![[lang]](/images/eye.png)
free clocks |
2007-11-21
|
Dobey, I'm not sure Jeff ever said that these distributors are not free to exercise their right to fork the code and "innovate" in that way (it's a frigging clock, so not sure what the innovation was, but anyway).
I'm definitely not contesting Novell's, Red Hat's, Sun's, or anyone else's right to modify some GPL code and ship a new version of it, with source code, obviously.
But I think Jeff was quite right to respond to Federico, saying "you exercised your freedom rights as a distributor, but you did it in a way that created the very problem you complain about."
I completely agree with Jeff there - using or not using a distributed VCS in that discussion is merely a technical implementation detail. I have strong doubts that any of the distros involved would have in fact taken the time to go look for all of the patches in the various (probably completely local anyway) distributed branches.
It's not the tools that need changing, it's the mindset.
In any case, correct me if I'm wrong, but it looks like you're trying to make the point that anyone's allowed the freedom to fork, while all Jeff was saying is "if you break the collaboration model, you get to keep the pieces". It's merely pointing out the difference between "the community that a license can enable" and "actually being a community".
I think a distributed VCS would still help. It won’t stop anyone from working on a diverging branch, _but_ it will definitely help merging back with upstream, because the distributors can simply file a bug report saying “hey, I added cool feature xpto, feel free to merge it back from my branch that is at URL dvcs://foo.bar/path”. Compare that with current situation where 1. distributor needs to create a patch (and manually keep track of its own changes), and 2. the upstream maintainer has to manually merge everything (manually running a patch commands creates a lot more conflicts than letting a DVCS do a branch merge).
Comment by Gustavo Carneiro — 2007-11-21 @ 12:02
Maybe you should add some lines to tell the reader what all this is about. I did not understand…
Comment by U. — 2007-11-21 @ 13:10
“The tools we use have a profound (and devious!) influence on our thinking habits” – Edsger Dijkstra
Comment by fraggle — 2007-11-21 @ 14:06
Good to hear Dijkstra’s taking responsibility for his creation. :-)
Comment by Jeff Waugh — 2007-11-21 @ 14:48
Dist VCS is not the issue at all and confuses the real issues. How about, as you said, “it’s a frigging clock” – why in the hell would that many people devote that much time and energy to maintaining a separate version? There’s something seriously wrong with the decision making that went into that decision. Also, if these people and companies cannot cooperate on “a frigging clock”, how can they be expected to agree on anything? In an ideal world, that clock wouldn’t even be an issue and the real issues would have been solved long ago, but all the distros still use different package managers, different interfaces for those, different system management tools, etc. Hell, it goes farther than that and brings the GNOME v. KDE differences into it, but I digress. GNOME as a platform or desktop or whatever it’s being called these days, needs to be more cohesive so as not to encourage the diversion that is happening. At least KDE, from an outsider’s perspective, doesn’t have several versions of the basic clock and has more standard tools that can found in any KDE realm and has a file manager that can actually manage files and offer some flexibility.
Comment by Erick — 2007-11-21 @ 16:41