Linux 4.2 release 'possible' for next week, if Linus feels good The world may be on the threshold of another Linux Kernel release. Or it might not. Whether or not Linux 4.2 emerges next week depends on how Linus Torvalds feels next Sunday. On Sunday evening, US time, the Linux Lord gave release candidate 7 the thumbs' up and let it be known he'd quite like it to be the last such release …
"The instruction set so bad even Intel has tried multiple times to kill it."
Did I read somewhere that x86 chips are really RISC under the hood but run an on chip emulator to appear as an x86 to the outside world?
If true, maybe Intel just need make the underlying architecture available to programmers to begin a transfer.
> we had some more fallout from the low-level x86 entry code rewrites
For those of us that don't have time to trawl the kernel list, it would be interesting to hear a technical background what this problem was, what the "low-level x86 entry code" is, and why it was being rewritten.
Edit: my google-fu returns some incomprehensible patches from Ingo Molnar but sheds no more light than that.
Actually from what little I can tell, it's reorganizing the crufty old x86 assembly code to make room for crufty old ARM assembly code and other such architectures.
Apparently x86 is a sorry bag of special cases because "it was there first", "this is the way we've always done it" and "this needs to be fixed someday" and Ingo is tired of it and making that "some day" now.
Just a SWAG from looking at patches.
> Does anybody actually look at this code outside the select few who can read uncommented and undocumented kernel code?
Sadly this seems remarkably true in either open or closed source. Reviewing code is expensive in time and money so we see things like bugs going back decades.
"Geez.
I like how if a corporate CEO would say this, they'd get well and truly shafted by the media, the board and the shareholders and the Linux community is like, oh well."
That is because Linux has been built from the ground upwards, not to sell, but to be the best operating system in the World
Bill Gates never cared what kind of poor software came out of Redmond because knew he could sell just about anything as most users did not have much knowledge, or an alternative. Luckily things have changed!
Welcome to the fold. *BSD is great and much closer to true UNIX but generally doesn't come as polished out of the box for the desktop and requires more work/research and general UNIX expertise but the end result is so worth it (absolutely love running my home box like my HP-UX boxes at work (can do also with Linux but not as well) so my muscle memory works regardless, ksh forever, fuck bash). Sadly thanks to Red Hat now we get to watch more and more FOSS go Linux only as its already getting tough to port some stuff. Linux is trying to basically replace POSIX and will probably succeed eventually. Mac, windows, or windows lite will be your choices.
" Mac, windows, or windows lite will be your choices."
You're being a bit unfair. GNU/Linux is way more modular than the BSDs, so while you're right about Red Hat and a lot of the crowd, it's still possible to build a no-nonsense Gnu/Linux system that avoids the "windows-lite" crap. I've got 5 such systems at home. There are several ways to do it, the "out-of-the-box" approach would be Devuan, but for the more fiddle-oriented / control freaks among us distros like Slackware or Gentoo are build-as-you-go.
Ironically the only system I have at home that uses systemd is the media center SD card for my Pi, because I got the lazies.
> ... ksh forever, fuck bash
Have you looked at zsh? Maybe your disliking of bash would carry over to zsh. If not, give it a shot!
> ... avoids the "windows-lite" crap.
There are many plain window manager and each can be configured.
I use a Linux system that is _very_ bare bone Unix-ish.
@ Medixstiff
The immediate takeaway point is this is a philosophy of ship when it's fit to ship vs ship on a given date (say July 29th?) whether it's fit or not.
Another point is that new kernels are released on a cycle of a few months, not a few years. It's not a case of some new shiny that has a big marketing machine cranked up to go on it.
You should also realise that this is just a kernel release. Only a few people waiting to pounce on this: people who like to keep a bleeding edge box to play with, kernel devs taking this as a new baseline and distro builders. Of the latter those building rolling latest-everything distros will incorporate it. Others will do so if it fits into their time-line for a next release.
For most users it's the major distro releases that matter. They do tend to release to a fixed schedule for the simple reason that they're building from components that, like the kernel, have a release when it's ready approach. And for those of us who've been round the block a few times the distros we prefer are those that have the most conservative release cycles which can run to years; we're not sitting chewing our fingernails in anticipation of a new distro release let alone a new kernel.
But the real issue is that the comparison between Linus and a corporate CEO is utterly false. His role is that of gatekeeper of what actually goes into the product free of any external concerns. That means that the Linux kernel is a product determined entirely by the organisation's QA authority. It's the absence of media, boards and shareholders that allow that to happen. The stakeholders here are just the developers and users; that is, the people who really matter.
So, taking your statement that you like this at face value, I have to agree with you.
"Geez.
I like how if a corporate CEO would say this, they'd get well and truly shafted by the media, the board and the shareholders and the Linux community is like, oh well."
That is because Linux has been built from the ground upwards, not to sell, but to be the best operating system in the World
Bill Gates never cared what kind of poor software came out of Redmond because knew he could sell just about anything as most users did not have much knowledge, or an alternative.
And the *nix crowd actually think this is a better leadership/management than MSFT? Really, 1 person controls your destiny.
A basic tenet of fixing sw defects is being able to reproduce them, and then knowing the defect is fixed in code. Wondering or hoping is not the way...
Laughing....... Ego centric control freaks who when things don't go their way fork the code and split the limited resources some more. Sticking with MSFT thank you very much.
"A basic tenet of fixing sw defects is being able to reproduce them, and then knowing the defect is fixed in code. Wondering or hoping is not the way..."
"Sticking with MSFT thank you very much."
And you're comfortable you're getting the former by doing the latter...
Fair enough...But I'll stick with the code base with the largest peer-review pool.
http://www.theregister.co.uk/2015/0817/microsoft_replaces_windows_10_patch_update_wont_say_why/
Oooo - you're a brave boy AC, maybe you're just ribbing us?
I'd sooner have a foul mouthed, ranting person that strived for perfection heading the production of my OS of choice than a bunch of 'is it good enough yet?' sales men.
If you were ribbing us, well done - if not, well stick with your three-wheeled lorry (truck) of an OS, my OS is ready for work in less than a minute, shuts down in 5 seconds, doesn't need AV, does its updates in seconds and doesn't need a restart afterwards. F*ck Microsoft and f*ck the horse they rode into town on.
And the *nix crowd actually think this is a better leadership/management than MSFT? Really, 1 person controls your destiny.
Most projects have one person (or a small team) who review the current state and decide if/when a release should happen. This is true for both commercial and Open Source projects; the difference is that with most commercial projects (eg MSFT ones) this review process remains hidden, so you don't get to see the hums and has although you might see a slippage from an announced schedule and not know why.
One of Linux's strengths is that the review process is open to public gaze; also Linus releases ''when it is ready'' and not when some marketing team or accountant says he must.
Please learn about the Open Source processes before criticising - don't flaunt your ignorance.
I know about the Open Source process, which is why I made my comments. I tried, long, to use Open Office but it was lacking in documentation, spell and grammar checks and other issues. Instead of that team fixing the issues, some went off in a snit due to disagreements in direction and forked the code. 4 forks later it is still a weak product.
If the peer review process were so good then Linux would be perfect by now. The reality is that 99% of its users never look at the source code, they just use it. Bad security defaults and all.
I also use FreeSwitch where the basic sequence when you report a defect is 'try the latest'. The fact that your symptom goes away means nothing, it could just be hidden by new cruft.
> ... when you report a defect is 'try the latest'.
I am not aware of any project where you would not be asked that.
Have you ever programmed for any project whatsoever?
> The fact that your symptom goes away means nothing, it could just be hidden by new cruft.
Oh, I see, the answer is "no".
Yes, I have done lots of coding and design over the past 48 years of work including safety sensitive stuff like DICOM image transfer protocols and even helped write the MR and CT storage chapters. You know, things like when you get left and right reversed on an image it actually alters someones life, like doing surgery on the wrong knee. So when I hear 'just try the latest' it means they are too lazy/incompetent to figure out what the problem is and then apply a fix. If you can't reproduce it, then you won't know if you fixed the problem. As the US military found out years ago, Hope Is Not A Method.
If sw developers built bridges there would be a lot fewer people on earth. Well meaning, no doubt motivated but not disciplined.
"And the *nix crowd actually think this is a better leadership/management than MSFT? Really, 1 person controls your destiny.
A basic tenet of fixing sw defects is being able to reproduce them, and then knowing the defect is fixed in code. Wondering or hoping is not the way...
Laughing....... Ego centric control freaks who when things don't go their way fork the code and split the limited resources some more. Sticking with MSFT thank you very much."
Okay, you've said your piece. Buh bye now. Go back to the Tech Republic - Windows fanboys where you belong. We know who you are.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
