Not necessarily. In the short term, I'm guessing Oracle will just make support, patches[1], and so on contingent on having a new-from-them license. Other "enterprise" vendors will do the same.
For stuff like Windows and whatnot, sure. But the business model Oracle were suing to prevent isn't that hard to cripple.
[1] Not that Oracle actually do meaningful security patching, e.g. TNSpoison.
LARC is a huge benefit in the Solaris world, where the comparison is UFS. Solaris UFS is unbelievably primitive, and the reason why job 1 in many SOlaris shops is to buy a third-party filesytem for any serious work.
In the Linux world, where aggressive use of caching has been standard for over a decade it compares a lot less favourably.
That said, btrfs management is a shit sandwich. ZFS's management tools actually let you, you know, use the advertised features of the filesystem.
I failed to see what the big benefit of GRUB was in the first place. It adds a huge amount of complexity for standard Intel boxes, minimal benefits, and when it was first jammed into distros, regressed all sorts of use cases (such as booting from broken software RAIDs).
Much like the Linux audio subsystems, it's a tail of throwing out something that works for 90% of users, replacing it with something of dubious virtue, and then declaring the remaining problems too hard to solve and moving on to the Next Big Thing (GRUB 2 in this case), while giving you a pile of new and insane problems to deal with.
I've dealt with RedHat, and their support has been superb for the most part, working through weekends for crticial problems and shipping us kernel patches for feature enhancements when we've needed them.
They've certainly got a better track record of getting stuff into the mainline kernel than Google, of late, and I'd hope nobody from Ubuntu-land was throwing rocks at them, either...
Boy, I guess it's just as well that "Windows" covers one operating system with one API for writing games with, not DX9/10, XP/Vista/7, and multiple additions of each operating system release, plus multiple processor and GPU architectures. That would be a nightmare.
No, SunFire V440s and V490s, which have been an exercise in terrible since they were new - long histories of kernel panics on memory errors that only go away after motherboard+CPU swaps.
That's based on David Boyes' endorsement of an anon comment in an interview here. Given David's spearheaded the System Z port, I assume he knows what he's on about.
YMMV. We have more problems with our Sun hardware than we ever do with our HP Lintel boxes. Hell, we had an M5K dead within a week of delivery due to a single point of failure with a fan stopping and frying a backplane. And let us not speak of the 4xx series machines, whose memory controllers appear to be made from components eMachines rejected as too crappy.
If you read more closely, you'll see OpenSolaris is poisoned by dependencies on key closed components (including libc) that Sun never released, only providing binary builds for. That fork's a non-trivial task.
If it's that big an issue for you, I'm sure you'll be happy to stand up for those ideals; set up a hosting setup and provide SourceForge quality hosting to anyone, irrespective of local laws.
I can either run my home Fedora 3 server unsupportable, upgrade to a recent version of Fedora (a fairly fraught process if the crappy FC 1 -> 2 -> 3 process is any indication), or bite the bullet and go to Debian or Ubuntu server.
The answer isn't looking good for the last RH box in my house.
And it's a killer or their international reselling; eg, in New Zealand, Battlestar Galactica is just starting its first season airing. Everyone who cares has watched the US torrents already. 10 - 15 years ago, fans just had to take what they were given, when they were given, and like it.
This sort of thing is breaking a lot of profitable distribution models.
I have no idea why they've let this slip though in the XP.
Because when they re-implemented it from scratch for the NT code line, like all the slashweenies suggested, someone made a cock-up not in the original code.
Plenty of people have linked to Joel Spolsky's essay on why rewrites from scratch can be a bad idea, so I won't bother. But he addresses precisely the problem of people re-creating already solved problems when you rewrite.
Slashdot Mirror
Not necessarily. In the short term, I'm guessing Oracle will just make support, patches[1], and so on contingent on having a new-from-them license. Other "enterprise" vendors will do the same.
For stuff like Windows and whatnot, sure. But the business model Oracle were suing to prevent isn't that hard to cripple.
[1] Not that Oracle actually do meaningful security patching, e.g. TNSpoison.
LARC is a huge benefit in the Solaris world, where the comparison is UFS. Solaris UFS is unbelievably primitive, and the reason why job 1 in many SOlaris shops is to buy a third-party filesytem for any serious work.
In the Linux world, where aggressive use of caching has been standard for over a decade it compares a lot less favourably.
That said, btrfs management is a shit sandwich. ZFS's management tools actually let you, you know, use the advertised features of the filesystem.
I wonder how long it is before Canoncial make their own fork, claiming to be contributing code back to an upstream no-one's heard of.
I failed to see what the big benefit of GRUB was in the first place. It adds a huge amount of complexity for standard Intel boxes, minimal benefits, and when it was first jammed into distros, regressed all sorts of use cases (such as booting from broken software RAIDs).
Much like the Linux audio subsystems, it's a tail of throwing out something that works for 90% of users, replacing it with something of dubious virtue, and then declaring the remaining problems too hard to solve and moving on to the Next Big Thing (GRUB 2 in this case), while giving you a pile of new and insane problems to deal with.
I've dealt with RedHat, and their support has been superb for the most part, working through weekends for crticial problems and shipping us kernel patches for feature enhancements when we've needed them.
I'd like some evidence that you actually have more understanding of how to run a profitable game company than the people you're flinging poo at.
They've certainly got a better track record of getting stuff into the mainline kernel than Google, of late, and I'd hope nobody from Ubuntu-land was throwing rocks at them, either...
Most likely. But RedHat are now driving a decent chunk of Java business (making more from JBoss than RHEL these days, I believe).
Always completely clusless, huh? I guess you've got a successful PC games company going, then?
Boy, I guess it's just as well that "Windows" covers one operating system with one API for writing games with, not DX9/10, XP/Vista/7, and multiple additions of each operating system release, plus multiple processor and GPU architectures. That would be a nightmare.
No, SunFire V440s and V490s, which have been an exercise in terrible since they were new - long histories of kernel panics on memory errors that only go away after motherboard+CPU swaps.
They are focusing on the people who pay money for their games: console gamers.
I don't expect that reality to make the freetards happier, but whatever.
That's based on David Boyes' endorsement of an anon comment in an interview here. Given David's spearheaded the System Z port, I assume he knows what he's on about.
YMMV. We have more problems with our Sun hardware than we ever do with our HP Lintel boxes. Hell, we had an M5K dead within a week of delivery due to a single point of failure with a fan stopping and frying a backplane. And let us not speak of the 4xx series machines, whose memory controllers appear to be made from components eMachines rejected as too crappy.
If you read more closely, you'll see OpenSolaris is poisoned by dependencies on key closed components (including libc) that Sun never released, only providing binary builds for. That fork's a non-trivial task.
Nokia N900.
Yeah, taking it all the way to the Supremes worked really well in killing off infinite copyright extension, didn't it?
ceph, XIV, and other distributed storage controller models are available today, and avoid controller bottlenecks.
Microsoft is more my friend than Apple, yet they seem to escape such suggestions in the free software world.
If it's that big an issue for you, I'm sure you'll be happy to stand up for those ideals; set up a hosting setup and provide SourceForge quality hosting to anyone, irrespective of local laws.
C'mon, you love freedom, right?
Install Ubuntu's openssl build.
Still won't be fully fledged computers? That must be why you can't run Fedora Linux on your PS3.
Oh wait, you can.
The modding scene will die? I guess that's why PS3 Unreal players can get mods that let them have Master Chief as a model in UT3.
I can either run my home Fedora 3 server unsupportable, upgrade to a recent version of Fedora (a fairly fraught process if the crappy FC 1 -> 2 -> 3 process is any indication), or bite the bullet and go to Debian or Ubuntu server.
The answer isn't looking good for the last RH box in my house.
And it's a killer or their international reselling; eg, in New Zealand, Battlestar Galactica is just starting its first season airing. Everyone who cares has watched the US torrents already. 10 - 15 years ago, fans just had to take what they were given, when they were given, and like it.
This sort of thing is breaking a lot of profitable distribution models.
Because when they re-implemented it from scratch for the NT code line, like all the slashweenies suggested, someone made a cock-up not in the original code.
Plenty of people have linked to Joel Spolsky's essay on why rewrites from scratch can be a bad idea, so I won't bother. But he addresses precisely the problem of people re-creating already solved problems when you rewrite.