The Active Maintainer's License a new license I'm developing

Hi there,

I’ve been dissatisfied with the current trend in software business models towards open core SAAS and I’m trying to figure out how local first software could be funded while still being developed openly and collaboratively. I’ve come up with a new licensing model that I’ve tried to embody in a new license titled the Active Maintainer’s License. I’ve tried to explain how this license works in a 30 minute presentation which you can find here:

Or on youtube here:

You can read the license text here:

If you have time to watch my presentation, I would be very grateful for your feedback as I plan to use this license for my work once the rough edges have been sanded over.

1 Like

I didn’t finish watching the video but I did try to skim it from the captions.

There are a number of real-world issues with this.

  • Once you get past small deployments of computers, software is not installed on computers one by one. They are deployed wholesale. It is also virtualized so it isn’t even installed on the computer but it can pretend as if it was (to manage licenses).
  • Not everyone’s time is equal, and yeah I know that you acknowledged this, but this isn’t about money. Some people are faster and more capable of working with a given codebase over others. So four hours for some people might be a pullable change, whereas, for another, they barely scratched the surface.
  • There may not be four hours worth of work to be done. Or that you as a company have somebody capable of working on that project.
  • Not all companies are going to allow you to install software that will take a video of your desktop. That is a potential security risk. Not to mention those files are going to be big and you will have to store them forever or risk losing the license.
  1. “Once you get past small deployments of computers, software is not installed on computers one by one. They are deployed wholesale. It is also virtualized so it isn’t even installed on the computer but it can pretend as if it was (to manage licenses).”

Why is this a problem? Maintainers can use any licensing terms they want when granting the right to use.

  1. “Not everyone’s time is equal”

The point is not just to ensure that you valuable work gets passed back, but also to ensure that your competitors costs are non-zero :wink:

  1. “There may not be four hours worth of work to be done.”

I presume you’ve studied at UNI. If you haven’t I’ll give you a valuable lesson and ask you to find 4 hours of work to do on the cat utility. This could be writing documentation or finding new ways to test the performance, or even finding ways to reduce the number of times memory is passed between the kernel and user-space. If you cannot find 4 hours of work, I’ll fail you.

  1. “Or that you as a company have somebody capable of working on that project.”

You can always pay an existing maintainer or baring one exists, pay someone to become a maintainer. If they don’t know how the software works, they’ll learn on the job, nothing in the license prevents spending that time reading source code.

  1. “Not all companies are going to allow you to install software that will take a video of your desktop.”

The license doesn’t force you to do this. You can always pay someone else to maintain the software. You only need to get a license from someone who can, you don’t have to do it yourself.

  1. "That is a potential security risk. "

I use a VM for development. I already do screencasts.

  1. “Not to mention those files are going to be big and you will have to store them forever or risk losing the license.”

I did the math. It’s manageable: https://github.com/gradesta/gradesta/blob/default/SOLVExTHESExPROBLEMS/PT3_2022_01_live_coding_stream_hosting.md

Comes out to about $0.80 per hour when stored for 10 years, probably less (when I did that math I was working with a much higher file size estimation than is possible with better encoding.)

I like the broad concept. The problem identification part is good… the freeloading market failure of FLOSS is the #1 issue. I ALSO agree that dual licensing simply divides the problem in the wrong way and from a systems archetype POV I reckon is some kind of ‘fix that fails’, it seems to fix the problem but just stores up problems further down the road.

That’s the good part. I don’t like the low tech implementation details though, screencasts. Just in passing, a labour theory of value is a nuclear option, politically. Software industry is 100% wedded to market fundamentalism and right libertarian economic ideology, not theories of value based on labour time… which is classically Marxist and classically what I think is the right way to go though.

I also like the non-copyleft stance, again for political reasons and to encourage small scale experiments and so forth.

This is why I like your project, because it is at least looking in the right place for the solution, not in some cryptocurrency or decentralized autonomy or FLOSS evangelism or dual licensing sects.

I think it might be possible to prove you are an active maintainer and deal with the code sharing problem using AI… my punt would be to maybe use something like a zero knowledge proof plugin via the IDE whether it’s VSCode or one of the others?

I would strongly suggest you drop the ethical angle… I think the ethics of maintaining software dev. can be computed.

Hello Tammy,

thank you for your detailed feedback. I’ve tried to understand it and respond to your concerns.

Honestly, the idea that this could be Marxist didn’t occur to me. Rather than approaching the design from the standpoint of political theory, I was only trying to find an objective measure. That is all. Time was the best objective measure I could come up with. I was only able to come up with three potential objective measures:

  1. LOC
  2. Monetary investment
  3. Time

  1. I personally try not to produce a lot of LOC and believe that turning LOC into an objective creates distorted incentives that would be harmful to code quality and product health.

  2. Using monetary investment as the objective measure has the potential to disenfranchise independent developers who might for entertainer’s collectives and develop the software without a formal wage system. Also, it is easy to come up with accounting tricks in order fake or inflate monetary investment. Finally, it would be difficult to account for the angle that the result of that investment must be returned to the project under the AML. Companies could claim to invest millions yet only publish a few lines of code, how would you know if that was fair or not.

  3. Time obviously is problematic, but it’s hard to fake, easy to account for, and there is a clear binding between a concrete screen-cast and the results of that labor which makes the mandatory reciprocity very easy to measure and enforce.

I did not consider the possibility of using AI to preform the objective measuring, but now that you mention it, it could work. However, there are a few challenges:

  1. It is easy to produce an adversary for an AI using AI.
  2. In order to respond to 1 AI must be updated, how would you manage an ever changing objective measure? Licenses work best when they change slowly.
  3. I am working on my project now, I need a functioning license within 6 months and for limited investment. My company is not an experimental business models development firm, my company is a software company building software unrelated to the AML license. I cannot make the development of this license a priority. Given these circumstances, I must make the license as simple as possible.
  4. Simple licenses are easier to understand and therefore easier to put one’s trust in.

All that said, the license itself is released under the creative commons license. You are free to create your own version so long as you change the license suffix. So if you want to create the AML-AI license, you are more than welcome to do so. Who knows, if you do a great job, maybe I’ll even use it. As I said, I don’t have the resources or the time-frame to do that myself.


I’m not quite sure what you are referring to in terms of “the ethical angle”. I do not mention ethics in the license text at all. That said, ethics is important to me personally. I will not stop thinking about ethics and talking about it.

Once again, I’d like to thank you for your feedback. In the spirit of the AML-GR0 I can truly say that I value your time investment :wink:

Timothy

One bizarre problem with the AML-GR0 that just occurred to me is the following:

If I were maintaining a piece of software licensed under the AML-GR0 I would likely spend a lot of time watching screen-casts released by other maintainers, so that I could stay up to date on development. Would I be able to count that time, watching the other developer’s screen casts as maintenance time? Would this not lead to a potential feedback loop? Furthermore, that would be something that would be very easy to fake, as watching a video is naturally a passive activity.

I think you will want to check https://rovas.app. It is a generic tool - not software specific - but seem to solve some of the problems you identified in the video. It was developed to enforce (or aid in enforcement of) reciprocity in production of volunteer-made goods or services, but I think it is actually more universal then that. It also uses labor time (represented by a custom currency*) as a proof of labor. I think the “maintainership” requirement could be delegated to just being able to pay with the custom currency. The verification of reported work is also taken care of and even the meta-problem with reporting time doing verifications can be handled in the system. There is also a skills/reputation-like Merit reward that takes care of the differences in value of work between skilled and unskilled workers. There is much more. It is being actively used for reporting and rewarding work in a crowdsourced web portal and for Openstreetmap activities, but also documentation writing and software development are being reported as well as other activities.

  • not crypto!!
1 Like

The best political theory relies on objective measures… like the Labor Theory of Value (LTV).

I agree that labor time is the most reliable measure. My point was that the LTV is fiercely resisted[1] by bourgeois economists that are unashamedly pro-business. My objection is not the LTV, but your proposed method of proving it.

I think I would call the possibility of using AI to preform the objective measure ‘surveillance socialism’.

Using similar methods to automated DevOps functions used by corporates for scientific management to suit their own private wealth interests, very similar tech could be used to encourage collaboration among developers and users.

I think it is less likely that an adversary would be created, and more likely that some collaborators may use automation to ‘game’ the system to their own advantage, or ideological antagonists trying to shut it down.

None of this ought to discourage us from developing it though.

The license text does not have to contain the detail. It’s quite possible to achieve this through project governance, a bit like a ‘code of conduct’ where the implementation details are updated dynamically while the license text refers to the objective measure through an active link.

Your license could be simplified by removing the implementation details and simply providing a link to them.

The implementation could become a part of your product portfolio.

I read the license and my first impression was ‘it’s complicated’. I am suggesting you simplify it by actively linking to the implementation detail.

Licensing is possibly the most important detail of your working life, not something to rush IMO:

Focusing on the problems of justifying intellectual property is important not because these institutions lack any sort of justification, but because they are not so obviously or easily justified as many people think.

We must begin to think more openly and imaginatively about the alternative choices available to us for stimulating and rewarding intellectual labor.

E. C. Hettinger, ‘Justifying Intellectual Property’, Philosophy & Public Affairs, vol. 18, no. 1, pp. 31–52, 1989.

This would possibly require a small web app that handles the current protocol, which could at first be simply an agreement to post output from IDE extensions that monitor time spent. On it’s own, this would be a blunt instrument but over time, with blending data from LOC and possibly also financial contributions you would end up with a very smart measuring system.

In terms of “the ethical angle”, I am referring to your presentation where (from memory) you do implicate readers in ethical considerations. I had a similar problem when I wrote a license with economic restrictions. I started out using ethical rhetoric but that attracted very low effort rebuttals from FLOSS evangelists which have an eccentric allergy to ethical licenses.

I reviewed the way I presented the license to attract better calibre critique and so far, it has at least stopped the hate from FLOSS zealots… since I am using the same argument for economic restriction that FLOSS advocates use around vague and idealist notions of freedom and openness - namely that licenses ought to enable the outputs we value… quality, fairness, performance, security and so on so forth.

[1]: Adam Mossoff, ‘Saving Locke from Marx: The Labor Theory of Value in Intellectual Property Theory’. 2012.

Conceptually, good. This project looks as though it has stalled… fallen into non-maintained status? Also I reckon it tries to do too much by creating a new currency which looks excessive to this use case… AFAICT the dev. hasn’t released the full codebase, which is a bit of a problem. My guess this is possibly a headless drupal thing maybe… so will have scalability issues maybe if it takes off again. Weirdly, the dev also situates it within his academic proposition of ‘merit signalling’ rather than mainstream economic practice which I really don’t get either.

Sorry, @Tammy, I haven’t been notified about your reply - I noticed it by accident just now.
The project is not stalled really, but I am switching my focus from development to marketing this year (trying to get some investment to push it little faster :slight_smile: ).
Yes, the dev. hasn’t released the codebase and it is for several reasons. The main one is that indeed, the app is built on Drupal, with lots of custom php (modules and such) and I simply haven’t investigated how to create a project in github (say) yet. Yes, scalability is a concern, but I always thought about what I have as a POC.

Now about the fun details…

The merit signaling is the key to understanding what I am doing there. Yes, it is different, but in fact based on a simple idea - that most people in developing developed countries work to satisfy primarily their psychological, not physiological needs. That’s what I call merit seeking. The dominant form of merit seeking out there is status seeking, which is competitive (“outward-oriented”) and one that leads to zero sum game outcomes. Some economists (like Robert Frank) consider status seeking pathological and believe that there is an opportunity for Pareto improvement by suppressing status seeking. IMHO the medicine he prescribes (luxury goods taxes) is not really the best approach.

The desirable form of merit seeking is what volunteers, children, or the really good scientists do (also some entrepreneurs). It is inward oriented, non-competitive. It is about scratching an itch - solving a “personal” problem - and any thought about monetizing one’s product comes later if at all.

I claim that it is the merit seeking what actually motivates people to supply labor and innovate, not a promise to get exchange value reward.

Capitalism rewards people with a reward that is at the same time - a merit signal, as well as the means allowing the recipient to buy stuff. The problem with this arrangement is, that - as I write above - the really creative folks do not need the exchange value to be motivated to work. However, in capitalism they are getting it anyways and that leads to income (wealth) inequality.

NEO basically makes a contract with workers - “I will pay you ahead for any work you want to do (a flat, hourly wage regardless of what you make), but I own your product. You can still sell it, but instead of profit, you get a profit size amount of merit points”. In NEO therefore virtually everybody earns the same exchange value reward for any activity they want to perform (just like volunteers or scientists). The workers remain motivated to innovate, because they are getting unequal merit signal, specific to their field of expertise. Its amount is market determined, scarce and therefore as powerful as monetary reward in capitalism (we will see, but there is plenty of evidence it should work like this).

I am aware that the system is not easy to grasp, but people do not have to get it to use it. Most existing NEO participants work there because they were forced to do so. People come for Chrons (the system currency), but stay for Merits.

NP. About those fun details…

…most people in developing countries work to satisfy primarily their psychological, not physiological needs.

I would enjoy some empirical support for this claim since I suspect this not to be correct. Maslow’s hierarchy predicts economically disadvantaged populations to be primarily concerned with meeting their physiologial needs first, although that hierarchy has been influential it has become disreputable from closer scrutiny AFAIK. Either Maslow is correct, in which this statement looks like an error, or Maslow is wrong, in which case empirical support based on an alternative model would be required.

Quantititative data collection initiatives about developing nations is of course very difficult because the information infrastructure in those regions are, by definition ‘under-developed’ and thus non-representative of populations that are living in some of the more remote areas of the planet with logistic weaknesses.

Merit seeking has been the subject of scholarly attention in fields like psychology, economics, management and organizations but (AFAICT), pathologizing our motivations for work is based on the theory of methodological individualism, which tends to deny any macroeconomic theories. Macroeconomic theories are contested, but macroeconomics as a field of research has a fairly reputable status which the type of interpretation you prefer would have a hard time to unseat, even on a a good day?

Your claim that merit seeking is ‘what actually motivates people to supply labor and innovate, not a promise to get exchange value reward’ may be true in outlying cases, like an immiserated poet, artist or programmer who may work ‘for the love of it’ but the world works for the most significant part due to people wanting to improve their material conditions, not their affective ‘mood’.

Capitalism only rewards people with private property interests, everyone else gets paid less than the value of the work they produce. This is the bourgeois mind trick, it convinces us to go to work while camouflaging the value we produce from our cognitive/affective apparatus, so we are led to work longer hours for less pay and consume more as we go.

I would also be suspicious of the idea that there are ‘really creative folks’ and thus ‘really un-creative folks’ since creativity is socially constructed and has a large historical component steeped in the values of property owners from at least the middle-ages.

That is the problem I have with the NEO part… I think it is based on unreliable assumptions about the world as we know it, but I do like the concept of rewarding workers for the actual labour they are supplying rather than on ‘market rates’.

That seems like a fairly uncontroversial way or organizing a fair economy, my preference would be to ensure that the workers keep all of the value they create because that would provide adequate motivation for everyone to create a better world, rather than just profit for corporations.

…there is plenty of evidence it should work like this.

I don’t think the evidence you have will outweigh my claim that people work, primarily to improve their material conditions. The only people that do not are property owners who do not have to work and can afford to idly spend their privileges on a whim, such as luxury goods and services

I think I do understand your system, but I am not convinced it is supported by sufficient evidence.

Thanks.

Sorry, I meant to write “developed”, not developing. A big difference of course. The paper you cite has references, but I think my correction might address (some of) your objections.

I would also be suspicious of the idea that there are ‘really creative folks’ and thus ‘really un-creative folks’ since creativity is socially constructed and has a large historical component steeped in the values of property owners from at least the middle-ages.

I meant to suggest that ability of (intrinsically motivated) people to produce socially useful products follows the normal distribution - there are some who are better at it and some who are worse. Of course this is not a moral judgement, but a question of luck (I believe).

my preference would be to ensure that the workers keep all of the value they create because that would provide adequate motivation for everyone to create a better world, rather than just profit for corporations

I believe NEO achieves this goal, as everybody gets paid the same, which means the most disadvantageous get the most value (due to floating exchange rate between Chron and national currency). I think NEO can be considered a “Rawlsian economy.”

Your correction resolves my initial objections, yes.

One remaining query I have is how you prove a reliable link between ability and intrinsic motivation to produce ‘socially useful’ products in the normal distribution.

A person’s ability will be constrained by material conditions, e.g. if they are impaired by social structures (educational establishments, language, culture etc.). A visual impairment for example is only an impairment when visual cues are necessary to produce work. An opera singer, strictly cannot be visually impaired in singing an opera if they can reproduce the music without the need for vision, and so on and so forth. An opera singer is evaluated on their ability to evince enjoyment in the audience of listeners. However, if an opera could be enjoyed without sound, perhaps by visual representations, the opera singer would be unable to contribute since the production would require visual elements.

I have selected these vignettes to highlight the assumptions you are making about ‘ability’/intrinsic motivations reduction to the Bell Curve, and also the idea that some products are more useful than others without thinking about what makes something ‘more useful’ than something else, which again - is largely a historical question, not a scientific one.

Your solutions assume an individuals ability and motivations are linked (which again isn’t obvious since there are many counterexamples such as champion paralympians - an example of humans without the means to compete against some others due to missing limbs etc. yet train and work just as hard as other athletes, whereas there are numerous examples of young, prodigious students ending up dropping out of society and contributing very little utility for others). This sort of thing cannot be reduced to a single curve, ability and intrinsic motivations do not admit ranking or ordering in a linear way, and to think that is tantamount to saying our ability to contribute social goods is genetically based, which is observably wrong.

Success for us as a species is not about optimizing efficiency in the distribution and production of resources according to the needs of property owners, but in ensuring we avoid the excessive strains of mental and emotional commitment to that kind of ideology. The idea that some of us are ‘just better’ and more deserving than others based on a history dominated by the interests of the most powerful and wealthy in society, the idle, corrupt and undeserving aristocrats, media proprietors, industrialists and financiers just doesn’t fit my vision of a good economic system.

Peoples abilities and motivations are not immutable and so the mental model we need to make useful interventions must be nonlinear.

I am sure you do not wish to make a moral judgement from your intervention here, but the consequences for contributors to that system will create asymmetries in life that I think would be indistinguishable from the barbaric and crippling system you are setting out to critique and reform/revolutionize.

If I am right, then it’s hard not to draw a moral conclusion since a system built on similar assumptions being made by capitalists might not be the best candidate for a possible solution.

Not sure who said it, but there’s a saying that goes something like… ‘you can’t dismantle the masters house with the masters tools’, which to me means you can’t create a fair society based on tyrannical judgements about merit seeking and merit signalling.

It is not the aim of NEO to fix all that is wrong with the world. If it achieves the main goal - provide fair redistribution of exchange value the economy produces while keeping it at least as efficient as it is, it will achieve its goal. There still will be people born with various degrees of luck, but NEO is not going to be able to alleviate all of it, certainly not directly.

NEO does not want to be a celebration of merit in any form, it is really a response to what I see as a source of inefficiency - that exchange value and “merit signal” are fused in money as we know it and that leads to undesirable distribution of exchange value.

Can we bring in some external investors for open source projects?

As an act of bootstrapping a project I see it could be beneficial, but in the long term - do we need to? You yourself are pointing out how much value FOSS generates that the developers just leave on the table.

My idea is to create a new license to force cloud computing giants to give back part of their profits to the community, without restrictions on other companies or individuals using the software internally, and introduce a mechanism so that investors (including developers and non developers) can receive dividends from the profits of the project by purchasing securities

For example, Amazon sells redis based cloud hosting services, competing with similar cloud hosting services provided by redis labs, an open source project sponsor

According to the new license, only the company that has purchased the securities of redis project can sell the cloud hosting service developed based on redis, and the community will receive a dividend of some profits from the company according to the proportion of securities held by the company

Here is a bit of history that I haven’t found the willingness to post, but I think that it will interest you.

https://dri.freedesktop.org/wiki/DriHistory/

1 Like

@sano Zdravím z prahy! Budu na tebe reagovat, ale ještě jsem neměl čas.
(Translation: Hi from Prague, I’ll respond but I haven’t had time…)

@Roy The challenge I see with the shareholder system is that pricing is decided in a centralized fashion and thus access to the software is not guaranteed. How would you solve pricing/decide how much to charge. Pricing by profit is not possible because profit is just an arbitrary number that can be changed depending on the desires of the accountant. Revenue can’t be changed, but it is problematic for low margin industries. I see that you are wanting to limit your money charging to “cloud computing giants”. That might be a perfect model for a database or something that runs mostly in the cloud. It also doesn’t suffer from the problems with using a percentage of revenue since cloud is high margin. However it wouldn’t work for my usecase as I’m creating a mobile app.