Slashdot Mirror
Should Aunt Tillie Build Her Own Kernels?
DeadBugs writes: "Linux Weekly News is reporting on a new linux controversy. The inclusion of a Kernel Autoconfiguration program that would make it easy for almost anybody to build a custom Kernel on their computer. Eric Raymond supports this idea saying that this will bring Linux to a wider market. Those that oppose this idea mainly think that only those educated few should custom build their own Kernels. I for one hope this gets included if only to make standard installations and upgrades faster."
yes.
Why the heck is this a controversy? It seems to me that anything that makes good technology accessible to more people is a good thing.
I'd like to hear good arguments in the other camp, though.
I can see it now from the vendors "Compiling your own kernel voids any software support." Can you imagine trying to keep up with all the changes as a software vendor? So maybe as a test system, yes. Supported? No
"If you are on fire you can just stop, drop, and roll. If you fall into Lava you are just dead." - my 5yr old daughter
Why not throw out all the cruft your system doesn't need? It's just arrogance to oppose things like this, which make it easier for the masses to get performance boosts. I say put it in all the distros!
personal attacks hurt, especially when deserved
Am I reading correctly? Is this a debate over limiting vs. allowing certain behavior? What part of the Open Source philosophy got suspended while I was at lunch?
Let some distribution try this. It may take off, it may fail-- that's what it's all about...
davejenkins.com |
Everyone should have a chance to build their own without a lot of work.
Expect to see a lot of "This software only supported under the standard Red Hat v7.2 Kernel."
I don't blame the software companies one bit either.
This is a good thing, is it not? I think that this might make people who have some computer skills but aren't exactly willing to set up something like Linux a little more open to trying it.
This is the NFL, which stands for "Not For Long" if you keep making those bulls*** calls.
Get real. Aunt Tillie thinks kernels are something you find on a ear of corn.
Lost: Sig, white with black letters. No collar. Reward if found!
where do I sign up? I need to compile a two screen driver into my kernel but i don't understand how to the man pages weren't great in that regard and how about some kernel configs for a usb mouse wouldn't be bad!
--
WHO ATE MY BREAKFAST PANTS?
A Resounding YES to that..
Isnt this what we call a choice ? A choice to muck around with what we are given. Isnt that right of a hacker to mess around even when he is warned with dire consequences ? The reason Windows is so popular (reason why we hate them so much) is because Bill Gates makes those choices for you. Should we do the same too ??
Rapid Nirvana
...that this would just make things easier for a Linux newbie to break the OS. Then they can't fix it and are screwed. Then you lose a new Linux user because they don't want to feel stupid using their computer.
You are standing in an open field west of a white house, with a boarded front door. There is a small mailbox here.
No, Aunt Tilly is a retard. She's best off not knowing what a kernel is (besides something on an ear of corn).
IF Linux ever comes close to being comfortable enough for the Aunt Tillys of the world to use, modules should be loaded and unloaded dynamically via a friendly, Windows-esque GUI.
"Would it kill you to put down the toilet seat?" -- Maya Angelou
If people expect to make linux a desktop OS, then this will probably not fly. The sheer number of total borkages compared to the gain is not worth it.
If people expect to make linux a server/embedded OS then it *would* be nice if powerful things could be done without scaring off PHB's and NT admins.
Though of course it could be argued that PHB's and NT administrators are just as likely to screw themselves as Joe User...
- Get the source
- Probe for devices
- Configure, make, make install
more people might consider using Linux because one of the major hassles is removed.Overrated / Underrated : Moderation
Most Linux users are already familiar with the caveats and reprecussions of customizing your kernel. This kind of tool would just make it easier to get to.
There aren't all that many "casual" Linux users. That market is dominated by Microsoft. And if you've deployed Linux to a work environment, chances are you won't allow a tool like this to be used, because you'll probably want to lock down the configurations (making your life as a sys admin a lot easier).
Assuming Linux continues to proliferate to the consumer market, I still wouldn't be worried about people tinkering with their kernel too much. Most people, especially at the "average Joe" level, don't understand the inner workings of their OS. Heck, most of them fear their OS and assume that they'll break something if they tinker with the OS's inner settings. I wouldn't conclude that simply because the tool is there that most people would be interested in using it.
My sigs always suck.
Those that oppose this idea mainly think that only those educated few should custom build their own Kernels.
This I just do not understand. Should that attitude prevailed when it came to PCS or ISA cards pre Plug and Play days when you had to be an expert and getting interrupts set correctly or your system hung (and yes I realize problems still happen with PnP, but its still a billion times better than the old days). What an elitest attitude.
*Make it easier*
Should we get rid of the './configure && make' cycle because its too easy for those of us who don't know the ins and outs of the compile cycle?
(Man, am I in a snippy mood today or what!)
And that's all I want.
Can I build a kernel with only those pieces I need to make my Pentium 233Mhz a kickass firewall AND THAT'S ALL?
If so, I'm all for it.
Especially if I don't have to learn the whole kernel to do so.
You shouldnt have to rebuild your kernel at all.. It's about time Linux, if it wants to be in the desktop market, moved toward a more modular design where adding things to the system is a matter of loading a module.
I really can't emphasize strongly enough that I believe that if Aunt Tille has to build her own kernel, we have much bigger problems that Eric's autoconfigurator will solve.
Linux isn't intended for a certain person, its a conglomeration of problems that others have solved and been good enough to share the solution.
The reason that some would object to such a tool would be the pride that comes with being a 'true linux hacker'. They take pride in knowing that they just don't load up KDE and surf the web, they are in there tweaking the nuts and bolts of the OS.
Well, this pride is a flaw. Do the best you can and try to help others who aren't as far along as you. If you are tweaking the kernel and downloading every patch and upgrade just so you can puff out your chest over what a linux god you are, you're only fooling yourself. Want to really be a guru? Help someone else. Show a newbie how to patch his kernel, or maybe even write a program to make the process easier. Wait..someone already did that. Well at least you can tell them about it.
-- When a fool hears of the Tao, he will laugh out loud.
Most of the kernel configuration is simply a matter of which drivers to include in the kernel instead of as modules. Distributions put most of the stuff out as modules, so all that a kernel autoconfigurer would do is notice which modules are loaded and build them as part of the native kernel. The advantages of this are minor--slightly better memory utilization, no need for initrd.
On the other hand, there are some areas where an autoconfigurator would be handy. That's when determining which chipset features/bugs to compile for. Hopefully this project will focus on the areas of configuring that are more complicated than (y/M/n).
i remeber setting up my first Linux installation on a laptop and the hell that ensued when trying to figure out how to put together a custom kernel that would support PCMCIA. Yes, i did learn a lot about the kernel, the Linux boot process, compilers and all sorts of other stuff. Problem being, the average computer user has no desire to do any of these things. This is why the average user won't use Linux. If the goal is really to get Linux on more desktops, we're going to have to see WAY more wizards and configuration tools.
I think the beauty of linux is that I can manually edit config files to my hearts content, or I can fire up Linuxconf and do the same thing.
No one forces me to do either.
Choices are good.
Where was this before the question was asked?
:-)
Always nice when Slashdot can answer it's own questions
No flame intented, but what kind of elitest attitude is that? Linux is never going to gain mainstream acceptance by the average 'Aunt Tillie' until she can upgrade/install things easily. Given that most 'Aunt Tillie' types can barely do it on their Windows systems, Linux needs to be making every effort to become extremely user friendly. Deciding for someone else what they should be able to do easily sounds ridiculous to me...
"Compiling a kernel" means collecting seeds for aunt Tillie.
This seems like a bad idea if it's a desktop icon or an easy to access program. Let aunt Tillie mess with the kernel and watch how fast the computer will grind to a halt, heck, we're lucky if aunt Tillie knows that a computer mouse is not a rodent !
- sigs are for wimps.
Well, if it wasn't NECESSARY to recompile for every little thing you want to do I'd agree. But any time you want to do ANYTHING even remotely out of the ordinary you need to recompile. Got a new Wacom USB tablet? Recompile. Want to use a webcam? Recompile. Want to do anything interesting? Recompile. Until things are a little more modular (i.e. the kernel can change without breaking every module) ALL users are going to need to recompile.
How much easier can it be?
tar -xzf linux-2.4.17.tar.gz
cd linux
make xconfig
make dep
make clean
make modules
make modules install
...and make it boot...
I mean, If you think that is hard you probably wont be able to give any useful instructions to a kernel configuration program at all... Maybe not even know you need a new kernel...
What is nice with linux (compared to Windows) is that very few things happen "behind your back". The system does not change itself. I find this very comforting.
And modern distributions tend to make it quite easy anyway... I installed Redhat 7.2 from isos today for the first time in over a year. All hardware was autodetected and worked without any tweaking at all (then I felt like compiling an own kernel to play DVDs well, but that another history)
While 2.4's module support is excellent, and modularisation is become more and more prolific throughout the Linux architecture, there are still several important features which need to be excised from the kernel core and made available as runtime modules. Trivial features such as APM support, SMP and Unix sockets shouldn't require a full recompile to activate. Why do we insist on prolonging the life of "make config" and its brethren when we could very well do without it altogether?
Although I use FreeBSD, building a custom kernel is good for Linux or any of the *nix's. You can get rid of device drivers that slow down the boot process, and you can tailor optimization for your specific uP. That will be especially true if we ever get a gcc that has decent Athlon optimization. I'm also told that taking out the plain Jane i386 support speeds up things considerably.
My first reaction was, no way, don't allow everyone to do this. But after reading posts like yours I realize that my bias is based the the incredibly poorly documented and ill-conceived kernel configuration programs available in the early days of Linux. I tried to use one, and despite being a computer engineer, I managed to screw up my distro.
If a kernel configuration program was more like an "options dialog" that actually made sense and worked, then the user doesn't have to know that they are actually recompiling the fundamental execution unit of their operating system. In computers, appearance is reality.
personal attacks hurt, especially when deserved
Why not? It follows the basis of Open Source Software in that anyone can build off/with the code. The same thing has been done for programming. There are many languages out there that cater to diffrent groups and ideals but they all do one thing in common, let anyone willing to learn a little bit know how to make their own programs.
If you want zero-effort working systems, the distributions are the way to go at the moment. If there is anything that can be done to help that unfortunate situation, I am all for it.
:-)
... I actually *like* learning this sort of thing, I just prefer to choose the time of my enlightenment :-)
I know lots about unix in general and linux in particular. I've written kernel drivers. I've designed embedded CPU's and PCI plugin cards. I am generally regarded as being very technically minded
The flipside: I have also been mystified as to why one of my (admittedly more esoteric) kernels just gives up the ghost at inopportune moments. It's almost always my fault, and I almost always learn more from the mistake, but it's sometimes a non-timely learning experience
In short, why would you want to make it difficult ? Use the time you save to solve other problems instead - ones that someone else hasn't kindly provided a solution for...
Simon.
Physicists get Hadrons!
What's the problem if Joe Rube decides to build a new one? I mean, if he smegs up because he didn't ask Jane the Ubergeek to help him, all he has to know is to boot the prior kernel and no damage done for the most part. If he's using Mandrake, he doesn't even need to worry about how the LILO prompt works as he'll be able to select the old Kernal from a list at bootup. Force a timeout for LILO and keep the old Kernel and you're ALMOST Idiotproof, IMHO.
Devo Andare,
Jeffrey.
Time Lord, Dark Horse: The Techno Mage of Gallifrey
...introduce me to Penelope?
I thought the geek-girl was just a myth created by lonely geeks, day-dreaming by the light of their Amigas.
I have found there are just two ways to go.
It all comes down to livin' fast or dyin' slow. -REK, Jr.
Ponder:
Should Aunt Tillie Build Her Own Kernels?
Should Aunt Tillie Install Her Own OS?
Should Aunt Tillie Install Her Own Applications?
Should Aunt Tillie Run Her Own Applications?
Should Aunt Tillie Produce Her Own Documents?
Should Aunt Tillie Think Her Own Thoughts?
Configuring a kernel from source simply ain't that hard. That's why so many pimply-faced youngsters do it. For Eric Raymond to characterize the prospective users of his "make config with training wheels" as something that is needed to get kernel updates to females (his examples are Aunt Tillie and Penelope Power User) is just sexist bullshit.
*****WARNING. USING THIS TOOL CAN SCREW UP YOUR COMPUTER, BIG TIME. IF YOU WANT TO USE IT, DONT BE MAD AT ANYONE BUT YOURSELF IF YOU FUDGE THINGS UP*****
.. I mean, seriously. The only thing you could end up with is some fucked kernels (who should get along just fine with the fucked registries) and some users who will learn and be cautious, and end up having a better understanding their computers.
If MS can include regedit, you cant tell me that we can't inlcude autoconf
"Old man yells at systemd"
This is what the purpose of MODULES was suppose to do .. or lack thereof Compile the entire beast and then have what is needed get loaded up when it is needed or installed.
.. no.
.. yes.
Should Aunty be compiling kernels
Should Aunty be able to install new pre-compiled modulated kernel with a few clicks of the mouse
-- Knowing too much can get you killed, but knowing who knows too much can make you rich.
You don't need to tinker with the kernal... This old lady proberly uses Windows ME or Mac OS 8.1... I don't see here hacking here kernal, nor do i see here wanting to...
"You win again Gravity!" -Futurama (Zapp)
...doesn't everyone build a custom kernel? I've been using Linux for years, and I always assumed that the prebuilt, "Christmas Tree" kernels were just to make installation easier. People actually run with those things? Heh..
I'm sure a bunch of others will post similar sentiments, but...
The average computer user - who knows nothing about the pros and cons of various choices - would gain nothing from having this ability. Now, it would be nice to include this program, so that it does not take as much knowledge or effort to build a custom kernel if one wants to, but a significant amount of attention should be paid to the defaults, because those are what most people in practice will use. Even the knowledgeable: if I know Web services like the back of my hand, but I don't know about setting up email servers, I might install an insecure version of BIND if that was the default (say, because mail servers were listed in a dropdown with "None" at the end, and the others sorted by name and version, placing "BIND 0.9" at the top of the list).
Most users do not, and do not wish to, know how the guts of their computer works, no matter what. They want black boxes. Building one's own kernel - unless it is done automatically, to the level that it isn't really one's own kernel but a recompile of a standard build - goes against this.
some cocky hackers will think its stupid prolly, but it sounds like a brilliant idea, it might put the less educated users (like me) to learning more about the kernel
But I don't think "Aunt Tillie" should accidentally come anywhere near a kernel: Users should not care about kernels because they have to, but because they can. That means that most hardware configuration tasks should be accessible without touching the kernel, including installation of new drivers. So include lots of warning signs -- optimally a normal user will never have to log into his box as "root" except for installing new software with a graphical apt-get like tool.
I mean honestly, what claim can linux hold over windows if not that the availabillity of the source code allots the user more freedom? This is, as far as I'm concerned, what linux is all about. I am totally unable to understand any argument against making one of the most important benefits of linux more accessible to a wider market.
If she really wants to consider herself a hacker...
This puts us one step closer to "Automatic/Random Update" feature found on XP!
Yes this type of tool is a good thing, as long as alternatives exist, that's the point of using Linux isn't it?
John, I'm Only Dancing!
-- Erich
Slashdot reader since 1997
Sure, we all want to build our massively customized kernels for server use, etc. etc.
Take 2 boxes, one running 2.2, one running 2.4. Throw KDE on there, and I am willing to bet that the average Aunt Tillie can't tell the difference which one is faster/better.
The kernel as packaged by distros does a good enough job of running the typical desktop system that Aunt Tillie uses to browse the web. Leave the custom build stuff for the experts.
even if all this will do for me is save me 5 minutes while compiling the newest flavor by not having to have to search through the list of chipsets and network cards.. I am for it 100%. Heck, that 5 minutes might be just enough to make me not feel tired in the morning when I decide to compile my kernel in the wee morning hours.. I can get to bed at 2:55 am insted of 3 ;)
:)(smile)
put it as an advanced option... As long as you give a warning - "This could severly f***-up your computer if you don't know what your doing."
I mean come on here - I was trying to get my funky (old) linux laptop to tell me what the signal to noise ratio was on the wireless card, and to get it to work it turns out I have to recompile the kernel, just to set a single flag. WTF!? Install the source, make and compile an entire operating system, all just so I can get a stupid little program to run? That's more than a little counter productive, and a royal pain in the ass. It was literally easier to install win95 and use the default dos program to monitor the link quality. What a pain in the butt - something's smells rotten there.
Well if I have to explain all the kernel parameters to auntie, like all the networking modules "no Auntie, you don't have any IPX network game installed, well anyway you don't play network games, you don't need it" and "on your machine you don't have any SCSI card but you should use some SCSI module if you want to burn CDs" and the "no, no SMP... I won't explain to you what it is but you are certainly not running a biprocessor machine"
My questions are: will the module and kernel configuration description will be written in Aunt Tillie english? does some options will be masked to simplify it?
--- Bouh !!! ---
I think most people, even if they could, simply wouldn't bother recompiling the kernel. Many people barely know how to send an email: and asking questions like, "Would you like the USB to allow for IEEE support through your firewire hub?" (or, whatever) would only confuse people.
Besides, most people still think kernels are those things that you pop to get popcorn!
Make it simple, make it intuitive, make it make sense: and you have a product that people will use. Forget recompiling the kernel: have the kernel poll the hardware and reconfigure itself based on the hardware it's run on.
I guess the Linux Support/Services market might see new demand from a general deployment of a tool like this, but unless it's part of the installation, I cant imagine alot of folks using it who couldn't already build/update/install a new kernel from the shell prompt or one of the tools provided with the source, like menuconfig or xconfig oh and sorry for the runon sentence.
Why in the world would a *user* install a kernel. Yesterday, I tried to explain a colleague of mine (we both work in the Geographic Information Systems industry, and he's an NT system administrator) what a kernel is. The user doesn't care if he can get another extra 0.0000000002 bogomips with a rebuild of the kernel. The user wants the software to *work*, just browse the web, type some documents.
And besides, why has the kernel to be recompiled? Why can't I just use my moduler kernel? apt-get install kernel-image-2.4.18-686-smp is a lot simpler for me than downloading a kernel source, untar'ing it, and so on..
And what about hardware swapping? What happens when my aunt tillie's IDE controller dies and she replaces it with another.. OOOPS.. You compiled a kernel for a specific piece of hardware, now you can't boot anymore..
Modular kernels are great.. Let's use them...
(*): a real world user is not an average slashdot reader who has a pretty good knowledge of software.
xer.xes -- 4181
Well... why not.
Grandma can built her own website using wysiwyg editor. Yes, it's create a net pollution with ugly stuff on the net, but, this bring grandma more closer to the net.
Giving her to opportunity to built her own kernel would bring her closer to her computer.
They'll still get the standard binaries, right?
That means they'll have to go out of their way to tweak with the kernel. It should be easy to throw up a disclaimer. Think of this: even Micorosoft includes tools for editing registries, with the standard boilerplate.
When in doubt, have a man come through a door with a gun in his hand.
I dont even know what a kernel is. Why do I need to compile my own? Because otherwise my printer won't work? I'd say that an OS needing that sort of work is not fit for the masses yet. Maybe great for those who do understand (and have spare time to spare)
The Bigger The Headache The Bigger the Pill
make xconfig
make dep
make modules
make modules_install
make bzImage
make install
reboot.
trained monkies do more than this when they fling their poop at you.
I suppose I'd have no trouble believing this. I'd still like to know what the requests are and why they are ridiculous.
Aunt Tillie shouldn't have to
But.. she does. We don't have runtime autoconfiguration that works in every case. If an autoconfigurator is easy to build, and won't impact the people working on runtime configuration, then why stop them from doing it? My computer should read my mind (or at the least, the pointer should move to the thing my eyes are looking at) but I'm not going to tell people to stop working on improving mouse support.
The autoconfigurator is bound to be an imperfect job
True enough, but this is true for runtime autoconfig as well.
The kernel people are already drowning in bogus bug reports
Kernel bugs are reported via email on the mailing list. This is described in marginal detail in /usr/src/linux/REPORTING-BUGS. Furthermore, it begins with the following dubious paragraph:
What this document highlights more than anything is that kernel developers are drowning in bug reports because linux kernel bugs are reported in an informal format on the mailing list. Get a proper bug tracking system and it will be much easier to keep track of real bugs. This should be done regardless of whether or not we make "kernels for the masses". I hadn't heard about the bug report problem until you brought it up, and it's frankly amazing that it hasn't been addressed in this manner already.It's rare that you're presented with a knob whose only two positions are Make History and Flee Your Glorious Destiny.
My emotional reaction is Noooooooo! Not because I'm elitist and arrogant, I can always find another thing to be arrogant about, ("You use the newbie tool to rebuild, loser" ) but I don't want to field a hundred questions each from a hundred people. I don't want my mother calling me and asking me if she needs iRDA modules, or constantly answering questions at the bar from people who probably have no need to get into that stuff. It's bad enough now fielding questions about windows... I gotta get this shirt from thinkgeek.
I'm the big fish in the big pond bitch.
That's such an elitist viewpoint. We think quite a lot of ourselves when we know how to "compile a custom kernel", don't we? It makes us feel like ameteur mechanics the way we so smoothly "trick out" the insides of our computers.
Plus, since "normal" computer users don't know how long it takes to compile a kernel, the configurator can build everything into modules, who cares if they never use the drivers from 20 year old hardware ;-)
Bye!
(Tillie? Tilley? I foget...)
I don't think the biggest benefit of the new autoconfig module esr added to the new kernel build system will be for little old (non-tech) ladies. From what I understand of esr's latest achievement, it could be coupled with a distro's installer software to simply weed out un-needed drivers, to clean up the distro so it matches up more with the machine it's running on.
I think distro makers might make use of this way more than Aunt Tillie (sp?), because IIRC esr's announcement on lkml, what he described is a form of kernel/OS self-maintenance. For some reason, I see this completing/augmenting Red Hat's (?) kudzu.
I do not recall, though, if this system can keep track of patches applied to a kernel (pre-emptible, real-time, etc.). I imagine a scenario where a distro maker could provide a list of patches to apply to kernel rev. x that this new subsystem could use to ensure the end-user has a kernel always in line with what his/her distro is supposed to have. If user upgrades to new version (say downloads a new version of the stock kernel because it's quicker to download from www.kernel.org than from distro X's web site), the autoconfig system could auto-detect this, check with the distro maker's website and fetch all relevant patches that applies to this new kernel rev.
I haven't fully thought this out, but somehow, I feel we are getting distracted by Aunt Tillie and loosing sight of bigger benefits this could bring. Maybe I'm wrong, but I have a gut feeling this might be *big*.
It would be great if anyone could build a custom kernel.
Imagine this... Let's say it's 5 years down the road, and the hot new computer is the 72 ghz Apple Pentium G7 with 64 gigs of on-chip ram. Hard drives have been totally eliminated because new, memory based permanent storage technology has been invented and proven over the past 2 years. An entire meg can be recorded in under 1 microsecond. The only remaining mechanical component of a computer is the standard Glass-RW drive (the 2 terabyte recordable successor to DVD), so the whole computer is now a small single board, and most of the electronic hardware is inside the main processor, an inch square in size. In fact, the plugs on this board take more room and cost more than the computation hardware.
Now imagine that a build world takes 4 minutes to complete. Here's how installation of FreeBSD 9.8-RELEASE takes place. (Yeah, I know this was a Linux thread.) You pop the Glassdisk in the drive, choose a few options, and all your software is configured, optimized, checked for security vulnerabilities, compiled and installed within 2 or 3 minutes.
In order for that to happen in 5 years, Granny needs the ability to custom configure her own kernel right now.
As long as the old-time hackers and wannabees can do it the Hard and Right Way <tm> I see no reason to limiting others from trying out the wonderful world of kernel compilations.
If the not-so-tech-savy compile a kernel that doesn't work they could always go back to running the one from their distro. They should be told to always keep that one as a backup.
The negative karma that surrounds recompiling your kernel.
:-).
I recompiled my kernel the very first day I installed linux (which was Slackware loaded off of floppy disks that I downloaded over a 28.8 modem - that was a weeks worth of fun
Sure you screw it up every now and again - but if you just read the damn help things and have a little bit of sense it isn't that difficult. And its not like all of your data suddenly goes *poof* when you do screw it up. It just means you need to boot off a recovery disk, and retry it.
make menuconfig
make dep
make bzlilo
make modules modules_install
I don't know what is so damned hard about that.
Derek
Some people just aren't happy unless they start an arguement.
And that issue is timely turnaround of getting updated kernels from the commercial stewards.
;-)
For example I've given up on Redhat ever providing a newer kernel for RH 7.1 than 2.4.9. If I wanted 2.4.16 it was clear I was going to have to build it myself.
The single biggest reason that dear old Auntie needs this ability? Because dear old Auntie doesn't want the dear old script kiddies leaving porn on her computer when she's not using it, that's why. Building newer kernels (with their security fixes) helps make sure this doesn't happen.
But building kernels is *hard* to learn. Yea, sure make xconfig makes it easier. But, hell, I have a Comp Sci degree, wrote my own OS in college and *I* had a hard time figuring out now to build a kernel with the same base feature set as my stock kernel with RH 7.1. So if *I'm* left scratching my...uh.. head.. I'm sure dear ol Auntie is going to be totally overwhelmed, and why should she be? Doesn't it benefit *everyone* if people's computers aren't being exploited by cyber-vandals???
If we're going to get linux distros in more hands it needs to be more available to people. Even if it's only an 'easy config' or 'auto config' for a limited option set to allow the less adept to build 'stock variant' (for lack of a better term) kernels. Over time these people *can* learn the more 1337 t00lz out there like make xconfig...
Polymorphism -- It's what you make of it.
Did anyone else see the article about the Apple II IDE/Compact flash interface?
This is much more exciting news than an autoconfig'ing kernel. Where did the article go?
Know what I like about atheists? I've yet to meet one that believes God is on their side.
Should there be a pop quiz before this program in order to enable it only for /. ers ?
look my sig changes!!! nrrt mf oci jdabi.o!!! z..a ir kot gh-ntbk{{{
Under capitalism man exploits man. Under communism it's the other way around.
I wish I could use some tools like that on Linux to learn faster. Of course you have all those conf files, but having a single app to mess around would be so cool. Just click here-and-there, reboot, crash, restore, reboot, play again... It's much easier that way to build a mental image of your system. Then, when you know everything by heart, you can start using that keyboard again (and pretend to be part of the 31337 since the beginning). Just my .02
Only if Aunt Tillie did a complete install...
Aunt Tilley bought her PC and it came with Windows XP preinstalled, has a web browser preassigned and an email program automatically assigned to her preassigned ISP built in already - so guess what..
she doesn't give a crap and wouldn't know what the hell all of you geeks in here are talking about.
A kernel to Aunt Tilley is a piece of corn... Jeez...
Talk about being out of touch....
guns kill people like spoons make Rosie O'Donnell fat.
i can smell the buttery goodness now.
Granda's answer to compiling her own kernel: "It's not near as good as my kettle korn".
I compiled my kernel 2 weeks after I first started using Linux (94 or so). I followed the instructions. make xconfig (GUI), read the HELP. It was pretty easy. The help even tells you whether you should fsck with the setting or not (If you have no idea what this is,leave it alone)
then do make dep clean bzImage modules modules_install. The kernel is almost all modular now, so you can make almost everything a module (don't know if you have a Goober Radio Card? Make it a module) I am surprised (especially with the Linux users here) that Linux users think the kernel is hard to configure. It's like overclocking. Sure, you CAN, basically just move a jumper, but unless you know what you're doing, you SHOULD'NT. 'Nuff said.
If you're not a Liberal in your 20's, then you have no heart.If you're still a Liberal in your 30's you have no brain.
Yeah, but what happens with 'make xconfig'? Does that not pop up a menu with about 20 different sub-menus, each with menus below that?
How is that easy?
(and you forgot make bzlilo or make bzImage)
-W
As much as I like the fact that the kernel can be so easily customized in many ways this has hurt it's adopted by the general public.
The fact that I might have to recompile my kernel just to include support for a new piece of hardware is a real pain in the ass.
Any chance that this would become a thing of the past becomes less and less the easier the rebuilding of the kernel becomes.
As much as I like linux, WindowsXPs incredible plug and play capabilities are always going to be better than recompiling the kernel no matter how easy it becomes.
Pro Coder: "So Aunt Tillie, how would you like to compile your own custom Linux Kernel?"
Aunt Tillie: "What the hell are you talking about?"
Pro Coder: "You know, compile a custom Linux Kernel, so you can have a very customized OS."
Aunt Tillie: "Why would I want to do that?"
The conclusion we draw from this interview is: your average user doesn't have any idea what a Linux Kernel is and that they don't need a custom kernel, at least not yet
...when elite hackers said that only elite hackers should have Linux, and all of these "Red Hat" guys are polluting the user base. They are, of course, full of shit.
The whole point of Linux is having a stable and friendlier version of UNIX that is GNU and doesn't have any ties to MS. We now have Average Joe User with their own copy of Linux/X and they are using it just fine. Why should we limit ourselves because we need to do it the "old fashioned way"? Let them (and us) have a easy-to-run auto-config script for building kernels. Are we going to delete our "make menuconfig" scripts and tell everybody to replace it with "vi Makefile", just for being elitists?
Personally, I think these are the "10 miles in the snow, both ways" people, who still believe that the best way to configure PPP on Mandrake is rolling your own scripts. (Uhh..."netconf"...duh!)
Zodiac Survey
Personally, I am of 2 minds on the subject: if there is a demonstratable speed/performance increase vs. modules using this new "autokernel" system, then I think it would be worthwhile.
However, if I have to support the box, I don't want Joe User hitting "Get new kernel" and then having data corruption problems because that bug accidentally slipped into the new kernel.
If distro's set it up for root-only (or perhaps a "power user"-like setting), then that's great. But, I think it also needs some sort of safeguard - similar to RedHat's updater, by which Red Hat has compiled custom (and presumably tested) releases of software. Again, I really don't want to have to try and support some user's data-corrupting kernel setup. So, perhaps a "Windows Update"-like central repository of "good" kernels/options may be in order in addition to this??
Glenn
You are 100% right but, you stop short of the big picture. While you are correct that Aut Tille should not have to build a kernel, ever. The fact is that Aunt Tille *should* be able to take advantage of the later fixes and upgrades.
In Microsoft's world Aunt Tillie can click on setup.exe and she is updated, patched or upgraded, what ever. But, in the Linux world, Aunt Tillie must wait a year or, likely, more for the next release of her distro (most distros are still shipping 2.4.2 while the present stable is 2.4.17) and even that upgrade is not so simple.
Runtime configuration is a major advantage, as you said. But, it is still not enough. Aunt Tillie should also be able to upgrade her kernel and take advantage of the lastest developments with just a click or two. Perhaps a kernel rpm and runtime autoconfig.
If it is properly thoughout and written this will definitely make Linux installation/administration easier. After all how many of us actually still compile any UNIX source ourselves when installing a new system? Its just the natural trend of evolution. As long as we are still left with the option to go in and manually edit the kernel config, I cant see how this could be negative.
Just make sure that there is an ironclad *required* boot disk step in the process so when system.map is inevitably overwritten, and LILO brings up screens of 8s or #s, Aunt Tillie can log back in.
Having two or three LILO prompts of backup kernels that can't be passively overwritten would also be a good idea.
I don't think it is smart to have Aunt Tillie care about kernel recompilation, even if almost everything is done by an autoconfigurator.
But what about auto recompiling the kernel while aunt tillie's screensaver is running ? The kernel could collect usage and performance data while it runs and automatically make kernel configuration changes that suits the usage of aunt tillies kernel.
Modules are a nice thing, but there is a small performance lose when you use modules. Why not ship linux distribution kernels with almost everything compiled as a module and then let an autoconfigurator compile an custom new kernel every few weeks until the kernel gives aunt tillie near optimal performance ?
ISA Cards are a problem, but there aren't almost anymore ISA cards in the aunt tillie systems out there anymore and normal distribution kernel have the same problems, they also need to try find all the isa cards in your system and normally it doesn't work that bad.
Jan
http://www.aunt-tillie.org
Blame Alan
Autoconfig at runtime? Hmm, I hate the bloat of prebuilts, but given that future versions of Linux are going to replace XP, perhaps it's time to think less about bloat, and require memory, like the other guys (it'll already be there). However, Redhat and others install these days with just about everything turned on, (and detects automatically lots of peripherals) so perhaps the reason for this tool is to reduce bloat or allow Auntie to tune her kernel, or add a very special device driver, if it's well documented.
"Compiling a kernel is hard and should only be done by the select few."
Thats sounds like what programmers used to say when they wrote things in machine language.
The goal is to make it easy enough for anyone with a brain to do it. Hell they don't even need to know that they're recompiling a kernel.
"Oh you want to do that? Ok give me a sec and then I'll reboot and you'll be all done." *compiling*
Thats the goal. The user doesn't HAVE to know just 'what' they're doing.
Do you really know exactly what that for statement you just wrote compiles to in machine? Do you care? If you want Open GL do you have to know ANYTHING about the kernel?
Sure it HELPS to know these thigns but for the end user it's not a must and should never be a must.
In Soviet Russia, the television watches YOU!
Well, think of it this way. The registry editor (two of them actually) live in Windows, but you don't have the masses actively going into it to tweak settings. Yeah, you hear some horror stories about people going into the device manager and mucking stuff up, but in general, most people turn on their computers, log into AOL, and surf. I don't see any reason why having a nice easy kernel editing program is any more an issue than having regedit or DELETE on a Windows system.
But I doubt it's going to make any difference to Aunt Tillie that she can compile her own kernel in choosing Linux over Windows. Either it runs AOL or it doesn't. Either it runs Master Cook or it doesn't. Either it runs Family Tree Maker or it doesn't. You can say until you're blue in the face that there are compatible programs, but all her friends use Master Cook and she "just can't swap recipies without it". Linux on a desktop? First you gotta get past Aunt Tillie and her recipies.
-- If god wanted me to have a sig, he'd have given me a sense of humor.
There are quite a few posts of the form: make xconfig && make dep && make clean && make bzImage && make modules && modules_install && reboot
Depending on your hardware that's a 5 - 30 minute process. Most of the time spent on kernel configuration is in make menuconfig (xconfig -or- config). This is where you make decisions on what drivers/features you will need/include. Keep in mind that the above won't always boot the machine if one of the decisions is made incorrectly.
When a distro is first installed, it always has a working hardware footprint by definition. Running through the options/features that are working for the install process to create a
www.dedserius.com
VB != VisualBasic
Well, on one of my systems, EVERYTHING is a module (Mandrake 8.1). No recompilaton necessary
If you're not a Liberal in your 20's, then you have no heart.If you're still a Liberal in your 30's you have no brain.
If you were to implement such a system, along with some sort of QoS monitor (a la Mozilla), you might be able to analyze the data from all the different builds and more easily figure out which kernel modules were interfering with each other and causing the instabilities in 2.4.x (the Kernel of Pain).
Anything to make Linux more complicated, cumbersome and fragmented! After all, it's the best way to set it apart from the competition... :)
Easier to use tools are great.
I just hope we don't start designing things such that people say "oh, to do that just reconfigure your kernel with the foobar option". Feature sets should generally not require kernel recompile imho. For a long time, this was a UNIX weakness.
If we can avoid this (which is after all worse than the old "reboot NT to configure something"), I'm for it 100%. I'm not saying that you have to recompile the kernel much nowdays (I had to once to get an unsupported Ethernet driver working), but kernel recompile gets really easy, I'm nervous that people would start to rely on that way of doing things. Which would be bad.
--LP
is to understand it completely to begin with.
Personally, I don't care one bit if this happens or not. As long as I am not forced to use it, it doesn't matter. It isn't my responsibility to decide how other people should use their computers. Although my suspicion is that this will suck supremely. But whatever, give the newbs the device to shoot themselves in the foot with.
----
All of whose base are belong to the what-now?
It's hilarious to me that this is even being considered. Linus' ludicrious obsession with having a monolithic kernel is fully at odds with having a stable and flexible desktop system on par with MacOS X or Windows. Loadable modules or not, this is a fundamental weakness in Linux and will be it's Achilles' heel in the desktop space.
Good luck with all that. Any sentence that includes the words 'compile' and 'kernel' CAN'T be in the manual for any OS you expect to have significant popular support.
. . . well, not really. But, isn't building your own kernel part of what makes Linux so usable and unique? I've only recenlty become familiar with building a kernel. I was really surprised to find out how easy it was and how safe it was. I think that if you give users a tool that will make them more comfortable with recompiling the kernel, than more people will realize how easy it is to customize their OS. My sister uses Linux every day at her job (network engineer), but she's scared to death of compiling a new kernel.
Any utility that makes the non-coder linux-user more comfortable with building their own kernel is a good thing! It won't bring more users to linux, but will make linux more usable to the casual user.
The problem with the whole idea of computer elitism and the idead that only certain people should be able to try something on a computer is what always gets companies in trouble.
In the early days of computers, only elite people with technical knowledge bought, or indeed could afford a computer. Apple brought it to the average Joe and 'lo and behold computers took off.
Then Apple got elitist with it's GUI, providing only to apple users (I'm not bashing apple here, I am a mac fan). That was a mistake, soon realized when M$ released their GUI to the masses, and took the market by storm.
Now, here's a chance for the concept of creating custom kernels to come to the masses. I say give it to em, let's see what Uncle Fester can come up with.
T Money
World Domination with a plastic spoon since 1984
If I am not a kernel man, then not will I a kernel can.
Linus is a kernel man, so yes does he a kernel can.
_________________
EBAY SAFETY TIPZ!
Aunt Tillie doesn't need this. But, as a computer consultant and VAR, I need the ability to easily make these kinds of changes based on what my customers need.
Sure, I can do this myself the old-fashioned way. But this is the kind of thing I prefer to delegate to someone with a lower billing rate so I can focus on the things that really bring in the bucks. It is easier to train someone to use Eric's AutoConfigurator than it is to explain make files and such...
Jack William Bell, who likes the KISS method in most things.
- -
Are you an SF Fan? Are you a Tru-Fan?
It can be talked to death but everyone knows the bottom line - I know what to expect with MS Windows or an Apple OS out of the box. I don't know what to expect with Linux and I have no idea whether I'll be able to resolve any problems I have, and if I can't suddenly I have a very expensive paperweight on my desk.
(Please don't flame me about how Linux is just as good or better or tell me all I need is to get this or that. I don't dispute this. I say only: I've been using Apples and Windows PCs for 16 years, I know what to expect, I know will be able to make the computer do what I want it to. I don't know this about Linux. I don't know what a kernel is, really. The last thing I compiled was som crap Pascal code I wrote in college about 9 years ago. And I would guess about 90% of the computer buying public knows less than me)
So the question is framed wrong. The real question is, do sophisticated users, already capable of operating in a Linux environment, want this? To me the basic concept sounds great - streamline my computer's OS to maximize efficiency for my particular needs? Wonderful. But don't sell to me that I'm "building my own kernel..." I don't know what that means so I don't want anything to do with it. Offer me a supported service to "customize my computer..." if I can get it in a box, if it loads and runs out of that box, if there is a number I can call that will help me when I have problems, I'll not only use it... I'll pay for it.
It Is the Nature of Information to Transgress Artificial Boundaries
I have multiple engineering degrees and several years' experience building and using Linux and BSD, and *I* have trouble configuring, building, and installing the Linux kernel. Forget Aunt Millie -- I want a good kernel autoconfig tool for myself!
This is a step in the right direction, but true kernel compilation for the masses is still a ways off. I think for custom kernel compilation to hit the mainstream there needs to be an automated tool that will inspect your machine configuration and hardware and then automatically configure most options (eg... don't compile in X.25 drivers if you don't have an X.25 device). Just my 2 cents.
Aunt Tiley would likely not need to upgrade her kernel all that often once she's up and running anyhow. We don't need to establish the same kind of upgrade-frenzy that Microsoft and others have pushed. Linux isn't a purely market-driven product, and so frequent updates aren't a requirement just to stay in business. Short of any major technical or security issues, Aunt Tiley would be fine running the default kernel for quite some time. No need to upgrade to the latest and greatest kernel in order to get more eye-candy. Hell, the eye-candy's in a wholely seperate bunch of non-kernel packages anyhow.
So, Aunt Tiley could as easily grab the new kernel upgrades and new kernel modules via red-carpet or Redhat's FTP site. A simple upgrade using RPM since she's never installed anything from a TAR, she should never run into any kinds of conflicts.
I don't see Aunt Tiley or anyone out side of us hairy-chested geeks wanting to recompile their kernel anyhow. When was the last time your Aunt Tiley insisted she needed to use regedt32? They're users, not tweakers. A normal user shouldn't have to recompile the kernel. Ever. Thanks to loadable modules which have been around forever. It's just a matter of a good installation system that sets things up well to begin with.
--- http://foo.ca
Years from now, when people wonder why in over ten years linux failed to do anything but run some servers and preoccupy some geeks, they may also conclude what I have: that the linux community wasted time on worthless ideas like trying to enable Aunt Tillie to compile a kernel, when the OS wasn't usable for people at home who lacked a tech department or strong specific technical know-how. Anyone who doesn't have hundreds of hours to devote to learning such a user-unfriendly system would have to be pretty dumb to select linux, when they could buy Windows and get back to having a productive life.
This is why microkernels rule. The smaller it is, the less bugs. The less bugs the less need to update the kernel.
;)
Oh and QNX rules
Programming is the act of creating automations of complexities that are made up of simpler things.
Does the programmer re-write open() every time they need to open a file?
There is not only nothing wrong with making it easier to build a custom kernel, but in fact there should be a growing interest in doing this sort of simplifying, given the GNU Hurd is about not only modularity but about servers/transltors and creating such, even custom as is needed.
This can be taken even further in that autocoding tools can be and should be built for the GNU users.
In a hundred years from now, how do you suppose programming will be done (given programming today is only about 50 years young)?
As things are being done today, it is not possible to do such a program of complexity as can be imagined of what would be a holodeck program (And we do have such virtual reality cudes today in university labs).
It won't be untill the general programming field realized the need to genuinely and honestly address and do the automation of the field of programming. Certainly everything else can be automated, including human balance and movement (segway).
It's fooling to continue the illusion that programming is not itself automatable. And to begin making it happen, where better than on higher level like autoconfiguration system that allow custom kernels to be done? (Or at least one place for it to begin)
A recent research paper on autocoding presents the current/recent mindset on autocoding. It's worth reading to see how young and admitedly immature the field is. Open system and Open Source Software such as the GNU efforts (Linux, the Hurd, etc..) with their open community has far better ability to do what needs to be done than any private effort which will be biased away from doing the things that need to be done.
Soooo, anything that automates computers and their use is inherently a good thing, for iot will allow us all to reach and achieve much more advanced systems and the benefits of.
Aunt Tillie can do anything she wants, as long as the software she's using to do it will automatically back up the current, working kernel with *no prompting* (because you know she'll just click past it if it takes too long, right?), and if the OS has the 'newbie' flag set, it also won't let her do it without a working boot disk. This isn't so much for her, more for the poor person who has to make the computer boot after Tillie's had her way with it.
Newbie flag, you say? I say Linux should have a newbie flag set in the system somewhere (selectable on installation) which is basically linux with training wheels. It provides detailed prompting, hand-holding, and will try to let the user be able to recover from Really Stupid Mistakes.
Applications which support it will also have additional prompting, plus a training mode, perhaps. This adds usuability for newbies and stays out of the way of those of us who vaguely know what we're doing.
Hey, it's a thought.... at the moment, there are no gradations. It's Win9x or WinMe for newbies, and linux for us. If linux had a training mode, wouldn't it be better for both Aunt Tillie and Slashdotters in training?
-- INTX Grouch. http://www.midnightblue.net
that loads and unloads sercices from memory without rebooting, that has enough sense to remember where it patched the CALL addresses or whatever these systems do nowadays...
Too much configurability is already a problem for some.
My parents, for instance. I didn't believe it at first when the inevitable tech support came:
"There's something wrong with my word."
"Your Microsoft Word?"
"Yeah, that one."
"What's wrong?"
"There's nothing on the screen and I can't control it."
"Ok, well click on the file open icon... looks like a file folder with a curved arrow on top."
"There isn't any."
"Hm.. ok well click on where it says 'File'."
"It doesn't say that..."
It was true. They'd managed to get rid of every icon and command on the screen, and it was completely blank.
How did they do it? I don't know, but you can do it pretty easily by right clicking and then checking or unchecking things.
So as we add configurability to Linux and its applications, we should hide it and protect it behind selections which say "are you sure you want to..." -- similar warnings can be placed as comments in the top of all the important config files.
I know this will rub many the wrong way, but this sort of protection can be turned off or on as an installation option (just make it on by default).
Linux can work for the mainstream the way cars work for the mainstream: in both cases, the ugly guts (or beautiful guts, depending on your attitude) can be hidden, locked away under the trunk.
Aunt Tillie told me to quit worrying about what she was doing and to eat my nice apple pie. Seriously though, who is he to say? If people are that interested, then they will. If not, they won't. Time for coffee
Sent from your iPad.
After all, God knows how many times my mom asks me, "Should I up my ODDDEFNAME to 256?"
That this question is even being asked shows that Linux is not ready for the desktop. Or, more to the point, Linux zealots aren't ready to even develop a desktop OS.
Potato chips are a by-yourself food.
and please remove whatever crawled up your ass.
If I understood him correctly all he was saying that Aunt Tillie should have the choice to compile her own kernel if she wanted to instead of having a bunch of jackasses telling her no it's too complicated for you.
I want a distribution that has a similar GUI installer that RH and Mandrake have, but instead of invoking "rpm -i" for each package, it would build all the install packages from source drops. The "installer script" could be a large XML file that describes how to compile each package, what its dependencies are, and provides a mechanism for tweaking the packages configuration. Most of the packages out there can have their runtime configuration configured via their 'configure' script (wow that's a lot of "configures"), making it a fairly uniform approach. In addition, at the beginning of the install, it would be neat to see controls for your *exact* hardware configuation that get turned into CFLAGS like -march=i[my]86 and -O3, etc.
The only drawback I can see is that it would increase install times by a *lot*. However, in the end you would end up with a *highly* optimized distribution.
The idea came to me while building my own.
std::disclaimer<std::legalese> sig=new std::disclaimer; sig->dump(); delete sig;
call it my wacky british sense of humour but i think that this picture sums up the entire story perfectly.
free (as in mp3s) electronic music
Its a good idea to suggest that no home body should be encourgaed to violate the sacred kernel. Only suitably trained and spiritually pure programmers, after a period of fasting and absolution, should be entrusted with this sacred task. The last thing anyone wants is for everyone to have a chance at learning about Linux, programming or creating their very own programme. It should be left to Mandrake, Debian, Redhat, SuSe, etc. People who can make money out of it.
Imagine the horror of someone compiling their own non-functional kernel and then having to learn how to fix it, or worse giving people a non-functional kernel (doesn't M$ do that already)and dragging the sacred Penguin into disrepute.
...and who decides what her place is? Shouldn't it be her?
>Well, on one of my systems, EVERYTHING is a
> module (Mandrake 8.1). No recompilaton necessary
How can EVERYTHING be a module? You need boot devices and filesystems built into the kernel or you can't boot. Say I use an all SCSI system and don't need those crufty IDE drivers anymore (the IDE ports on my system are disabled), I can't remove them in the standard kernel because someone thought it would be a good idea to boot from them. The only way to get rid of devices like that is to recompile your kernel.
I think an easy to use, Joe User can make his own kernel is a great step as far as making linux more usable for the masses, i.e in a desktop environment. You can make a more compact kernel without knowing ALL the minute system details and random hardware that most people don't have or know what it is. Definitely a step in the right direction.
~.Evanrude
what his business is...
has it already been made? or is it just being debated whether it SHOULD be made? btw, mandrake already has a snazzy kernel compiler in what i think is its standard install. Havent tried it, but it looked a lot more friendly than the standard graphical compile tool in each kernel release.
"Computer games don't affect kids; I mean if Pac-Man affected us as kids, we'd all be running around in darkened rooms,
LOL... the only thing ESR is good at is writing an essay (nice essays though)... after OpenSource.Org failed and VA Linux (Board Member) decided Open Source was not the future he wants to weasle his way into the Linux Kernel.... stick to improving fetchmail, or I might be forced to spend 2 hours writing a replacement.
LOL
How would this cover up the Oracle debacle?
I guess you may mean that with this sort of thing going on, Oracle supporters hope that Oracle will be seen as "not so bad" after all, despite promoting a system as rock-solid that turned out not to be.
How to make it easy for applications to provide this sort of function. i.e. multiple application images...
..we'll also probably want a special "boot" filesystem, which automatically re-runs lilo whenever someone copies a new kernel to it. Yeah, it's for Aunt Tillie, not me, yeah, that's the ticket...
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
We're proud of our mad programming skillz and our ability to wrestle with arcana.
Which is on the bottom of the Aunt Tilly link. If you want people to stop confusing the two groups ESR you could at least try to make them appear somewhat seperated by, say, not talking like them. Maybe ESR can hook me up with some "leet private sploits" too.
Personally I think autoconfig is a great idea, especially if it can detect my video card (which only happens to be the most popular video card on earth btw, an nVidia GeForce II).
How we know is more important than what we know.
...the hell is wrong with xconfig?
I think a small script to detect obvious hardware configurations should be available (is there one available now?) as a separate package, but this "feature" shouldn't be included in the kernel source.
Besides, don't you think people are better off *learning* anyhow?
If i remember right, this "flamewar" started when ESR wanted to change the output of each of the drivers to contain their CONFIG_ symbols. This got a lot of people irritated, and then the reasoning he used was "Aunt Tillie."
Otherwise she won't notice a blind bit of difference. IRL I never have, & I've recompiled the kernel many times.
Full plate and packing steel! -Minsc
to have Aunt Tillie care about kernel recompilation?
why would you be AGAINST someone compiling their own kernel? seems kinda elitist to me...
That's meant to be >4yo. Sigh, HTML formatting.
Full plate and packing steel! -Minsc
I think if Aunt Tillie can create a swap partition during installation, pick a window manager, download compile + install the latest mozilla browser update (or maybe she prefers Opera), configure her firewall, and set up lpr for her printer, she can recompile her kernel. I just don't want to be around when she starts looking for "Freecell."
_______
2B1ASK1
Now THAT'S the spirit! (No, that wasn't sarcasm).
Being politically correct should be outlawed. What do you call fat people?
Gravitationally challenged?
Horizontally challenged?
All these "challenged" expressions gives me this idea about politically correct people:
"They are intellectually challenged."
(Posted anonymously, because this is an offtopic, trolling flamebait of a post).
that would give you the list of what you would be typomg if you were doing by hand.
That was way back in the 2.1.x days. Then, I knew all the caveats of the minor revisions, and I knew which particular revisions were more stable than others. Now I'm nearly the opposite. I'm happy to leave my system running for months on end without checking the status of the kernels. I actually have to "cat /proc/version" to see what revision I have fired up.
That attitude is only reinforced with the 2.4.x tree. Pondering a kernel upgrade is like pondering if I want to step into a minefield.
Reading the comments in "2.4, The Kernel of Pain", I know there's still Version Whores out there. They know the obvious stuff like "don't use 2.4.15". And I'm sure there's less obvious stuff, too. Aunt Tillie or whomever isn't going to keep up. If she steps onto the 2.4.15 mine (or its equivalent in the future), she could do damage to her system.
To that end, we could use short, digestible ranking/summary system of the kernel revisions. (Or does one already exist?) Which kernels in the stable branch are really unstable? Which are the most robust? Many, Aunt Tillie and myself included, would find value in such a system, regardless of a Kernel Autoconfiguration program.
Of course something can and will eventually go wrong when auto-configuring a kernel.
A distro's installation could do the following to make this recoverable for most users:
Prompt the user on whether or not to run the auto-config. Compile the new kernel, and add an entry to lilo (or whatever boot loader) in addition to the precompiled kernel. That way if it's broke, Aunt Tillie has something to fall back on.
Mmmm, -funroll-loops
Compiling the kernel is easy. Make menuconfig has everything laid out and easy to find with a little exploration. Compile a kernel a couple of times and it becomes second nature. Really, if you've never tried it, read one of the 200,000 pages out there on how to do it and jump right in. Don't forget to run lilo.
The real issue is that the kernel deals with concepts that "good ol' aunt tilly" doesn't comprehend. Does she know what chipset she has for her IDE controllers? What about old CDRom drivers? "That's a 12x... it's pretty old... Bluetooth? What's that? Better put it in just in case... MTRR support? Oh, I'd better put in math emulation so I can use my calculator..."
There's only two ways to solve this; One, put better help in the kernel configuration. This is being done, actually, and I'm all for it, but some things will still be very vague to the non-techie no matter how much help you put in. The other way is to "user-friendlyize" it, which is usually done by taking away options and hiding the real technical details from the user. Do that and a lot of people are going to be pissed off (like me).
A new configuration program I'm not opposed to; just don't take away our options - keep in mind who the majority of people out there who compile their own kernels are (i.e. people who know what they're doing). And for the gods' sake don't give aunt tilly the root password.
Those who can't do, teach. Those who can't teach either, do tech support.
Actually you can. You just need an initial Ramdisk containing the modules. (mkinitrd something something creates such a ramdisk) Just set the correct parameter in lilo.conf or grub.conf (rerun lilo if you use lilo) and you're set!
(2) Like most of us, Aunt Tilly has a brain. She is able to spot the difference between a real issue and a quasi-troll dreamt up by ESR. She knows, for example, that no-one wants only "the educated few" to be "able" to compile their own kernels. All the tools to do it are already there. But who should have to unless they want to? And why on earth should anyone forbid them if they so wish?
(3) Like some of us (though not me), Aunt Tilly has an iMac and thus already has a powerful Unix-based OS with a beautifully designed GUI and proper software management. She knows she can tweak it if she likes (see [2] above), but it works so well she doesn't need to think about it, which suits her fine (see [1] above).
Take a giant step backwards with ESR if you like. I hope for the day when Open-Source software is as well-designed and functional as it is well-motivated and worthwhile.
IF the compilation of a new kernel can be made easy and foolproof enough - I do not see the problem. It's just a different label on a progress bar, instead of "copying files" to "compiling kernel". The problem is if it goes wrong. Of course updating a distro *might* go wrong as well and leave your system in non-bootable state (not that I've seen that happen - ever). Compiling a kernel for a specific machine might have some special advantages (speed/memory optimizations, modules support)...
It's not about naming the kids, "Tillie" is the aunt's name...
In the bad old days of Unux the geek control was high. Unix was real he-man stuff. Can't use vi, awk and sed... go away. No wonder MSDOS got a pat on the head with its simple concepts. Apart from the aunt tillies, there are a lot of people who are technically savvy, but don't want to learn the kernel config tools yet. The same goes for kppp etc. Sure I could learn how to set up ppp, but I don't need to. kppp makes it easier to do something that would otherwise be a grind to learn. Why put up barriers that don't need to be there? Barriers and Open are multually exclusive. IMHO any tool that makes it easier for someone to get going is a good thing.
As many others have pointed out either here or in other conversations, until Linux users can do ANYTHING with a Linux system that Windows can do, but without ever having to recompile the kernel or even know what a compiler (or, for that matter, a kernel) is, then it's not ready for prime time.
But I do agree that the terminally geeky should always have the option of compiling the kernel and hand optimizing their configuration, if they care to. They should never be forced to do so, though.
"...Under the hood, the machine is downloading the tip of the stable branch from a kernel.org mirror site..."
First we got stuff going on under the hood, then we got stuff going on behind our backs, and soon we got stuff calling home to the mothership every time we start up a particular app...
Wait!?
Doesn't this sound familiar?
Should...
It's what Window$ has been doing, more and more, for years.
As Mr Horse used to say: "Hmm.. Nope. I don't like it. I don't like it one bit."
t_t_b
I'm on PJ's "enemies" list! Are you?
Your analogy, too, is false. If someone new to computers, someone without the "respect" for the technology you mentioned, tries something dangerous that could delete all his or her files, don't you think that there should be at least one little modal dialog that pops up asking "Do you REALLY want to do that?" What you are suggesting is nothing more than classic Slashdot eliteism: "The user shouldn't touch anything they don't understand." Where's the problem in that statement? Think about how you learned to use a computer. Did you ask someone "So, can I really double-click here? You're sure about that?" I doubt it very much. Users should be able to mess around a bit with there computer, and have a "safety net" underneath them to catch any mistakes before they do serious damage. You should realize that restricting the access of users to their computers is a bad idea. Make the interface usable without an extensive knowledge of C, Perl, asm, and Forth, but if someone wants to mess around, let them. How would you feel if every time you tried to access your BIOS the computer said "Sorry, can't access that, you're just a moronic user who probably will wipe out his entire HDD with one misplaced command. No way you're getting in here, dumbass!" instead of letting you actually change settings? Perhaps it's time you started examining the roots of your elitism.
That's it. I'm no longer part of Team Sanity.
Democrat delenda est
I don't think that anyone genuinely interested in linux/open source/what have you, and who doesn't have their head stuck up their own asses in conceit would honestly say that an autoconfiguration tool for the kernel is bad. Let's look at this objectively.
First, such a tool would only make linux easier for people that are not knowledgeable with computer workings, and make it a more viable option for those who don't want to mess with, or aren't knoledgeable of, the inner workings of the computer. I've run into many people (online) who don't have support for xy device with #.#.# kernel, don't want to install another distro, and need to compile a kernel.
Second, (as far as I know) this would be something fairly easy to do, provided that the device that wants to be used is already attached to the system - the kernel seems to have a decent detection system already, just have, say, a 'kernel compilation disk' which would have the kernel you want to compile, with all the possible modules compiled in, which would use your system. it'd have it's own initscript, which would have a step-by-step process, walking you through the configuration (eg., Is the kernel source tree untarred already?, Is the kernel source tree in a location other than the standard location? etc)
Just some ponderings.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
This thread's funny.
I put together a kernel rebuild guide a few years ago ( Kernel Rebuild Guide ). I'd guess that for perhaps 95% of Linux users, there's absolutely no need to rebuild a kernel. For those that do, it's usually to enable a feature or to tweak just an iota more performance from the system.
Sure, anything that makes the system easier to use is good. It would be wonderful if guides such as mine were obviated. At the same time, should we really be wasting time on what's essentially a band-aid? By this I don't mean that Aunt Tillie shouldn't re-compile her kernel, only that if Aunt Tillie (a regular user) requires the feature then the distribution should already support it through other tools.
The main problem I see is that no matter the frontend, a kernel recompile will invariably ask a lot of questions that Aunt Tillie may be unprepared to answer. And if she can answer them I strongly believe that she would have absolutely no problem with the current configuration tools such as xconfig/menuconfig.
I've been programming for 23 years. I've used Linux off and on for about 5 or 6 years, but not daily.
I used to be able to do kernel builds. My last several attempts have been disasterous. The main thing is that the config just has too many damn options and the language is too technical, even for a programmer. That's insane.
I'd like a simpler utility, but Aunt Tillie doesn't need to be rebuilding her kernel. That's insane. This is the dumbest question I've seen on Slashdot in a long time.
that people understand precisely by messing with things they don't understand. In fact only for very unimportant things does anyone stand a chance of understanding prior to messing with.
This is perhaps the dumbest flame war I've yet seen on discussion. This is one of the reasons Windows is leading Linux by FAR in the OS wars - if you even want to call it a war.
The answer is very simple. Of course allow this autoconf. Autoconf's are great, people should be able to run a program that makes life easier.
BUT
If you're building a Distro of Linux for end users like your fictional Aunt, don't include the feature. Just don't include it. There isn't enough of a performance increase that you'll see from a kernel optimization in almost every case. Truly if Linux wants to make it onto the mainstream, they are for all intensive purposes going to have to dumb it down a bit. People who just want a simple environment to write their reports, file their taxes, surf the web, and email friends are not going to give a crap about optimizing their kernel. That is best left to hackers. Why not create a distro that speaks to the masses? So don't put it on your 'enduser' distros. That's why distros exist, isn't it?
Now let's face it, the majority of people who use Linux are using it in server environments. If I'm a sysadmin and I want to setup this new distro of Linux quickly and easily without having to search through lines of what ends up looking like a bunch of code, I'd easily take autoconf. I don't see what the argument is about, really. What it comes down to is, there's a bunch of little Linux brats (no better than 5cr1p7 k1dd13z if you ask me) who are trying to protect their little clique of windows-bashers and Linux advocates (who probably don't use Linux anyway), who would rather dismiss the general public as idiots than work with something innovative and smart that makes life easier. These are Syds of the world who insist that the world was better when people did their programs using punch cards.
Is this what the gnu-linux/opensource community is coming down to? Whether we should include a dialog box that says "Compile custom kernel... experts only!" ...that opens up the configurator screen? Because, aren't you the same people that resented the absence of such features on proprietary software packages not so long ago? Since when do we go around "not-selling-scissors-because-somebody-might-run-w ith-them"
I mean, I have no delusions of nix-guruhood, but I tinkered with linux and I built my own kernel, and it worked just fine, not because I needed to, but because arrogant *nix-users quipped "I compile daily!" to prove their *nix-hood. And when I was done, I was thoroughly disappointed with how little I actually got out if it, and how easy it actually was, but at least now I know what I could do to the kernel if I had to.
And you know what? If aunt-tilly accepts that she may bork her system if and when she checks that box, more power to her. I mean, for god's sakes, every windoze installer now has a "custom" install box, and you can still use uncertified drivers on XP if you really want to, and as others have said, nobody is stopping anybody from running regedit and borking their doze box.
Dialog boxes don't kill *nix-boxes, lusers do!
Supposedly.
Programming is no more automatable than mathematics. The two fields
are highly akin. Please refer to Gödel's Incompleteness Theorem.
If you mean to say that SOME of the programming that is done today --
the lower level, already explored and understood coding which
is so much reinvention of the wheel -- can be automated
I'm with you all the way. But that just moves the act
of programming towards the higher level, towards building
on that which is already written. It doesn't automate
programmint itself.
-josh
With those new Ghz machines, better hardware detection, and some luck, The install process could now make the entire system, or at least the biggest resource users (kernel, X, KDE/Gnome, audio/video crap) completely customized and optimized for the machine. And with a simple autodetection program run at boot, new hardware could be taken into account and stuff re-compiled as needed. It might extend the install time a bit, but if packages are already optimized for 686 and above on consumer distro's (as it should be I think), It might not be that much time at all. Even a simple question at the start of install might do it: Are you a user? (compile for celeron, and budget graphic card). Are you a geek? (compile for ridiculously overclocked Athlon and GeForce 3). This really could work...
Shift happens. Fire it up.
Well, unless they make it user-proof, this could be a big mess.
We had a kernel auto-config program at one of the places I used to work, at if the users screwed up on selections (and they always did) you can end-up with a nice doorstop with disk drives and a graphics card in it.
Imagine the bad press Linux would get then.
I believe you mean 'elitist' instead of 'egalitarian'. An egalitarian attitude would be that it's OK to let anyone go recompiling their kernel. :-)
Well, if all the Unter-Geeks start flooding the lists with why doesnt my kernel work then it sucks..
UNLESS that leads them to learning to do it themselves.
Ive often wondered why DISTOS didnt have an autocompile script for their kernels so at install it builds one to suit your system
Sig went tro...aahemmm.....fishing........
> Eric Raymond supports this idea saying that
> this will bring Linux to a wider market.
This is one of the silliest things I've ever heard, and shows a remarkable lack of understanding of people. Allowing people to customize the kernel will not lure ONE new person to Linux; it would probably just turn people off, adding an extra layer of complexity
I think the file would only get larger with prebuilt modules. If there were not prebuilt modules, the compile would only take longer. Maybe you should try to make it smaller || compile faster || get a new computer before adding to it's work.
What is it with ESR and coming up with the dumbest fucking arguments about practically everything? Does he enjoy running up hill with ankle weights pulling a Eurovan by a rope between his teeth?
"Hey I know...I'll pitch my kernel autoconf to people by saying it is so easy their aunt can use it, I'm a fucking genius!"
An automagically config for kernel compiling would be a really nice thing to have but not becsause my extended family is going to fucking use it. It also isn't going to do anything to make Linux more popular. Joe Sixpack doesn't know what the hell a kernel is let alone what compiling one would do. Linux users naysay stuff like Wizards because they feel it takes away from their freedom somehow which is an entertaining thought because anything done with a wizard can be done by hand if you want. However having something that knew what it was doing but let you pick some parameters would be a good thing for something complex...like compiling a kernel. ESR's shit about Aunt Tillie doing kernel compilations is entirely moot and too damn stupid to even discuss. The point of discussion is ought an autoconf be included in the stardard kernel package. I think so but if not distros can always add it anyhow as they fucking should. Most distros pack a system configuration utility to pick what software packages you want installed by the same token they ought to have their own (or a standardized) kernel config utility that does most of the picking and choosing for the users.
I'm a loner Dottie, a Rebel.
I have problems with Eric Raymond's scenarios. Forget about if it's a good idea to make it easy for anyone to build a custom kernel, my question is, why should you need to recompile the kernel just to install a device driver ? That's just stupid. Installable drivers, that's the way to go.
I mean should Aunt Millie be allowed to chop and channel her Chevy? Sure if she's up to it. That's what Linux is all about - a computer replacement for making hotrods. Not many Millies did this - but hey a few did I'm sure. More now-a-days than ever before.
Get out those torches and chop away!
According to LWN, Alan Cox said that Linux 2.6 will not have compiled-in modules.
From: Alan Cox
To: babydr@baby-dragons.com (Mr. James W. Laferriere)
Subject: Re: ISA hardware discovery -- the elegant solution
Date: Mon, 14 Jan 2002 18:08:32 +0000 (GMT)
Cc: alan@lxorguk.ukuu.org.uk (Alan Cox), esr@thyrsus.com,
cate@debian.org (Giacomo Catenazzi),
linux-kernel@vger.kernel.org (Linux Kernel List)
> Hello All , And what mechanism is going to be used for an -all-
> compiled in kernel ? Everyone and there brother is so enamoured
> of Modules . Not everyone uses nor will use modules .
For 2.5 if things go to plan there will be no such thing as a "compiled in"
driver. They simply are not needed with initramfs holding what were once the
"compiled in" modules.
Alan
cpeterso
An observation - If a person wants to build a customised kernel for their machine, that person would probably be someone who is interested in learning more about their machine and/or OS (excepting particular cases). If such a person wants to learn more, how do they do it?
If they feel as though they don't know enough to recompile a kernel without some kind of extra assistance (and most new timers would), but they don't happen to know anyone good at using Linux, then using a piece of software to help them would be the obvious third step (right after obtaining documentation on the matter, which can sometimes be confusing).
How did the "educated few" become educated, if not from the assistance of someone else or from some kind of documentation when they first started getting interested in the deeper aspects of linux? Linus didn't provide much documentation when he first released the kernel(s), and it was only through the efforts of some highly intelligent people who had to pour over the code to work out what it did, but that isn't necessary now. (And many thanks to those people, while I am on the subject.)
If a person is interested in learning more, and the "educated few" decide that since this person isn't educated enough they shouldn't be playing around with what they don't understand, then how do the uneducated "majority" (apperently) learn?
Education happens like it does.
Any comments?
-JB
"I love deadlines. I love the "whooshing" sound they make as they pass by." - Douglas Adams.
"I do trust Red Hat's binaries...I _don't_ trust kernel.org sources." Wow, that's terrible. Several kernel hackers on the mailing list admit they don't trust the production kernels, and joe-user is supposed to? The production kernels should be rock-solid. We shouldn't require people to do research before chosing a kernel. That is the real problem that must be overcome before the Aunt Tillie problem can be answered. If she can't trust the kernel, then she can't compile it in the first place.
Granted, I've never used the CML2 tools, but I followed along on the discussion on LKM for quite some time. It seems as though every post is way off the mark.
First, ask yourself this. Is 'make xconfig' a bad user interface? Nope, it ain't. What sucks about kernel configuration? The dependency resolution crap. Linus has a nifty little program in place that does a pretty good job of figuring them out too -- but it's ugly, and admittedly a kludgey solution. CML is more "elegant and flexible" which is a damn good thing -- but last I knew the bugger took 2x longer to do it's job than the old system. Kernel developers do probably 99% -or more- of kernel builds so why on God's good green earth would they want a system that's going to slow them down right now? They don't and I can't blame them in the least bit.
CML2 is nice, and it seems like it's a really good little system, nobody on LKM is opposed to it really (that I saw) they just don't want something that's going to suck minutes out of their programming day. "Aunt Millie" can't answer kernel configuration question anyway, period. Heck, most Windows users don't know if they have 95/98/NT/2000/Me/XP some of the time, let alone if their processor is Pentium III, Pentium IV, or K7 based.. unless the sticker is still on it. Shoot, they don't know if their mouse is ps/2 or serial, or what USB is. Do they know if their USB host system is UHCI or OHCI? Hell no.
CML2 is about making kernel configuration easier in terms of expandability -- not usability. The current interface is very usable, just not very flexible. Because of it's inflexibility and complexity it leads to un-bootable systems sometimes when depency stuff get borked up in strange configuration situations. CML2 takes care of -that- and nothing else. It doesn't keep you from having to know your hardware inside and out. End of story.
I get paid by the hour
An article by Michael J. DeMaria over at networkcomputing.com.
My Indigo2 does it quite nicely, but as Linux does not yet recognise all hardware on install (my ess onboard sound, darn it) it will be a hit or miss thing, untill hardware vendors conform to general standards not proprietary ones.
Crap, when did that happen? Why doesn't anybody ever tell me these things? Here I've been recompiling kernels since the first day I started using Linux and nobody ever told me I couldn't do it.
I also wrote my own kernel HOWTO that thousands of people use daily. I've been doing kernels for almost a decade at over 400 a year, sometimes more. It surpasses the linuxdoc one in simplicity. It also uses an incredibly simple solution to the annoying "/usr/src/linux" being a real directory problem.
This is nothing new...
In Netbsd there is a perl script to reconfigure a kernel to just compile in the hardware that you have. It is really handy to be able to just type perl adjustkernel GENERIC > MYKERNEL and have the correct hardware selected. I sat at my desk and `configured' the hardware for a kernel recompile in Netbsd in less than five seconds. Try that with Linux. I can code, I've looked at the Linux source code, but I have better things to do than try and remember whether the 5 year webserver that is sitting at work has a pci bus or not. Netbsd has this, why doesn't Linux?
I read documentation, I experimented, I'm 15.
I think make menuconfig/make xconfig make it easy, personally.
For the "average" joe user, custom kernels are not really needed anymore.
With the advent of dynamically loadable kernel modules, these days the distribution vendor can ship a fairly flexible kernel.
A custom kernel these days may save 200-300 kilobytes of RAM by removing un-necessary drivers - at the expense of support from your distribution vendor. when you consider that the "average" user these days will have at least 64 MB, its pretty insignificant.
there is of course the issue of specific CPU optimizations, however I believe this can (and should) be handled by the distribution - just give the user a choice of kernels on install (or ideally autodetect CPU type and give the option of "standard 386 kernel" or "xxx cpu optimized kernel").
The only real *need* for custom kernels these days is if you are doing "funky" stuff with your network, and this is really a situation in which you would hopefully have someone who knows what they're doing.
of course, geeks will always want to play with the kernel for fun and amusement, but it shouldn't be necessary for everyone...
just my 2c
smash
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
i love how i got modded twice as redundant with a first post :-D that's logic, that.
slashdot: where everyone yells sarcastic metaphors to themselves to understand the issue
Won't the people who know what they are doing be the ones writing a kernel autoconfigurator? I don't see what problem there is with letting a program probe system hardware, select the appropriate options, and build a kernel for Aunt Tilli.
Afterall, half the things you might need a user to know, like what kind of mouse he or she has, is already asked by most installers.
Yes...you use mkinitrd to make the ramdrive(initrd.tar.gz). This is Mandrake's default boot mode, but you can do that on ANY distro.
The only reason to compile is to install the new kernel b4 your vendor distribs one. (or to test the 2.5 series..crikey!)
If you're not a Liberal in your 20's, then you have no heart.If you're still a Liberal in your 30's you have no brain.
It depends on how you see the question. Should Aunt Tillie ever be required to build her own kernels to get certain features? I think not. However, if Aunt Tillie wants to optimize her kernel for her particular system and configuration, I don't see why it would be a bad thing to make this easy for her.
Lame, Taco! Why'dja slap all the responses to First Slashdot Troll Post Investigation!?
There will always be a select few who think that Linux should remain obscure to the average user which is fine, everyone is entitled to their opinion. That being said I think this is a good idea. It would make my life as a system administrator easier and it will also give lesser skilled users the ability to tweak their kernel if they choose. For me make menuconfig is very simple to use but why not have another choice? Isn't choice one of the strengths of Linux?
More people fall off cliffs with fences simply
because more people visit cliffs that have
fences, that is why the fences were built.
...
More people eat ice cream in the summer and
more people drown during the summer, but to
assume there is a direct connection between
the two is misleading and wrong.
Charlie
Recompiling the kernel means a reboot to install the new kernel. Having to reboot to reconfigure stuff is bad (see: Windows NT). Why go backwards to a way of doing things that is as bad as Windows NT and worse than Windows 2000/XP? If you could dynamically switch to a new kernel (beware: deep magic), fine, but AFAIK, Linux is a *long* ways from being able to do that.
He's saying that only expert drivers should be allowed to drive. ;-)
Of course how in the hell you would become an expert in his world, we will never know...
I went from Windows-junkie to Linux geek and the more and more I used Linux, the painful, stabbing-in-the-eye feel I got everytime I used Windows.
What kills me is when NT Admins talk about "user friendly" (for a server OS?!?), I still cannot understand how a point-and-click is faster and simplier than command line interface.
Like changing the IP addresses or editing files... you mean I have to use a mouse to do this? Don't even get me started about click Control Panel or each and every directory to get to the file.
NT Admins fire back with "you can use the Run command" but is there a vi for Windows?
ChozSun
ChozSun.com
i was a young and cocky kid
dont you mean a young cock eating kid?
your penis is covered in far too much hair and feces to be any sort of beacon
Hey adlam.bor, I took a look at some of your recent posts. It's quite an impressive posting run at score 1 you did on January 18! Each post is different, inventive and sometimes even clever. Keep up the good work! I shall watch your trolling career with great interest.
Yeah, this is a great idea. Especially considering the average computer user doesn't even know what the hell a kernel is. (Ain't that thar the guy in chaaarge of a buncha guys in that thar aaaarmy?)
Anyone wanna become filthy stinking rich?
Here's a brilliant idea! Sit down in front of a bunch of average (As in *average*) computer users, and ask them for input on a user-friendly operating system!
Sure. It'll suck ass for everyone who has a clue, but we're the minority, and you'll be rolling around naked on a pile of cash.
I'm with Cox on the matter that I think Aunt Tillie would be better off with the distro's kernel (where she might have lm_sensors, nVidia, TV and Radio drivers), but !
I'l defend Aunt Tillie's *right* to chose !!
That's what freedom is all about, options !
echo '[q]sa[ln0=aln80~Psnlbx]16isb572CCB9AE9DB03273snlbxq' |dc
A very large percentage of the population of the US is full of negros, mexicans, femals, or otherwise intellectually challenged people.
I know you're trolling, but I must vehemently protest against your description of negroes as "people"!
I think people should calm down a little. I presume you're going to need to be root to do this, yeah? So if Aunt Tilly had root and wants to recompile her kernel, then seems to me she's gonna screw it up anyhow.
I've recompiled the odd kernel myself despite being a bit of a Linux newb, and it's not that hard to do. OK, I broke stuff with a few of my tries, and one or two wouldn't boot, but it was always easy to switch back to a working kernel. Now, if I had a utility that made it all smooth, with it able to semi-automatically swithc me back to a working kernel, I'd be doing it every day.
Look at it as a leveraging thing. Here's something that Linux can do that Windows, MacOS, etc. will never be able to - recompile the kernel for optimization. It's a feature. MAke it easier to use.
~~~~~ BigLig2? You mean there's another one of me?
only use this way :
1 - Probe for devices
2 - Get the source
3 - Configure, make, make install
Also this would mean a "Linux Update" that could have RPM + Kernels Update...
Please remember then to add an option in LILO that allow you to get back to "last stable kernel", because doing it manually is definetly beyond Aunt Titties...and most of the time beyond my own competence...
It takes 40+ muscles to frown, but only four to extend your arm and bitchslap the motherfucker
Windows 2000/XP (which is a microkernel client server architecture based on Mach and VMS, IIRC) does this really well, basically because they have to because they don't give you the source. With Windows, you don't have to restart the system, let alone recompile anything to add a lot of new hardware devices. If Microsoft can get this right (for the most part, anyway) why can't the Linux community?
1 -"Eric has requested several in-kernel facilities solely to support his autoconfigurator"
/.../at least not in the presence of ISA "
Even if most of the requests seem ridiculous, we have seen stranger things happening (distro kernels ?). A modified Kernel Patch is just another kernel patch...
2 - "Aunt Tillie shouldn't have to build a new kernel"
Right. And this is the whole point of an auto config thingy. Modular autoconfig at runtime isn't either something for her. I don't see my Auntie Hot Swapping hdds or raid cards...
But why not have BOTH.
3 - "The autoconfigurator is bound to be an imperfect job
My last computer (Abit VP6) has no ISA. and this will come more and more. And the Isa configurator I know in Linux is quite good as it is. So what is the problem....
I prefer an automated job working fine 99% of the time that everybody can use than no tool at all so you have to do it all manually...
4 - "kernel people are already drowning in bogus bug reports"
Then maybe an unified tool to configure the kernel would help reducing all those people that tried a new kernel and got something wrong and think it's a bug...
+ Do you implie you prefer they keep their free time for you ?
" if Aunt Tille has to build her own kernel, we have much bigger problems that Eric's autoconfigurator will solve."
=> "If Aunt Tille had the right to vote and decide for herself, we're in big trouble"
Who are you to decide what's good for Auntie ? And what if she needs that preemptive patch? Is it your problem ? Shouldn't you be pleased that Auntie Dear get's an interest in Preemtivevness and Frame buffers (* shudder *)
As I often said in my family (to my dad mostly) : I prefer to teach you how to do it yourself than just solve the problem and be on my way, caus next time you got the same problem you won't need me...Yes teaching you is harder, but not in the long term.
So, imagine all the trouble we can save to future generations who won't have to help Auntie with the new kernel config... 8)
It takes 40+ muscles to frown, but only four to extend your arm and bitchslap the motherfucker
IMPORTANT: Please read the whole post
I'm sure many of you are aware of this thread already.
If you are interested in helping against the moderators who have been "editing" the thread, please read this.
Please do not moderate this post down. It is good for the long term, but if you still feel like being someone who denies the horrible truth, give me your best shot. You will help hold all of Slashdot users back in the long term.
For more info, read this piece from an apparently superior news site.
This really has nothing to do with aunt Tille, Grandma, or what ever other prototypical uninitated user you might be talking about. Although it is a good idea to keep them in mind when developing things, there is far more potential than just to make it easier for a novice to configure a Linux Kernel.
/boot or wherever else someone might want to put it, add the proper entry in lilo or grub config files, and get everything ready to run the new kernel on reboot. No this shouldn't be the only way to do it, but it would make things simpler.
Every time a new kernel comes out, I have to go through every option, and check what it is, I usually need to activate and follow every sub-menu, just to make sure that there isn't something bizarre down there that I need. Usually I don't, but I've been bitten more than once.
If you recompile your kernel every week or so, you get to know the menuconf menus pretty well (does anyone use the original make configure anymore?), but when something new is added or changed, everything changes, and you have to go through it all over again.
If you recompile occasionally, though, it becomes a very time consuming task to check and recheck that everything you have done is correct. Even then it usually takes me a couple of tries because I overlook something stupid.
This makes a lot of unnecessary work, that isn't difficult, but does take far more time than is really necessary.
Something very like what is being suggested would remedy this. It could detect the hardware, go out get the newest kernel, check for other drivers or patches that might work with the detected hardware (this has the advantage of making it possible to add support for some strange hardware), patches the kernel, then optionally allows the user the opportunity to do some basic or even full configuration tweaks on the config file.
This has the added benefit of making it easy to compile and install a kernel for unknown hardware. Laptops and the like are notorious for strange hardware configs. Without research as well as a lot of trial and error, it is usually not possible to have a fully functional system. You might have a usable one, but some components might not work quite right.
An automatic config generator would resolve this as well. Forget Aunt Tille, this would make my life easier. Sure, give me a way to tweak the configuration later, but something that would generate an intelligent config file for most hardware would make things a lot easier.
How about trying to create some sort of Linux network, say a hetrogenous lab. Something that would allow me to select some configuration options the kernel config that have nothing to do with the hardware, then automatically determine the hardware and compile a custom kernel for each machine. This would save a great deal of time and energy.
For that matter, it should be rather simple to create a utility that would take a newly created kernel image, put int in
If I were updating the kernel on say 30 or more hetrogenous machines, I sure wouldn't want to sit at each machine and configure it independantly. I'd want to be able to specify some config stuff, run a command, let the machines sit, and come back to check them later to pound on any that didn't work with the new kernel.
Some might fail, but a reboot into the old kernel, a little tweaking of the config file, and a retry would still be better than doing each one by hand.
What if I am a sysadmin, and I need to keep someone elses machine running? (Best is to keep a running kernel, but sometimes you need to upgrade.) Unless I know their exact hardware, I'm going to have to spend some time figuring out what they have. Why should I waste that time and effort (that could be used more constructively), if the machine can do most of it for me?
Yes, it would be nice if a novice user were able to get a new kernel automatically, and yes, it would be nice if all novice users were to learn how things work, but neither really cuts it for me as the best motive for going either way.
For me this or something very much like it could make life much much easier.
There is a civil war coming in the United States. Remember which side has most of the guns
You just have to write it AutoKonf for Kernel 8)
It takes 40+ muscles to frown, but only four to extend your arm and bitchslap the motherfucker
Kernel compiling is a rather easy task. Most books are only taking 2 chapters for it and not more is needed. When you have a GUI with checkboxes and stuff it wont get any easier because if you dont know about your hardware the gui cant help at all.
I think using the console is fine in that case, people can use the custom built kernels that come with thei distro anyway...
cu,
Lispy
Why are we only talking about the kernel? Should the average ..?
user have the opportunity to modify OpenOffice? Mozilla?
This should have a much higher impact on one's Aunty than the
kernel has.
To me the answer is yes. Open Source should communicate
technology on all levels and is as such a medium of education.
I agree that one should allow the user to break things,
however, the user should then be aware that (s)he is on
dangerous grounds. To a certain extend this is already
governed by
* the option to include experimental code
* the "choose N when unsure" messages in the help system
* dynamic offering/hiding of options
My personal opinion is that at the current stage not
much more is needed. If Eric's new installer is more helpful,.
fine.
From Apple's hypercard I remember that different user levels
where offered - ones were allowed to program, others could merely
use a system and intermediates. But we have this too as there
is root, world, groups and users and the ability to change can be
assigned respectively.
The "use a key and drive a car"-philosophy was said to have driven
the development of the closed-case Macintosh as a successor of the most
successful and most open Apple II. Learning from this they then had the
open MacII and learning from that they had the closed-again iMacs.
So, apparently there is a market for closed systems, but it seems to me
as if we need the open system to start with since this
is our model of development.
At least my aunties will never want to compile the kernel. People have too many
difficulties to distinguish file systems from the kernel from the programs
anyhow, they don't care. Root should always know what root is doing or know
that (s)he is risking to do it again.
Any open source system should come with all that is needed
to make somebody understand the system if these are willing to learn.
Today this is the README, the documentation, FAQs and of course
the source code. The kernel has it all and it can always be better. The kernel
should serve the whole community and the community should grow, but it should grow imcrementally.
HOW is this not a good thing overall? Not to mention the fact that it would be extremely easy to put a huge, flashing, bright red WARNING at the beginning of the kernel autoconf routine (which most people would probably ignore anyway) saying that "Recompiling your kernel can lead to serious damage to your system. We suggest that if you are not familiar with how and what this does, that you read up on the subject before attempting ANY recompile. Thank you, the (company of choice) Team." Then put in a disclaimer in the software package that doing a compile on the kernel and fscking up your system will cost you extra for tech support as these problems are very hard to diagnose.
Same as with labels on the bottom of Coke bottles in other nations stating "Open other end." :-p A little common sense and a warning/instruction can go a long way.
Dream as if you'll live forever.
Live as if you'll die tomorrow.
~Anonymous~
all I got was told to RTFM!
and use man.
and it worked pretty good.
And the man called Stallman came unto them and he said "Thou shalt use info" for info was begotten of emacs, and emacs was begotten of Stallman.
And the gnu tools moved their documentation unto info, and the man pages were deprecated.
The users were mightily angered by this, for they already had man, and it worked pretty good.
Info, the bastard son of emacs was no better, but did devour great multitudes of megabytes in the manner of it's father. And the users wailed and gnashed their teeth, and rent their robes in anger.
The users went unto the house of stallman, and demanded that he deliver unto them the gnu tools documentation in a manner that could be read by the venerable man program. And stallman refused, saying "all of you, let the light of emacs into your heart, and you too shall know the joy of info". But the users did not, for their hearts had neither the megabytes nor the megahertz to satiate the appetite of emacs.
Stallman saw that the users were not swayed by his words, and he sent them away, saying "begone, before I order thee apped "GNU/" to thy names, and thy hostnames, and the names of your wives, and children, and cattle, and pigs, and goldfish.
And the users left in haste, for they did not wish to be known as the GNU/users.
I don't see why allowing accessibility is a controversy. I could see how this tool would allow new users and soon to be guru's start off and gain confidence, so later they will be able to build their own kernel.
Maybe the patronizing tone of the post is the reason.
I've been using Linux for two years now. I've manged to use it successfully, in a production environment, implementing Samba, Novell's eDirectory, Apache, Tomcat, Squid and more. I also use Linux at home, if that matters. Yet, with all this I had never compiled a kernel.
My home machine was running 2.4.2 and when 2.4.17 came out I decided it was time to upgrade. I gave it a whirl and failed although, I did save the old kernel so all I lost was time. After my initial failure, I read the documentation, the how-to, newsgroup postings and more, repeatedly trying to compile the new kernel. Nothing fancy mind you, just a new kernel and leaving out a few totally unecessary items like HAM Radio suuport. After at least 8 failed attempts I gave up and continued to run the 2.4.2 kernel that my distro had intalled.
After seeing your post and the ones about what you left out, I thought that maybe I had missed something. It really doesn't look that complicated. 30 minutes later, shutdown -r now.
The system restarts and the kernel still doesn't boot. Screw it! Obviously, I am doing something wrong. But, the point is that I am an engineer with years of experience in several OS (NT/2000, OS/2, Netware, OS390, OS9, IOS) and I am unable to get this simple compile to work.
I'm thinking that if I can't do it, there is no chance that Aunt Tillie is ever going to get past tar -xzf linux-2.4.17.tar.gz
A quick search at rpmfind.net gave quite a hefty listing of extremely easy-to-install kernels. When 2.6 comes out, shouldn't I just be able to install the rpm over 2.4?
I only posted a brief description of the article. I included a link to the article. If I had posted the entire article that would defeat the purpose of posting a link and providing the due respect the original website deserves
However had you clicked on the link and read the whole story you would also read about "Nephew Melvin" who uses the autoconf tool to quickly and easily modify a setup he had previously created. Or "Geek Girl Penelope" who needs a custom kernel to support a driver but has no time to become a kernel hacker.
The main point is not wether there should be an autoconf tool. There will be...regardless of what anybody thinks. The point is should it be included into the mainstream, available in the standard kernel and in the popular distributions.
http://www.kubuntu.org/
I say we all ignore E.S.Raymond.
He's full of shit.
we should be working on moving more and more kernel functionality into modules, which are loadable and configurable at run-time.
You mean like Mac OS X kernel extensions?
That that is is that that that that is not is not.