Value networks for software
Bob Haugen
Public
Seen by 99
This is to separate the discussion of reciprocity licenses from the general idea and development of reciprocal economic commitments among networks of software developers. In other words, even if you are not interested in the license tactic, you might be interested in the strategic direction.
I'm using value networks as defined by Tiberius Brastaviceanu of Sensorica as an example of reciprocal economic relationships in networks. @ahdinosaur wrote about the software version
here.
The short version is that your project would be connected in reciprocal relationships with all of your open-source dependencies, and all of the projects that depend on yours would be connected to you.
So if anybody in the network got any income from their project, they would distribute some of it proportionally to all of their dependencies. And some of that income would flow to all of the contributors of each project that got some of the income.
Now, many questions in my mind about this idea, and probably more will arise in yours.
Some of the questions might include:
- Does this really mean all dependencies, or would it only include those who commit to reciprocal relationships?
- How to value software contributions? Or, will this lead to useless pull requests begging for credits?
- In other words, how can this idea be gamed, because it will be?
- How does such a network make money? Or are we talking about a bunch of people who are already have trouble making money being expected it to dribble it away along their network? Or can such a network possibly make more money than any participating project make alone?
This last question because I think that the Open App Ecosystem has potential for being more valuable together than any of the apps alone. I'll explain in a comment.
Bob Haugen Wed 26 Nov 2014 12:40PM
@simontegg - good point. I wonder to what extent value networks will be able to escape such biases.
Lynn Foster Wed 26 Nov 2014 1:46PM
@simontegg yes interesting thought. Reminds me how much I like being "retired". :)
So far what people we work with are doing are:
- hours times rate (R&D),
- words times rate by type of work (like translation, editing),
- % of sale price of the deliverable by process type (like grow, harvest, process).
All pretty objective actually, although of course any of these may or may not be "fair" for various reasons. Looking back at BetterMeans, I understand they used weighted estimates for worth of various deliverables.
Very interesting to see the experiments people are engaged in as all of this unfolds.
Bob Haugen Wed 26 Nov 2014 2:09PM
@lynnfoster one place where subjectivity comes in there is in different rates for different types of work. I remember your arguments for higher value of administrative and cleanup work...and the Guerilla Translator's problems getting people to agree with the value of administrative work. Whereas Sensorica can now see the value of the work that Francois used to do...
Simon Tegg Wed 26 Nov 2014 7:30PM
It seems that there's 'external biases' an 'internal biases'.
In a company I worked at I remember arguments against performance pay metrics because of concerns about how these play out socially as in the linked comment. Unfortunately, the mainstream alternative is to use 'market value' for work on the perception that this is 'fair'. There are consultants that specialise in telling you the 'fair market value' of particular roles.
I find it useful to remind myself that since wealth and spending power are unequally distributed (and controlled largely by men) 'market value' is just what rich people (men) value and are willing to pay.
For ValNet, perhaps the partial solution is a UI one. Make the different rates for different work transparent in the UI -"senior admin work is valued at 90% of 'junior X'" and then easy to get metrics by gender and other demographics, "70% of women are performing roles XYZ valued at W% of the average rate". This would lead the OVN towards a dialogue on pay rates and gender biases. Not necessarily something that software should solve directly.
Bob Haugen Wed 26 Nov 2014 8:49PM
I like the idea of transparency and visibility as a stimulus to dialog on issues like this.
Caroline Smalley Wed 26 Nov 2014 10:53PM
it's key. Our values unite us
Bob Haugen Fri 19 Dec 2014 3:08PM
I wonder what the assembled multitudes will think about this: The Road to Turking.
I agree with some and disagree with some of it, but thought it was good brain food, and on topic for this discussion.
Simon Tegg · Wed 26 Nov 2014 7:07AM
This reddit comment on the gender pay gap reminded me of some of the issues tangentially related to value equations.