-
Facebook Joins the Rust Foundation developers.facebook.com
A brief history of Rust at Facebook engineering.fb.com
via
dayanruben
2 years ago
via
dayanruben
2 years ago
As someone who wants Rust to succeed and thus be well funded but at the same time Facebook to be gone from this planet I find myself in a minority expressing concern about this. I find it weird that there’s very little objection to this from either the community or the foundation. The general consensus is “it’s free money”, “they’re doing great with open-source”, “who else should pay for this?” but I rarely see people questioning where this money comes from. Not to mention the concern that they could always threaten to remove the funding if x, y or z doesn’t happen.
Are there open source projects Facebook has become a part of that went downhill as a result of them? (I get your point, I was just curious if there was a bad track record of their interaction with the open source community…)
Not that I know of.
When the founding members are Google, Huawei, AWS, Microsoft, and Mozilla I don’t know if there’s much point complaining that Facebook is joining in.
Personally, I think it’s probably one of the better ways for these companies to spend their money, and I don’t have any strong feelings about Rust one way or the other.
I have about as much of a problem with Facebook as I do with every single other company on that list (including Mozilla itself). That is to say, they all do things I disapprove of, and there’s nothing particularly special about the things Facebook does that I disapprove of.
First off, the funding is probably a rounding error in FB’s finances (and is almost certainly written off as charitable).
Second, sneakily enforcing FB’s will through subtle arm-twisting of the Rust Foundation is incredibly inefficient. This is a company that can literally decide a country’s election, should it wish to do so.
Third, Freedom Zero. As long as an organization follows the license and the code of conduct (if applicable) it can be as “evil” as it wants.
Frankly, FB hate is probably a small minority voice. If you’d ask the stereotypical person on the street, they’d probably rate FB as a very trusted company/brand, at the level of (say) Sony.
Is this meant to be a convincing argument? Shouldn’t the last twenty years have put the myth of moral compartmentalization to rest?
Aside from that, this is not actually responsive to the concern it’s ostensibly addressing. “Hey, it’s a free country, they can do whatever they want,” is not a reasonable or helpful reply to, “Facebook’s involvement makes me uncomfortable for many obvious reasons.”
It was the devil’s bargain the Rust project signed when they chose a license.
I don’t think that picking a license that allows corporations to use, fork, contribute to etc your work requires you to put them on your foundation.
Though, given the list of founders, I don’t think you can get much worse by adding FB. And I’m struggling to imagine how Facebook’s badness could influence Rust in a negative way; I don’t think they’re trying to pull some weird ‘sabotage Rust from the inside because ???’ move.
I think is responsive, and the important point, to my eye, is that evil corporation X is benefiting from Rust already. Because it’s released under a license that upholds freedom zero.
If there’s discomfort, I think it should come from the fact that this community project is helping to further the agenda of a corporation that you consider evil.
The fact that they’re now participating in the funding of the project in addition to benefiting from its output doesn’t alter my level of discomfort. The bad thing is that they were already enjoying the benefits.
When you release software under a license that lets it be used for any purpose, that discomfort is a sunk cost IMO. Contributions that benefit the entire community seem like they should offset that.
I don’t see how compartmentalization comes into play here.
I don’t see the connection. From The Free Software Definition on Wikipedia…
How is the argument “As long as an organization follows the license and the code of conduct (if applicable) it can be as “evil” as it wants.” related to Freedom Zero?
My take: it isn’t.
I think the argument above is making a different claim. The argument above suggests as long as an organization is complying with legal requirements, no other considerations matter. This, of course, is not true. Particularly with the Rust Foundation, they are free to have additional criteria. It is also in their interest to do so.
I think I see where you’re coming from.
You’re not trying to prevent FB from using Rust or the tooling around it. You’d prefer the Rust Foundation to not accept funding from FB, under the assumption that the reputational hit from doing so would outweigh the monetary and other benefits deriving from FB’s support.
I agree, that has nothing to do with Freedom Zero.
Not exactly. I distinguish between Rust’s open source license and the Rust community’s governance model. I have confidence that the Rust Foundation is setup in a way that reduces undue influence from contributing organizations. Putting aside its societal impacts, Facebook has many well-meaning individuals and does a lot of good engineering work. I think FB’s contributions and support will likely be a net positive.
Rust Foundation is a 501(c)(6) business interest group, not a charitable foundation. To the extent that FB can write off tax against its donations that would be as a business expense and not as a charitable donation.
(Almost every “we think Free Software Foundation sounds serious and we want to be taken seriously too” technology Foundation created in the USA after the FSF, most notably the Linux Foundation which is the legal home of many of the others, is a 501(c)(6) which is more a not-for-profit cartel than a charity. Notably 501(c)(6) are allowed to lobby in the interests of their corporate “donors” which 501(c)(3) may not do. If you want your public software to be managed in the public interest you have to look beyond corporate associations to groups like FSF, Conservancy, or SPI.)
Thanks a lot for clearing that up. US non-profit law is confusing to outsiders.
You aren’t the only one who reacted that way. I have the same sentiments. Though given the other members….Rust could easily be in jeopardy of lost funding even without FB.
Nice, yet another reason not to use Rust!
Be careful, they also use C (e.g. in zstd). Better not use C either!
It is one thing to dislike Facebook having influence on the Rust Foundation. However, it is another thing to not use Rust because of this event. Would you care to unpack your reasoning? Does it have to do with moral principles? Predictions about where this leads?
Also, for the language(s) you prefer, to what degree are they “free” from the same concerns you have about Rust and/or Facebook’s involvement in Rust?
I admit that I’m prodding a bit here, because when people say “reasons” it is only fair to dig in and get, well, the reasoning as opposed to a one-liner.
I don’t write Rust, but most it seems like most developers have had an overwhelmingly positive reaction to the emergence of Rust. I’m curious, what (other than Facebook’s involvement) makes you hesitant to get on board? What are some alternatives that you would consider instead?
I’ve been told many, many times, “Why are you using Rust? You should use Rust!”, and it feels very much like the push for everything to be Java and OOP years ago. I actually like writing Rust and I’ve made a few things with it. I might have burned out of learning updates to it since I started writing it around 1.0 until around async/await getting stabilized.
Ada 2012 is a viable mature alternative, and is very close in feature set and many conceptual constructs to a Pascal version of a safer C++, with extra features for type safety and error detection. Going from C++ -> Ada is actually a two week process or so because there’s many similarities, despite the languages being in different families.
With all due respect, I was around for the “everything Java” push (maybe you were too!) and the movement around Rust is much more “grassroots”. Java had the backing of a big “hot” corporation (yes kids, Sun was hot once!) and there were endless streams of books and tutorials and people cold calling companies and seminars and stuff. Not to mention tons of back-end development of standard libraries, etc.
Rust to me still feels like an enthusiast movement, but coupled with an actual, delivered product.
My tone may not have come across well, I’m not trying to disparage Rust or be combative, I’m just giving my opinion. Rust usage would likely affect me at work, so I’ve been trying to predict it’s adoption by reading about Theory of Reasoned Action, the Technology Acceptance Model and similar concepts. At the time people were just trying to solve development problems, just like they’re trying to do now. I’m curious if Github and more widespread internet and computing resources existed back then, if there would have been more parallels, or if Smalltalk would have had more sticking power.
I take your point. There are (at least) two underlying aspects:
Distinguishing initial enthusiasm from longer-term suitability.
Distinguishing ‘disinterested’ (i.e. unbiased) recommendations from marketing-driven hype.
Another factor to consider: Around Java’s first public release in 1996, the main popular alternatives were C and C++. (Fair? I’m not a historical expert, so please add to this – i.e. let me know if others were considered both popular and situated similarly.) In contrast, compare 1996 to 2015, when Rust 1.0 was released. The diversity and language competition is much deeper now. My point is to say that Rust’s popularity is even more impressive given the choices available; e.g. C, C++, Java, Python, Go, Scala, C#, F#, Haskell …
I think Delphi was a popular alternative back then while Smalltalk was close to where Rust is now in industrial mindshare.
A lot of this was told to me by a Smalltalk developer at the time. Smalltalk was somewhat already established. It wasn’t just Sun, but it was competing against IBM, a giant at the time, that made the independent Smalltalk tool developers feel like they had to merge together and resulted in many internal issues. With the smaller folks struggling, Smalltalk effectively died when IBM pivoted to Java. This is a bit oversimplified, there were other technical concerns involved.
I think Facebook joining up, along with other big names is the equivalent of Sun and IBM with Java. In this way, I feel Rust vs Ada seems to parallel Java vs Smalltalk in some ways and I think Ada (and possibly C++) is in big trouble, if it wasn’t already. However, AdaCore is in full swing promotion and marketing mode right now, so things might get interesting over the next year.
Be careful. We the Lobste.rs are crustaceans.