does things only slightly different.
You hit the nail on the head regarding Microsoft implementing their own for the sake of control. It’s the same reason that Oracle spent $5.6 billion to acquire Sun despite the fact that Java was open-source – Oracle wanted control over it. I imagine this is going to make open-source projects even more valuable as companies adopt them to catch up to competitors and then attempt to buy the primary contributor(s) to maintain that competitive advantage: control.
Microsoft will need the copyright to that code, all of the code, and that isn’t happening by simply forking the codebase. This is a simple legal requirement, not an ideological position.
sj87,
That’s not right though. They prefer to own all of the code, but there’s no “legal requirement” for microsoft to implement it’s own code instead of using & modifying appget under an open source license. That would be perfectly acceptable if they wanted to go that route.
Nah, they could have had that for peanuts. They were talking about buying the rights to the software and hiring the developer. In my opinion, they probably evaluated his potential as a community leader/manager and found him lacking. Instead of paying for the rights and transitioning him to a role unrelated to the reason why they bought the software, they just copied it. It kinda make sense. Its an awkward conversation to have ” Hey great ideas here, but the implementation we aren’t happy with and we’re not sure we want you leading this project or having much of a role in it at all, have some cash and build a new movie maker beta download tool” .
Bill Shooter of Bul,
Yeah, I’m curious what microsoft’s intentions were before the interview. Obviously they knew the quality of appget before they interviewed him, so it seems that either the interview went poorly, or it’s possible they never intended to hire him and just wanted to extract his ideas. It’s hard to unravel their motives as both scenarios are plausible. If microsoft considered it a genuine interview, then it’s quite unprofessional of microsoft to leave him hanging in radio silence after the interview. If microsoft considered it a form of intelligence gathering from a rival tool, well then he was totally outwitted by microsoft.
https://daringfireball.net/2004/06/dashboard_vs_konfabulator
iampivot,
Good observation regarding apple’s rip-off of konfabulator. I’m sure this has happened to a lot of developers. Unfortunately things like this highlight the inequality in market opportunities, who you are can be significantly more important to your success than what you do
One of the reasons I’m so critical of huge mega-corps is because companies that exist at such large scales are inherently detrimental to fair competition and meritocracy.
https://www.howtogeek.com/297651/what-does-it-mean-when-a-company-sherlocks-an-app/
avgalen,
Those seem like good examples too. I’m sure we could find cases of this on all platforms including mobile.
Markets are neither fair nor based on merit.
1) People are random, and they will latch onto things for the dumbest reasons. Fitness of the product generally has very little to do with their decisions, and if it did, we would be living much different reality than we do.
2) It’s all about marketing and knowing people who can get you sweetheart deals.
I’m also convinced fitness of the product is detrimental to adoption. If something works, consultants don’t get to charge customers to build elaborate and fragile Rube Goldbergian contraptions with fat support contracts, and ecosystems filled with add-ons to cover the deficiencies of the product don’t form.
Economics and finance are pseudosciences. It’s the modern equivalent of alchemy.
Don’t get me wrong. I agree with you. Large corps wield an outsized influence because of the amount of money and resources they have. Where we disagree is if the game was ever straight. You say it is fundamentally straight, and I say it’s always been crooked.
Flatland_Spider,
Obviously I agree with your assessment, but I think ideally it’s something we should strive for in the same way that ideally we strive for democracy. A market that makes incumbent kings is wholly incompatible with the philosophy generally used to defend “free markets”. Calling for a free market that is not a meritocracy isn’t too different from a autocratic ruler suggesting he believes in freedoms as long as he’s in charge. This should rub people the wrong way.
I think this has always been a problem over the entire course of human history. The americas were a “modern” experiment in self rule rather than having rulers, but it’s still notoriously difficult to negate the unfair effects of the powerful becoming more powerful. The robber barons of the 19th century are a prime example of that, we managed to mitigate it for some time with progressive taxes and enforcement, which lead to a fairly prosperous middle class.
https://web.stanford.edu/class/polisci120a/immigration/Federal%20Tax%20Brackets.pdf
However most of these mitigations have been completely undone. GDP has gone up significantly due to skyrocketing worker productivity, yet more than 100% of the resulting gains have gone to owners rather than workers, leaving the middle class looking at lower economic quality of life today despite their increased productivity. If we don’t try & succeed in curtailing the growing inequalities then we absolutely will regress back to having rulers ruling over the rest of society.
The current whitehouse administration opened my eyes as to how dangerously close we could be to dictatorship. I’m thankful for the checks and balances provided by the constitution, but with enough bench loading of the federal and supreme courts and a corrupt congress, it opens a viable path to dismantling democracy. Neither party has done a good job protecting the middle class, but one party is much guiltier of actively making the powerful more powerful at every turn, so I hope things go better in november (the next election).
Personally I fail to see what’s wrong in this case. Obviously Microsoft is allowed to copy the appearance of an application, depending on the license also copying some of the code. They even tried to do the “right thing” by contacting the AppGet developer and considering employing him to potentially straight-up buy his source code and rebrand it as a Microsoft product. But it did not work out for some reason.
The guy of course wants to see it as Microsoft being an evil big corp, but it could just as well be about his personality not fitting what Microsoft needs in their giant organisation. Not everybody who write great code are great personalities, sorry but not sorry.
Yea, they still could have handled it better though. Like a bigger shout out and thanks.. Small but would have been nice. I suspect you’re right about it not being a good fit. I mean I don’t know this guy. Maybe he’s great, maybe he’s the next Hitler or Markus Persson. If he was you don’t want his name anywhere near your company. .
Jerry Kaplan’s book Startup tells essentially the same story of how he pitched a working GO Corporation pen tablet to Microsoft in the early ’90s, hoping that Microsoft would be interested in developing application software. Instead, Bill Gates and Jeff Raikes examined it very carefully, going through every single menu item and gesture of the UI, then proceeded to walk away and produce a software-only knockoff called Pen Windows under Raikes’ direction. Kaplan didn’t want to be in the hardware business so he licensed GO operating software to AT&T, but their EO tablet fizzled in the marketplace. Meanwhile John Sculley and Apple, who had been alerted to the GO project by Kaplan’s unsuccessful attempt to hire away Bill Atkinson as lead engineer, had launched the Newton project, with Atkinson in charge of development. As a postscript, all three competitors failed.