A different perspective on dependencies and core critical software
We live in a low-trust society. A society in which reliability is scarce due to economic and political reasons. Reliability should not be a commodity, any commoditization of reliability makes the commodity unreliable and susceptible to economic and political forces.
We do not live in a vacuum. Economic and political system encompasses and deters any decommoditization efforts, including reliability. This is why more and more of FOSS is transitioning to partial commoditization, offering “advanced features”, an increasingly larger set, for an increasingly larger price.
In such an environment depending on “FOSS” or other commoditized projects is a bad idea. What if they “introduce a subscription model”? What if they cut free access for some feature you rely on? Maybe they do it to you just for political reasons because they’ve read your blog and determined your opinions do not align with “company values”. What if they buy a project in entirety and cease its FOSS activity?
Many of these projects, even if not yet commoditized, further depend on commoditized “FOSS” potentiating the unreliability in a chain of single-point failure.
Projects whose entire chain of dependencies is not yet commoditized are what I call POSS - pure open source software. However these is an increasingly vulnerable and smaller set of these because if at any point any dependency in the chain becomes commoditized, all downstream software becomes “FOSS” or worse.
The greatest software project, The GNU Project known for its reliability and strong ecosystem, featuring free tools, free C/C++ compiler, free OS kernel (Linux) is an example of POSS maintained with the philosophy (a decentralised governance) that the sanctity of POSS must be maintained. That will not work long-term because it ignores economic and political reality.
The first problem to solve is clearly the dependency chain. The dependency chain has a property called direction. The chain that “FOSS” and POSS models employ is a mixture horizontal and vertical chains. The horizontal chain is the single-point of failure chain, it exists largely outside your control and is particularly vulnerable to commoditization. The vertical chain is your friend, this is software immune to commoditization because you own it.
Let’s examine what is roughly happening to chains in different modes of software:
POSS tends to minimize horizontal chains.
“FOSS” tends to minimize vertical chains.
Closed-sourced software tends to maximize horizontal chains.
We want to introduce the 4th way:
The 4th way along with not having the vulnerabilities of the 3 old ways with its sovereignty, allows for high control and customization of such software infusing the much needed richness and innovation into the space.
The 4th way also forces software into decommoditization, there will simply be no reason for people to pay for “FOSS” when they can get the richly-featured 4th way for free.
The second problem to solve is, we do not live in a vacuum. The 4th way requires time, time is money today, money you don’t have. Remember everything is in the context of the existing economic and political system. The 4th way will face costs because it will bootstrap itself in a highly commoditized space, these costs will not be able to be upheld by an individual.
The 4th way community will need to eventually step in through contributing to such an extent that the use-value of 4th way projects becomes indispensable. This is achieved with the customizable part in self-dependency customizable software.
Here are proposed ways to overcome economic difficulties when operating under the current economic system:
The 4th way must introduce a discrete, capped and democratic bounty system, in which individuals (or small non-profit organizations) who use the products of 4th way software, should they want a specific feature implemented but haven’t the ability to do so, can place a monetary bounty on the feature to be developed by some contributor.
The 4th way must seek funding from friendly governments who recognize the need to foster locally-grown 4th way software that protects government’s sovereignty as a use-case.
The 4th way must embrace Tayloristic principles when distributing money:
Contributors are divided into Tiers (best to worst): A, B, C, D, E, F.
Every contributor must strive to become A Tier.
Every contributor must only work on PRs of his Tier. Among many equally-tiered contributors, the most suitable is selected by Manager.
Difference in monetary compensation between A Tier and others is at least 30% and at most 100%.
Every contributor starts at F tier and works his way up by successfully merging PRs (SMPR).
Contributor tier promotion table:
F —> E after 3 SMPR.
E —> D after 7 SMPR.
D —> C after 18 SMPR with Manager approval.
C —> B after 47 SMPR with democratic peer approval.
B —> A after 123 SMPR with democratic peer approval.
D to A tier receive equal base compensation if between 2 consecutive SMPRs there is no more than 6 months time.
Manager of a project is highest Tier contributor assigned to that project. If there are many with same Tier, Supervisor chooses manager or Supervisor himself manages the project.
Manager has B tier compensation.
Supervisor has A tier compensation.
Manager can manage no more than 3 projects.
There will certainly be a backlash against the 4th way by the benefactors of “FOSS” and Closed-source software. There might even be a backlash from POSS benefactors. The 4th way community must stand ground.