On Mon, Sep 22, 2014 at 7:21 PM, Johannes Schlüter <johan...@schlueters.de> wrote:
> Hi, > > On Mon, 2014-09-22 at 14:36 -0700, Kris Craig wrote: > > > > Slightly provocative: Why should I be forced to maintain code by > > people who > > > don't want to maintain it themselves? > > > > Nobody is forcing you to do anything. You choose to contribute to PHP > > in the manner in which you do, just as other people choose to > > contribute in different, sometimes less obvious, ways. > > Right, nobody can truely enforce me doing something, still I gave some > form of promise/commitment to less so since 5.3 reached EOL but still > this might require me to do something. > > > Probably even due to votes by people > > > about whom I don't know anything? Mind that most maintenance work by > > > most contributors happens in free time on a voluntarily base. > > > > > > And no open source doesn't mean democracy as governing model. > > > > It can. Every project is governed differently. > > Well democracy can mean so many things - in ancient Greece, the origin > of democracy, only the men of a social group had a vote. Even in > Switzerland, which is famous for its direct democracy, women weren't > allowed to vote till 1971 (in the canton Appenzell Innerrhoden even only > till 1990 for municipal issues) in others the voting power is unequally > distributed (see i.e. the EU parliament where larger countries have less > MEPs than smaller ones and different voting system's in different > countries give different weight to citizens of different countries) > > Anyways this is a way different debate. > Fair enough. > > > Winston Churchill once famously said that democracy is the worst form > > of government, except all the others that have been tried. > > While this depends on your view on what is good - Louis XIV of France > was quite happy with his, I assume. But government of a society is > different from governance of a software project. One case leads to a > revolution, the other to a fork. > Also fair enough. > > > The > > > democratic part is that people who don't like it can fork the > > project and > > > eventually receive a higher traction. > > > > And then we can have dozens of competing PHP codebases floating > > around. > > That's were the social aspect comes back in - even people without a > formal vote have ability to impact the project. > But that's assuming the threat of fork will be enough, thereby keeping forks to a minimum. I'm not sure I can concur with that assumption. > > > The problem with that model is that history has consistently shown > > that those in power may listen, but will ultimately just do what they > > want, anyway. > > If those with power will "ultimately just do what they want, anyway" the > official form of governance doesn't matter at all. Thanks for agreeing > to that :-D > I think you misunderstood. Ignoring vote results derived from a legitimized process that was agreed to is much more difficult that ignoring a request made by some person without karma, with or without the threat of a fork. > > But as this went to a path through European history let me reiterate and > clarify what I said in a different post in this thread: The strict > dependence on a vote impacts the constructive feedback for proposers > negatively. It also provides no feedbackloop for leading to constructive > critic being ignored, it becomes less clear whether voters were aware of > that. It also makes simple contributions hard, adding quite some > transactional cost for small improvements by newcomers. (then again here > is no clear and objective measure what "small" includes) This is > demotivating for all sides. > I wouldn't be against modifying the voting process to require everyone to state a brief reason for their vote in order for it to be counted. The current table could be modified to add a text column easily enough, I'm sure, and the results could display the reason next to each vote in the row. I think that would at least help mitigate the concerns you're raising here. > > The approach I have in mind is going back to a consensus model by > default, allowing truly everybody to participate and giving the > opportunity to call for a vote if consensus can't be reached. Given our > social diversity I however think that this hardly works out as there > always will be somebody calling for a vote ... obvious consequence would > be a quorum for calling for a vote .. wich ends up in even more > bureaucracy hell. > I've noticed that minor changes are already made all the time without a vote being called and I don't have any problem with that, nor am I aware of anyone else who does. Perhaps we could clarify exactly when a vote is required and when it's not, but since that does not appear to have been an issue thus far, it would probably just be a solution in search of a problem. > > > > I feel it's also worth reminding everyone that VCS accounts generally > > aren't given away like candy. Most people who have that access have > > done something or another to earn it. > > It depends on the time of day and position of the stars, sometimes they > are thrown on people unless they run really fast, sometimes nobody looks > after requests ... :) > Heh true. Perhaps we should focus our energy toward improving the process of giving out VCS accounts, instead. > > johannes > > > > > --Kris
