I wonder if software really deserves its own economics.
If you haven't read Hal Varian's Information Rules, I highly recommend it. Check the publication date, then read it anyway, then reflect on the publication date when you're done. I found it very worthwhile.
Yes, this is a great read. After that many years it still influences me. However it is not that kind of book you read before going to bed. It requires intense studies to take something out of it.
Such a business model exists and it's extremely well proven, and it powers the majority of major open source software: build a proprietary product or service, and open source any component that is more of a cost than it is a unique selling point of your system.
Do you need a faster compiler, or a better OS, or some cluster operator just to get your widget factory working? Don't build those in house, instead find others with the same problems and create an open source project together to work on them.
But don't try to sell open source software. It's essentially impossible to do that, it has been tried time and time again and success is rare, and huge success is basically unheard of (RedHat being probably the one single exception).
> Such a business model exists and it's extremely well proven, and it powers the majority of major open source software: build a proprietary product or service, and open source any component that is more of a cost than it is a unique selling point of your system.
Sure, it is the commoditize your complement strategy [0]. But that doesn't help get complex open source products to market, it only helps with tooling.
Maybe you are right and there's no way to directly pair the freedoms of OSS with the capitalism of VC backed startup.
Maybe we instead need a model where FOSS is not about profits for anybody, and is just a passion of love, from a large community of amateurs doing it for the technology and fun.
Projects could still be funded by community users, but "venture funding"? That's how projects turn to shit.
I agree, but what I think is curious about the whole situation is that you can also see it strictly as a market failure.
It's a very pure example where parties in competition each that have a use for some kind of software can shortsightedly develop their own versions of it in-house, but that duplicates a lot of effort. They'd be better off getting together with their competitors and collaborating on a shared version that suits their needs, avoiding duplicating effort and all benefiting from each others' contributions. They could do this by direct collaboration or by funding an independent organization that fulfills their needs.
Sure, this can go badly if there's a large difference in scale between the different parties and some can muscle others around. But it and similar models do work out at the scale of the Linux Foundation, Khronos, down to Mastodon, GitLab, Blender, Krita, Forgejo, even arguanly projects like Bitcoin Core.
There isn't the structures to facilitate this kind of regime shift. But there should be.
In such a world most of the open source software you’re used to wouldn’t exist (or would be much less complete) and you’d be forced to work with and use proprietary systems most of the time.
>In such a world most of the open source software you’re used to wouldn’t exist
As part of that world, I also want livable wages and work-life balance for developers, so they can work on their passion FOSS off-call. And for students and programming enthusiasts to be more passionate about FOSS. Like in the 90s before the corporates took over FOSS.
If some FOSS still wouldn't exist then, I'm fine with that.
Developers need a salary to pay the bills. Let's say that covers the first 40 hours of the week.
Those who are searching for significance outside their day job offer free labor as their "hobby". Maybe 10 hours a week?
For projects that want to move forward with some velocity it makes sense to make some of that development into paid day-jobs.
As projects get very large, there's a fair amount of overhead in just "keeping up". That erodes the 10 hours quickly. Further reducing the time to contribute.
So where is all this cash to pay employees coming from? Certainly not end users (as anyone who's tried funding an OSS project from users knows.) No, it comes from commercial companies (MS, Amazon et al) or venture capital.
This is the cognitive dissonance that underpins OSS development. The very people OSS treat as the "enemy" are the people funding OSS in the first place. As much as say RMS rails against big tech, Linux and the rich Linux economy system only exist at the level they do -because- of big tech.
Of course, I painting with a broad brush, and there are exceptions, but the point remains. It's turtles all the way down, and those turtles are not funded by users.
Those turtles didn't need to use funding pre-doc-com-boom, they were passion projects and people with time devoted to the "cause" of FOSS.
>This is the cognitive dissonance that underpins OSS development. The very people OSS treat as the "enemy" are the people funding OSS in the first place. As much as say RMS rails against big tech, Linux and the rich Linux economy system only exist at the level they do -because- of big tech
Perhaps that's the problem: that they exist "at the level they do", meaning most of it is corporate focused, and not enthusiast and user focused.
Even ourselves, as devs, evaluate FOSS as to whether it's "useful" for our corporate/startup needs. This wasn't exactly the case, or at least not the main case for a FOSS project.
Gnome, for example, wasn't created to give RH and co a desktop shell for corporate installs...
The ‘cause’ of oss? I doubt many people ever were dedicated to a cause outside of GNU diehards. For most other people it was about curiosity or fun, a hobby etc.
>> Those turtles didn't need to use funding pre-doc-com-boom, they were passion projects and people with time devoted to the "cause" of FOSS.
Except they kinda did. The foundations of FSF are born by academics working at institutions, getting paid salaries. The were devoting time certainly, and certainly in the case of RMS with passion and cause, but that work was definitely funded - usually by the university.
>> Perhaps that's the problem: that they exist "at the level they do", meaning most of it is corporate focused, and not enthusiast and user focused.
I think we can drop the term "enthusiast". It implies tiny niche group with little practical value. I'm thinking of classic car "enthusiasts" who spend all their time under the car, and precious little driving it.
So let's talk about users. Users want full-featured reliable software. I would suggest all software, if successful, is user focused. (To he honest, I'm not sure what you have in mind with "corporate focused".) Firefox, to pick one project at random will seemingly live or die based on the individual user experience.
Equally take databases - there are s plethora of options to suit every use case. Need big powerful fast enterprise scale - Postgres is for you. Need small footprint with easy install - try Firebird. And a gazillion others. Surely such quality is a good thing?
>> Gnome, for example, wasn't created to give RH and co a desktop shell for corporate installs...
Um. Sure it was. It was designed to offer a gui desktop on top of Linux. Who did they think would use it if not Linux distributions? Given that for decades "the year of Linux on the desktop" was a meme, I'm not sure it's fair to claim that distributions using Gnome to create desktops for business users was a surprise.
The historically 'good' open source companies like Sun got bought but the ones that weren't like Oracle. The selling support model alone does not seem evolutionarily fit for the market.
Now we have these VC-backed 'open source' companies that have a playbook wherein they appear open source at first. But when you dig deeper, you find that the heart of the thing is a closed binary.
The investors are going to want to be paid back somehow. And the business model of VC means that one of two things happens:
1. The company finds a way to 100x the return. Which, if you're a customer, might be a scary prospect.
2. The company makes an amount somewhat lower and, while it would be a good business for a non-VC company, they're considered a zombie by their investors. So, they are killed leaving you as a customer in a bad position.
I therefore trust non-VC backed companies substantially more to keep alignment with their customers long-term.
A workable model could be for instead companies that have legally-enforceable promise not to enshitify their closed sourced product. So that the product will always be aligned with the paying customer. The customer cannot be made the product at a future date.
Sun was mainly a hardware business; you bought their workstations and servers. And oh, they also had this unix-y thing that came with that. Later software did become a bit more important with Java and MySQL and all of that, but it was still primarily hardware company.
I think it's pointless to even compare it to the Redis company; just about everything is different.
You go homeless so Bezos can make his yacht a foot longer.
I find it amazing how much money is being spent to ensure open source code doesn't end up in the hands of users and how many people are blaming the ones trying to increase user freedom.
Their definition of "fair, reasonable, and nondiscriminatory terms" seems... incredibly vague, and with a big chicken-and-egg problem for the first license.
> If the licensor advertises license terms and a pricing structure for generally available commercial licenses, the licensor proposes license terms and a price as advertised, and a customer not affiliated with the licensor has bought a commercial license for the software on substantially equivalent terms in the past year, the proposal is fair, reasonable, and nondiscriminatory.
My thesis is that when you don't have the pressure of VC funding (gotta hit the revenue numbers you promise to investors sooner or later), alignment between the business and the OSS community isn't as tough to find.
I'd agree with that. Your message sounded to me like you thought VC funding was desirable for software projects. I wonder why we can't just fund software like a regular business- why look for venture returns?
That works great! I think the best money to get to run a business comes from customers. Bootstrapping is great.
However, just like fewer homes would be owned if you didn't have mortgages, less software companies would exist without VC. It's basically a subsidy from the rich, endowments and pensions, to the rest of us (consumers because we get stuff for free, developers because it increases the demand and thus salaries for us).
I think VC is a net benefit to the world in terms of software delivered and companies built. I think OSS is a net benefit to the world because of the explosion of possible ideas and the leverage it lets developers have as they build on it.
I would love to see these two huge innovations in building software work together well. Haven't seen it yet, hence my original comment.
This is very good use case of micro-transactions. If AWS makes $100 off Redis, they should be pay back X% to Redis project, from which the money is distributed to contributors based on how important their contributions were. Also Redis project is also supposed to pay back to the software components and 3rd party libraries it uses, so C project gets a fair share of the pie contributed back to them as well.
A grand unified theory of software goods funding, if you will.