A letter to open-source maintainers

https://xuanwo.io/2024/10-a-letter-to-open-source-maintainers/

by xuanwo • 2 years ago

178 82 2 years ago

zabzonk• 2 years ago

I think there is a bit wrong with this, IMHO:

“only I can do” Sorry, friends, but this is incorrect.

No, not incorrect - if you have the vision for a FOSS project then you are probably the only one that can steer it. That's not to say you should reject contributions out of hand, but no need to accept them, either.

The project lacks enough users, which might suggest it doesn’t provide sufficient value to the community.

What "community" would that be?

I also feel angry when I see maintainers abandon a great project without providing any explanation.

Doesn't the rest of article contradict this? What right do you have to feel angry?

“I’m changing the world” Accept my respects, friend. Please remember to revisit my post whenever you feel tired.

What?

Now, you are tired, busy, and hurt, so you finally decide to leave

This is all just nonsense.

Ensure your projects can continue to operate smoothly even after your leaving.

You are under no obligation to do that.

bee_rider• 2 years ago

Doesn't the rest of article contradict this? What right do you have to feel angry?

You don’t need a special right to feel angry. We all come with a right to feel whatever we feel in our heads. Of course, there are limits on the ways we should express these feelings.

abeppu• 2 years ago

I agree that maintainers shouldn't need to feel committed or obligated to keep working on a project.

And yet:

- it's also not reasonable to expect that anyone should use your project if it's not clear it's going to maintained etc

- OSS contributors should not over-state the capabilities of their project, or make untruthful comparisons to paid alternatives

Yes, it's unhealthy and unsustainable to feel like you're on the hook for the health of a project, which may not sit easily alongside your day job, school, life, and your own changing interests. Yes, it can feel like a community of users has high or unending expectations. But also, projects often invite and encourage users to use their cool new thing, develop processes or workflows that depend on the project, make claims that their project does everything that existing commercial project X does but cheaper/faster/with greater data privacy, etc. Announcements are made that "the much requested feature X is underway and is planned to be included in release y.z". These all contribute quite understandably to the expectations of the community.

The view that every contribution is a one-time gift and the world isn't entitled to your future time and attention only works if we're all clear and honest about that up front.

bawolff3• 2 years ago

Do you feel like open source maintainers have this attitude? I feel like its exceptionally rare in real projects. Most open source maintainers fix things and what not, but don't really care if you actually use the code or not. No money changes hands so there is no marginal benefit of one additional user.

brabel2• 2 years ago

A lot of newer devs seem to have this attitude. They start a little hobby project and think they solved all problems of the world and start shouting about their project on social media and forums... in reality, the project is just a toy, of course, much of the time. But the users may not have a clue and may believe the author if they get enough hype. I think that at least one example of that which had very real consequences is the V Programming Language. It attracted quite a lot of hype, and there's not so much "hype budget" to go around with PLs. They started getting quite a lot of donations and sponsors, despite never delivering on most of the points that made their project unique, while other more serious projects, like Zig and Nim, which were really delivering good stuff, for some time didn't get as much attention. I think Andrew Kelly himself expressed this feeling at some point (sorry if it wasn't him, my memory is haze).

phoronixrly• 2 years ago

Sharing the source code (under an OSS license) with its users is one's only obligation from an ethical standpoint.

If its only user is the developer themselves, then they're already doing more than they're ethically required to by sharing the code.

People must come around to this view and the 'volunteer open-source maintainer/developer' role must be left only to people who enjoy doing it. Require payment for your services otherwise.

pessimizer• 2 years ago

The problem is that a lot of people are doing OSS to get work, or to get better work. Almost all of them really, lately. So although they don't have a boss (irt the software), they're desperate to impress, and to please, and to network, and they feel a lot of pressure. Every request is like a demand, so when those requests are petty, or entitled, or feel like a nuisance, it's like they have a horrible boss that's yelling at them and trying to destroy them. They're terrified of a fork; it's like everybody getting up and moving to another table in the cafeteria when all you want them to do is like you. Also, it's a lot of work, and can stack up, and you're doing it for free as basically specwork. They also half the time have to do their real jobs that actually feed their kids, unless they can work on FOSS that their job is willing to pay for.

They shouldn't do it for free, they should use every means and trick available to make it worthwhile for them to do and to get themselves paid. They need to survive for FOSS, for the software. Which means that they have to make sure the license is open and the source is available, and that it's ethical, because that's why it's worth fighting for. So everything that's ethical and open, very preferably GPL, do. If you have to write closed software to get yourself paid for your GPL software, do it. Eventually you can open that closed software up too, when (and not before) you can afford to. All FOSS licenses have different business implications, and ultimately income implications.

Morally, it'd be better if there were no software copyrights and everybody was allowed to inspect and modify all the code that they interacted with, and we all shared. You also have a moral obligation to feed yourself and your family and not go nuts, though, to have some security.

If you're not one of those people, and it's just a hobby and a charity thing for you, that's wonderful and ideal, bless you. You may have already made all your money. If you have to make a living from it though, don't let yourself be bullied into compliance and submission to the users because you need them to like you, to get work, to survive. Do whatever you have to do to be self-sufficient. It's hard out there, and you might have to be hard. Not mean, just firm about the amount of care and interest you have about individual strangers' problems with what you gave them for free.

You can leave a lot of people disappointed that they're not going to get more from you for free, yet still be ethical. Being ethical, being open, leaving people in control of their devices and in control of their lives, that's the point. Keep ethics at 100%, and put pleasing the most people as a (far) lower priority than paying your rent, and taking a vacation every once and a while, with the kids, and having health insurance.

Taking care of yourself means more FOSS in the world, more people free of black boxes in more areas of their lives. The OP seems wise.

rikroots• 2 years ago

The problem is that a lot of people are doing OSS to get work, or to get better work. Almost all of them really, lately.

This is exactly why I started my canvas library back in 2013. Hundreds of job applications were generating near zero interview invites. Then a recruiter told me that I was failing because I didn't have a portfolio of code to show prospective employers. I was so desperate (mid-40s looking to break into the tech industry) that I grabbed something I had been fooling around with (trying to make pretty animations in a canvas which I could share on a web page), slapped it on GitHub and started learning about PRs and stuff ... all to impress prospective employers.

It took over a year, and hundreds more applications, before I finally got a job offer.

If you're not one of those people, and it's just a hobby and a charity thing for you, that's wonderful and ideal, bless you.

I still work on the canvas library. I'd like to think its a hobby/charity thing. The reality is that I go through bouts of obsessive computing that can take over my life for weeks at a time - I lost last Christmas to building a better text layout engine for the library.

The one thing I'm grateful for is the library's unpopularity. No issues or PRs for me to triage/manage; no users making demands etc. It's just me making stupid demands on my energy and time, which I do my best to control.

bruce5113• 2 years ago

I would suggest that the goals of the author matter here. Obviously across so many authors you get different goals, and hence different requirements.

Part of this is understanding what your goals -are-. It's really important because without that target it's easy to be subsumed by other people's goals.

Are you writing this for you? Or you want to build a large user base? Or you want lots of contributors? Or it's a marketing exercise for your startup? Or.....?

For example, if your goal is to write code and post it, then you're done. If your goal is marketing then really you'll spend more time marketing the project than writing it. If your goal is a huge user base, well you better fix bug reports Pronto and so on.

If your goal is to get an income from this and go full-time then that's the problem you need to solve first (before you start writing code.) Have a plan for income - nake that clear from the start.

Of course goals change, that's OK. But make sure what you are doing lines up with your goals. That's the only way to succeed.

dmezzetti• 2 years ago

I've written two articles on this topic over the last few years:

https://medium.com/neuml/grow-your-open-source-project-5b439...

https://medium.com/neuml/grow-your-open-source-project-2-0-5...

In 2024, it's hard to get anyone to know you even exist. You can work all you want on GitHub but it likely will be a couple star project unless it trends on a site like this.

With how many developers oversell what they've built, everyone is generally skeptical with any new project. People are even more skeptical if a project mentions "AI".

Then you have VC-backed companies who have their ways to get their projects noticed. They'll get stars, articles and publicity but there is no direct correlation with project quality and funding. People use these projects, they're terrible and the cycle of disillusionment continues.

Given all that, it takes a certain resolve, belief in what you're doing and commitment to keep pushing forward. There will be little carrots along the way to re-energize you. Perhaps that is good enough, perhaps it's not. It's not for the faint of heart.

I fully agree that when an open-source maintainer doesn't want to do it anymore, it's like anything else in life. We have free will and the right to decide to move on to other endeavors. Just like one does with any other job.

joshka2• 2 years ago

It’s entirely acceptable for someone to use your MIT-licensed projects to make money without even acknowledging your name.

Well, actually...

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

-- https://choosealicense.com/licenses/mit/

bruce5113• 2 years ago

Sure, that's in the copyright notice, maybe in the docs no-one reads.

But from a marketing and sales point of view, it's nowhere.

For example, I can start marketing and selling a new image editor (blimp), and on my web page, in videos, Facebook yo face, I need make no mention that it's based on Gimp.

And that is -explicitly- allowed by Open Source licenses. Which is part of the Ooen Source deal - others can build on what you made, take credit for it, profit from it. That is not a bug, it's a feature which was explicitly allowed for.

thayne• 2 years ago

I need make no mention that it's based on Gimp.

You would however, need to make the source code available under the GPL license, since GIMP is licensed with GPL.

joshka2• 2 years ago

That is not a bug, it's a feature which was explicitly allowed for.

I agree entirely :) It's why I prefer to use the MIT / Apache Dual license (and generally to avoid the GPL in certain contexts).

Apache's condition is more explicit and nicer from the attribution perspective

       (d) If the Work includes a "NOTICE" text file as part of its
          distribution, then any Derivative Works that You distribute must
          include a readable copy of the attribution notices contained
          within such NOTICE file, excluding those notices that do not
          pertain to any part of the Derivative Works, in at least one
          of the following places: within a NOTICE text file distributed
          as part of the Derivative Works; within the Source form or
          documentation, if provided along with the Derivative Works; or,
          within a display generated by the Derivative Works, if and
          wherever such third-party notices normally appear. The contents
          of the NOTICE file are for informational purposes only and
          do not modify the License. You may add Your own attribution
          notices within Derivative Works that You distribute, alongside
          or as an addendum to the NOTICE text from the Work, provided
          that such additional attribution notices cannot be construed
          as modifying the License.

I've actively said to people as a result of looking at my own Apache/MIT licensed software (as well as others) to go fork themselves. You don't need my permission to do so as by choosing that license I already gave you the permission.

ranger_danger• 2 years ago

Any time I bring this up to people it's immediately shot down with "but it goes against the spirit of the license, and/or the developer's wishes, or is somehow otherwise unethical."

And at that point you basically can't disagree without looking like a ghoul.

xuanwo2• 2 years ago

Thank you for pointing that out. MIT is not a good example to convey my opinion. I will add an update at the end of the post.

edfletcher_t137• 2 years ago

Thank you so much OP for writing this, and I hope it resonates not just with maintainers (or potential maintainers) but to users of FOSS too. You needed to write this - specifically, the "Baseline" section, which I agree with vigorously - because too many users of FOSS consider themselves entitled to continued maintenance, support, feature development, community fostering, etc. etc.

We (maintainers) owe you NONE OF THAT. We gave you our code. That. Is. Enough.

yaris2• 2 years ago

I believe your last line should be written in any OSS README file first, maybe twice (in the same vein as another poster suggested because "no one reads the license").

ericyd• 2 years ago

It's always nice to see things on the internet with which I agree unequivocally.

jandrewrogers3• 2 years ago

Great article. Like many other people, I largely stopped contributing to open source years ago because people don’t respect boundaries. If you point out that they are not respecting boundaries, many will try to argue that they are entitled to this behavior because reasons. Long term, this dynamic is what is going to kill most OSS by people doing it for the love of it.

Unless you are making a lot of money by doing it, the act of sharing source code often has a negative return.

boris5• 2 years ago

Everyone agrees the unreasonable entitlement of many OSS users is bad but I don't see much analysis of where this entitlement comes from. Maybe it's the loss aversion: you first gift them the software but then by refusing to maintain it or add the feature that they need, you effectively take it away or, at least, suggest that they will have to pay for this gift after all (by implementing what they want themselves or paying someone to do it).

It could be the same dynamic with the "license rug pull", like in the Redis case discussed the other day. The new Redis license seems like it would be a non-issue for most users but because they had a more liberal license before and it was taken from them, they feel irrational sense of loss? (I realize there are also valid reasons why the BSD license was objectively better than the new custom license, like the ease of compliance.)

So maybe what Redis should have done is not to pull the existing rug, so to speak, but to leave it be and lay next to it a new, better rug and offer the users sitting on the old rug to move over, if they find the new features of the new rug outweigh the drawbacks of a new license.

jandrewrogers3• 2 years ago

Everything you are talking about is precisely the undue sense of entitlement that drives people away from contributing to open source. They owe you and everyone else nothing. Not their time, not their attention, not anything. The sense of entitlement to other people’s labor is grotesque.

You are free to fork the code and make it your problem to do the things you think the original code authors should be doing. That is your only right. Yet curiously many of the complainants don’t do this.

akalsey2• 2 years ago

My wife runs an estate sale company, helping families dispose of property. Sometimes the family wants her to give something away. She hates that. It’s way more work for her than selling it for even $1.

She says, “There’s no one more entitled than someone you’re giving something to for free.”

People responding to an ad for a free item will ask her to deliver it, load it, store it, repair it, and warranty it.

My theory is that folks who seek something out because it’s free tend to have more time than money. And that means they don’t value time very much.

eviks3• 2 years ago

not to pull the existing rug, so to speak, but to leave it be

How is this different from just using the old version before the "rug pull"?

pkaye• 2 years ago

Each of the previous versions the Redis released was an existing rug.

throwaway20373• 2 years ago

    > So maybe what Redis should have done is not to pull the existing rug, so to speak, but to leave it be and lay next to it a new, better rug and offer the users sitting on the old rug to move over, if they find the new features of the new rug outweigh the drawbacks of a new license.

Can you provide more specifics about your idea? It sounds like Redis would need to maintain two version. Can your ideas be implemented with the same cost structure? I doubt it.

rob74• 2 years ago

I don't see much analysis of where this entitlement comes from

Might I propose one possible reason? Take, for example, the Node.js ecosystem. Someone, a while back, built some library, added it to npm, then people used that module in their projects (maybe as a dependency of another module), along with thousands of other modules, and built an application on top of it. So they don't even really know that they are using the project, have no idea about who contributed to it, never saw its page on GitHub etc. Then, maybe years later, when the project is abandoned and flagged as risky by various codebase scanning tools, some poor developer (who maybe wasn't even involved in the initial decision of using Node for the project) will have to find an alternative for the module (and possibly others that depend on it and are also unmaintained). So maybe in this case it's frustration rather than entitlement?

NGRhodes• 2 years ago

Some people treat all open source projects as products that they use, when many are labours of love and charity, sometimes pet projects that have become useful to [many] others. They expect product support at a business level like all the other software they use. They see the project's website as a place to get support for their product, not a place to contribute.

block_dagger• 2 years ago

I think if someone invests the time to learn and use (depend on) an OSS, they feel like they paid for it. Not only that, but they chose it over a competitor, thereby propping up the OSS author. Therefore, irrational or not, they feel entitled.

perrygeo• 2 years ago

I've not experienced too much entitlement from users of my open source projects.

What I have experienced is the long tail of changes unrelated to the core competency of the library. I have some libraries where the fundamental logic and subject matter hasn't changed at all in a decade, yet I get a few dozen PRs a year to keep up with churn in the ecosystem. Sadly there are software ecosystems that don't value backwards compatibility at all - which causes massive pain for anyone who wants to build something that last in those ecosystems.

Running in place to keep up with your dependencies is not why I code! Maintaining these projects has become 99% busy work and I regret releasing them as open source.

GoblinSlayer• 2 years ago

There are two infinite things in this world: universe and human stupidity, and I'm not sure about the former. Why people are surprised when somebody is dumb on the internet?

vouaobrasil2• 2 years ago

I agree with this article entirely, except I have one counterpoint that should be more emphasized: open source programmers and maintainers should be more explicit and clear that the product can be discontinued, have problems, or just stop working with any updates beyond saying it in the license. Because otherwise normal people will have some ambiguity and have the mistaken idea that there is some social contract, however slight, that what you are offering has some promise of functioning.

icambron• 2 years ago

I don’t get it. The document setting out this fact is the license, which also tells them they’re allowed to use it at all. The MIT license is short and simple and very explicit about lack warranty.

If we were to say “no one reads the license; it’s just boilerplate!” and add it elsewhere — say the bottom of the readme or in another file like disclaimer.md — it would quickly just become other, different boilerplate that no one reads.

You can’t solve the problem of understanding norms by writing copy. I’m resigned to everyone learning the hard way.

gyan• 2 years ago

You can’t solve the problem of understanding norms by writing copy

You can mitigate the problem by adding a short summation before or in between something they're likely to read.

damethos2• 2 years ago

I always want to know what is the "intention" of the author/maintainer before I invest time in an open source project I want to use. I would love to see that with big bold letters at the top of each README.

yaris2• 2 years ago

And when the "intention" changes after you invested a few years/$ thousands into using the project, then ... ?

sabbaticaldev5• 2 years ago

also known as due diligence. that’s the users of open source burden, not the developers

vouaobrasil2• 2 years ago

The thing is, I am not talking about the burden or morality. Just what will happen.

growse• 2 years ago

The social contract arises from social relations. Ie: the dynamic between maintainer and contributors. It doesn't exist just because stuck something on GitHub with an MIT license.

This is why there's no conflict between the people who are aggrieved at OSS maintainers suddenly closing their software, and the people who say "it says no warranty right there in the license".

The default position is that there's no social contract. It's only through interacting with and building a community does a social contract (and associated expectations) arise. It's not codified in the license.

carlosjobim• 2 years ago

It is crystal clear by the fact that the users didn't pay anything for it.

dismalaf• 2 years ago

Open source is just a license. Anyone who makes it more than that is looking for some sort of religion or something IMO.

Here's your obligations: if you're the author, none. If you're a user, whatever the license spells out. Nothing more, nothing less.

Open source works because individual incentives can be mutually beneficial.

Ascribing too much personal meaning to OSS is what causes burnout, just touch grass and if you need money, start a real business or get a job.

eviks3• 2 years ago

If you want to reduce the sum total of human interaction re code to a license, you're part of the religion where the priests are lawyers. But it's not a common religion, so plenty of people behave differently

avighnay• 2 years ago

Yes this is exactly what it is. I think there is so much haze built around it for last 15 years because corporates started using OSS as a means of viral marketing. Individual developers on the other hand saw it as a status enhancer. In this, the original principle of FSF, 'Free as in Freedom, not Free as in Free Beer' was forgotten completely. This in principle is what you have stated above.

The freedom to choose, not the entitlement to demand was the original promise of OSS

marc_abonce• 2 years ago

“only I can do” Sorry, friends, but this is incorrect. You are not the only one capable of maintaining this project.

Technically true and a good thing to tell people who are burning out, but like always, different people need different pieces of advice at different times.

In my case, I mostly stopped my (sporadic) contributions to free software partly because I felt a sort of bystander effect where I assumed that other people would step forward and implement the features and bug fixes that I wanted.

However, in most cases this has not been the case and I think we've all seen issues that affect us that have been reported but unsolved for over a decade, if not forever.

So, in my case it is far more motivating to think that I have to step forward and do the work, because otherwise nobody else will.

Again, this is specifically for people in the opposite extreme of the open source contributor (de)motivation pipeline, so I'm not even disagreeing with OP, just hoping that other people do not get demotivated with this bystander effect.

xuanwo2• 2 years ago

Thank you for the comments. I agree with you as well. That's why I always show great appreciation for the heroes who stepped up to fix the bugs. I want to motivate people to do the same, but with an established baseline: don't burn out, don't overwork, and prioritize self-care.

bawolff3• 2 years ago

Re "i'm the only one that can do it"

I think we in the open source communities have trouble accepting that sometimes projects fail.

In the business world, companies fail all the time, in the open source world a project never really dies until the last person stops using it.

As a maintainer it can be really hard to accept that something you put your heart and soul into didn't work out. However sometimes that is what happens, and you need to allow yourself to move on to new and better things. Taking the lessons learned and applying them to some other project.

brabel2• 2 years ago

What do you mean by "didn't work out"?

Like, it didn't manage to implement the features the author set out to?

Or more like "no one is using it", or even "it's not making any money"?

I feel like, for a passion project, only the first one applies. I have many projects that as far as I know, I am the only user of, and I am happy about that! As far as I am concerned, my project is very successful because it solved a problem I had and that no existing project addressed.

bawolff3• 2 years ago

I mean more - if you are not getting value out of maintaining it anymore and there wasn't enough interest to make the project self-sustaining without you.

But also sometimes things change, what was a good solution once might not be a good solution now.

There is no one answer, all im saying is that there is nothing wrong with moving on if the project no longer interests you.

throwaway20373• 2 years ago

See also: "Open Source Maintainers Owe You Nothing"

https://mikemcquaid.com/open-source-maintainers-owe-you-noth...

karel-3d• 2 years ago

Just give the commit rights to Jia Tan

demarq2• 2 years ago

After I left a job I wrote a library to accomplish something we really struggled with at my previous company.

I realized that if I open sourced and maintained it and my old company used it, then I’d basically be working my old job with no pay.

It completely changed the way I view open source.

stronglikedan• 2 years ago

Surely you could have had a clause in the license along the line of "free to everyone... except Old Company, they can get bent."

demarq2• 2 years ago

Haha, yes that’s a possibility . But in all seriousness I don’t hold a grudge against the company.

It’s just the first time I realized this is thousands of dollars worth of work and a hefty commitment to support a profit making entity.

It just doesn’t make sense.

mxuribe• 2 years ago

That's certainly one perspective....But, another could be that you could now be considered an expert in your library, and your old company's direct competitors are free to hire you for this expertise. This of course needs to take into considerations whatever agreements you have/had with your old company. Yet another perspective is that your old compoany could hire you as a highly-paid consultant to help them support your library, etc...In this latter case, you "win" because you're not helping competitors, and your old company "hurts" ever so slightly by needing to pay you hopefully more than when they had employed you....i'm sure there are other opportunities for you - that have and have not to do with your old employer/company. Then, there's also the more philosopihical perspective: don't look back, just forget them, and move on with your life. :-)

dolmen3• 2 years ago

Except that the company might be interested in paying you to use the library, not to maintain the library itself.

"don't look back, just forget them, and move on with your life" is a better advice.

nickpeterson• 2 years ago

Does GitHub have an option to turn off most interactions on a repo unless you’ve had pr’s accepted? In effect, a public repo that is read only (no comments/issues/etc) unless you’ve contributed?

I’d imagine it can be done through various settings but I wonder if there is a term for such an arrangement?

A letter to open-source maintainers

Related Stories