In my opinion, Debian has the best open-source governance model out of all the projects without a BFDL. Because it basically emulates it by appointing a single person to be responsible for the entire vision, yet making him still accountable to others[1]. Thus the horror of the "design by committee"[2] is avoided.<p>[1] <a href="https://www.debian.org/devel/leader" rel="nofollow">https://www.debian.org/devel/leader</a><p>[2] <a href="https://en.wikipedia.org/wiki/Design_by_committee" rel="nofollow">https://en.wikipedia.org/wiki/Design_by_committee</a>
Councils, committees, multiple obtuse voting systems, governance documents, 'courts' of appeal... Why are they so keen to construct such a grim bureaucracy for themselves?
I find it fascinating that programmers like to discuss governance at such great lengths without any reference to the vast knowledge of Political Theory that lies at their fingertips.<p>At the end of the day, I suppose this is simply yet another form of NIH
I have to confess, I was always envy of Python. Look at those PEPs! Yes it may be a little too formal, but there are lots of PEPs, many discussions, and progress. Well the word progress might not be entirely correct, but lots of activities. Look at Guido! He was always on the front battling, arguing or trying to steer the ship. Look at Python and Data Science, it went from good programming language to learn to must have for Data Scientist and Machine Learning. PyPy!<p>I have to confess, I was very envy of Python. I wish Ruby had something similar to PEPs, many more suggestions, Matz making more decisions, Ruby entering more Domains. Ruby could have grown itself 5x and its usage will still possibly be counted as a niche language. MJIT!<p>That was until past few months.<p>It wasn't until Guido Steps down from being BDFL for Python, before the simplest question pops, What happens to Ruby if Matz suddenly step down because of a similar "lively" discussions on features? And now looking at this "governance model", the bureaucratic nature of it. Makes me appreciate a lot more of how Ruby is being handled.<p>This isn't to say the Python model is wrong, far form it. Java have a similar model and it is brilliant.<p>During one of the recent talks Matz said he is already starting to work on Ruby 4.0, which is not about features or speed, but testing a model of future Ruby without Matz when he retire. He is enjoying life, and he is still having fun, but it will come a day when he retires, so he is preparing for it. Despite their syntax being somewhat similar, Ruby's values and culture that makes it a lot different to Python.
It's curious that there was so much discussion about <i>how</i> to vote, while there seemed to be no question <i>who</i> would vote. The article mentions there was a pool of 94 eligible individuals but doesn't explain further where they come from. Does anyone have more info about that?
> Steering council members will serve for the length of single Python feature release; after each release, a new council will be elected.<p>Won‘t that get old really soon?
OP posted a subscriber link to LWN... if you like the content then subscribing to support LWN is well worth it! They are one of (if not THE) highest quality Linux news sites out there.
It's incredible how infectious the democracy meme is. It makes most sense when determining how violence will be used within a society (i.e. the fundamental function of government). It makes a lot less sense for running projects and organizations that people are free to come and go from. But the idea's intellectual stranglehold is just too strong to resist.<p>The merits of singular vision, unity of design, counter-majoritarian good judgment, predictability, and decisiveness should really get more credit in these contexts. As should the downsides of bureaucracy and democratic decision making: infighting, politics (the sacrifice of sincerity for popularity), gridlock, disunity of design, etc.