The question of portable code vs. taking advantage of the best possible system interfaces is a very hard problem indeed.
On the one hand, the number of people using non-Linux POSIX compliant OS's have dropped significantly in the last decade. As a percentage of the installed base, either by machine or by user, what's left? The *BSD's, where I'm not sure there is hard evidence that they are growing even by absolute numbers, never mind as a percentage of the total market. Solaris, which Oracle is busily killing by increasing support costs to the point where people are quickly bailing to other alternatives --- in the end I won't be surprised if aside from the one or two big iron databases servers, there are very few other Solaris installations left. AIX? HPUX? Don't make me laugh.
The only POSIX compliant OS that is growing is MAC OS X --- and Apple isn't afraid to add new incompatible features to their kernel, and then release software which only runs on MacOS --- and usually people don't wag their fingers at Apple complaining that they are releasing non-portable programs. Instead, Apple is usually lauded for keeping a solid eye on usability and writing programs that have a deep sympathy for the end user.
So why is it that a lot of people are kvetching at GNOME for wanting to use kernel-specific features in the name of improving the user experience, when the same complaints aren't leveled at Apple? I think there's a bit of a double standard being applied here....
I think the difference is "interoperability" v. "integration".
The OSS community has typically favored the former, while denigrating MS for the latter.
There's benefit to the integration path. MS' tools play very nicely with one another, and it would be nice to see improvements on the OSS side.
However, I don't think the Gnome project should go that route. Part of its success has been its modularity - it has been long-lived, and used in unlikely places. Choosing the integration route significantly reduces future options. Heck, what if Oracle decides they don't want the expense of Solaris, and spin it off into its own company (yea, yea, I know)?
If Canonical (or other downstream) chose to do something like this, I would think it was a great decision. Ubuntu could end up with a superior experience, if done right.
I don't see the point of a core project doing this. There's nothing about the current model which prevents implementing this downstream, or as an add-on for those that are interested.
Do you really think GNOME owes its "success" to it being able to be used in unlikely places? How are you defining "success"?
If you define success by its ability to run in obscure places, then sure, any change which limits this is going to cause GNOME to be a "failure". But that's not the only definition of "success" and "failure". Personally, I think the ability to run GNOME on obscure operating systems isn't all that interesting. For similar reasons, it really isn't impressive to me that NetBSD can run on ancient Amiga hardware. And in, fact, NetBSD's obsession of supporting legacy hardware slowed down their ability to run on modern hardware, I'd consider it bad and a direct cause of their "failure" in terms of market share as compared to say, Linux.
There's a similar issue going on here with GNOME and obscure/irrelevant operating systems. If that slows them down, then they should (IMHO) figure ways of not slowing them down. And in fact, it's not a complete abandonment of those lesser/legacy systems. There would be cut down systemd releases that would support what ever limited functionality that might be provided by that operating system. If Solaris doesn't have cgroups to provide resource isolation, then there's not much GNOME or the systemd/Solaris adapter layer can do to ameliorate the situation.
On the one hand, the number of people using non-Linux POSIX compliant OS's have dropped significantly in the last decade. As a percentage of the installed base, either by machine or by user, what's left? The *BSD's, where I'm not sure there is hard evidence that they are growing even by absolute numbers, never mind as a percentage of the total market. Solaris, which Oracle is busily killing by increasing support costs to the point where people are quickly bailing to other alternatives --- in the end I won't be surprised if aside from the one or two big iron databases servers, there are very few other Solaris installations left. AIX? HPUX? Don't make me laugh.
The only POSIX compliant OS that is growing is MAC OS X --- and Apple isn't afraid to add new incompatible features to their kernel, and then release software which only runs on MacOS --- and usually people don't wag their fingers at Apple complaining that they are releasing non-portable programs. Instead, Apple is usually lauded for keeping a solid eye on usability and writing programs that have a deep sympathy for the end user.
So why is it that a lot of people are kvetching at GNOME for wanting to use kernel-specific features in the name of improving the user experience, when the same complaints aren't leveled at Apple? I think there's a bit of a double standard being applied here....