Who controls glibc? [LWN.net]

Who controls glibc?

Posted May 8, 2018 1:20 UTC (Tue) by interalia (subscriber, #26615)
In reply to: Who controls glibc? by spacemachine
Parent article: Who controls glibc?

It's all in the article title, which isn't "Who thinks this is too PC?" but "Who controls glibc?".

It doesn't actually matter that this disagreement is about some joke in documentation. Suppose Stallman insisted that a function should return EAGAIN in some obscure corner case despite the maintainers thinking it was fine that it does not. Do you think they should defer to Stallman simply because he said so? Is he actually the BDFL he claims to be if the maintainers don't recognise that authority?

to post comments

Who controls glibc?

Posted May 8, 2018 2:12 UTC (Tue) by spacemachine (guest, #124210) [Link] (3 responses)

RMS ultimately controls Glibc, whether he's benevolent or not. Whoever thought otherwise was deluding themselves. They are free to fork if they want but the GNU project is under the authority of RMS, no question about it.

Who controls glibc?

Posted May 8, 2018 3:43 UTC (Tue) by rra (subscriber, #99804) [Link]

Do you think it's a good thing for the GNU project to be under the single authority of RMS?

Speaking as an FSF associate member, I don't, and I say that despite a great deal of respect for what RMS has accomplished. Any governance model that empowers one person that way seems rather dangerous to me. It's also not compatible with how tax-exempt non-profits should be run in the United States. (I realize that many of them, particularly ones founded by charismatic leaders, are run that way, but I think that's a bug, not a feature.)

The Free Software Foundation, like any other tax-exempt US non-profit, has a board of directors who are legally responsible for the actions of the FSF as a whole. They certainly are, and should be, reluctant to override RMS, but they should be capable of doing so if the situation warrants. (Jokes in manuals definitely don't; questions of maintenance authority for well-run GNU projects might.)

If RMS wants an organization that he solely and exclusively controls, well, don't make it a tax-exempt non-profit. Receiving preferential treatment from society and government because your organization supports the public good comes with an obligation to be responsible to the public and a board of directors for one's actions.

And even short of that (I certainly hope this particular issue doesn't get escalated to that level), devolution of authority is generally a good thing. GNU maintainers should be empowered to make decisions about the software they maintain in areas that aren't foundational. And this fairly obviously isn't foundational -- how many people reading this even knew that passage was there? Given that, how much influence could it have possibly had over the political question it tries to address?

Who controls glibc?

Posted May 8, 2018 3:53 UTC (Tue) by interalia (subscriber, #26615) [Link] (1 responses)

Maybe it depends how we define "control" in this situation. No doubt the FSF has ultimate power to appoint and replace maintainers of GNU projects, but does Stallman have sole power to do this on the FSF's behalf?

Even if he does, it would be interesting to know the political ramifications of doing so. For example, the Queen of Australia (a title held by the Queen of the United Kingdom) appoints and replaces our governor-general. Formally she has the complete power to do so, but de facto she does so only on the advice of Australia's Prime Minister, and it would be a constitutional/international incident if she "interfered" by declining to follow our PM's advice. So in practice she does not have control of who Australia's governor-general is.

Stallman may or may not formally be in control of GNU libc, but is he willing to replace all its maintainers in order to keep this one line?

Who controls glibc?

Posted May 9, 2018 4:05 UTC (Wed) by lxoliva (guest, #40702) [Link]

I don't get why the FSF is even brought up here. The FSF doesn't control GNU any more than GNU controls the FSF. One is an independent incorporated foundation, the other is an operating system, and an independent non-incorporated operating system development project. Richard founded and presides the FSF, and he's also founder and leader of GNU (Chief GNUisance), and he's also the founder of the Free Software movement, and more.

Saying the FSF has ultimate power to appoint and replace maintainers of GNU projects is not true at all. That who appoints and replaces maintainers of GNU projects is the Chief GNUisance. The FSF has nothing to do with it.

The FSF manages copyrights over some GNU projects, publishes software licenses used by most GNU projects, but it doesn't control GNU.