Propogation and glorification

Propogation and glorification

OK, so...

The Russell King quote was definitely marginal. The point was that many developers don't see much reason to work on documentation because they don't think it will be read. There was a better quote in that conversation that I could have used, though, granted. It's not like this was even an amusing flame.

For this one, Linus was having fun with the code, yes. Was it really that bad? (I'll note that we're now a year in the past, incidentally).

In this one, Steve was expressing his frustration with users who want a problematic (and arguably nonsensical) system configuration to work. It was not aimed at anybody in particular. Who was offended by this one?

Here is really just a restatement of the no regressions policy, along with an expression of frustration with people who ignore that policy. It communicates a clear point.

This one talks about the evils of rebasing in the context of a particular problematic pull request. Among other things, it seems like a proper service to LWN readers, who have a relatively high chance of being the authors of pull requests, to let them know how Linus feels about rebasing of trees.

This one, you said, is "a good example of the kind of thing not to encourage". I don't think that Linus is amenable to encouragement from me in any direction. In this case, we had the definitive end to an ongoing disagreement that had been documented here; I felt the need to document that as well. Should I pretend that Linus's involvement didn't happen? I thought that Linus's disinterest in the alleged bypassing of normal kernel procedures was an important part of the story.

The last one was from a G+ post; it was already highly visible. If nothing else, it documents how Linus claims to feel on this topic. You're not the first to call that one out, though; perhaps it would have been best skipped over.

In other words, most of those quotes were chosen for a specific reason that was not "it's a creative flame." Doubtless some of the choices could have been done better. But I have to ask what you think should be done differently in general; should LWN pretend that these discussions don't exist? Is it really your position that we should not have quoted any of those seven messages?

You seem to have looked at about two years worth of history, during which, I would guess, LWN has posted probably 5-10,000 messages from the kernel mailing lists. Even if all seven examples were a mistake, it's still a pretty low rate, I would think.

Anyway, I will try to be more careful in the choice of quotes. But not, I don't think, to a point of presenting a distorted picture of how specific conversations go. I'm not sure that would help anybody.

Thanks for pointing these out, it really does help to know what readers think.