Sure. But thing is, there’s software out there for which FOSS doesn’t even make much sense.
I’m talking things that are so niche, the total amount of potential users (not customers - that’s a much smaller number) is in the hundreds of thousands, not even millions - most of whom have no say in what software they use, nor does it affect their pay checks.
If I was building, say, accounting software that every company can use, that’d be different, because while still business focused, there’d be a lot more grass roots interest in it. But I’m talking about software where you have to sell it to a bunch of execs, along with support contracts and uptime guarantees, because their entire business is dependent on it functioning properly. I’m also talking about software for one niche of one industry in one country.
The project isn’t useful enough to you, an engineer, to reverse engineer the backend. Nor are there any open alternatives that work. It requires keeping up with regulations, including some that change every year. It’s not that the software itself is super complex magic, it’s that it stops being useful if not well-maintained.
What I have considered, though, is making parts of it open source, and keeping only the “secret sauce” proprietary. The open source parts would be stuff that could be used to build similar software for other niches of the same target industry, whereas the super specific niche stuff and all the regulation compliance stuff (much of which is just for that one niche anyway - other niches have different regulations) would be proprietary. Essentially building a set of FOSS libraries, and a niche proprietary application that uses them to service a specific market. Again, good reason for using a forge where you can have both public and private projects - but of course I could just use CodeBerg for the open source and host the rest of it privately.
I’m only building this in my spare time and fairly slowly because I have to do work that gets me paid though. I don’t know if I’ll ever have an MVP I could show investors or clients.
What I have considered, though, is making parts of it open source, and keeping only the “secret sauce” proprietary. The open source parts would be stuff that could be used to build similar software for other niches of the same target industry, whereas the super specific niche stuff and all the regulation compliance stuff (much of which is just for that one niche anyway - other niches have different regulations) would be proprietary.
This seems perfectly reasonable and I wish you the best of luck. Just don’t expect anyone to provide the infrastructure for your proprietary secret sauce for free!
Well, github would provide it for free. Their business model is that just hosting shit is free, but costing them actual server resources means you gotta pay 'em. And that’s a sensible business model IMO, but unfortunately they’re also owned by Microsoft, which I didn’t even like 2 decades ago, let alone now that they’re pushing AI.
Guess what I’m hoping is for Github alternatives, potentially based on Forgejo, to adopt a similar business model (free storage, paid runners beyond a very limited free tier essentially), without the whole using everyone’s code for AI training part.
I also have no problem with a small recurring donation. But the ironic part here is that I wouldn’t want to use a forge that’s so small that it NEEDS the donations. I don’t want it to disappear after a year.
Someone might spin that up, but it feels unlikely. Github was always kinda subsidized as a power play on MS’s part, and now that it’s well established enough they’re squeezing it for ROI. An instance that doesn’t need your donations still needs resources to perpetuate itself from somewhere, I’d personally rather depend on infrastructure that was transparent about that (whether paid or donation based) than be treated as the product
Sure. But thing is, there’s software out there for which FOSS doesn’t even make much sense.
I’m talking things that are so niche, the total amount of potential users (not customers - that’s a much smaller number) is in the hundreds of thousands, not even millions - most of whom have no say in what software they use, nor does it affect their pay checks.
If I was building, say, accounting software that every company can use, that’d be different, because while still business focused, there’d be a lot more grass roots interest in it. But I’m talking about software where you have to sell it to a bunch of execs, along with support contracts and uptime guarantees, because their entire business is dependent on it functioning properly. I’m also talking about software for one niche of one industry in one country.
The project isn’t useful enough to you, an engineer, to reverse engineer the backend. Nor are there any open alternatives that work. It requires keeping up with regulations, including some that change every year. It’s not that the software itself is super complex magic, it’s that it stops being useful if not well-maintained.
What I have considered, though, is making parts of it open source, and keeping only the “secret sauce” proprietary. The open source parts would be stuff that could be used to build similar software for other niches of the same target industry, whereas the super specific niche stuff and all the regulation compliance stuff (much of which is just for that one niche anyway - other niches have different regulations) would be proprietary. Essentially building a set of FOSS libraries, and a niche proprietary application that uses them to service a specific market. Again, good reason for using a forge where you can have both public and private projects - but of course I could just use CodeBerg for the open source and host the rest of it privately.
I’m only building this in my spare time and fairly slowly because I have to do work that gets me paid though. I don’t know if I’ll ever have an MVP I could show investors or clients.
This seems perfectly reasonable and I wish you the best of luck. Just don’t expect anyone to provide the infrastructure for your proprietary secret sauce for free!
Well, github would provide it for free. Their business model is that just hosting shit is free, but costing them actual server resources means you gotta pay 'em. And that’s a sensible business model IMO, but unfortunately they’re also owned by Microsoft, which I didn’t even like 2 decades ago, let alone now that they’re pushing AI.
Guess what I’m hoping is for Github alternatives, potentially based on Forgejo, to adopt a similar business model (free storage, paid runners beyond a very limited free tier essentially), without the whole using everyone’s code for AI training part.
I also have no problem with a small recurring donation. But the ironic part here is that I wouldn’t want to use a forge that’s so small that it NEEDS the donations. I don’t want it to disappear after a year.
Someone might spin that up, but it feels unlikely. Github was always kinda subsidized as a power play on MS’s part, and now that it’s well established enough they’re squeezing it for ROI. An instance that doesn’t need your donations still needs resources to perpetuate itself from somewhere, I’d personally rather depend on infrastructure that was transparent about that (whether paid or donation based) than be treated as the product