> ... we have raised over $10,000 towards open source sustainability.
In October, they brought in $7,437.20 from sponsors. They paid out $3,413.20 to open source developers.
40% was to be paid out to developers. I realize there are some "administrative costs" and it takes a while to ramp things up but that seemed like way too little being paid out.
On the other hand, that's $3,413.20 that otherwise might not have went into the pockets of developers so I'm still not entirely sure how I feel about Code Sponsor.
This isn't directed at the commenter, but the concept above: Can we all come to an agreement that a person running an endeavor be permitted to feed, cloth, and accomodate themselves reasonably?
Specifically, this crab mentality of dashing hopes of someone ending up right side up on a side project or even a nonprofit. Why pick on the little guys who are just trying to get by, especially if they have a noble intention of funding open source projects.
This is something that we really need, and I feel larger organizations that fund open source - not naming names - aren't transparent or egalitarian enough with who and how they fund. So many people do open source and don't see a penny from it. In 2017, if you want to receive any funding for an open source project, you have to be in a certain social circle, very popular, or very fortunate.
There are organizations that horde millions in cash reserves and give six figures to people just for sitting on a board. Then, there are people just trying to bootstrap themselves on a shoestring budget to buy groceries and pay rent to their roommate. It is my sincere hope we are more generous to people just trying to scrape by and not trying to unduly hound them and guilt trip them into not being able to pay their basic living expenses.
$3,413.20 over a period of four months? Thankfully he also had a day job, because we also don't know if he had other developers, or what were his true needs were to be able to sustain his project.
If the split happened to be 80%/20% or 90%/10%, then no one would complain even if the author was pocketing $1M/mo. In fact, I and many other people would likely be ecstatic.
However, knowing the actual split, and knowing that the majority of funding didn't actually go to developers, it's a lot more difficult to make the social argument at this time. It necessarily falls back on the practical question of whether it is effective to advertise in the READMEs. Most of the previous sponsors were drawn into the social cause but not necessarily the practical effectiveness of the ads themselves.
Even though it's not a charitable cause, they are trying to make the same appeals. And if a charitable organization was eating away more than 50% of the donations in administrative costs, you would likely be very wary.
That evades the point and reintroduces the same thought pattern I've criticized in my original post.
> Even though it's not a charitable cause, they are trying to make the same appeals. And if a charitable organization was eating away more than 50% of the donations in administrative costs, you would likely be very wary.
The funding is low, less than a fast food worker. The person running the operation couldn't even pay his own rent to run it out of an apartment.
Before we start slicing up the pie, we need to have the funds reach a point where they can sustain themselves. For instance, basic living expenses in midwestern US. Funds after administrative costs should be the pie.
We have to stop shaking down people who want to help open source when they ask for funding. It's as if whenever someone tries to fund an open source project or create a way to fix this problem funding in open source, they're hounded and dogpiled.
I've seen this sort of entitled, stingy attitude on the part of open source users, that despite using open source software to their benefit, those doing the work behind the scenes are shamed for trying to sustain themselves.
It's this strange double bind: first you want your software working, new features, and that bug fixed, but the one's facilitating that very thing for you can't have the material means to do it without being shook down.
When I see 10k in funding, less than poverty wage in USA, and people talking about it as if it was a prize pool instead of operating expenses, that's the mentality I'd like to see done more thoughtfully.
You make some great points. I do have one correction though:
> Most of the previous sponsors were drawn into the social cause but not necessarily the practical effectiveness of the ads themselves
This is true for a few of the sponsors, but not the majority. The majority were drawn into Code Sponsor because it can be as effective as Google Ads, while 40% of the marketing funds go back into developers pockets.
Code Sponsor was never presented to sponsors as a charitable organization, nor did we try to present ourselves as such to the community.
Here is the gist that I shared with sponsors to explain how Code Sponsor worked:
Hi. This is Eric, the founder of Code Sponsor. I appreciate your concern here. Please understand that for the last four months, I’ve been getting up at 5am and working till 7:30. I then would go to my day job every day and then get back to Code Sponsor in the evenings. I also worked every weekend. It was a huge struggle for me and my family to put in the time.
The reason I had a 40% rate was in hopes that I could make the business sustainable enough to pay me. Not once did I take a distribution.
It was always a slippery slope. The last thing Github wants is banner ads all over the place. Sure, this one was tasteful, but if they allow it the door is open for less savory actors.
Eric, how is your service tied to GitHub? As far as I understand it, you simply are a matchmaker for developers and advertisers. And you track clicks to the advertisers urls. Where the developers put the links to those urls is not your decision, right?
If so, I don't think GitHub has an angle to attack you.
A developer with a Commission Junction account could very well put CJ links in their README. That would not give GitHub a vector to attack CJ.
But why not get Github out of the equation?
I know it might seem a lot to ask but there are alternatives, Github has their right to ban you but it don't need to be the end. Or am I missing something?
> 60% of all traffic through Code Sponsor came from Github. 15% came from ad blocked traffic (I have no idea the referer). About 15% came from websites and then the remainder came from different repository websites (dockerhub, npm, etc).
From the monetizable repo views, 86% of all views were from Github. Barring a massive cultural shift, Github has to be part of the equation.
Yes, we are a way for sponsors and developers to connect. We do not tell developers where in their code to place the banner. In hindsight, this may have been part of the problem. Perhaps we should have been more rigid in how we allow developers to present the banner.
TAForObjReasons is exactly right. Github has to be a part of the equation.
Unfortunately, Github is where open source lives. The success that Code Sponsor saw was due to the minimal steps it took for a developer to start getting paid for open source. Asking developers to move their open source projects to Gitlab would not work.
Because most projects are not already there. Moving a community from GitHub to GitLab isn't trivial, not to mention all the work required to update material and avoid broken links.
When Eric is describing having to wake up at 5 am and work weekends to make Code Sponsor work even with the proceeds of 60% of all OS projects his ad network funded, it should be pretty clear the other 40% divided among many OS coders couldn’t be enough for their project’s development to feel sustainable.
As much as I commend Eric on trying to find a solution to OS maintainability, I think he actually just reinvented something already available to OS developers, a niche tasteful ad network, I.e carbon ads <https://carbonads.net> something which many OS projects with a large enough audience already include to support development.
Carbon Ads is a wonderful company. They provide an excellent service. However, per Rollbar, Code Sponsor performed better than they did at about 1/3 the cost. What does this mean for the developer? Scaling. Rollbar was able to scale their budget because Code Sponsor provided more interested developers to their product along at a much lower rate.
What Code Sponsor is is a mashup of Carbon Ads, Open Collective and Read the Docs (with a hint of grassroots marketing demonstrated by Wesbos). The 40% was enough for developers. I paid one developer last month over $600. Two others earned over $400.
Finally, Code Sponsor was available to every open source project. The popularity of the project is what self-governed the amount of money they would receive.
Carbon Ads also does not provide any visibility into what is going on. I have taken extreme effort to give sponsors and developers full visibility into their traffic patterns. Daily statistics and charts. This allows sponsors to immediately know how their campaigns are doing and shift if necessary.
Carbon Ads is great, but they are a black box and sponsors want more.
But it doesn’t mean that Codesponsor needs to shut down. They could easily check the referrer and show the ads only on npm, dockerhub, gitlab, hackage etc and they would still get good coverage.
60% of all traffic through Code Sponsor came from Github. 15% came from ad blocked traffic (I have no idea the referer). About 15% came from websites and then the remainder came from different repository websites (dockerhub, npm, etc).
I agree that this is a good idea, but open source exists on Github. Without their support, it just won’t happen.
Nevertheless, couldn't you just leave it running without GitHub' approval (perhaps, block requests with referrer from GitHub) and see if it gains further traction?
The personal cost it would be for me (time from family) to continue to grow Code Sponsor on the side is not an option. I love open source and want to see it be sustainable for developers, however, I can't do so anymore at the expense of my family.
This is too bad to hear, I only recently learned of Code Sponsor on JavaScript Jabber[0] - a pretty excellent podcast. I’m hoping there are good things to learn here and maybe iterate on to find what can work in the space.
Good. I would hate to see OSS go the route of online content, where it's generated primarily for clicks. GitHub is no saint but I'm glad they have this policy.
Totally agree, I see more and more ads on Github. I even get a bunch of SPAM to put ads on my projects since I got a couple of fairly popular projects.
Monetization is still a huge problem, just this is not the way to go about it.
What other solutions do we have to provide scalable funding to open source projects? This comment really frustrates me. Do we put more buttons on our repos asking for money from other developers? Do we try to find corporate sponsorship and hope that they don’t try to influence the project? And if we did want to go that route, how do we do it?
Why is this not the right way to go about it? The “ads” were relevant, unobtrusive and ethical. There was never any click tracking, never any remarketing, nothing.
Code Sponsor provided the FIRST and ONLY means of scalable funding for open source projects. It was not depending on charity from others. It was a business transaction between sponsors and developers. The sponsors saw a return and the developers got paid to continue doing what they love.
If anyone else has a better idea, please share. Until then, I claim that it’s IS the right way to go about it.
I understand your frustration: in an ideal world, we should be able to easily reward people who contribute to the greater good, and open source developers are certainly in that category.
However, this reward may not always be money: community recognition and extended network, better chances of hiring, or attribution from projects that depend on you should also be considered.
Yeah, it is valid to say that funding needs to be improved. But while ads provide a way, it is a stretching to call it the only or the right way. Maybe open source development is just fine with corporate sponsorship + people doing it in their free time. Maybe developers are repelled by ads so much that it actually hinders contributing to interesting projects. Maybe charity and donations work better we just haven't figured it out yet (see Patreon, which had a surprising success, despite the skepticism around microfunding).
> Community recognition and extended network, better chances of hiring, or attribution from projects
This has been the case since day 1. It is a great motivation for people to get into open source, but there gets to be a point where this is no longer a motivation. As their projects become more and more popular, they are having to work nights and weekends to keep the project alive and maintained.
> May be open source development is just fine with corporate sponsorship + people doing it in their free time
Again, this is how it’s always been. Most open source projects do not look for corporate sponsorship because they don’t know where to start. Working in free time is great until it isn’t.
> Maybe developers are repelled by ads...
This is true. Some are repelled. But ask yourselves if the Code Sponsor banner looks like an ad? It was crafted in a way to look enough like the documentation to not distract developers from the README, but different enough to not be considered deceiving.
> Maybe charity and donations work better
Ask Kent C. Dodds this question. He had the charity buttons on every one of his repos and he received $0 in donations. Scalable funding cannot come from charity, it has to come from marketing funds.
It is a great way, just hard to pull off on a platform like github combined with the general distrust of adtech that's common with devs/HN crowd.
Ads are still the most scalable, fast, easy and passive way to generate revenue but unfortunately the adtech industry has poisoned the well so people forget all the smaller companies trying to do the right thing. Perhaps opencollective can pickup something like this. Good luck on your next project.
A lot of people are saying that Github is shutting down Code Sponsor. This is not the case. Github is not forcing Code Sponsor to shut down. They are only forcing us to remove all banners from their README’s. The decision to shut down is because the impact Code Sponsor could have on sustaining open source is minimal without Github’s blessing.
So, GitHub isn't shutting down Code Sponsor, just taking action that makes it impossible for you to function effectively which is, as a practical matter, forcing you to shut down.
That seems to an extremely fine distinction, bordering on one without any difference at all.
What about the repos which have documentation off github and do sponsorships. Would code sponsor work in that space? Sorry I don't know the details of how code sponsor works but I do see lots of sponsorships on the main docs for lots of solid open source projects.
Ah got it, so in respects to your business github's value was as a channel of distribution... and without it your costs of marketing and user acquisition sky rocket?
Donate buttons don’t work. They are asking for money from other developers. Imagine 100 developers in a room and each one takes out their purse or wallet and hands a dollar to the left. It’s not scalable, and it feels like begging.
Code Sponsor took the approach that scalable funding should come through marketing budgets, not charity. Companies pay developers to place a non-obtrusive, relevant and ethical banner on their repo or website and the developer gets paid on a per-click basis. The larger the repo, the more clicks they get. It is self-normalizing.
I believe there is a chance of making it happen through donations like Patreon did for creatives. A Patreon for OSS would be a good pivot, don't you think?
Patreon already does OSS if the creator wants to list there, but there's also https://opencollective.com that has gained traction with many top projects for official funding.
In October, they brought in $7,437.20 from sponsors. They paid out $3,413.20 to open source developers.
40% was to be paid out to developers. I realize there are some "administrative costs" and it takes a while to ramp things up but that seemed like way too little being paid out.
On the other hand, that's $3,413.20 that otherwise might not have went into the pockets of developers so I'm still not entirely sure how I feel about Code Sponsor.