Open source funding is still unsolved. I have tried plenty of the fixes, and the tip button on my GitHub profile has earned exactly one dollar in eleven months. So when a cold email offered to sponsor s, my terminal search tool, I said yes - not because I need the money, but to see whether this was something the wider ecosystem could use. I had never used pump.fun and I am not a fan of memecoins, which is exactly why I treated it as an experiment. They launched a coin with my project’s name on it, set me up to collect a cut of the trading fees, and roughly $11,000 arrived inside a day. It also produced zero pull requests, zero code suggestions, and not one person willing to test the Android build of my game. This is the whole story, good, bad and ugly.

Why I Said Yes⌗

I do not need funding for my work. That matters for everything that follows, because it means none of this was about the money. What I wanted to know was whether this was a rail the wider open source ecosystem could use.

The problem is real. Paying maintainers has been broken for as long as I have been shipping code, and I have tried a lot of the answers. Gittip came and went. Flipstarter funded real work on BCHD. I have a tip.md button on my GitHub profile supporting Ethereum, Solana, Base and Celo, one scroll above pinned repos with thousands of stars between them. In eleven months it has taken in one dollar. As I write this, their GitHub sign-in is broken, which is a real problem for a service built entirely around tipping through GitHub. tip.md may be joining Gittip.

My GitHub Sponsors profile is live, active, and has met every requirement GitHub asks for. It has one sponsor and an estimated monthly income of one dollar.

Two of the most legitimate funding rails available to an open source developer, and that is what they pay.

For a sense of scale on the problem, omniauth-google-oauth2, a Ruby gem I maintain, has been downloaded over 120 million times. In its entire life it has received zero donations. Not one, at any amount, from anybody.

That is the state of the art, and it is why I say yes to these experiments. What I am looking for is simple: a platform where people fund a project because they use it, believe in it, and want more work to happen on it. Every new attempt might be the one that gets it right, and somebody has to try them and report back. I would rather it be someone who can afford for it to go badly.

This was my first time on pump.fun. I am not a fan of memecoins, and I said so out loud more than once while it was happening. I have been in crypto a long time. Five years at Coinbase building send and receive infrastructure across a lot of chains, and years contributing to Bitcoin Cash. None of that has made me any warmer on tokens that do nothing.

Which is why experiment is the honest word here. I went in with no expectation of success.

The Email⌗

It arrived on July 28, from someone who had clearly read the README.

I came across s and wanted to reach out directly. The idea of turning web search into a lightweight terminal command, while letting users choose from a huge range of search providers, narrow searches by tags, use partial provider matching, and even define custom search engines, is such a clean way to remove friction from everyday research.

Then the offer:

We’d like to back the project with a sponsorship so you can keep building it out. No strings attached, we just want to see more of this happen.

I said I was open to it and asked what they had in mind. The reply pointed me at pumpboard.dev, run by a self-described team of four, and leaned hard on transparency:

Developers and creators we’ve previously supported. The sponsorship amounts each developer received, all verifiable on-chain. Links to their social profiles, so you can reach out to them directly if you’d like independent feedback about their experience.

The site looked like a platform getting started. A leaderboard of developers, amounts next to their names, links out to their profiles. It felt incomplete, but every crypto tipping platform I have ever looked at felt incomplete early on.

Then came the practical part: download pump.fun, sign up, send over your wallet address. I did.

That was my mistake, and I will own it. I handed a wallet address to strangers on the assumption that I was signing up for a donation platform, without asking a single question about how the money would actually reach me. One question would have surfaced the entire mechanism. I did not ask it.

The actual terms arrived in the next message. A sponsorship profile, an X community for the campaign, an introduction post about the project, and a contract address to temporarily add to both my X bio and my GitHub bio. Two or three posts in the community. Stay reasonably active. Roughly one to two hours in total.

Note the order there. “No strings attached” was in the first email. The strings showed up one message after I handed over my wallet address.

There was also a small comedy of errors about where to talk. The email said to DM @samuel_sol on X. That account looked like a bot and had DMs closed. It turned out he meant Telegram.

Donations That Are Not Donations⌗

This is the part that took me a while to see clearly, and it is the most important thing in this post.

Nobody donated anything.

pump.fun users cannot donate to a developer. That is not what the platform does. What actually happened is that a token was created and I was set up to receive creator rewards, which is a share of the trading fees generated every time somebody buys or sells that token. The money is a cut of speculation volume. It scales with how hard the coin gets churned, not with how much anyone values the software.

Created by whom is a fair question, and the answer makes my role here smaller than it probably looks. I did not create the coin. I did not launch it, name it, fund it, or choose its ticker. I did not create the X community that carries my project’s name, and I was never even able to administer it. What I did was sign up for what was presented to me as a donations platform, hand over a wallet address, and write some posts about my own software. Everything else was done by other people, on a platform I had never used. That includes the part with my name on it: the creator rewards were pointed at my wallet without my knowledge. Nobody asked me, and nobody told me. I found out after it was already in place.

A cut of trading fees is a very different thing from a donation, and PumpBoard calls it a donation anyway. The site brands itself “Decentralized GitHub Donations for Developers.” It reports a “Total Donated.” Its walkthrough tells you that you will “Receive donations in SOL.” A donation is somebody choosing to give you money. This is a percentage skimmed off strangers trading a ticker with my project’s name on it, and some of those strangers lose money.

I want to be direct about this, because it is the critical failure in the whole arrangement. Describing a trading fee split as a donations platform is not a marketing shortcut or a simplification for a non-technical audience. It is the single thing a developer needs to understand before agreeing to any of it, and it is precisely the thing the pitch obscures. Had I been told plainly that a memecoin would be launched under my project’s name and that I would earn a cut of the churn, I would have known exactly what I was agreeing to, and it would have been my decision, made with the facts in front of me. I probably still would have said yes, because I wanted to see what happened.

To be straight about my own position, I did suspect a token was coming. It is pump.fun, and that is what pump.fun does. What I did not have were any of the specifics, and what I did not grasp at all was that the money would be a cut of trading fees rather than something anybody had chosen to give me.

I would also have said yes expecting almost no volume. s is ten years old, it sits in maintenance mode, and it has no new features planned. That a coin attached to a project like that would get traded heavily never crossed my mind. That assumption was the whole reason I was comfortable, and it was wrong by a wide margin.

There is an irony in here worth pointing out. Their emails called it a sponsorship. Their website calls it donations. The email was the accurate one.

A sponsorship is a commercial arrangement. Somebody pays you, and in return you promote them. That is exactly what this was, right down to the deliverables: posts on a schedule, a contract address in my bio, a pinned announcement, stay active for the duration of the campaign. A donation has no deliverables. That is what makes it a donation.

So if the coin is what paid me, and promoting the coin is what I did in return, then the honest description is that my sponsor was a memecoin. That is a strange sentence to write about a terminal search tool, and it is still closer to the truth than anything on the website.

Once you understand the mechanism, the leaderboard makes more sense. One project on it, mezo-lite, has a single GitHub star and had claimed $4,500. As donations, that makes no sense. As fee revenue on a token that happened to get traded, it is exactly what you would expect.

I want to be fair here, because one thing about the setup is genuinely solid. The creator rewards were permanently locked to my wallet at 100%, and pump.fun states plainly on the coin page that the admin has removed their own ability to modify that. They gave up the power to change the split later, which is a real structural guarantee and more than the email ever promised.

It was not done purely for my benefit, though. A permanently locked split is also a selling point. The coin page shows my name, my GitHub, and that 100% figure right there in the interface, which tells anyone looking that the developer is attached and that nobody can quietly change the terms later. That makes the whole thing look more legitimate to somebody deciding whether to buy. The guarantee was real, and it worked in their favor at the same time. Both of those are true.

There is a third thing that is true as well. The rewards were attached to my wallet before I knew anything about it, and by the time I found out, the lock meant nobody could change it. Not them, not me. The arrangement was a done deal before I ever saw it, and the only choice left to me was whether to claim.

Launch Day⌗

Communication moved to Telegram, and then into a five-person group named “S Sponsorship Launch.”

The scheduling was courteous. They asked what time worked for me, said they would work around my schedule so there would be no delay, and settled on late morning my time. Then on July 30, at 9:36 AM:

Stay active bro, I’ve pinged every other members and in about 30 minutes or less we launch.

Twenty-six minutes later:

Hey Josh, can we do this tomorrow, same time? After having a quick meeting my team, we have to get the infrastructure right and the market is not really bullish now.

That line is the tell, and it is better evidence than anything on the website. You do not postpone a sponsorship because the market is not bullish. Sponsorship does not have a market. The moment the timing depends on conditions, the thing being scheduled is a trade.

The next day did not go much faster. I was told to get ready in about an hour, then that he was out getting groceries, then that he was almost home. By 11:31 AM I asked if we were doing this or not. At 11:37 AM I was told to give it a couple of minutes. The next message came three hours later.

None of that is fraud. It is just much smaller and more improvised than “a team of four” and “we have to get the infrastructure right” suggest. Operationally, it was one guy driving home from the store.

In between, I was asked to draft a writeup about what s is, what inspired it, and the value it has. I sent one over. It was too long, so they sent a sample of what they wanted instead:

I built S, a lightweight CLI tool that lets you search the web directly from your terminal, no need to break your workflow by opening a browser.

The project has grown to 2.6k+ GitHub stars, and it’s been recognized by developers across the industry. I’m also followed by leaders like the CEOs of Coinbase and Y Combinator, which has helped bring more visibility to the project.

Excited to be onboarded and looking forward to claiming.

I wrote five replacements. Every one of them left out the name drop. It happens to be true, but I hate name dropping and it had nothing to do with this coin.

This is the one I actually posted:

Hello everyone, glad to be here today.

Starting a search used to mean leaving my terminal. Reach for the mouse, find a window, click into a field, then finally type the thing I already had in my head.

Once, that’s nothing. Fifty times a day it adds up. s collapses all of it into four keystrokes and a query.

Glad to be here, and ready to claim. =)

Theirs sells the author. Mine sells the problem. That difference is the whole distance between marketing a coin and talking about a project.

That sample also tells you exactly why I was approached, and it is not because s is good software. It is because I have a reputation that markets well to pump.fun users. Ten years of commits, a couple thousand stars, a long history in crypto, a following with some recognizable names in it. That is a story a trader will buy, and a coin needs a story. The pump.fun page for the coin makes the same pitch in its own words, describing me as a developer who worked on Coinbase’s Solana integrations. Accurate, and entirely beside the point of a terminal search tool.

Which project they chose says the same thing. s has the most GitHub stars of anything I have written, and a star count is a number you can read off a profile page in about a second. It is nowhere near my most used software. That would be the gem with 120 million downloads, which nobody has ever offered to sponsor.

What I kept, without thinking much about it at the time, was the closing line. Their template ends every post with the author announcing they are about to claim, and my rewrites carried some version of it too. That was their framing rather than mine, and I would leave it out if I did this again.

The rest of the day was a sequence of instructions. Join the community. Post. Add the contract address to the GitHub and X bios. Make a short post showing support for the launch, include the contract address, and pin it. When I should claim was on their schedule as well.

The first claim came to 72.9 SOL, a bit over $5,000, in an afternoon.

For the record, the email had described all of this as roughly one to two hours of staying reasonably active. I was in that group from 9:41 AM to 4:07 PM.

The Space⌗

I opened an X Space and ran it for about two hours. A thousand people tuned in over its life, with more than two hundred in the room at once. I did not record it, so everything here is my own recollection rather than a transcript.

I talked about s, why I built it, and how people actually use it. I talked about my time at Coinbase, why I am in the crypto space at all, and the open source funding problem that got me into this in the first place. I dropped some early details about the coming Android launch of Perfect Shot and the work I have been doing on Nexus, my BBS client and server.

I also said, plainly and more than once, that I am not a fan of memecoins. I said the coin was not going to be used for anything, and that I had no plans to give it utility. When one of the speakers got onto the subject of memecoins generally, and which ones were worth buying, I told them not to risk money on memecoins at all, and that coins with real utility are the only ones I find interesting.

I was earning a cut of every trade while I said it, and I would say it again. I never supported the coin. It was a mechanism for funding open source work, nothing more, and pretending otherwise to a room full of people putting money in was not something I was willing to do.

That was not a performance for the blog post. The same thing happened privately in Telegram that evening, when I was asked whether I would really incorporate utility for the coin. My answer was that it was doubtful, the coin is just for people to support, and I do not plan on any direct utility.

What I Did With the Attention⌗

I made eleven posts in the community over two days. They picked up about 63,000 views between them.

Only one of them carried the contract address, and it was the one I had been asked to write, a short post supporting the launch. The contract address also went into my GitHub and X bios for a short while, then came out. Copycat coins showed up almost immediately, and once it was obvious which one was real, there was no reason to keep pointing at it from my profiles.

Everything else I posted was about actual work. What is everyone’s favorite terminal browser, and why w3m pairs so well with s. That Presearch died recently and s now defaults to Brave Search. That s can run as a server, and that you can see one running at jumps.io or just run it locally for privacy. Nexus. Perfect Shot. A link to my last post about s, and a note that 0.8.1 has shipped since.

I also spent a fair amount of the time moderating. There were wallet drainers being posted in the community and I banned people for it. I asked to be made an admin so I could clean it up properly, and was told I already was. I was not. I was a mod, and the only admin was somebody named Ackerman who I had never dealt with directly. So the community carrying my project’s name was one I could police but not control.

Who Was Actually There?⌗

Here is where the experiment produced a result.

Sorted by engagement, the notable ones went like this:

  • Presearch dying, plus s running as a server: 1.9K views, 8 likes
  • Favorite terminal browser, w3m or links: 2.4K views, 12 likes
  • Nexus, my BBS client and server: 5.3K views, 34 likes
  • “Super excited about the launch,” with the contract address: 8.5K views, 57 likes
  • Thanking everyone for the Space, plus Perfect Shot’s Android launch: 11K views, 83 likes
  • The Space itself: 15K views, 52 likes

My two most technical posts sit at the bottom of that list. The post with a contract address in it did three and a half times the views of the terminal browser question.

The pattern held everywhere I looked. GitHub traffic to the repo jumped from a baseline around fifty views a day to roughly three hundred and sixty at the peak. Clones did not respond at all. Both clone spikes inside that fourteen-day window landed before the launch, and the launch itself barely registers on the graph. Of the sites referring traffic in that period, the only crypto one sent nine views.

People opened the page. The clone graph never noticed.

The final tally, against 63,000 views and a thousand people in a Space:

  • Zero pull requests
  • Zero code suggestions
  • Zero signups to test the Android build of Perfect Shot

The Android testing one stings the most. I talked about the coming Perfect Shot build during the Space and again in the community, in a post that pulled eleven thousand views and eighty-three likes. Not one person sent an email.

Plenty of the replies did not read like people either.

So I am left with questions I genuinely cannot answer, and I mean them as curiosity rather than judgment. Who are these people? What makes a coin attached to a ten-year-old shell alias replacement worth buying, with the developer of the software in the room saying it is a bad idea? I would like to understand it, because from where I sit it does not add up.

I do not think many of them could have used s, or wanted to. The moment I said something that required knowing what a terminal is, the attention went somewhere else.

Which leaves the question I most want an answer to. What did they think they were buying?

The coin is not equity in s. It is not a claim on the repository, the roadmap, my time, or the fees. It carries no governance, no revenue share, and no promise of anything at all. Holding it entitles you to precisely nothing beyond the hope that somebody buys it from you at a higher price. That is not me being harsh after the fact, it is what I told the room while it was happening.

So either they understood that and spent an afternoon trading a ticker, or they did not understand it and believed the coin represented some piece of a ten-year-old terminal tool. I have gone back and forth on which of those I find worse.

The Numbers⌗

The coin hit an all-time high of $308K in market cap and, as I write this a day later, sits around $69.5K on $1.34M of 24-hour volume. That volume is the whole story of the payout. Roughly twenty times the market cap changing hands in a day is what generates the fees.

I have claimed about $11,000 so far, and it is still accruing.

Set that against what the two rails on my GitHub profile have paid in their entire lifetimes, and you have the uncomfortable center of this whole thing. The honest options have no liquidity behind them. The speculative one has real money. Not because it is better, and not because anyone involved cares more about open source. Because there is a token.

And it is worth being precise about where that money comes from. It is fees on trades, which means it comes from the people trading. Given that fall from the high, anybody who bought near the top is holding a loss of roughly 80%, and it dropped 40% in six hours at one point on the way down. The coin page puts all of this in the open, with each trader’s position and profit and loss sitting right next to their posts, and plenty of those numbers are red. Mechanically, that is where my money came from.

At least the copycats do not seem to have caught anyone. We spotted those quickly, and I do not know of anybody who bought a fake.

Was It Worth It?⌗

As an experiment, yes. It answered its question clearly, and the answer was not the one I would have guessed.

Being that wrong about the scale was genuinely surprising, and I say that as someone who spent five years building payment infrastructure in this industry. If I misjudged it by that much, most developers will too.

What I got wrong in the other direction is more interesting. The pitch was that this gives developers visibility for their projects, and my own post at the time said as much. The visibility was real. Sixty-three thousand views is real. But it was not developer visibility. It was trader visibility, and it was worth nothing to the software.

On PumpBoard specifically, I want to split the verdict. I think they legitimately tried to get funding to me, and the locked creator rewards were a real commitment rather than a promise, even if that lock also sold the coin. But the whole approach is built to cater to traders and gambling rather than to find people who actually want to support open source, and calling that a donations platform was not honest.

I will probably not do this again. It was still neat to see what actually happens on pump.fun.

The honest summary is this. It produced funding. It did not produce funding that is a net benefit to open source developers.

The money is real and it will do real work. But it came from people gambling on a ticker, not from anyone who wanted the software to exist. It arrived because I have a marketable story, which means it is not available to the maintainer of a critical library who does not have one. And it bought no code, no contributors, and no users. If every open source developer did this tomorrow, open source would not be better funded. There would just be more coins.

None of it was ever about the code.

What I Would Tell Another Developer⌗

Be careful, and do a lot more research before jumping into anything on pump.fun.

Understand what you are actually agreeing to. If someone offers to fund your project on a launchpad, ask them directly whether people are donating to you or whether you are collecting trading fees, because those are completely different things and only one of them is what the word sponsorship means. Ask who controls the community that is about to carry your project’s name. Ask what happens to the contract address in your bio when the copycats arrive. And understand that you are being approached for your reputation rather than your code, because your reputation is what sells the coin.

And know what you are getting. The pump.fun audience is not the place to promote open source software if you want an audience that actually cares about your projects. What I saw was people trying to make a quick buck.

If you go in anyway, be honest with them. Tell them plainly if the coin does nothing, because it probably does nothing. Do not promise utility you have no intention of building. That costs you nothing and it is the least you owe people who are putting money at risk on your name.

Giving Something Back⌗

I should be clear about what I actually want out of this, because it is not money.

No pull requests, no code suggestions, nobody willing to test a build. That is the whole point of the post. I would trade the entire payout for a handful of good bug reports and a couple of pull requests, and I mean that. Money keeps infrastructure running. Contributions and feedback are what make software better, and they are exactly what this experiment failed to produce.

Which leaves the money itself, and a fair question I would rather answer than leave sitting there. I do not need it, and a share of it came from people who lost on the trade. Quietly pocketing all of it would be the wrong move.

The coin is still generating fees, and I intend to keep claiming them. Leaving them to sit in a vault does nobody any good. What matters is where they end up, and that is open source development, whether mine or somebody else’s.

Where it starts was never in question. It was clear from day one that the coin existed to support s, so that is where it goes first, along with the Kubernetes infrastructure I run and the work behind everything else I maintain: plank-reloaded, Nexus, celestial-gtk-theme, dream-prompter, fortunate, omniauth-google-oauth2, and the rest of the list.

I am also open to sending a good portion of it to other open source projects that need it more than I do, and I would rather not choose those on my own. If you know a project that deserves funding, tell me. I would rather the people who showed up for this help pick where some of it lands than pick it all myself.

As for the people who bought in, what I can offer has nothing to do with the coin.

The door has always been open and it stays open. Open an issue on any of my projects and it gets my attention. Send a pull request and I will review it properly instead of leaving it to sit. Bring a feature request and I will give it real thought, and tell you straight if it is not something I will build. Want to test the Android build of Perfect Shot when it lands? Email me and you are on the list.

None of that needs a wallet. All of it needs you to show up.

Open source funding is still unsolved. I understand it better than I did four days ago, and the three zeros taught me more than the payout did.

The only true donations in this entire story are the dollar on tip.md and the dollar a month from my one GitHub sponsor. Two people who decided the work was worth supporting, and gave. No ticker, no trade, no deliverables. A platform full of people like that, people who believe in open source and want to help the developers doing the work, is the one I have been looking for all along. Nobody has built it yet.

The tip button is still on my profile. It is still sitting at one dollar, and I am no longer sure it works. That is fine. There are a lot of ways to support open source: pull requests, issues, translations, donations for development. I will gladly take any of them, and so will every maintainer you know. What matters is the one thing you cannot verify on-chain: that the heart behind it is in the right place.