I wrote the first line of code for /kbin on January 14, 2021. Around this time, I started working remotely and decided that the time I used to spend commuting to the office would be devoted to /kbin. Throughout this entire period, /kbin has been a hobby project that I developed in my free time. It was also when Lemmy started federating. The full history is available on GitHub. The Polish instance - or rather its prototype - was created on 2021-09-08.
By the end of 2022, I decided to take this a bit more seriously. The work that had brought me much satisfaction began to tire me out - anyone who’s experienced burnout likely knows what I’m talking about. I needed a breather and a sense of doing things my way. I had some savings put aside, so I could work on this full-time. The amount of code might not reflect this, but it’s only a small part of the things that need attention in such a project ;)
I don’t know if it had any impact, but on January 4, 2023, I received information that the project had qualified for the NGI0 Entrust program. I had applied for funding a few months earlier. Currently, I have outlined my milestones in the Roadmap. The plan was to gradually complete each stage (after finishing one of them, I can apply for a funds release). However, due to the situation and how /kbin has developed in recent weeks, I had to completely change my priorities. As a result, I have started each stage, but none is polished enough for me to honestly apply for a payout. I’ll need to address this promptly.
The fact that I could take certain steps amidst all this confusion is solely thanks to your support. The kindness I’ve encountered here will be remembered for a lifetime :) My buymeacoffee account currently has 818 supporters, who have donated $11,320. This is a lot of money, and for a while, I’ll be able to sleep peacefully, not worrying about maintaining kbin.social.
Nevertheless, this money is meant for project development. Every expense will be documented in monthly reports. If necessary, I can also provide insight into the invoices. Things have been so heated recently that I consider the spending over the past months to be a failure. Most of the costs need to go to S3+Cloudfront, where costs due to the traffic increased from $2-3 per month to $1,000. This is about half a year of basic servers in the current stack. But in hindsight - so much has happened that faster migration was impossible. However, this has certainly accelerated the process.
None of this would have been possible without the contributors and project guardians, and without Piotr, with whom we spent many hours and sleepless nights trying to stabilize the situation and bring it to its current state. This time we’re much better prepared for potential surprises. I hadn’t set the terms of collaboration before and I admit, I had some concerns when we arranged a call to discuss this. However, it turned out that within the foundation, Piotr introduced a “Pay what you can” financing model, whether it’s $1 or $100 a month. As I mentioned earlier, this is a huge relief for me and we started from scratch regarding security matters.
Many of you asked me about the possibility of recurring support. I wasn’t entirely convinced, especially since the current account balance should maintain the instance. However, I think it would be irresponsible of me not to consider it. /kbin has grown to a level where I can’t foresee everything that will happen. It would be great if we could cover monthly costs with Patreon / Liberapay. All funds from Buy Me a Coffee will be transferred to this pool, but from now on, I’ll treat it as buying me a coffee… or a beer… literally ;)
For me, this also means maintaining critical zones for the project. I see this as a long-distance run, so I’ve decided to allocate:
$100 monthly - donation to Piotr’s foundation “Fundacja Technologie dla Ludzi” - I really encourage you to support it, they’re really doing a lot for the fediverse.
$24 monthly - donation to Codeberg - a great ecosystem for free projects. We’ve been making quite a buzz there recently.
I also want to support contributors and creators around /kbin as much as possible - but I’ll do this privately, and for now, I can only afford symbolic amounts.
| | | | $ |
| ------------------------------------------------- | ---------- | ------------ | ---- |
| Hetzner Jun 2, 2023 | €131.63 | one-time | 145 |
| Hetzner Jul 2, 2023 | €246.74 | one-time | 271 |
| OVH 24 cze 2023 | 2246.66 zł | 6 months | 553 |
| OVH 1 lip 2023 | 904.63 zł | monthly | 223 |
| OVH domains | 116.43 zł | annually | 30 |
| AWS (S3+Cloudfront) July 3, 2023 | $1079.21 | one-time | 1080 |
| AWS current | $320.45 | one-time | 321 |
| Mailgun 2023-07-02 | $49.76 | one-time | 50 |
| Testing enviroments, demo instances, landing page | $130 | monthly | 130 |
| FTDL | $100 | monthly | 100 |
| Codeberg | 95.33 zł | monthly | 24 |
| Yubico 2x YubiKey 5C NFC Jun 22, 2023 | €135.30 | one-time | 149 |
| Accounting and legal advice | $100 | one-time (?) | 50 |
| Taxes in Poland | ??? | | |
Thank you once again for that. I will respond to your questions, but it may be delayed as I have a few important tasks I want to focus on. Soon we will also write more about the cluster and the conclusions we have drawn from creating infrastructure with Piotr. Then it will be time for the first release of /kbin.
@ernest, if Kbin starts making okay money, don’t be afraid to give yourself a salary. It’s important that you get to eat too.
I agree … one of the greatest things I’ve seen in FOSS has been #HomeAssistant growing to the point that Nabu Casa can employee 25 people to work on the project (I have no idea if they’re all full time or what, but I know at least a decent chunk are).
If I spin up an instance, whether it stays afloat is between me and the people on my instance, but if we want the flagship to stay up and for our dev to have the time/willingness to make improvements, he needs to get paid. Even just project managing a project of this size is an immense undertaking and just accepting PR’s from others can get to be crazy.
I’d honestly prefer to not have to decide between “I want this to go to /kbin” or “Ernest is ‘allowed’ to buy a beer with this”. I’d prefer to donate to something that ensures /Kbins needs are met for x amount of months and then the rest is split between employees of the org at whatever ratio is agreed upon. That’s just my $.02 … I really do appreciate that Ernest wants to be so careful with the fund though, I just don’t want the /Kbin account to be sitting multi-thousands of dollars in the black while Ernest is struggling with basic subsistence.
As I’ve been lurking around the fediverse, running instances seems to be universally a hobby project, and it’s a little concerning. It kind of gives the impression of all being idealistic young kids embarrassed to ascribe value to their own time. I mean, you can do a lot with volunteer labor, especially if it’s a good ecosystem with appropriate recognition and gratitude, but the people are absolutely the most valuable parts of kbin.social, lemmy.world, etc, and they do have to eat, pay rent, go on vacation. It’s tough to respond to a 3am message about your instance being hacked if you have a job to be at four hours later, and leads to a whole different kind of burnout.
It’s early days yet, but I hope the bigger instance teams get some input from people who’ve managed growth spurts in non-profits, and especially the transition to their first paid staff members (even when that staff member is the owner).
Maybe this was also how the internet was intended.
I’m not sure if you’re joking but it’s not that long that’s how it used to be.
@elscallr@kbin.social Well some history. IPv4 (and later IPv6 now) was meant to connect computers together, ideally without any router/modem in between but each device directly on the web (but ipv6 came too late). So we got an interconnected web.
Later Tim Berners-Lee just want to have a human-readable documents to be linked together, with a distributed architecture that would see those documents stored on multiple servers, controlled by different people, and interconnected. I think the fediverse comes pretty close to this idea.
I also think big companies and centralized solutions might make it easier for the user, but we also now know all the downsides of those solutions from Google, Meta, Amazon, Microsoft,… you are the product.
@ernest@kbin.social @jon@kbin.social @tburkhol@lemmy.world
ipv4, and the structures that came before it, were meant for academics and military commands to talk to each other on government funding. It was the definition of an elitist space and filled with idealistic kids and dilettantes who didn’t need to worry about rent. Nobody in the public would even know it existed for a decade.
Hear, hear