Linux 3.10 Officially Released 157
hypnosec writes with word that "The Linux 3.10 kernel has been officially released on Sunday evening which makes the 3.10-rc7 the last release candidate of the latest kernel which yields the biggest changes in years. Linus Torvalds was thinking of releasing another rc but, went against the idea and went ahead with official Linux 3.10 commit as anticipated last week. Torvalds notes in the announcement that releases since Linux 3.9 haven't been prone to problems and 3.10 is no different."
Pass (Score:5, Funny)
I'll wait for 3.11
Re: Pass (Score:5, Funny)
Linux for Workgroups is the best version
Re: (Score:2, Funny)
Re: (Score:2)
Fucking floppies for the win!
Re: Pass (Score:4, Funny)
Re: (Score:2)
And me without mod points
You without mod points is the best version of you? Don't worry, having mod points is a metastable state anyway.
Re: (Score:3)
screw that, the 3.0 (Warp) is the shitz
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
OS/2 is still around and the latest version of 2011 is for sale. Serenity Systems bought the rights to it, now calls it EcomStation 2 and continues development mainly intended as a point of sale system.
http://www.ecomstation.com/product_info.phtml [ecomstation.com]
Re: (Score:2)
Re: (Score:2)
not competitive? it's essentially the same price as windows, and in my opinion a superior OS to windows. if one had OS/2 applications or the development tools that can emit OS/2 code (which incidentally I do) it's not unreasonable.
though I'll be sticking with BSD and Linux myself, thanks.
people pay $140 for their windows home premium plus more than that for office software...spending $500 on home computer software is not uncommon
Re: (Score:2)
Re: (Score:3)
Linux for Workgroups is the best version
Does it come with LinSock support out of the box?
Re: (Score:3)
Re: (Score:2, Funny)
3.14 will be the 'geek' release the mainstream press will notice.
Note to kernel team: so try not to screw that one up.
Re: (Score:1)
Re: (Score:2)
It's not that it "crashed", it's just that we expect it to take an arbitrarily long period of time to start running again...
Re: (Score:1)
They could always move up to 3.1415 if they screw up 3.14.
Re:Pass (Score:4, Funny)
Re: (Score:2)
Re: (Score:1)
Re: (Score:1)
I tried the Bing challenge. Google won.
Two of the searches I tried were "Bing sucks" and "Google sucks". Figured I had to do both to be fair.
Did they put the 'Start' button back in? (Score:1)
Re: (Score:2)
Nvidia drivers (Score:5, Interesting)
Nvidia drivers should be available for the new stable beast by the end of next month. They will get around to it when they are darn good and ready. 3.10-rc1 broke the latest driver. They released a driver about two rc releases ago, but it was still borken. I actually think they released it so that they could say 'see, see, we released a driver just a few weeks ago, so you shouldn't see anything new from us for a while!' It was a fluke that my current hardware build included an nvidia video card (the radeon card I originally bought was borked from the computer store: it wouldn't display video), so I took it back and the only thing they had that was close was an nvidia. They have worked hard to lose me as a customer. I suspect next time they will be successful.
Re: (Score:2, Funny)
Re: (Score:2)
Re: (Score:1)
Generally speaking you are correct. If you want a pain-free Linux install stick with Intel. (This also applies to WiFi cards as well.)
Re: (Score:2)
Re:Nvidia drivers (Score:5, Insightful)
Re:Nvidia drivers (Score:4, Insightful)
I'm with you in this sentiment generally, though I'm also away of Linus' "fu** you, NVIDIA" moment. Apparently, NVIDIA are annoying collaborators with devs, and not only for video drivers.. so let's not cut them too much slack.
My pet peeve: people complain constantly that NVIDIA "refuses" to open source their drivers. But these people don't understand that it's not a matter of merely deciding to do so: the NVIDIA drivers contain a whole bunch of 3rd-party code that NVIDIA cannot legally open source. It would require either 1) a lot of legal agreements (and likely lots of royalty and lawyer fees) to make 3rd-party agreements, or 2) rewriting the 3rd-party code from scratch, without referring to the original code. Both of these tasks are monumental and very expensive (for task #2, they would have to hire new programmers that have not been "tainted" by having seen the original code).
Specs can't be "just" released for similar reasons: like the code, they are encumbered by patents and copyrights.
NVIDIA have expressed a general will to open source the driver, but it may take years to take it to the next step.
Re: (Score:3, Insightful)
Given that patents has already been sent to the patent office and are public accessable from there; patent are never a reason for not open up specs.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
Yeah, but the hideous thing about patents is that, even with the published specs, it's illegal to implement them without paying up. So only the patent holder or a direct licensee gets to write a driver for the patented shit. Any roylaties for a patent-encumbered piece of hardware should be included in the price of the hardware - end of story. And that includes your PC - if you paid for an OEM Windows license that covers various patents, that should cover any implementation of those patents on the device.
Re: (Score:3, Interesting)
So what's to stop them opening up those bits that they do own, and then allowing the community to fill in the blanks?
Considering people are willing to try writing a complete driver from scratch, replacing a few missing bits in an otherwise complete driver isn't much of a stretch.
Re: Nvidia drivers (Score:2)
Perhaps they can't open source the stuff they own because they're not exactly sure what they own?
Think about a decade of legacy code which may not be completely documented of who each individual author is and what license each line is under. It could be a mess that they'd rather avoid by just helping the open source community write their own code from the ground up.
Re: (Score:3)
for task #2, they would have to hire new programmers that have not been "tainted" by having seen the original code
Uhm, why?
Specs can't be "just" released for similar reasons: like the code, they are encumbered by patents and copyrights
That makes no sense. Patent encumbrance can't possibly matter for releasability (is that a word?) of specs; patents are public by definition. And copyright is yours if you write the spec yourself.
Re: (Score:2)
They are dumb. Release the driver as opensource, and depends on the closed 3rd party code, shipped as the current binaryblob.
Open source devs would waste their time implementing open source version of those components, freeing nvidia ofof paying royalties in the future.
Re: (Score:2)
Re: (Score:2)
I have the same video card (in my case, the 1GB gigabyte card, whose fan has failed and which I've replaced with cooler master) and the same logical basis for my decisions. Best video card value I've ever bought. It's slow now, but it worked when I bought it even though it wasn't officially supported since it's derived from another card and it's still working today, many moons later.
The latest AAA games aren't on Linux, so unless you want to use it for GPGPU you just don't need the latest video card. You ca
It's much better, but I still have serious bugs (Score:3)
Re: (Score:2)
Ask anyone who has struggled with marginally-supported graphics
Re: (Score:1)
Agreed.
HD3200 works splendid with Mesa 7.11 or later (GL 3.1 currently), and that's a few years old. Anything new enough before "GCN" (the new architecture that the upper-end HD7000 chips use) has GL 3.1, though GCN is still at 2.1 plus GLSL 1.3 (the version for GL 3.0).
HD5xxx up through HD8xxx currently have hardware VDPAU via UVD on Mesa.
HD4xxx and up to GCN have better OpenCL support via clover than any other FOSS driver.
Power management just got added, and it works.
Fedora 19 has good enough support to u
Re: (Score:3, Funny)
Re: (Score:2)
Wrong. It's impossible to use higher revisions of OpenGL on the open source AMD drivers, and the AMD binary drivers are complete shit.
I recently tried the open source Radeon drivers and to my surprise they exposed a OpenGL 3.0 core profile.
Re:Nvidia drivers (Score:5, Insightful)
don't blame nvidia for not supporting the ever-mutable internal API of the Linux kernel. it's your fault for trying to run bleeding edge crap; stick with stable polished mainstream distros and you'll always have an nvidia driver
Re: (Score:1)
you can run drivers for XP on Windows 7
Correction: You can run a select few drivers from Windows XP on Windows 7. Most drivers will not function, even in 32 bit mode, which is why most drivers you download have a separate xp/2k folder. Microsoft pulls this crap all the time.
Windows 3.11 drivers worked in 95, but they changed the framework in 98, then again for 2000, then again in Vista. I'm quite surprised they didn't do it again in 8, because it's been a constant "Every other release" change. Your whole "Windows uses a consistent ABI" thing
Re: (Score:2)
wrong, mature unix kernels have stable interface and backward compatiblity in their internals. it's immaturity of Linux developers that is issue
Re: (Score:3)
Given that 3.10 is not a release, getting new drivers for 3.10-rcX is better than you can expect with WINDOWS so I'm not sure what your bitch is. When I've upgraded Windows (RELEASE software) I've had driver issues for weeks or months while the vendors catch up. This has happened to me every single OS upgrade in Windows land, save for the jump from Windows 95 to Windows 98.
Having a cry about Nvidia's shitty Linux support for this is a bit off the mark, IMHO. They don't even put drivers out for rc vers
Re: (Score:2)
Re: (Score:2)
How Long Before Showing up in Major Distributions? (Score:1)
How long before it shows up in major distributions such as Linux Mint?
Re:How Long Before Showing up in Major Distributio (Score:5, Informative)
How long before it shows up in major distributions such as Linux Mint?
Don't know, but Fedora 18 has 3.9.6-200.fc18.x86_64 and that was a week ago. A quick check of the updates indicates that the 3.9.6 kernel is still the latest. As far as getting the 3.10 kernel goes I would say within a week or two, however it really depends on your distribution and how up to date the maintainers like to keep the repositories.
If you are the repository maintainer for a customer that is using say Redhat Linux (you would be crazy to install a non supported Linux distribution on a production or even development machine) you may have a two to six month delay offset on updates and that is assuming that the customer or company allows 6 monthly updates. In my experience many companies don't like to do any updating once their systems are up and running and it is allot of work on the IT managers side to even get critical patches applied and without the appropriate sign-off's and agreed outages (normally 10 minutes) nothing gets done.
Re: (Score:2)
(you would be crazy to install a non supported Linux distribution on a ... development machine)
No, I strongly disagree.
You should be able to smash up your development machine with a hammer now, and be back up and running in a few hours. I've run all sorts of stuff as development machines, including distros far out of support for various reasons, and others like Arch which are totally bleeding edge.
Also, hardware aside, I've never screwed up a developement machine so badly that I couldn't put off fiing it un
Re: (Score:1)
That includes accidently killing an ubuntu upgrade part way through.
You've done that? Far out.
:)
I've never been able to stick with Ubuntu for more than an hour, much less upgrade it...
Re: (Score:2)
You've done that? Far out.
It's kinda fixed, with judicious use of cargo-cult apt and dpkg incantations.
I've never been able to stick with Ubuntu for more than an hour, much less upgrade it... :)
Well it helps if you install FVWM and a few other nice tools :)
Re: (Score:1)
Re: (Score:2)
Fedora makes available new kernels within a few days, for those that want to play with the latest and greatest. The 3.10 kernel should be available within the next 24 hours using the Fedora rawhide kernel nodebug repository [fedoraproject.org].
Re: (Score:3)
What does kernel development have to do with UI design?
Re: (Score:2)
You were rather vague about just what you were complaining about, but from context I presume it to be GUI. Probably Gnome3. If so I certainly agree, but it doesn't have much to do with the kernel.
OTOH, there was a time when the scheduler varied a lot between releases. That seems to have stabilized, though, over a year ago. Otherwise, I can't guess what you are talking about.
(FWIW, there was a C library change a few years ago that broke some games I have installed. So I run them on a virtual machine. I
Re: (Score:2)
You clearly aren't an Ubuntu user. Neither am I anymore. (I went back to Debian.)
Re: (Score:1)
It's funny because you pretend like Mint isn't a major distribution. I mean, it's not like it's the 2nd most widely used or anything, right?
Thank you, Dr. Linus Torvalds (Score:5, Insightful)
You are one of the greatest and most generous people on Earth. Thank you for all your work!
Re: (Score:1)
Re:Thank you, Dr. Linus Torvalds (Score:5, Funny)
why, he didnt do shit except scream like a tyrant
Perhaps he's channeling Steve Jobs.
Re: (Score:3)
why, he didnt do shit except scream like a tyrant
A bit of applied tyranny can be just what certain situations need... (And, by historical standards, Torvalds provides tyranny services at extremely reasonable rates)
Re: (Score:2)
Linux is obsolete. HURD is coming (Score:5, Funny)
GNU Hurd is going to reach stable status very soon! At that point, Linux will be essentially obsolete.
Re:Linux is obsolete. HURD is coming (Score:5, Funny)
just in time to run the Perl 6
Re: (Score:3)
GNU Hurd is going to reach stable status very soon! At that point, Linux will be essentially obsolete.
HURD? Heck, why? I mean: what's wrong with EMACS OS? You can even tweet [emacswiki.org] from it: try this using only the retarded Linux or HURD kernels!
(ducks)
Re: (Score:2)
Yea, sure, EMACS is a great OS.
It just lacks a decent editor. :!duck :x
Re:Linux is obsolete. HURD is coming (Score:4, Funny)
It just lacks a decent editor
No it doesn't. Try the following:
M-x term
vi
Re: (Score:2)
Re: (Score:3)
I wish they would fix khugepaged (Score:2)
Re: (Score:3)
Yeah, someone should invent "echo madvise > /sys/kernel/mm/transparent_hugepage/enabled" and maybe make the default a config option or something... oh, wait.
Except that crappy workaround doesn't fix it, you have to set it to never as a workaround or simply disable transparent huge pages on boot.
Haven't been prone to problems? (Score:2)
The 3.9 releases haven't been prone to problems? Half of the 3.9 RCs panic'd my Phenom II X6 1045T system.
I call shenanigans.
Re: (Score:1)
A 3.9 RC is not a release version...
Kernel Newbies? (Score:5, Informative)
Do we not like Kernel Newbies anymore? I've always looked to them for a synopsis of kernel features: http://kernelnewbies.org/Linux_3.10 [kernelnewbies.org]
Re: (Score:1)
Do we not like Kernel Newbies anymore? I've always looked to them for a synopsis of kernel features: http://kernelnewbies.org/Linux_3.10 [kernelnewbies.org]
Even Linus's release mail suggest that "As usual, I'm sure H-Online and kernelnewbies will do better writeups of the details.." :-)
http://lkml.indiana.edu/hypermail/linux/kernel/1306.3/04336.html [indiana.edu]
But why aren't people using it? (Score:2)
So here's a question: why aren't SBC manufacturers keeping up with kernel versions? Shipping product is often stuck somewhere in the middle of the 2.6 series.
Re: (Score:1)
Re: (Score:1)
Makes perfect sense, the xpad driver is for Xbox controllers... The Xbox and Xbox 360 (as well as PS3 and probably next-gen) controllers already interface via USB so they make great PC controllers as well.
Re: (Score:1)
only over a decade late on that post!
Re: (Score:2)
Re:Ok (Score:4, Informative)
No [semver.org].
Re: (Score:2)
That (totally non-standard) spec you point to has a severe downside: it recommends for pre-releases to have a patch level. That's no only wasteful (it will be always 0), but also makes pre-releases sort AFTER the final:
3.0.0-rc1 > 3.0.0
3.0-rc1 < 3.0.0
(because - < . in ASCII).
Most projects I know of, including Linux, use 3.0-foobar for versions leading to 3.0.0.
Re: (Score:3)
If you're doing version comparison based on ASCII codes, you might as well give up now, because you have a terminal case of stupid.
Re: (Score:2)
So I'm supposed to have a magical sorter that hires an oracle for every pair of version numbers, right?
There's a common scheme of comparing them that almost everyone agrees with: sort lexically, taking a string of digits as single symbol: 3.9.2 compares as less than 3.10.1, 3.10a as less than 3.10b. That "semver" proposal doesn't work with that.
Re: (Score:2)
Simple string sort says 3.10a > 3.10. "a" and "b" can work like version numbers as in OpenSSL, but they don't work as designations for "alpha" and "beta" unless you're already finagling your sorting algorithm, in which case checking for semver's "1.0.0-alpha" or "1.0.0+a1b2c3" and placing them behind "1.0.0" is hardly more effort. If you're really concerned about making the string sort slightly simpler, then why not put in a change request to make the patch version optional on pre-release versions? There
Re: (Score:2)
That's a good point. And yes, I agree that semver is in no way standard. In fact, I was mainly lazy and picked the first semi thorough reference I could find on the classical versioning number scheme, though to be honest I'd rather just distill it to:
A version number is a tuple of integers of decreasing significance separatade by a dot. Whenever one of the integers is incremented by one, the subsequent ones are reset to 0 or removed.
Other shenanigans (such as -rcX) varies between projects, and is usually ea
Re: (Score:1)
Someone pulls idea out of their ass, makes website about it, then claims everyone else is wrong. Film at 11.
Re:Ok (Score:5, Informative)
You are wrong. You are assuming software version numbers are numeric, following a decimal number system. They are not.
They are strings, in this case, of the format : '(major_iteration).(minor_iteration)'. Such a pseudo-numeric format is used for several other denotations. A commonly used one is the date. A less common one is chromosomal locations of your genes. To parse such a string, you must know the rules of the format.
Print this, paste it on your wall. And never whine about software version indicators of any kind ever again.
Re: (Score:1)
I've been writing software since Gerald Ford was president, son.
3.9 > 3.10
Take your pseudo-numeric format and park it.
Re: (Score:2)
If you haven't picked up that software versions are not decimal numbers in 2013 , you should have stopped writing software since Gerald Ford's term was over.
Re: (Score:1)