[imiric]

I do think this is the most sustainable approach for monetizing open source projects, as long as the developer is fair about it. That is, truly make features that only large and/or enterprise users would have a use for, and only charge for those, instead of arbitrarily deciding to put core features of the software behind a paywall.

In the case of a web framework, that choice is a bit difficult. But if the software is fully functional for the large majority of users, then charging for niche features, or those that are actively discouraged, sounds like a fair approach to me.

[jonathanstrange]

I'm skeptical about this business model because it can become worse than proprietary software dependency over time. You get lured in by thinking it's open source and free, and might end up paying as much as the developer wants for essential features a few years later.

[imiric]

Like I said, it's crucial for the developer to be fair about it. Running a sustainable business built on open source is entrepreneurship on hard mode. Some companies do a better job at this than others, and, unfortunately, open source has been abused as a marketing strategy on many occasions.

No matter how you look at it, though, any business model that enables users to use open source software is a much better option for users than any proprietary software. There's no comparison. Given the choice, I'd much rather use OSS that is eventually rugpulled or enshittified than proprietary software, which carries those same risks, while also restricting my freedoms from the get-go, and having additional risks I might not be aware of at all (exploiting my data, security issues, etc.).

[nchmy]

They're not even running a business on this - its literally registered as a 501c3 nonprofit. Its a labour of love that theyve shared with the world, and they actively tell people that the pro features are unnecessary for most projects. If you want those features or just want to support the project, the cost is quite fair, and just goes to things like traveling to do conference talks etc... They are not getting rich off of this.

Also, they intend for v1, which will be released soon-ish to be essentially the final version of Datastar. There wont be a need for much further development. So there's minimal risk of "rugpull" or even abandonment.

[imiric]

That's admirable. They just gained another sponsor. :)

I do think that they should be paid for this work, and be able to sustain themselves from it. So I wouldn't be against it being a business, or the project having a subscription model. The idea of open source being gratis, and products in general being "free", has done enough harm to the world.

[andersmurphy]

100% so many projects have died because they didn't have a longterm plan to keep the lights on other than hoping for donations and or offering support that no one needed (the better the project the less likely you need support).

[limagnolia]

If they are charging for features, those features are a business, even if its a nonprofit business. There is also precedent for the IRS to consider it unrelated business income and to make them pay taxes on it, but that seems to be a huge grey area in US tax law.

[withinboredom]

LOL. What US tax law?

[limagnolia]

See for eaxmple the Mozilla IRS dispute as an example. Mozilla resolved this by putting their money making endeavors into a for profit org, which is in turn owned by the nonprofit, but the for profit org pays tax on their income. https://www.irs.gov/charities-non-profits/unrelated-business... See also

[andersmurphy]

Yeah it's a weird one people complain about open source projects having extra features you can pay for. But, then they also complain when you make open source GPL.

I'm sure it would be the same group complaining if it was GPL too.

[nchmy]

Except the pro option is not "a few years later" it's available in pre-v1 and is a one-time lifetime purchase, not a subscription.

And the framework and docs are so small and simple, that you can read the entire site in an hour or two, in which time you would have noticed the pro features and pricing many times.

[jonathanstrange]

At the risk of repeating myself, I was talking in general terms about the business model and the uncertainty about the licensing it creates concerning possible future features.

I wasn't aware that they are a non-profit organization and agree that my remark doesn't necessarily apply to them since they're not a business. If the pro subscription helps them to maintain the open source project without making any profits, then that seems alright to me.

[nchmy]

This is exactly what is done here. They're emphatic that most will/should never need the pro features.

[simonw]

Looking at https://data-star.dev/reference/datastar_pro#pro-features the features that stood out for me as things I would want to to use (and that don't feel particularly "pro" to me) are the animation features, data-replace-url, data-on-resize, data-scroll-into-view, the @clipboard() mechanism and, most importantly, the debugging tool.

My problem with this business model is that it has a chilling effect on open source contributions to the project, because it incentivizes the core maintainers to not accept community contributions that overlap with the paid features.

[nchmy]

Datastar is currently in the Release Candidate phase of v1, which they seem to be certain will be the final form of Datastar. So, they're really not looking for "community contributions" to the core of datastar, other than for feedback to polish it off.

Moreover, ALL of the features are "plugins" - there's nothing stopping anyone from building and sharing their own. It is actually encouraged. In fact, in their next release they'll be releasing and documenting some sort of public API for plugins. Not sure exactly how it will differ from the current form though, since you can already make your own plugins.

They also intend to have some sort of free, open-source "marketplace" for community-built web components, based on their upcoming Rocket web component and Stellar css framework. You would just need to have the Datastar pro license to be able to use any of them.

I can see how this final part might be criticized, but it does seem quite fair to me. Sustainable open source is a big problem - in recent months ive had some important fully-open source dependencies disappear - even ones that were backed by well-funded companies. Moreover, Datastar is registered as a 501c3, and they really don't intend to "make money" from all of this. Its just to pay the bills, travel to conferences etc...

They're very reasonable people and open to all discussion in their discord. Im sure that whatever concerns you still have could be explained quickly if you went there.

In fact, i just posted there about this thread and how its overflowing with nonsensical complaints about how there's no link to or mention of Pro on the homepage - they said sure, we'll add a link.

[withinboredom]

Naming things really is hard. I can’t even with these names.

[nchmy]

what's wrong with the names? they seem perfectly fine to me...

[nchmy]

Also, btw, some of these - specifically data-replace-url - are strongly discouraged by the devs, to avoid footguns. I dont have experience with it, but people have spoken lots in their discord about how things like Hotwire Turbo are too magical and end up causing a lot of issues. Datastar strong recommends just using hypermedia as it was designed (HATEOAS https://htmx.org/essays/hateoas/) and doing full page reloads when you change a page. And for more ephemeral in-page changes, you can use the sse fragments morphing etc