Slashdot Mirror


Linus Torvalds Isn't Looking 10 Years Ahead For Linux and That's OK

darthcamaro writes: At the Linuxcon conference in Seattle today, Linus Torvalds responded to questions about Linux security and about the next 10 years of Linux. For security, Torvalds isn't too worried as he sees it just being about dealing with bugs. When it comes to having a roadmap he's not worried either as he just leaves that to others. "I'm a very plodding, pedestrian person and look only about six months ahead," Torvalds said. "I look at the current release and the next one, as I don't think planning 10 years ahead is sane."

108 comments

  1. I'll look ahead for him by Anonymous Coward · · Score: 1, Funny

    Systemd

  2. Linux and Bloat by Anonymous Coward · · Score: 5, Interesting

    If you actually read TFA you see a little bit of nostalgia from Linus about how lean the kernel used to be and how modern Linux may be a little too bloated for some IoT applications. The truth is that Linux can certainly be less bloated that a full desktop Windows 10 installation, but it is nowhere near as lean as it used to be. Not much of an issue in larger hardware where even smartphones have more power than powerful desktops did 15 years ago, but there are definitely areas where the modern Linux kernel is a little too big for its own good.

    1. Re:Linux and Bloat by Anonymous Coward · · Score: 1

      modern Linux Kernel is a little too big for its own good.............such as?

      Bloat yada bloat blah blah. I use linux for some IoT stuff. Strip the shit out of your kernel and you can get it pretty small. Maybe not XT-era-omg-20KB but yeah, pretty small.

      Remember, LInus doesn't look ahead. In 10 years, our laptops will be as powerful as current moderate-sized super clusters. Or something. The IoT world will be running a Linux 4.20 kernel because the Linux 19.23 kernel is too bloated.

    2. Re:Linux and Bloat by phantomfive · · Score: 2

      He's been complaining about bloat a bit more often recently. The thing is, all the features that have been added are used by someone: they're not useless.

      --
      "First they came for the slanderers and i said nothing."
    3. Re:Linux and Bloat by Anonymous Coward · · Score: 4, Insightful

      That's the thing about bloat. It's always used by someone, somewhere.

    4. Re:Linux and Bloat by phantomfive · · Score: 2

      That's the thing about bloat. It's always used by someone, somewhere.

      Well said.

      --
      "First they came for the slanderers and i said nothing."
    5. Re: Linux and Bloat by Anonymous Coward · · Score: 0

      If it's used it's not bloat.

    6. Re:Linux and Bloat by Gr8Apes · · Score: 2

      He's been complaining about bloat a bit more often recently. The thing is, all the features that have been added are used by someone: they're not useless.

      But if they're useless to more than 90% of users, perhaps that should be an option for the remaining less than 10%? I'd go so far as to say that perhaps that split should be at 70/30, because I'll bet a whole lot of features falling lower than 70% usage are only used by a small subset of users.

      --
      The cesspool just got a check and balance.
    7. Re: Linux and Bloat by Anonymous Coward · · Score: 1

      Today's laptops are around four times as powerful as a decade ago. A typical cluster today is 1000 times as powerful as a current laptop.

    8. Re:Linux and Bloat by phantomfive · · Score: 1

      Yeah, maybe eventually he'll come to that kind of solution. A lot of features can be left out at compile time, after all.

      --
      "First they came for the slanderers and i said nothing."
    9. Re:Linux and Bloat by phantomfive · · Score: 3, Insightful

      I would strongly argue that if the feature is used by less than 30% of the users, maybe it shouldn't be in the OS at all.

      Linux is used by so many different people, and by such a diverse group of people, that almost everyone probably uses a feature that is used by less than 30% of the users.

      --
      "First they came for the slanderers and i said nothing."
    10. Re:Linux and Bloat by phantomfive · · Score: 5, Interesting

      To add to my earlier comment, note that most Linux installations are probably on Android, or in other embedded. If we used that as the metric, we may end up removing all mouse support, since most people don't use it.

      --
      "First they came for the slanderers and i said nothing."
    11. Re: Linux and Bloat by Anonymous Coward · · Score: 2, Insightful

      Today's laptops are around four times as powerful as a decade ago.

      Far more than 4 times as powerful, my child. Try more like 50 times as powerful as 2005.

    12. Re:Linux and Bloat by Dutch+Gun · · Score: 4, Interesting

      I remember a reading about metrics collected by Microsoft on features most commonly used in their Office products. What they found was that, apart from a handful of the most obvious features such as "Open", "Copy", "Paste", and "Save", the use of features flattened out very, very quickly, in terms of percentages. As such, it would really no sense to create a version of MS Word that only had the "most commonly used 70% of features", because that subset of features would tend to differ wildly from user to user.

      I have a suspicion that you'd find the same to be true of features in the Linux kernel. There are obvious features that everyone has to use, but among all the "optional" features, I wouldn't be surprised to find find that the usage curve tends to flatten out fairly quickly.

      I think there's probably a reason the most popular Linux distros are *not* the stripped down models, but the more fully-featured distros.

      --
      Irony: Agile development has too much intertia to be abandoned now.
    13. Re:Linux and Bloat by Dutch+Gun · · Score: 2

      I would strongly argue that if the feature is used by less than 30% of the users, maybe it shouldn't be in the OS at all.

      Like accessibility features for disabled users?

      --
      Irony: Agile development has too much intertia to be abandoned now.
    14. Re: Linux and Bloat by w_dragon · · Score: 3, Insightful

      Not true. There may be things in the kernel that could be in user space, or in a kmod, or done in a driver. Linus is concerned about the code that 100% of Linux users are required to have, not the stuff that can be easily added or removed from an install.

    15. Re:Linux and Bloat by ChrisMaple · · Score: 1

      Aren't accessibility features handled by the desktop - like KDE or Gnome?

      --
      Contribute to civilization: ari.aynrand.org/donate
    16. Re:Linux and Bloat by Dutch+Gun · · Score: 1

      Well, true, but AC said OS, not kernel. I'm pretty sure I could have found an equally valid example even if we restricted it to the kernel.

      It's just that the premise itself is ridiculous. I mean, a billion people use Linux nowadays, so it's fine to cut a feature so long as it only impacts 300 million or fewer people?

      --
      Irony: Agile development has too much intertia to be abandoned now.
    17. Re:Linux and Bloat by rubycodez · · Score: 1

      What percent of users are disabled users? less than 1 percent? why pander to them to inconvenience the 99%+ percent?

    18. Re: Linux and Bloat by Anonymous Coward · · Score: 0

      Except Android does have mouse support...

    19. Re:Linux and Bloat by Runaway1956 · · Score: 2

      AC posted just above - your kernel and my kernel may not have much in common. Simply compiling with the native CFLAG reduces the size by quite a bit. When I compile, I sometimes watch the screen scroll along, and one of the more common words I see is "stripping". I realize that optimizing the kernel doesn't make my computer 50x faster - but there are times when a native compiled system does seem a little faster.

      The features that you deem essential, can be stripped out by those who don't need or want them. That's one of the great things about Linux - you can do what you want with it.

      I've not fooled with compression. Maybe that will be my next fun project, to see just how small I can make my running kernel.

      --
      "Windows is like the faint smell of piss in a subway: it's there, and there's nothing you can do about it." - Charlie Br
    20. Re:Linux and Bloat by Anonymous Coward · · Score: 0

      And that's more a problem of confusing Linux with Android. You see as much as Linux is a kernel, what people are talking about here is specifically is Linux as a kernel for the desktop/laptop/etc. In many ways, Linux's real bloat problem is a side-effect of its success: it's really good at doing a lot of core things and it's generally well designed enough that compile time or run time compartmentalizing means whole subsets of users can have radically different kernels.

      So, taking a step back, please realize that neither Linus nor most the GG*P posts are probably talking so much about pulling a MS Windows Metro and blindly just reducing it all down to the lowest common denominator (or in MS's case, to try to shift and corner the mobile market). I think it's more about perhaps modularizing Linux more so that Android, Linux as a desktop, and Linux as an embedded system can still fundamentally have the same kernel but not have so many of their differences either (1) in the kernel or (2) in the mainline kernel.

      I mean, in a lot of ways we'd be better off in Google had became a sort of custodial kernel forker and pushed more to get Android drivers to be in their own tree (and not in mainline). Sure, that'd still mean a lot of code sharing both ways. But there's something to be said for not stuffing everything into mainline. Of course having said that, if Google had went that way we'd probably have a larger mess with the whole ARM systems merging that's being going on as some would follow Google and others would follow Linus and likely others would make even more forkers even further splintering an already heavily splintered platform.

      But, then, I honestly don't think the mainline kernel will ever get to the point of sufficiently covering ARM precisely because Linux at its core is too focused on higher-end systems that don't have to worry things like, oh, proper suspend support and battery life. *sigh*

    21. Re:Linux and Bloat by phantomfive · · Score: 1

      And that's more a problem of confusing Linux with Android. You see as much as Linux is a kernel, what people are talking about here is specifically is Linux as a kernel for the desktop/laptop/etc.

      This is a story about the Linux kernel. Linus Torvalds doesn't have much to do with the individual distros, he only works on the kernel.

      --
      "First they came for the slanderers and i said nothing."
    22. Re:Linux and Bloat by Anonymous Coward · · Score: 0

      Like every driver? Every architecture (I wouldn't be surprised if all architectures are under 30% usage these days)?

    23. Re: Linux and Bloat by Anonymous Coward · · Score: 0

      I think that's a valid concern and a better description than what we get usually. Something that can't be easily removed when compiling a kernel and can be concidered to either create more potential security threats or slow the overall performance and is *not* used by the user(s) is bloat. Most of what people concider to be bloat these days can actually be left out of the compiled kernel.

    24. Re:Linux and Bloat by Anonymous Coward · · Score: 0

      This is a story about the Linux kernel. Linus Torvalds doesn't have much to do with the individual distros, he only works on the kernel.

      Then I'd suggest you make up your mind and not mention Android since clearly it's not the Linux kernel. Or you accept my observation about how Linus actual behaves towards Linux kernel development (a statement he has clearly made in the past: a focus on high end PCs and an acceptance of code patches that are compatible, with acceptable performance penalties, and maintainable in that vein, even if they wildly diverge into micro embedded devices to super computers). At which point statements about "user programs breaking" as a serious contract the Linux kernel puts towards user space becomes clearer.

      PS - Seriously, read android.txt and note just how clearly Android is its own separate ABI which actually reliably supports Suspend (among other things). And Google is perfectly willing to break ABI (pushing the requirement for PIC which breaks old apps, for "security"* reasons). <<sigh>>

      * Because ASLR isn't just a potential mitigation technique, apps that don't use it are insecure....or not. That's a little overkill, heavy-handed PC security theater.

    25. Re:Linux and Bloat by shentino · · Score: 1

      How much of that bloat can be disabled through kernel configuration?

    26. Re:Linux and Bloat by DrXym · · Score: 1
      They've already created a version of Word with the features most people use. It's called WordPad.

      Strangely enough people don't like using WordPad. It may have the features most people use but it doesn't have the specific features that individuals need. One person might need outline mode, another might need mail merge, another might want table of contents and citations. The sum of all these needs is the bloat that is MS Word.

      I think bloat is okay providing it doesn't become dead weight - code which is so esoteric, ancient and / or broken but still used by influential customers that it undermines the entire product. Microsoft tends to keep Office fairly clean all things considered, but if you want to see an extreme example of doing it wrong, then look no further than Lotus Notes.

    27. Re:Linux and Bloat by Anonymous Coward · · Score: 0

      You can cut out all the fat, though. `make menuconfig` lets you remove or module-ize just about everything.

    28. Re:Linux and Bloat by Anonymous Coward · · Score: 0

      Not true. When the LibreSSL guys started chopping out code from OpenSSL they found, among other things, code to support for big endian x86 platforms, which don't exist.

      stratus.com decided to emulate big endian in gcc on x86, so maaaaybe you can argue this bloat is OK. It is the reason it was added after all. However, OpenSSL also had code to support big endian x64 platforms just in case someone does the same for x64, which has never happened. It has never happened because it is kinda insane to do that (as it was for x86).

    29. Re: Linux and Bloat by Zontar+The+Mindless · · Score: 1

      I just happen to have fairly 2 representative models handy.

      2005 laptop: 1.6 GHz Celeron. 1 core. 1GB RAM. 350GB disk. 100 Mbps Ethernet.

      2013 laptop: 2.6 GHz i7. 4 physical cores, 8 virtual. 8 GB RAM. 750 GB disk. 1 Gbps Ethernet.

      I could easily bump up the latter to 16GB RAM and add an SSD. (But I'll probably just buy a new one, as the warranty on this one expires in January/February.)

      --
      Il n'y a pas de Planet B.
    30. Re:Linux and Bloat by Anonymous Coward · · Score: 0

      I'd be happy to just use WordPad but it isn't completely compatible with Word files. It renders things incorrectly.

      Excel is a better example of the concept anyway.

    31. Re:Linux and Bloat by JazzLad · · Score: 1

      I know I'm feeding a troll, but every sane (and many other) country has legislation regarding accessibility. Someday perhaps you will have a child with a disability. Do you want that child to grow up without access to technology, or do you think those of us with a disabled child should separately fund accessibility features?

      Besides (and almost as importantly), just how big is the inconvenience to the 99%? I'd venture it's only inconvenient to a very small subset & now we're looking at a group of people that cannot do without versus the convenience of another small group. Since society as a whole recognizes the need to assist persons with disabilities, perhaps the other small group should fund their own version that is not encumbered.

      Finally, either double-check the definition of pander (perhaps you didn't mean what you said), or screw you.

      --
      "If you have nothing to hide, you have nothing to fear." - Every fascist, ever
  3. Re: Linus Torvalds is for cows. by Anonymous Coward · · Score: 0

    How dare you to insult the great benevolent dictator! He is the first king of the lkml!

  4. Re:In ten years... by whh3 · · Score: 1

    I thought that was going to happen "this year".

    --
    remove nospam. to email!
  5. Re:In ten years... by Anonymous Coward · · Score: 0

    Desktop is so last millennium.

  6. Re:Stable driver ABI by phantomfive · · Score: 2

    The driver ABI is unstable by design.
    Open-source your driver, and you won't have to worry about it: the kernel team will maintain it for you.

    --
    "First they came for the slanderers and i said nothing."
  7. Linus Torvalds Isn't Looking 10 Years Ahead by thinkwaitfast · · Score: 1, Offtopic

    Isn't this what everyone says is the problem with American corporations?

    1. Re:Linus Torvalds Isn't Looking 10 Years Ahead by Anonymous Coward · · Score: 5, Insightful

      Difference is:

      Corporation - We should lay off half the workforce, that would save us so much money.
      [6 months later]
      Corporation - Why is productivity so low?

      Linus - Lets get/keep things working
      [6 months later]
      Linus - Lets get/keep things working

    2. Re:Linus Torvalds Isn't Looking 10 Years Ahead by Lisias · · Score: 1

      That ones that are currently in the most rich country in the World?

      That ones that leaves the risk of ahead planning to others, and just buy who does it right?

      yep, I'm afraid it is. =/

      --
      Lisias@Earth.SolarSystem.OrionArm.MilkyWay.Local.Virgo.Universe.org
    3. Re:Linus Torvalds Isn't Looking 10 Years Ahead by Shimbo · · Score: 1

      Well, not exactly. If I plant a tree on the basis that I have no idea how it will look in 100 years time, it's not really the same as short term thinking.

    4. Re:Linus Torvalds Isn't Looking 10 Years Ahead by Richard+Steiner · · Score: 2

      Linux isn't a corporation. Also, for something as specific as an OS kernel, I'm not sure there's much to be said for looking out in a detailed manner more than a few major iterations in advance. Sure, there are probably features being considered all the time, but that doesn't make them a focus.

      --
      Mainframe/UNIX Bit Twiddler and long time Windows/Linux Hobbyist.
      The Theorem Theorem: If If, Then Then.
    5. Re: Linus Torvalds Isn't Looking 10 Years Ahead by Anonymous Coward · · Score: 0

      Which is why we can't expect true innovation from the Linux world. They only will build what they can already see around them. Finland, start your photocopiers!

    6. Re:Linus Torvalds Isn't Looking 10 Years Ahead by Your.Master · · Score: 3, Insightful

      I think when people are talking about corporations being shortsighted, they aren't talking about corporations failing to plan ahead, they are talking about corporations taking actions that clearly damage their own future potential. Linus taking Linux day by day, or in 6-month sprints, or whatever, isn't really the same thing because that doesn't hinder Linux's ability to compete. At worst, it helps it sub-optimally. This as opposed to killing your most profitable product line, or laying off the people who work on your next product instead of the people who sell last year's product, etc..

      It would surprise me though if he doesn't have at least some long-term goals that take over 6 months to complete and that he's not focussed on working on right now but has in his back-pocket, but maybe he really doesn't.

      I also think the statement about corporate shortsightedness is somewhat overused, although not entirely without merit. When somebody says something like that, I sometimes click on their posting history to see if they also make claims like "big Pharma will never release cures because palliative care is more profitable" and the like to help me determine if they're logically consistent and therefore might be worth paying attention to, or just reflexively take anti-corporate positions (likewise for pro-corporate positions). And yes, I know they could believe that all corporations *except* big Pharma are short-sighted.

    7. Re:Linus Torvalds Isn't Looking 10 Years Ahead by Kjella · · Score: 1

      It would surprise me though if he doesn't have at least some long-term goals that take over 6 months to complete and that he's not focussed on working on right now but has in his back-pocket, but maybe he really doesn't.

      Well, the kernel runs on everything from cell phones to supercomputers and that it doesn't run on desktops has absolutely nothing to do with kernel features. I don't see any huge glaring TODOs, by far most the changes are drivers (non-CPU hardware driven), followed by architecture (CPU hardware driven), followed by high performance in kernel implementations like file systems or network filtering. The number of patches that really touch core "kernel" functionality of managing other processes seem to be rather few.

      --
      Live today, because you never know what tomorrow brings
    8. Re:Linus Torvalds Isn't Looking 10 Years Ahead by allcoolnameswheretak · · Score: 1

      FYI, per capita, the US is not the richest country in the world. Its about 10th place.

      And if you factor out the wealthiest 2%, which in the US own almost half the country, and only consider average Toms, then the US would probably rank around 25th place.

      And if you factor out how the US likes to spend tax money producing F-35 instead of reinvesting it into the neediest parts of the populace via social programs and welfare, then the US would probably rank around 30th place.

    9. Re:Linus Torvalds Isn't Looking 10 Years Ahead by allcoolnameswheretak · · Score: 1

      Its probably worth mentioning, that the US is also no longer the richest country in the world by total GDP, when measured by purchasing power parity (PPP). It has been overtaken by China this year.

      It is only the richest country in the world by total, nominal GDP.

      GDP (PPP)

      GDP (nominal)

    10. Re:Linus Torvalds Isn't Looking 10 Years Ahead by gweihir · · Score: 1

      Hehehehehe, nice!

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    11. Re: Linus Torvalds Isn't Looking 10 Years Ahead by mr_mischief · · Score: 1

      Your reasoning falls short. Linus isn't the driver of new features. He's the head maintainer. Many other people contribute code they need to do new things. He vets it and makes sure it doesn't break the rest of the kernel. He enables other people to do the innovative things by giving them a well-maintained place to put them.

    12. Re: Linus Torvalds Isn't Looking 10 Years Ahead by Richard+Steiner · · Score: 1

      Your reasoning falls short. Linus isn't the driver of new features. He's the head maintainer. Many other people contribute code they need to do new things. He vets it and makes sure it doesn't break the rest of the kernel. He enables other people to do the innovative things by giving them a well-maintained place to put them.

      It's true that Linus is simply the head of the maintenance tree, but the fact that he isn't the driver of new features seems to support my thesis that he doesn't have much need to look ahead very far. That function is generally handled by others lower in the tree, or completely outside of it.

      I'm not sure I see my logic failure. Could you elaborate? Am I emulating Eliza? :)

      --
      Mainframe/UNIX Bit Twiddler and long time Windows/Linux Hobbyist.
      The Theorem Theorem: If If, Then Then.
    13. Re: Linus Torvalds Isn't Looking 10 Years Ahead by mr_mischief · · Score: 1

      First of all, I didn't reply to you. I replied to the AC who replied to you. That AC said " Which is why we can't expect true innovation from the Linux world. They only will build what they can already see around them. Finland, start your photocopiers!"

      How my telling that AC their reasoning falls short becomes me telling you that your reasoning falls short is beyond me unless you replied to yourself as AC and are now claiming that post.

      There can still be true innovation in the Linux world. Just because it doesn't come from the maintenance team doesn't mean it doesn't happen. Linus's role isn't to be the pioneer. He's too busy paving the roads the pioneers cut.

  8. This just in by Anonymous Coward · · Score: 0, Flamebait

    Linus Torvalds has no vision of the future. Linux doomed.

  9. Seems Reasonable... by fuzzyfuzzyfungus · · Score: 3, Interesting

    As much as 'Linux' keeps being dragged into assorted rambling-think-pieces as though it were a direct analog to the OS-and-also-a-big-suite-of-hardware-software-and-'cloud'-service-offerings referred to as 'Windows' and 'Apple' or 'OSX'; 'Linux', so far as it is Linus' problem, is a kernel. It's also a kernel that has succeeded largely on the basis of being widely supported, reasonably flexible(with greater flexibility available to those willing to do additional heavy lifting themselves), and an inexpensive implementation of mostly-unix-like behavior.

    That's not a role 100% free of strategic considerations(like the current 'beating on the ARM vendors to un-fuck the current fragmented hellhole of disjointed BSPs and embrace sanity' initiative); but it is one where "ensure continued cooperation among interested users and hardware vendors, integrate promising out-of-tree developments as demand and maturity suggest" is more or less the best strategy to take. It's not as though it would even be meaningful for an OS to "Embrace a cloud services strategy", since that happens at a different level of the stack entirely; and to the degree that OS development does need, and do, blue-sky cool-new-architecture-from-the-ground-up; that isn't exactly mainline Linux's problem; and Linux probably isn't even an obvious starting point(if your bold new OS concept makes use of some sort of exotic hardware capabilities, you'll presumably be prototyping on FPGAs or the ASICs you are developing in tandem with the OS; if it is designed to work with mostly standard hardware; but do some part of being an OS differently, you can develop against a delightfully small and stable collection of 'hardware' thanks to VMs.

    1. Re:Seems Reasonable... by NotARealUser · · Score: 2

      You are right on all counts. I especially agree with making a goal of fixing the BSP issues with ARM development.

      What a lot of people do not understand is that Linus does not have a 1000 person development team sitting at his corporate headquarters, churning out code and ideas. Code gets essentially given (through patches, Git pull requests, etc.) to Linus (or someone Linus trusts) and that eventually makes its way into Linus's branch of the kernel. The system is built on the idea that needs get fulfilled by those who need it. At this point, Linus just integrates the really good ideas into the kernel while filtering out the bad stuff.

      As far as I could guess, Linus probably still codes some stuff, but the genius of the whole system is that he does not need to. As ideas are implemented, the community evaluates, and Linux integrates. No one can know for sure what kind of great ideas people come up with in the coming years, so no one can know for sure what that roadmap looks like.

  10. Re: Stable driver ABI by Anonymous Coward · · Score: 0

    Really? And how do they test this for me? Okay, so maybe I donate hardware. Do I really trust these guys to keep using it and really really test it?

    It's an argument that doesn't really work

  11. Re: Linus Torvalds is for Swedish cows. by duckintheface · · Score: 4, Informative

    Torvalds is a Swedish speaking Finn. That's why he says "planning 10 years ahead is not sane". The Swedish word is "klok" which can be translated as "sane" or, more reasonably.... "sensible."

    --
    "He took a duck in the face at 250 knots." -- William Gibson, Pattern Recognition
  12. only to dumfucks by Anonymous Coward · · Score: 0

    Desktop is so last millennium.

    only to dumfucks jizzing all over their iTablets.

    1. Re:only to dumfucks by umghhh · · Score: 1

      Maybe the GP was being sarcastic and we just miss the massive whooooooooosh sound?

  13. Re: Stable driver ABI by Anonymous Coward · · Score: 0

    If you see it broken you submit a bug and wait for fix.

  14. Re: Linus Torvalds is for cows. by shentino · · Score: 1

    That AC could have been a hindu for all you know, so don't accuse him of making an insult.

  15. Re: Stable driver ABI by phantomfive · · Score: 2

    Really? And how do they test this for me? Okay, so maybe I donate hardware. Do I really trust these guys to keep using it and really really test it?

    Maybe they'll miss something, and there will be a bug. That's going to be a problem if you try it yourself, too, except no one on the kernel team will even attempt to keep it updated.

    The problem you were asking about was the unstable ABI. That problem is a solved problem, as explained in the earlier post.

    --
    "First they came for the slanderers and i said nothing."
  16. Hans Reiser didn't plan ahead either by Anonymous Coward · · Score: 0, Interesting

    And where is ReiserFS now? If only Hans had done more planning and been less impulsive, his wife and his project might both be alive right now.

    1. Re:Hans Reiser didn't plan ahead either by rubycodez · · Score: 1

      let me stab your straw woman argument for you

  17. I don't think planning 10 years ahead is sane by fustakrakich · · Score: 2

    Well, if you're building Hoover Dam or the Golden Gate Bridge, or even the Brooklyn Bridge, you might want to reconsider that thought..

    Linux, on the other hand, can get along by just 'evolving' via the natural process of bickering...

    --
    “He’s not deformed, he’s just drunk!”
    1. Re:I don't think planning 10 years ahead is sane by LaurenCates · · Score: 2

      Indeed. I've been out of college for 15 years. In that time, I've watched so much change in the cultural landscapes of IT, CS, OSes and even general computing usage that any self-proclaimed expert who claims to be a great prognosticator for anything greater than a couple of years out is only selling so much snake oil.

      --
      Some people don't believe in fairies. I don't believe in The Patriarchy.
    2. Re:I don't think planning 10 years ahead is sane by mr_mischief · · Score: 1

      The Linux kernel is already built. It already meets Linus's needs. Linus is no longer an architect. He's the head of maintenance. Other people create new things. He chooses which ones don't break the stability of what's already there and merges them. IBM, RedHat, RackSpace, and others do the long-term thinking. Linus provides them a stable common foundation on which to build.

  18. Re:Stable driver ABI by Anonymous Coward · · Score: 0

    The driver ABI is unstable by design.

    Open-source your driver, and you won't have to worry about it: the kernel team will maintain it for you.

    Aaaaaand THAT is why Solaris, HP-UX, and IBM mainframes still live.

  19. Who cares? by Anonymous Coward · · Score: 0, Flamebait

    Seriously, who gives a flying fuck any more what Linus thinks? He's been somewhere between indifferent and openly hostile toward usability issues for years, and all the important work on a Linux distribution, in terms of making it more appealing to mainstream computer users, happens well outside the kernel.

    If this article said that the heads of GNOME, KDE, etc. weren't looking 10 years ahead, then it would be much more newsworthy (and troubling).

  20. Re: Linus Torvalds is for Swedish cows. by Anonymous Coward · · Score: 0

    To be honest, a better translation, especially in this context, would be "smart" or "wise".

  21. Re: Linus Torvalds is for Swedish cows. by Anonymous Coward · · Score: 3, Informative

    Torvalds is a Swedish speaking Finn.

    Correction:
    Torvalds is a Swedish-speaking Finn.

    (You forgot a hyphen.)

  22. DUH! by Anonymous Coward · · Score: 0

    Being a tactical is a natural trait of smart logic people like Linus. You can see this behavior in his emails/posts.

    Being a architect, road map visionary is daunting to those personalities. Too much [mind] work for a research for instance. From his comments, just be happy Linus knows himself.

  23. Re:Stable driver ABI by Anonymous Coward · · Score: 0

    Actually yes. Knowing that my RAID card will work on Solaris even after I upgrade it is a reason why I still use it.

  24. Security is also about design by cant_get_a_good_nick · · Score: 5, Interesting

    "What I see is that security is bugs,"

    Pretty much all Outlook viruses were design issues, not bugs. They designed a mail system which, on a OS where files were executable by extension, attachments from unverifiable senders had their extension hidden so you didn't know it was an executable.

    This was baked in design. It wasn't an execution bug.

    There are entire classes of bugs you could get rid of by certain design choices. Address space layout randomization helps a lot. W ^ X, or if you can write to memory, you can't execute it. These are not infallible (there's lots of webpages on how to get past ASLR) but if we design these things as more secure, we will be more secure.

    1. Re:Security is also about design by phantomfive · · Score: 2

      Many IoT devices do not have ASLR. Because of this (and other reasons) the IoT looks like a hacker's playground.

      --
      "First they came for the slanderers and i said nothing."
    2. Re:Security is also about design by Anonymous Coward · · Score: 0

      The IoT is a hacker's playground because they go directly counter to the idea of a secure central authority to authorize interactions in a sensible fashion. Kerberos would seem like some sort of solution, but it's almost a given that with a sufficient amount of devices and a little physical access to a few that you can enter and own the whole network through just one device. ASLR doesn't really enter into it at that level.

    3. Re:Security is also about design by BBCWatcher · · Score: 1

      I completely agree. I'm afraid that was not one of the smarter things Linus -- an otherwise routinely brilliant guy -- has said. Outlook is a good example. As another example, if you look at the (continuing) evolution of mainframes and their operating systems -- and you must if you want to understand IT security fully, competently -- one interesting bit of history is that IBM had to rewrite OS/360 MVT in the early 1970s. That rewrite (OS/VS2 Version 2 MVS, later evolving through several MVS releases into OS/390 then z/OS) notably included adding what became SAF to get the security design right, though there were many other security design-related decisions in that massive rewrite effort. IBM, with all its resources -- and with ostensibly a less complex base -- couldn't stomp out all the bugs in OS/360, and there were lots of ongoing security and integrity problems in OS/360. (The two are closely related.) The z/OS security subsystem, RACF, uses SAF interfaces, and so do other security subsystems/providers such as ACF2 and TopSecret. Yes, you can choose your preferred security subsystem on this most popular mission-critical operating system. The security architecture is that clean and separated, as it should be.

    4. Re:Security is also about design by Anonymous Coward · · Score: 2, Interesting

      Don't forget to separate the data stack from the execution stack.

      Many of C's buffer overflow exploits would be meaningless if local variables weren't stored on the same stack as return addresses. If you think about it, a return address is half of a jump instruction, yet it lies in memory that the application can write to, and so it is a violation of W ^ X.

    5. Re:Security is also about design by cant_get_a_good_nick · · Score: 1

      IoNUEPT

      Internet of Never Updated, Easily Pwn3d Things.

  25. Re:Stable driver ABI by 0123456 · · Score: 4, Informative

    If the ABI was stable, they'd be stuck with supporting crappy, crusty old crappy crap that some crappy old closed-source driver for some crappy old hardware requires, and unable to rewrite the kernel to provide similar functionality in a better manner.

    Like Windows.

  26. Re: Linus Torvalds is for cows. by Anonymous Coward · · Score: 1

    It's well-known who that AC is, as he forgot to click Anonymous a few weeks ago.

  27. Re: Linus Torvalds is for Swedish cows. by swillden · · Score: 1

    Torvalds is a Swedish speaking Finn. That's why he says "planning 10 years ahead is not sane". The Swedish word is "klok" which can be translated as "sane" or, more reasonably.... "sensible."

    Linus Torvalds has pretty thoroughly mastered English. I'm sure he said "sane" and meant "sane", with full appreciation of the meaning and overtones of the word and of the alternatives he could have chosen.

    --
    Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
  28. Re:Stable driver ABI by Mr.CRC · · Score: 1

    What about drivers that require $100000 worth of test equipment and an electrical engineer to test?

  29. Re:Stable driver ABI by phantomfive · · Score: 2

    What about drivers that require $100000 worth of test equipment and an electrical engineer to test?

    In that case, you'll probably want to test it yourself if there are any code changes. But by putting it in the kernel mainline, you won't have to worry about ABI bugs anymore.

    btw, with super-expensive hardware like that, the vendor usually manages the OS and upgrades, so the unstable ABI is not really an issue.

    --
    "First they came for the slanderers and i said nothing."
  30. The future of L*n*? by aNonnyMouseCowered · · Score: 0

    In the future Linux will be spelled Lennart.

    1. Re: The future of L*n*? by Anonymous Coward · · Score: 0

      Maybe. If so I will install Lenbart on my machine just for the joy of secure formatting the disk 1000 times afterwards to make sure it's gone. Then I'll install FreeBSD and live happily ever after.

    2. Re: The future of L*n*? by Anonymous Coward · · Score: 0

      Maybe after that you'll have time to attend night school and learn about adverbs.

  31. Is it really an issue? by dbIII · · Score: 1

    but there are definitely areas where the modern Linux kernel is a little too big for its own good

    It's too big to booting off floppy disks now, but that is very rarely an issue these days.
    It's still small enough to run on an Nintendo DS FFS!

    So I'm curious - where are those areas? Don't hint, give an example or several (since there are apparently "areas") and let us see if it's as ignorable as the problem of not being able to boot it from a floppy disk anymore.

  32. Here we go again by dbIII · · Score: 2

    Well, true, but AC said OS, not kernel

    By the textbook definition the OS is the kernel and the software distribution has all the userspace stuff. By the "beige box is the hard drive" definition the software distribution is the OS - the definition the Judge threw out as stupid in MS vs Netscape but has gained traction since.

    So the OS can have drivers to handle hardware used for accessibility and the software distribution can have applications that take use of them.
    The Solitaire game isn't part of the OS FFS.

  33. Re: Linus Torvalds is for Swedish cows. by greenfruitsalad · · Score: 3, Insightful

    maybe he isn't american and therefore doesn't hyphenate everything

  34. Re:Stable driver ABI by Anonymous Coward · · Score: 0

    What you really mean is that it's Torvald's choice and it's one driven mostly by arrogance. Torvalds isn't infallible.

  35. Re:Stable driver ABI by phantomfive · · Score: 1

    No, I agree with the decision for both technical and ideological reasons.

    --
    "First they came for the slanderers and i said nothing."
  36. Re: Stable driver ABI by sjames · · Score: 1

    It's worked out very well for companies that have gone that route.

  37. Never liked Linus after.. by Anonymous Coward · · Score: 0

    In all the time after Linus had been nodding, while verbally answering "no", on question if he had been contacted by authorities, I have disliked him for having goofed about something as serious as direct interference from a state authority and I don't trust the man.

    Maybe if you were to ask if he ever colluded with the NSA, he would both nod and answer no.

  38. Re: Linus Torvalds is for Swedish cows. by Anonymous Coward · · Score: 2, Funny

    maybe he isn't american and therefore doesn't hyphenate everything

    Wow, I was not aware that this was a stereotype about Americans. You learn something every day.

    But so be it. My countrymen! There is a plague in America today. For too long have we harbored it in our very bosom! I speak, of course, of the plague of over-hyphenation! Only through far-reaching long-enduring hyper-vigilance shall eradicate this thrice-cursed word-sickness!

    (Oh, also, I think there've been reports of actual plague in Yosemite National Park recently, so maybe watch out for that as well.)

  39. Re:Stable driver ABI by WndSks · · Score: 0

    If the ABI was stable in some form, v1, v2 etc you could at least develop a translation layer to keep old hardware working.

  40. Re:Stable driver ABI by BadDreamer · · Score: 1

    And why would you do that instead of moving the driver for that old hardware into the kernel? And for that matter, given how much old hardware is abandoned in newer versions of Windows, who do you expect would actually do this?

    Would you? If so, why?

  41. Linux does not have a design: by Anonymous Coward · · Score: 0

    Linus Torvalds have said that "Linux has no design and never will have". Linus says that trial and error in evolution has brought Homo Sapiens, the supreme species. And if trial and error has brought us humans, then that is good enough for Linux too. And that is why there are large parts in Linux rewritten all the time, slowly evolving Linux to a superior design beating all other operating systems. This is why Linus Torvalds does not look ahead more than the next release, because he expects Linux to have large parts redesigned then. And that is why the driver ABI is unstable, because it has no design and never will have. That is also why Linux Torvalds some time ago discussed whether a release should just be clean up of the code, because of the bloat:
    https://en.wikipedia.org/wiki/Criticism_of_Linux#Kernel_criticisms

    Personally, I think Linus Torvalds has that much knowledge now to make a good design (freezing ABI etc) now. So he should do it. All other OSes have stable ABIs except Linux. Linux has some 150.000+ drivers and a couple of 100 drivers released every week. There are only so many Linux developers, so they will never be able to update all of them when Torvalds changes the ABI in the kernel. This is a problem I think, as even largest OEM on the planet have trouble supporting Linux on their hardware because of unstable ABI. That must be fixed if Linux is to conquer the desktop. The driver model is broken.
    http://www.theinquirer.net/inquirer/news/1530558/ubuntu-broken-dell-inspiron-mini

  42. Re:Stable driver ABI by Anonymous Coward · · Score: 0

    What do you want one for? Please cite an example of a current closed source driver you're maintaining on linux.

  43. Re:Stable driver ABI by WndSks · · Score: 0

    I was thinking about drivers that are not open. A bit like how you can use Windows NDIS driver on Linux. Ideally I'd like this for Android so all OEM camera, sensor and GPS drivers etc follow one ABI so you can upgrade your phone without waiting for anyone to make custom ROMs.