Funds for Open Source
blog.opencollective.com
blog.opencollective.com
Most efforts do seem focused on corporate open-source. The under-funding of that stuff is indeed an issue. But it doesn't result in real public goods that treat the public well if it's all upstream stuff that only serves to make proprietary downstream end-user products.
We could certainly use help from anyone who would like to do some project management. In general, we have https://wiki.snowdrift.coop/community/how-to-help
We have weekly audio meetings and have Matrix and IRC discussions, and people are welcome to pop in.
Not sure if these are the things you were asking about
In the mean time, if your service can only exist as something outside the public domain, I don't think people will complain much if you're honest in your communications about it.
ES may not give their contributors the same rights on paper, but in practice there is no discernable difference, other than that the contributors aren't allowed to directly compete with them (iirc).
In contrast to the patent on insulin being made open for all to inspect, use, and build upon - SSPL does not smell very "open source" if it's similarly opposed to forking or being the foundation for something someone takes in a different direction.
No, the wording in their SSPL is much more broad than that, such that creating any logging facet would put you in legal ambiguity at best - the worst case, which unfortunately now two licensing lawyers I've watched be adamant about this being the realistic interpretation in court (/dev/lawyer and one who advised our and other software companies), is that you really can't be in compliance without releasing your entire software stack.
I totally agree with the idea that a business becomes before project/idealism considerations though, if you're a company trying to finance an OSS project you need to think about your financials/survival before any licensing purity - which I don't think anyone is criticizing them about in this scenario anyways.
That being said, you can make changes to your project for your business but do so on terms that the internal and external community don't find disingenuous, don't characterize those whom are upset with a withdrawal from previous promises(1) made to them as conspiring to spread misinfo, or even to have taken many in the community's suggestion for dual licensing under copyleft license and commercial license that provides a hedge against large cloud providers.
Anyways, if Elastic's response to going back on a promise is to gaslight their own (reasonably) begrudged customers and contributors about how the company going public shortly is the real little guy/victim, then I think most already know whether to trust their lawyers or the company.
1: https://web.archive.org/web/20200120104750/https:/www.elasti...
The same guy who defended the Elastisearch exec by claiming that by turning an OSS project into proprietary software was equivalent to companies open sourcing previously closed software, despite the former benefiting from contributions+customers which had been misled by communications about it not being or becoming proprietary!
Just curious - why do you invent your own meanings for terms that have different meaning for the vast majority of readers?
This is a genuine question since I don't think the doublespeam comes off as either sincere to those in the community or helpful to those learning about the subject, who are still only just a Google search away from finding that no one else besides those benefitting from taking away user rights describe such projects as open source: https://opensource.org/node/1099
I just feel like the new license doesn't take away any of the freedoms I care about, the source is open, I can run and modify it without issue. I can even run it on a AWS instance. I want more open source companies to figure out sustainable business models so we can have more open source software. Amazon taking all open source revenue streams will lead to less open source software.
I feel like OSI plays all these word games that are in bad faith, calling SSPL "proprietary" which is just wrong.
How do you know that for "the vast majority of readers" open source means exactly the 11 point definition that OSDI came up with?
(I realize that by replying I'll probably garner another -4 HN points, but I'm willing to spend the karma since you said it was a genuine question.)
I agree no one should be religious about OSI's standards, but open source colloquially means "open for people to modify/fork/use" - I think the FOSS movement has created a similar colloquial definition even if FSF isn't the end all be all for free software. The differentiator "free as in freedom, not beer" is a great and uncontroversial litmus test for the community.
OSD 6 I think presents something similarly uncontroversial and closely tied to the colloquial definition of OSS (unlike other parts of their directive) which is "No Discrimination Against Fields of Endeavor" for licenses. It's not very open if one can't fork the project should their vision diverge, which is explicitly what both of Elastic's reasonably both paid and unpaid proprietary licenses will restrict - it keeps the software closed to their domain rather than fully open.
This is in contrast to open copyleft licenses which only require that you return contributions that you make to the codebase - something very comparable and that open source businesses have successfully worked with is dual licensing AGPL. With it I have no restrictions on how I use the software, but if I consider it a business need to begin running it, I have to be *open* about what I'm running.
It's not going against the broadly almost anyone's understanding of OSS to say "fork it, do whatever" or "fork it, just be open" - but it is going against most people's definition when it's "fork it, but you don't use it if you fall into a category that competes with our monopoly". Patent rights are an apt comparison - just because a patent is available for the public to view doesn't necessarily mean it's open for the public to expand on without paying a proprietor!
And what you say about the SSPL just being okay to run for your own purposes isn't true either - I've been advised by two lawyers at this point that section 13 I believe is so broad that it's viral enough to require you basically publish everything that you run in a stack because of how open they leave the language to interpretation, and their stipulation that it's "not that broad" in a FAQ does not supersede the language used in a license - Google attempted that defense early on in their battle with Oracle about how Java namespaces were previously advertised as open. It turns out the law prioritizes licenses over box stickers every time, and I don't think any business redistributing ES in any form can guarantee they're complying with the license (which redistributing in Elastic's SSPL's terms could be a logging facet for an application that's just backed by ES).
Anyways, I totally back companies making great software proprietary or closed, it pushes our industry forward - I just find it abhorrent the argument being out forward is that anyone can redefine "proprietary" if they're up against a massive company (even when they're a commercial and not so tiny business themselves) and especially if they're gaslighting the broader community about how they're doubling down on open without any irony or explanations about what the majority thinks open means in contrast.
Many of these other "open source" licenses have restrictions on who can use and how the software can be used. They're basically created out of fear to prevent competition.
GPL licenses mainly benefit the user of the software, and for the user these are not restrictions but rights (e.g. the right to see the source).
Much of the benefits the user has here is also what benefits good contributors ironically, in contrast to licenses like SSPL which restrict contributors from freely and openly diverging (and perhaps competing with for the benefit of users) from a proprietor.
If there is some free software I like, but it is is missing a feature I want, I can pay anyone to make these changes for me.
If the software is proprietary, I can try asking the owners. They say no. The end.
The article doesn't go into detail, but it also appears to streamline the workflow of requesting money from an open source project's perspective.
Each fund[0] appears to have a "Request Grant" button, which is great, as it'll mean maintainers will hopefully have a curated list of contactable organisations if they need funding for a specific initiative.
The barriers to entry here are extremely high: a few days ago I was looking into corporate sponsorship for a non open-source-related event, and after a couple of hours of Googling I was mostly unable to find contact details for relevant companies. It's definitely a natural barrier to entry which will be reduced with experience, but it will gatekeep a lot of interesting projects where organisers aren't experienced in fundraising.
Reducing this friction for Open Source projects should allow the money to much more easily go to where it's needed.
[0] Example fund: https://opencollective.com/indeed
I used to determine my self-worth through the number of stars my GitHub project had. Boy oh boy was that silly, looking back at it. 1.2k stars on GitHub... and god knows who profited off it. Me on the other hand? Meh. Would someone like to buy my 1.2k GitHub stars? :D
If you really don't want corporations to benefit off your work without contributing back go GPL or stop maintaining it.
Despite most everyone implicitly associating open source with gratis, it's not a requirement or even preferable in my opinion. The benefits of OSS to users are all still there with paid software.
Tangentially this makes me think of a recent phenomenon I'm sure many OSS advocates love to ridicule or trivialize, but which shows at least an attempt to solve the problem of compensation for freely replicable work: NFTs. If you manage to see past the mainstream view that people are "buying links to jpeg files", and rather understand that people are paying for public recognition of patronage, there's something to be said about their ability to get people to pay for "open source" work.
Of course, given the choice, myself included, users prefer the more permissive license. The question is what strikes the right balance of incentives for developers and respect of users.
MySQL/MariaDB, OpenOffice/LibreOffice, Cyanogenmod/LineageOS, a couple others had the IP go to someone who didn't have an incentive to develop it properly, so didn't. Source Available permits this with no real recourse.
Agreed on the NFT comment, BTW.
This is slightly off-topic, but how does volunteering work?
There is this sticky idea of imaginary army of open source developers who produce full time job worth of effort and live from thin air and don't have to pay rent. And who do all that boring routine and difficult work of maintaining, merging, reviewing, patching, testing and so on for passion alone.
On other projects that weren't started in companies from the start you see paid FOSS developers doing maintenance.
Whether that is money well spent or if these developers ever do anything creative or original is another question.
How is that not development paid by companies and how exactly is that exploitation by big companies? You literally claimed it is paid by them.
...If someone makes something and gives it out for free, the government can't retroactively force users to pay for it. It's up to the creator to choose a different lisence.
GPLv3 plus a commercial license will go a long way toward that.
Why do people choose to work for free on open source projects that are mostly used by businesses? I can completely understand making things that are hard to monetize, user facing applications, or trying to learn something. But when I see free contributions to things like infrastructure management I just don’t get it. Why wouldn’t you just get a job doing this if there’s obviously a business demand for it? If I made something for free that was primarily used by people who made money off of my work it would completely kill my motivation and I would feel taken advantage of.
for the former case: do you think former gnome contributors (who started the project exactly because their feared the ossification of window managers trying to copy windows, see their https://web.archive.org/web/19990224084927/http://www.gnome.... manifesto) would be happy with designers that ignore users and copy everything from osx?
for the second case, just look up how much linus fought against tainted kernel and still ended up giving up because of corporate/funding pressure. Also how google employees do whatever they want to chromium, for example, removing every single contribution to restrict referrer because that is how they made money from clicks on google search ads.
My impression is they're fairly similar at first glance: companies can pay open-source funding to one central service, who redistribute to each of the appropriate individual maintainers & projects behind the scenes. I'm sure there's a bunch of practical differences in the details though?
As far as I am aware Tidelift pays maintainers in exchange for a minimal set of commitments, which is used to sell the Tidelift subscription. They curate packages of ‘known good’ open source projects for enterprises and provide a set tools for users to better understand what software they depend upon.
Funds is a little more free-format, facilitating a relationship between maintainers and organisations on their own terms. No contracts, no promises, no agendas. We take on the work of administrating payments to projects and of ensuring companies have what they need in their procurement processes.
In doing so we hope to lower the barriers to the degree that we broaden access to funding for open source.
Maintainers are above the other developers, is there going to be some community oversight?
Who decides how much expense can be claimed?
Q: who is the primary demographic of funders this is aimed at? Am I correct in saying it's mostly aimed at getting more corporate funding for OSS? (Non-rhetorical question)
The FSF is in some ways a competitor to this for-profit company ... is that a fair statement?
I signed the agreement on behalf of >2,500 open source projects (and the communities who maintain them) because I agree the call to action would strengthen the community, not weaken it.
This is not about the FSF or the GNU Project, it’s about building a safe space for people to participate in and build a commons that is as diverse and welcoming as it is free and open.
please reconsider.
Did you ask them for a vote, or is there a general arrangement that you / OSC speaks for them?
(I'm not a fan of RMS btw, this isn't about him at all. It's about principles and protocols)
Related discussion by supporters of free software: https://forums.puri.sm/t/stop-sponsoring-the-free-software-f...
I recommend that your organization officially stays away from the internal politics of other free software organizations, because you do not represent 2500 projects and are not a spokesperson for them.
Ideal world - You declare your stack, from OS, distributor up to all dependencies used.
Non-profit/individuals non-profit don't pay anything.
Everyone else/companies pay $0.01 per dependency, per month. Each package (at each layer) describes its own dependencies. So a docker image, installing ubuntu (ubuntu declares its own dependencies - libc, curl, git, whatever), running npm, and all node packages, and you take a bill of cost from this "bundle". If dependencies share dependencies it only counts as 1 no matter how many times it's repeated in a single setup.
If there's 1million servers running linux, that's 10k for the linux foundation per month. Probably there would need to be a tiering of cost/distribution, as a left-pad node package isn't worth the same as a linux distribution, and companies with + a certain threshold should perhaps pay more.
The benefits would be that people would stop installing 10.000 npm libs as they gets expensive fast, so people would try to write more comprehensive libs and the quality would go up.
You don't need to poke holes in the idea, it's basically an open ocean, but it would be nice to see something totally automated like this.
The OS of my choice has this:
https://pkgstats.archlinux.de/
although it would probably be easy to inflate these numbers if your income depended on it.
You can also do this by just paying into a commercial stack. The beauty of OSS is that it’s easy to use and reuse.
I’d rather just pay Microsoft than use a scheme where I pay by the number of dependencies.
Also, not that I release a ton of packages, but I do contribute some here and there. I would not contribute to commercial packages where my labor benefits some org.
And I would still release packages under a permissive license that allows for reuse without any compensation to me.
> I’d rather just pay Microsoft than use a scheme where I pay by the number of dependencies.
So you wouldn't want to support OSS, you would rather support a company.
> I would not contribute to commercial packages where my labor benefits some org.
That would be your call for sure.
> And I would still release packages under a permissive license that allows for reuse without any compensation to me.
This would be your prerogative as well.
Let's pretend we have in place the infrastructure. A place where anyone can submit a bill of usage, and do payments, and this holds the payments and a person can register their software there. For your use case it would have even the possibility of you redirecting the funds you don't want into other projects. Perhaps a badge on your profile, "redirector", "open hands", wtv.
Let's imagine that every piece of software besides a "readme.md" file, has a file describing its dependencies. NPM would be able to do this like they can build a dependency graph. Your OS would be able to do this because each programs/lib would be able to do this/provide their own.
Let's say there's a piece of software that can pick all these little files and coalesce them into a single one.
Now you could build a "package" of what it would cost you, see what is in there and then just make it part of your monthly payroll.
There's no bureaucracy. If you are a company using software and not paying for it, you would be under breach of contract and could be sued legally.
If I’m paying for it, it’s not open source. I’d rather pay a single company with a “simple”license, than something that costs me more when someone uses a leftpad package than just writing their own. I don’t want to have to have cost decisions factor into my design at that level.
Having the legal support to plan out if I’m under breach or not is expensive.
One of the things I like about OSS is I can avoid that. Paying and still having that threat is the worse of both worlds.
Also, unpredictable prices are really hard in my org. Having my payroll vary month to month on what’s happening is really hard. Do I pay based on when I compile? When I run? What if I want to have 10 test environments, so I pay times 10. What if I need to archive and might never run it, but need to make sure I can run it, do I pay. Etc etc. There’s a million different permutations based on project needs that vary by people.
With OSS, I can clearly plan and address all these. With commercial licenses, I usually can since I get a perpetual license per seat or cpu or whatever.
This new scheme would be really complex and include a lot of latent risk, and require that I’m constantly open to audit by some org. And audits are expensive to receive and support. No thanks, I’ll just skip using it and use OSS versions.
I don't know if my english is rudimentary or something. No, you would pay per month, it would be automated. We could throw it on the block chain, or have an AI calculate it.
If you never run it you never run it, if it's part of your business backup plan you pay it.
With OSS for sure, you just go 0$ monthly payments, most straightforward payment plan ever, can't argue with that.
> I usually can since I get a perpetual license per seat or cpu or whatever.
Well, subscriptions are the future.
I hope not. Subscriptions for software are a pain and something I try to avoid.
Open source and free are two separate things.
There are many projects out there that require a license for commercial use of the software because it's an attempt to make the project more sustainable.
Don't be surprised if more and more projects go down this route. The current model of giving corporations big freebies to make lots of money off of, and then often getting no financial or developer support for that, isn't sustainable and there's a massive burnout problem because of that. Times are good now, but just wait until a few things in your core infrastructure don't get patched or worse, they get archived, because the maintainer is tired of spending evenings fixing your problems for free.
When I use the term open source, I mean “OSI license” as that’s what I think is most important. That means free.
Projects are free to do whatever they want, and choosing a closed source license that’s not free is their prerogative. More power to them. I probably will never buy anything from them, but I harbor no ill will.
But when it comes to supporting OSS, I’ll do it as I’ve been for the past few decades; using and contributing to it.
Note this is different than “fixing your problems for free” and I’ve funded a shitload of consulting to fix my problems with OSS. I don’t think it’s reasonable to expect a problem to be fixed for free. Just like it’s unreasonable to expect that I must pay certain developers to fix problems in open source.
I find it funny when people say things aren’t sustainable and there are 20-50 year old projects and communities that are clearly sustainable and on multiple generations of developers.
“Not sustainable for me to do it” is very different from “not sustainable.” It’s ok that not everyone who wants to make a living doing something can’t afford to do it, that’s not a problem we need to solve as a society.
Add my left pad library as a dependency pull request.
The damage this would do to OSS would be irreversible.
Next year, we each achieve 33% YoY growth by splitting one of our deps into two pieces 'for increased modularity' or some other reason totally not financially motivated.
Find a way to get something into a popular Linux distro and you can micro-split the downstream dependencies over time.
“Every time I apt-get update && apt-get upgrade -y, my monthly bill goes up; I better stop doing that...”
My network gear and proxmox server and TrueNAS and Synology all run open-source software and all four of those support both personal and for-profit activities to different degrees.
The Plex container is clearly fully personal, but the Unifi controller and backups use cases are mixed, and the standalone Ubuntu container is fully for-profit, all running on a mixed Proxmox. Which ones get charged and which ones are free?
Also been thinking about a similar concept for a while. For the negative effects some have mentioned, one could also argue the possibility of a positive effect to make things more open and result in better quality / funded / open projects.
As an idea an addition to this could be a community trust score where users also get to vote on a few aspects of a project, like ‘is it well maintained?’, ‘do you understand changes in version updates?’ etc. maybe bad examples, but a few simple metrics which can give a project a good or bad rep based on what the maintainers does as presented by the community - not just nit-picked twitter quotes. Yes, these will need to be crafted carefully so that maintainer cannot game them but the result could be projects getting paid well for doing things that builds trust and makes the project more accessible.
Of course I could just look at the PR history and issues etc before I use a new package, but I would trust community feedback more than my 30 min deep dive.
Today we only have very vague metrics like npm popularity or github stars, which is very hard to accurately judge what package to choose or not.
And open source maintainers would be open to refuse PR's like they're now?
The beauty here is if Microsoft was using OSS they would need to pay still on their own bills of usage.
I've voted several times and think it's a great model. It was pioneered by Indeed and is likely also adopted elsewhere.
Yup. Open ocean. Because that just creates incentive to make a looooong dependency list.
Also, don't you think the complexity of packages are very different and it is therefore not fair, to make them all equal?
> Yup. Open ocean. Because that just creates incentive to make a looooong dependency list.
Well, perhaps no? Since those things don't get into your list by themselves I would think that it would incentivise the opposite. If whatever lib you're using has 100 dependencies, one for left-pad, the other for right, one for switching underscores for hyphens, this would make someone come and say, I can make this with much less cruft.
But still, if the dependencies are repeated across dependencies they wouldn't increase. So if you use 2 packages that both use say "curl", for your "bundle" it would still be only 1 entry for curl.
You want to argue with people that their long dependency list is totally unneccecary? That will become the norm, if this is the metric on how much income everyone would get. The higher the number of packages - the higher the pay.
Does not reflect reality, where one package can be a million times more complex, than simple 100 packages.
Thankfully, that's not what OC is proposing.
10.000 would still be 10.000 more than 0 right or is my math wrong?
A good kernel engineer is paid well above $60K a year.
Unfortunately, money brings in risks around project quality and integrity.
Just like in academia and industry, our work should be reviewed by peers with relevant and proven experience.
The fact that some thing of NPM installs, github stars and reddit upvotes as a measure of merit says a lot about the immaturity of the software industry.
> Everyone else/companies pay $0.01 per dependency, per month.
I don't get that dichotomy? People can do what they want, no? So are you simply saying that individual would be far less likely to fund open source software, while companies and assimilated would?
Or you would just want it to be mandatory for companies?
If the latter, that would be completely incompatible with Open Source.
The difference would be if you're setting up a blog to share your permaculture posts, you wouldn't need to pay, or something like that.
If you as a software author didn't want you could always put that the cost to use your package was $0?
And I've nothing against people who want to invent new models. Just: this can't apply to the whole current Open Source / Free Software corpus and ecosystem. And this new model will never be able to mix.
Payed license does not need to be created anyway. This already widely exist and this is just proprietary software. And among proprietary software, there are also licenses that are incompatible between each others. And proprietary licenses that don't require monetary paiement from individuals.
So does creating new proprietary licenses that would be obviously incompatible with Free Software licenses, and probably incompatible with most other proprietary licenses, would achieve anything interesting? I doubt it.
OSS already has a number of available licenses of varying levels of freedom that lets the creator decide how their creation should be used.
This entire discussion reads like "you developers are idiots who don't realize you're being taken advantage of." Whenever someone starts their argument with the (implicit or explicit) assumption that they are smarter than everyone else, my experience tells me that there's a hidden agenda in there somewhere. In this case it seems to be as trite as "corporations bad, people good" which, not to put too fine a point on it, is bullshit.
The AdSense and other forms of monetisation would be indeed trickier, as well as more complex usage patterns. Like if a business spins up a lot of machines, how would that go, perhaps charging for each machine the same would be too expensive. But charging just once would be unfair. Perhaps that could be somehow split/incorporated into what is charged to the end user, where it would just be a cost of running the service, the providers would do it on their end. Perhaps the licensing would be flexible, so you paid for some cloud provider they did the payment and as part of the invoice you would have this bill of usage as well. I would also imagine that no one would try to prosecute a guy using a blog, at the same time if they're doing adsense that blog/domain would be registered.
It's not like I've spent more than the time it took me to write that thinking about it. If it was to work I also doubt it would be something that would come out right at the first iteration(s). It was more like, "wouldn't it be great if we had some sort of automatic distribution of credit that would somehow flow directly back to the source" but at the same time keeping the spirit of open source. It could possibly also align further, in a symbiotic way, service providers and open source as better software would mean lower costs, better integration, more time for tooling, etc. Or it could work the opposite and give rise to a gamed ecosystem.
If I decide to spend thousands of hours writing software (or building a boat for that matter) and give it away for someone to make a profit with, that should be entirely my decision. The entire discussion is so patronizing I'm having trouble believing that I'm reading this. People aren't stupid: don't assume you know what's good for them better than they do. We all have our various motivations whether or not they make sense to someone else on the outside.
> "wouldn't it be great if we had some sort of automatic distribution of credit that would somehow flow directly back to the source"
You mean like GPL2?
It seems you're the one saying that others shouldn't have the choice of doing that? What would it matter to you if someone decided to do that for their own packages. How would that be different from using GPL on their own volition?
(while making it clear on your package page that you were doing so)
>Everyone else/companies pay $0.01 per dependency, per month.
This isn't going to be both scalable and sustainable. You will find that people will calculate the cost of this vs. the cost of being a non-profit. It also isn't going to work where a small dependency gets the same payout as a more complicated dependency.
You mean they would become a non-profit so they wouldn't have to shell 50 bucks a month?
The dependency values wouldn't need to be $0.01. Maybe if the theory about free markets is true, then it would somehow be guided by the invisible hand, and we would let linux charge the amount of $0.10 per month. With 1M commercial servers that would be 100k per month. Not faang salaries I know. Maybe left pad could be paid in bitcoin, so we could have like fractionality towards infinity.
https://en.wikipedia.org/wiki/American_Society_of_Composers%...
So a private company is taking a cut of each transaction headed to an open source project?
> In the open source world, formal contracts and partnership agreements don't happen the way they do in the business world.
In a way what OpenCollective offers here is much needed. The OSS world has a really hard time to get funded, while at the same time big tech is thriving upon their works.
Another pain point not mentioned in the article is that big tech corporate world also doesn't like FOSS (copyleft) licenses. They favor permissive licenses, as this is what their empires are built upon.
I fear this funding initiative will favor OSS to the detriment of FOSS. I also feel that OpenCollective - which I found aligned with FOSS principles at the start, hence really attractive - is moving towards raking in the big money now.
FWIW I agree with you about copyleft licenses, philosophically. But we've taken the approach of being very pragmatic when it comes to working with corporate sponsors, and we can't make them accept certain licenses, etc. Our approach is to build up funding for open source projects so they can be stronger and healthier and have the power to advocate for what they want to see in their ecosystem.
For example my company pays Linode to do backups - I'd love to be add a similar "percentage of the VM cost" to each box for open source support (so it goes to support projects). Something like 10% of the VM cost would be great - easy to get approval for and doesn't have the "donation" wording around it.
Say, I have a small, permissively licensed personal project (bittorrent implementation), that has some userbase, including for-profit companies.
There is a lot of shit code in there, and many features are missing. Let us assume for the sake of the argument that given sufficient time I would be able to turn it into a state-of-the-art reference library. But given the current state of my financials I'm finding it hard to justify putting any effort into it for free at the moment and end up purchasing other opportunities not related to OSS.
I'm pretty sure it is not a unique situation and wonder if there are any solutions for this (e.g. angels, incubators, something like that?)
If it turns out that they have real use cases for your software they should be glad for the offering. Make sure to either request a high rate, as you are the leading developer and leading expert for the software. This high rate basically pays for further development (as a kind of training or investment into your offering).
If they don't have a use case, code quality might not be as important and you can thus justify to scale down your involvement until such use cases appear.
As for your situation, put a notice up on your project that you are willing to fix bugs / code new features for a fee? If these companies use your library they might be interested in sponsoring your work.
If they are not however, then I don't think there is a way to fund your development.
As for incubators etc., they are probably not all that interested in a BitTorrent implementation, considering how many are already around? Unless of course the library provided some value nobody else does.
What I'd like to see instead of see people encouraging developers and open source maintainers start companies and create positive sum dynamics in wealth created.
For background see https://wtfisqf.com/ and Gitcoin Grants. Quadratic funding has been used to fund Ethereum public goods for a few years and it works pretty well.
IMO corporate funding is a death trap in the long run, because then they choose what gets priority and they have all the wrong incentives - they'll do right by the user when they can afford to.
More corporate funding is useful as a stepping stone, but in long-term planning it's a mirage.
Almost all the corporate funding we facilitate is no strings attached, meaning the project creators and maintainers stay in the driver's seat when it comes to prioritisation. Some projects opt to make other commitments to funders but that's totally up to the project.
Although they may add value to OSS in the short term by bringing in the $$$ we wouldn't otherwise get, their ultimate purpose as a business is to maximize the return to the shareholders, who are not you or me.
What we really need is more competition in this space to drive the fees down.
this type of service should be a public common good imho
1. People open source for free. They already do that. So, there is no real need to pay them as they are willing to do so for free. Greatest minds work for open-sourced projects.
2. If you want to pay/fund/support open source. Great, but these numbers we're talking about (like $1k/mo) are nowhere near what devs make at corporations.
I think Funds for Open Source is a great initiative! I just do not think it can be a real alternative to a career path.
I think the best open source support is done by big companies who open source their work. For example, Airflow by AirBnb.
Funds from tech companies won't alone make open source development sustainable, but it'll help.
I think open source improves on this problem because now I can fork and extend because a company open sourced something. This is not perfect, but it’s better than when I can’t.
I think a better approach is trying to encourage more of a gift economy with people working on stuff for free and sharing it.
It’s funny how I’m willing to accept some hair and do work myself as part of a community, but if I have to pay for it, I lose that community motivation.
For example, I’m willing to spend time writing up an issue for an OSS project with a test script, etc. but I won’t do that for a commercial company.
Edit : This is not a fantasy about poverty being more genuine. Poverty makes you die sooner. My point is a life spent running after money and positions is a meaningless life.
Edit 2: I may have a bias here since I live with under 1k a month, under my country's minimum wage that is .
You have to remember peoples circumstances too - many people have families or live in very expensive areas, they can't just suddenly stop everything to live on a lower amount like an individual can.
As I said my point is not to say that everyone should be like me although I like to show off how not-wealthy I am for the high of moral superiority. My point is that thinking you need more money is often intricated with irrational emotions. It has happened to me various times to have someone I know is doing fine complaining about being "tight". When a person that's poor say "I have no money", it's means the account balance is dangerously close to 0,when a well off person says so, it means "I wish I did not have to cut down expenses so I can maintain my standard of living".
It's literally everything in one room. Everything.
Living in a small flat (essentially everything in one room) even though I would not have to economically, it taught me about sufficiency and the cost of buying things. Empty spaces tend to fill up with stuff you buy. If you do not have empty space, you buy less, which is good for your mental well-being and the planet.
You can always go the park provided you got into that discipline, that you there is a park less than ten minutes away or that it's not too cold/raining/snowing: Not ideal. And it might a shitty square surrounded by cars.
Mentally, it is also very helpful to have separate spaces for different aspects of your life. When your room is also your office and your kitchen, it doesn't make for good sleep, good work or good cooking.
Relevant videos : https://youtu.be/LO1mTELoj6o https://youtu.be/snAhsXyO3Ck
Well, we did move on from living in caves for a reason, you know.
I don't think there's any shame in wanting > than the lowest possible money to achieve that.
Among us, some "entrepreneurs" will produce 100 little projects and try to get them all funded and then Show HN how they make a living off of dozens of lightly funded projects.
Funds for Open Source will globalize some work inasmuch as some developers will get paid some money in places where no development career paths exist now. Look at the World Economic Forum's Global Competitiveness report and you can find at least 20 countries where tech companies do not hire.
To your point, the reason people open source for free is that they cannot produce free software and get paid. There is no way for them to be compensated for working in alignment with their own values. Funds for Open Source attacks that problem.
I don't know that Funds is a good organization, or that they are fair in their pricing, or whether they are the best equipped private, for-profit company to wedge themselves between free-thinking developers and the corporate entities that consume their work for free, but I think the business model for Funds makes a lot of sense.
I want more open source software and more great open source software. But I think the best projects (Apache, Linux, etc) aren’t in need of funds and function more as common goods with multiple companies paying for work and contributing back.
So I think their heart is in the right place, but the implementation won’t lead to better software.
> Open Collective takes 10% + credit card fees (usually 3% + $0.30/transaction) [0]
A few concrete examples (dependent upon payment processor chosen)
For $1, a collective got ~$0.56
For $100, a collective got ~$86.60
I was getting very close to the inter-bank exchange rate for payouts.
[0] https://medium.com/open-collective/what-is-open-collective-h....
That works out to Open Collective taking 1.5% overall, if the fiscal host takes 10% gross.
That's honestly a lot.
It’s also an odd legal structure to separate the platform from a non-profit to do admin tasks, since non-profits are not really that good at doing admin and compliance.
I’d rather go through things like GitHub sponsorships that take 0%. Or I’d rather see a community oriented process that is OSS and just tries to connect donors to projects so I can donate directly to projects and just pay credit card processing fees.
I think Open Collective has a new pricing scheme [1] where they don't take anything if the fiscal host doesn't either, but I think more people should consider the work that fiscal hosts have to do.
Unlike Vanguard, Open Collective and the fiscal hosts might not have enough diversified income sources to pay for all the dev work and other manual labour that comes with handling money.
Bookkeeping/accounting work and annual compliance filings are both tedious and expensive. :(
> I’d rather go through things like GitHub sponsorships that take 0%.
GitHub/Microsoft is absorbing the Stripe fee for GitHub sponsorships, which amounts to something like ~2.9% – 4%.
For that, there's Loadsharers, a true community-run distribution network, which I invite everyone to go and have a look: https://esr.gitlab.io/loadsharers/
Come help us build it.