Experimental ideas on financial sustainability of open source


#1

Hey everyone,

Finally I’m finding some quality time to open this thread.

I would like to discuss some ideas, “a thought process” if you will, on “financial sustainability” side of open source.

As I mentioned earlier, my main interest on open source is about re-usability of the knowledge and the freedom around it. And from a personal point, see me as a developer “who wants to get in”, “who wants to work for open source full-time”. So, I would like to focus on more “the outsiders/the potential that we miss” part.

Current state of the open source tells us that if you want to “share your knowledge”, you should more or less have to give up your income. And one could decide to be an idealist and still make the switch, but clearly this wouldn’t be sustainable in the long run (personally I made a similar attempt couple of years ago, and now I’m back to being a freelancer).

In other words, “idealism is not sustainable”, and open source shouldn’t be based on that. If we want to maximize the potential of open source, we should be able to create an environment that open source is not a temporarily, volunteer act but it should be a regular, long-term paid job for any developer, regardless of their motivations.

You can take these ideas as my starting point.

In the following parts, I will write some statements, hopefully as short as possible, starting with these points. In return, you could reply by saying “Agree/Disagree/Unclear” for each statement. Of course you can write a longer response too.

I have to say that I’m trying something like this for the first time, and it’s very well possible that it won’t lead to anything exciting, practical, or it might be completely off the track. But let’s try and see how it goes.


Part I: Intro & Definitions

  1. Open source license is a legal contract about “sharing the knowledge”; “here is how we build this software and we allow everyone to re-use our solution”. The other alternative that we have is the proprietary license; it doesn’t giveaway both the knowledge and the rights. Benefits of open source will be discussed further but between the two, we can safely assume that open source is not the lesser option.

  2. Every individual/organization has a cost to build their software and they aim to deliver a certain value with their solution. Choosing which license to use (sharing the knowledge or not) is a separate decision than rest of the parameters. It won’t affect the cost of the product, and it cannot affect the value that it delivers.

  3. This means, that if a commercial software company wants/decides to “share their knowledge” by switching from proprietary license to open source, they cannot/should not get less income for their products/services.
    The solution they created is the same, therefore the value and the income must be similar as well. If the company is going to get less income, the consequence of this would be, that the company simple cannot make the switch (and cannot share their knowledge), else they would likely to get bankrupt in the long run and be out of the game.

  4. Similarly, if a software developer starts working on a open source project as a full-time job, she cannot/should not get less income/salary than the market average.
    Again, the work is the same, so income must be similar as well. Otherwise the developer cannot get into open source in the first place, or this would be a temporarily position. The developer would move on, when they get bored/burned out/run out of idealism/find a better paying job.

  5. In other words, if using open source license affects the income of the companies/developers, and this prevents them to switch to open source, this means that we have a “financial sustainability” problem.
    Or, from the other way around, the bar for “financial sustainability” of open source should/must be similar to market average in the software industry. Only if open source gets to that level of income, this issue could be considered solved. And naturally, reaching this level enables any regular software company/developer to get into open source (it would become a two way street).

  6. We need better studies than this one, but according to Gratipay, Open Source Captures 0.02% of the Value it Creates.
    a. So, we spend money on software, but it doesn’t flow into open source.
    b. If a regular software company would like to get into open source, they would be losing majority of their income.
    c. Open source side, with its limited resources, cannot hire/keep best minds and talents for a long period of time.

  7. Therefore, we can conclude that there is definitely a “financial sustainability” issue in open source.

I’m planning to continue with the following:

Part II : Problem
Part III : Benefits
Part IV : Causes
Part V : Solution

Thank you for reading. And again, it would be great to have some feedback :vulcan_salute:


#2

I’m happy to discuss. Below are some thoughts I had in response to your thoughts.

FOSDEM Community Dev Room has two sessions on the schedule that speak to this point:


#3

There are many nuances in between and within the license groups. Open source has major groupings with copy-left and permissive. And then there is the source available licenses. … I know you only list a short bullet point but I want to make sure you consider the variety of licenses and not oversimplify this section.


#4

This is a bold statement that I am not convinced holds true.

  • Open source done right incurs additional cost for building a community. Documenting properly and establishing community processes is additional overhead to some proprietary development practices.
  • The value that a software has depends on the type of software and how it relates to the value proposition of a company. If a software is the product that is sold, then releasing it open source impacts the number of potential buyers (e.g. Microsoft Office). If a software is a complement to a service, then releasing it as open source can drive adoption and market share (e.g. Android).
  • If a software is an industry collaboration for infrastructure, then the cost and value are again very different from consumer-facing software (e.g. Kubernetes vs. LibreOffice).

I see cost/value are not universally the same for all software (you acknowledge that in the next bullet point) and license choice is as much a business decision as it is an ethical decision.


#5

I’m happy to discuss. Below are some thoughts I had in response to your thoughts.

Thank you @GeorgLink, you’re being really helpful!

FOSDEM Community Dev Room has two sessions on the schedule that speak to this point:

I was thinking about going to Fosdem, it’s settled then :slight_smile:
And how about you, are you going to join?

There are many nuances in between and within the license groups.

That’s certainly true. I only want to compare open source in general with proprietary software.

This is a bold statement that I am not convinced holds true.

About this part, I can ask the question like this:
If there are two versions (open source and proprietary) of the same product (same cost, same value), what would their incomes be?

It might be easier to understand if we focus on “consumer-facing” software. With your “Microsoft Office” example, if Microsoft would decide to make their product open source:

  1. Is the cost of building Microsoft Office would be different? No, cost is the same
  2. Does Microsoft Office still solve it’s users problem? Yes, solution is the same, still delivering the same value
  3. Should Microsoft’s income be lesser if they make their product open source? My answer is no. Again, value is the same, so why being open source should affect their income in a negative way?
  4. But would Microsoft’s income be lesser if they switch to open source? It seems the answer is yes, they wouldn’t get same amount of income.

I don’t want to compare apples to oranges, but do you think that this statement would be true?
Even if Libre Office and Microsoft Office would have same amount of users, Libre Office would never make the same amount of money as Microsoft Office?

Basically I just want to point out the income imbalance between open source vs. proprietary solutions, and the consequences of this situation.

And from here, I want to ask questions like, what is the value of a product, how do we evaluate it, and how do we distribute our resources?

Hope this makes it bit more clear. I will continue with the second part.


#6

Part II: Problem

A basic definition of the problem could be as follows:

Commercial / proprietary license approach

  • Limits the rights and the accessibility to knowledge (you can only use the product, nothing more)
  • Has a price tag for the product and extracts the value from the market
  • And if it is successful, it captures a high profit income

Open source license approach

  • Maximizes the rights and the accessibility to knowledge (you can use/modify/distribute the product)
  • As a part of the accessibility, there is no price for the product and expects the value from the market
  • Even if it is successful, it captures a very limited income

                                         Proprietary    Open Source    Hybrid?
  Has a cost                             Yes            Yes
  Delivers a value                       Yes            Yes
  Knowledge accessibility                No/limited     Maximum			
  Income                                 ~99%           ~%1

Two statements:

  1. In proprietary case, if your product becomes successful, without requiring an extra effort, you start accessing the money, which will give you the resources to hire best minds and talents, while open source fails to get the same effect.

  2. Open source achieves “maximum knowledge accessibility” by removing the price tag (because the product is free, everyone can use it). By doing so, it reverses the way we do transactions. In a traditional way, first we agree on the price (payment), then we can access to the product. Here, first we start using the product, and then, ideally at some point, we should be paying back.

If I’m allowed to make another bold statement: Probably there is no example in economic history of such a case, that people are keep creating something highly valuable without expecting anything in return, at an industry scale. And this puts open source in a very unique position.

Can we create a hybrid system, that allows us to “maximize the accessibility to knowledge” and still captures a value that is equivalent to proprietary case?

I hope I’m managing to make the text clear/understandable enough :crossed_fingers:

Next, what are the benefits of solving this issue?


#7

No, not this year. If you are FOSDEM, maybe you would also like to join CHAOSScon the day before.

Is this assuming that they only dump the source code somewhere without trying to build a sustainable and healthy community?

I guess, they might loose revenue from users who compile Office themselves or get a binary from a third party that compiled the software.

Depends. How are revenue and cost effected? Can Microsoft reduce cost for development because tasks (e.g., QA) are now done by people from community?

Depends. Can LibreOffice build a brand that people are willing to pay for? Are there complementary products (e.g., office subscription with cloud storage)? What is the level of marketing for each? … The price a company/community can charge for software depends on more than license.


#8

If you are FOSDEM, maybe you would also like to join CHAOSScon the day before.

It would be great actually, but I can be there only on the weekend.

Is this assuming that they only dump the source code somewhere without trying to build a sustainable and healthy community?

I would assume this; the codebase is on a private cloud repo (maybe it’s already on Github?). It has all the unit tests, all changes are done through pull requests and continuous integration / delivery flows are implemented. So, technically it doesn’t matter where the PR is coming from.

All issues are kept on the issue tracker, and anyone with a user account can create an issue. Everyone can see the current and planned progress. There is a fair documentation about installation, coding standards, deployment processes etc. As a company standard, there is “Code of Conduct”, “Diversity & Inclusion” policies and “Succession Plans”.

So far, this would be “software & company management done right”, and I think it should cover 80~90% of “open source software done right”.

For the sake of our comparison, let’s say there is no extra effort for “community management” (no mentoring, no task assignment, documentation is not extensive as it should be - and please add if I’m missing something major), making it open source would still be highly valuable.

You can watch the progress, you should be able send issues (since every project should appreciate feedback and bug reports), and if you can find your way by yourself, PRs should be open, at least for minor changes).

We are free to examine & learn from the code, fork & modify (no vendor lock-in - I believe the most important benefit), code auditing could make it more secure, or at least we can be sure about the security (feel free to add if I’m missing anything).

What do you think, is this a fair picture? And do you agree that this valuable enough?

Allow me to stop here, since I don’t want to write a longer reply if I’m misunderstanding you :peace_symbol:


#9

I don’t know. I have never worked in a software company.

We talked about the company (i.e., Microsoft) perspective. The value described here is from a community/user perspective. – I agree with the value, but we have to be careful who we argue for and who we want to convince.

It is not only about the technical side of open sourcing. I heard that companies who do open source go through a rigorous process of license compliance, checking for intellectual property, and otherwise making sure they are not infringing on anything when they release source code that was previously internal intellectual property.

There was a discussion on open sourcing on Changelog episode #310 for a “small-ish” project. I imagine it is much more complicated for larger, but I have no experience with this.


#10

Given the fact that open source has sustained itself for a good number of decades up until now, I would reframe this as “something appears to have changed in this industry that makes us worry that, in the future, people working on open source won’t have a job” and maybe list out what those things are. In other words, explain why it wasn’t a problem before and why it is a problem now. We all know plenty of open source idealists that have other motivations to contribute, and plenty of non-idealists that are involved in open source.

This is not inaccurate but seeing where you’re going with the argument, I’m not sure people contribute just because the license is about sharing the knowledge. What the license also does is ensuring exposure across markets and industries for your code or contributions and there’s intrinsic value in that. Maybe it’s even the motivation for some contributors, beyond idealism. A questionable part of “source available” is whether those also share the knowledge but limit the exposure.

How does this argument apply to companies with large portfolios and multiple business models? Can a company switch to open source and have a loss leader that favor something else? On the other hand, can a company have a business model that is less sensitive to the license and then unlocks them from going open source?

Is there data showing this is still the case? It seems to me that the hottest and highest paying industry trends today correlate well with open source (there are entire categories such as programming languages where I don’t think there is anything other than open source at all)

Is this the main reason blocking companies from switching? I’m under the impression that, out of the % of the world where licensing is top of mind and stuff like software patents are enforceable, the main blockers are regulatory and security considerations. I’m all-in for showing the business value of developing in the open and committing to open source licenses, BTW. It’s a necessary fiduciary task.

This one I get. However, reading your Part 2, it appears you blame this on the license. Revenue isn’t predicated primarily by the license, but by the business model. This can be seen from the VC money flows into open source companies. Of course, that money doesn’t get to, say, the 31M+ open source developers on GitHub. One argument I’ve seen, though, is that those developers can contribute because they’re still part of an industry which is trillions of dollars richer because of those money flows to open source companies.


#11

Thanks for the replies. Meanwhile I am about to change my job and move to a new house. It’s difficult to focus here at the moment but I’m hoping to find some time this weekend :+1: