I quite like the idea to separate the GNU and FSF labels and approaches to development. The GNU projects working together in more integrated fashion is closer to the original idea of an integrated "GNU system". But GNU has also accumulated a lot of more standalone projects, so instead of axing them or adding unnecessary dependencies by trying to graft them together, I like the idea of transitioning them to be "FSF" projects, each managed by their own teams, with the FSF serving in a support role to provide resources, publicity, and possibly funds.
I don't necessarily think it's about dependencies, but more about the fact that it's impractical to have 1 guy per project considering all the small software now under their umbrella.
Regrouping things together and saying, for example, this one guy is the head of text processing means that he becomes the go-to guy for all the 1-logical-unit software that goes into that group. While traffic per project might be small, over all the projects it would be enough to warrant assigning a guy to it.
I really think that GNU should cut off GNOME, not make it part of a pruned core group of GNU projects. Under his proposal (focus on development/build/POSIX/GNOME) GNOME really seems like the odd one out.
There's Gnome the window manager library and there's Gnome the "Graphic Environment" the shell and related applications (Gnome-shell, Nautilus, etc).
Gnome the window management library would make a great thing to put within a tighter GNU. It is the basis of Xfce and Ubuntu Unity as well as the Gnome Shell. Taking the underlying library out of the current tug-of-war over interfaces between these environments would be nice. And I realize there's also the Gnome 2.0/Gnome 3.0 issue in libraries but I'd say a more neutral umbrella would still be useful.
Unity also uses compiz instead of Mutter. He was talking about gtk and the support libraries, where you split off Gnome Shell the desktop and leave gtk the toolkit.
> It's that GTK, or the rest of GNOME for that matter, doesn't seem like a good fit for GNU.
Why?
I vaguely agree about GNOME, who are often rather insular and sometimes do things that are of dubious wisdom for the GNU project (e.g., their on-and-off love affair for MS "standards").
However, GTK is something very different indeed. Even if GNOME has grabbed control (or tried to, I dunno the details) of GTK, GTK predates GNOME, was created by different people, is used by a far wider community, and even has a very different feel. At its core, GTK really is just a good UI toolkit; GNU needs a good UI toolkit, and GTK seems to serve that purpose very well.
Indeed, to the extent that GNOME has taken over GTK development, it reflects GNOME's good points (and do I think technically GNOME is often fairly good).
So if there's going to be some sort of distancing of GNU and GNOME, I'd rather see this accompanied by GTK becoming more independent of GNOME as well—not to "screw GNOME" or anything, but to ensure that it continues to serve all its users well (by "users" here, I mean "projects which use GTK", of which there are many, most of which have no connection with GNOME at all).
[It would be fine it the same people continue to work on GTK, but there should be a clear line between it and GNOME.]
The point is that GNU is not promoting Gnome, but GnuStep. And Gnome goes against everything the FSF and GNU stand for, so I never understood why they are keeping it under the GNU umbrella.
They can throw out Gnome and just keep GTK and glib, which are more widely used, independent projects not under Gnome's control.
Do fsf make any impact on everyday computer users?
Take mobiles and tablets, all I see is walled gardens created by large businesses. The longer you use a device the more you invest into it and the more locked in you become.
I would have thought this would be an area which fsf could make a difference in. Instead of a solution I get the impression that fsf would just support a gnu mobile os which would do nothing but create another silo for locked in content and purchases.
Ubuntu Phone is going to be running on top of glibc and the regular GNU coreutils.
You can also stick Plasma Active from KDE on top of most phone hardware, given driver support, and get another mobile OS running on top of GNU.
The FSF isn't supposed to be an OS company. They are a foundation for free software. They wanted to make a POSIX compliant free *nix, and did. Their stuff also happens to work with Linux by design. You can take it or leave it, but they aren't a commercial entity to push consumer adoption.
> There are too many projects in GNU. Micromanagement is impossible at this scale, therefore nobody really tracks them. This has many problems: the signal-to-noise ratio in the list of GNU software projects is low, the different projects are not as coherent as they could be, there is little or no mentoring for new GNU maintainers, and so on.
Somebody should curate GNU and come out with the subset that's signal minus the noise. One could argue that this is what distros are for, but there really needs to be something upstream of the distros. Likewise, there's Debian, which is great, but perhaps there needs to be something else that does a different level of curation.
In the modern world most of those projects hosted by GNU should have ended up on Sourceforge / Github / etc anyway. A GNU hosted project should be a blessed core component of the GNU OS, not a peripheral like a chess program.
Github is maybe lacking for a semi-major free software project looking to create a web presence. It does have source hosting, projects pages, and a bug-tracking system, but it doesn't let you upload builds for downloading any more (without a bit of hackery). There are probably other features that would be useful too, like having some sort of forum for users. Github is also kind of cluttered with small projects and forks. It maybe has kind of a low signal-to-noise ratio.
And Sourceforge is . . . Sourceforge.
It might be useful for there to be a website that would fill in the gaps. It would provide semi-major free software projects with the tools and hosting to create real web presences. It wouldn't even necessarily have to provide source hosting.
You know, that might actually be a good idea. Sounding more hip (without sounding stupid) could actually help. gnu.org could also use some serious redesign. Brand is important.
Also, fsf.org doesn't immediately scream at you what the FSF is and why you should care. They might want to work on that.
I agree on the fsf.org observation. If you look at the EFF as a comparison, their website has the tagline "Defending your rights in a digital world," which makes their mission clear.
Isn't gnome in serious trouble right now, losing developers, users and distributions? Even leaving aside my personal views on the software, "the GNOME project is clearly a successful community, and we all should learn from it" is at odds with the impression I've got from e.g. http://blogs.gnome.org/otte/2012/07/27/staring-into-the-abys...
The first thing I'd do is rename it. I'm not kidding. Inside jokes are fine when they're inside jokes, but when they're the actual face of the project, all they do is alienate outsiders.
