Slashdot Mirror
Linus Torvalds Speaks Out on Future of Linux
SlinkySausage writes "Linus Torvalds has laid out his plans for the future of Linux, including the 3.0 kernel [there probably won't be one], problems with the Linux release cycles and which distro he personally runs on his home PC. '"Compile everything by hand" ones simply weren't interesting to me,' Torvalds says."
guessing he's not a gentoo user :)
But really at this point, even if he stops developing the kernel, someone else will just pick up where he left off. I don't think we can ever really expect to keep one final generation of the kernel. It'll always be changing and morphing to new cpu's, hardware, etc...
cb_is_cool knows where his towel is.
Oh come on!
3.0 is a perfect excuse to break everything and allow your imagination to run riot. That's the fun bit!
Deleted
already slashdoted
I am thinking that his love for writing low level code is something of a failure of what will allow Linux to come into the mainstream.
And that is, getting the GUI perfected, fixing software distribution (standardize it, so no RPM or whatever... just a single package)... Click N' Run seems to be going this route.
I feel Linux can make some serious inroads on the desktop, but it has to be presented as a unified system. When you tell your friends you run Ubuntu (or whatever), then they are like "What's that?" There's no unity... if you run Gentoo, RedHat, Ubuntu... it's not "just" Linux, because even within Linux you have fanboys for different distros.
Anyway.. just throwing it out there. I'm still a Linux neophyte compared to most here, but I enjoy learning about it. I don't think there is a centralized focus for the future of Linux, other than through Linus who it seems to me, is far too geeky to realize what it takes to bring Linux into the mainstream.
The price is always right if someone else is paying.
The one thing I really like about Linux is that it adheres more to a Unix tradition of doing things continually and incrementally. Like, it drives me nuts that on Windows, to talk to SQL Server in C++, one has had to go from db-library to odbc to either OLE-DB or ADO... whereas, in a Unixy type mindset, one might ask, what really needed to change about db-library that required a whole new way of talking to databases? And, the answer is, not a lot. It is absolutely wonderful that in Linux there is a core set of APIs that always work, aren't suddenly abandoned to make a new feature that frankly, most people don't need.
So, in my mind, to say that there won't be a Linux Kernel 3.0 or a Linux 4.0 or something like that, is actually a GOOD THING. If you want dramatic, shocking, breaking releases that require you to rewrite 95% of your code to do the same thing, if you want to find that what you used suddenly can't work largely because it isn't supported any more, then Microsoft has plenty of that.
So three cheers for point releases, and here's to the death of "major" releases.
This is my sig.
"The good news is that a lot of hw manufacturers are actually doing the right thing. Intel in particular has improved wrt open source a lot, and for that reason I tend to suggest that when buying a machine, just make sure that you buy one with Intel graphics and wireless. That takes care of the two biggest annoyances right there."
Hmmm.
SO YOU'RE GOING TO DIE: The Comic for Dealing with Death
Sounds like he is sticking with the programming model of doing a large number of releases with small changes type model. Glad to see it actually, as this is the approach that I have been using on all of the software I build for work. What this does bring up, though, is the unfilled need currently of having an auto-upgrader software package where new kernel packages can be auto-upgraded and then migrated too on the fly without requiring a reboot. This would be quite complex I would admit, and maybe not possible in all kernel releases, but this is definitely something that needs to be looked at... Just my 2 cents worth.. :-)
There is a lot of new stuff happening, but it's in the main not specific to the kernel. New things the kernels needs to do are thin on the ground now. Not to say it'll ever be finished as such, just that there aren't any needed big new features. It'll take a major new shift in computing to do that, I suspect. Something way bigger than extensions or tweaks to x86/SPARC/PPC/ARM etc. I'm not holding my breath.
I may be stating the obvious, but the site is slashdotted, so I can't see what Linus has apparently said.
http://apcmag.com.nyud.net/7012/linus_torvalds_tal ks_about
Funnily enough, the only distributions I tend to refuse to touch are the "technical" ones, so I've never run Debian, because as far as I'm concerned, the whole and only point of a distribution is to make it easy to install
:)
Maybe we should send him an Etch install CD
Linus Torvalds, creator of the Linux kernel, has, along with others like Richard Stallman, literally changed the world of software forever.
;)
Linux-based distributions seem to pop up every day, more and more devices now run Linux at their core, from mobile phones to inflight entertainment systems, to the world's mission critical server infrastructures.
The development of the kernel has changed, and Linux is just getting better and better. However, with a community as large and fractured as the Linux community, it can sometimes be hard to get a big picture overview of where Linux is going: what's happening with kernel version 2.6? Will there be a version 3.0? What has Linus been up to lately? What does he get up to in his spare time?
I had the opportunity to chat with the original creator of the Linux kernel, Linus Torvalds, in a number of email exchanges.
APC: Writing an operating system kernel is a hard job. Why did you write Linux in the first place?
LT: Kernels may be hard, but partly because of that they are also interesting. I've always been more interested in "down to the hardware" details than in fluffy stuff like user interfaces etc, and an operating system kernel is about as down to the hardware as you can get without actually building it yourself (which I've also done - I was at a CPU company for seven years, after all). So I'm not into soldering irons etc, but I very much enjoy working at a low level, and thinking about how my software actually interacts with the CPU and other parts of the system. Besides, I really didn't realize how hard it would be. I really never expected to be still working on it 15+ years later
APC: What's the Linux Foundation?
LT: Heh. I just work here, you should ask some of the people who are actually involved in all the other things that LF does. It's basically the combination of OSDL ("Open Source Development Labs") and FSG ("Free Standards Group"), and is a vendor-neutral place for different organizations to discuss the issues they have, and trying to help Linux along. Part of what LF does is pay me to maintain the kernel.
APC: What are you doing with the kernel now? Are you working on it full time? What parts of it do you work on the most?
LT: I very much work on it full time, but I no longer really work on any particular "part"of it - I end up spending almost all my time on not writing kernel code myself, but on working with the flow of code and merging it all.
In fact, the biggest amount of actual source code I've written in the last two years is not in the kernel itself, but in the tool I use to just track the kernel development (called "git" - a source control management system).
So I still get to write code (and I send out suggested patches quite often - but usually they are along the lines of "so here's how we could handle this issue..." in order to prod others to actually do the final patch and testing). But what I do a lot more is go through other peoples changes and say "yes" or "no".
APC: The 2.6 series kernel has been around for a long time. Why?
LT: We used to have these big and painful development releases that took several years, and it worked reasonably well and people got very used to it ("2. is stable, 2. is development"), but it had serious downsides too.
In particular, the release cycles were so long that all the commercial vendors effectively had to back-port a fair amount of new code from the development kernels, and so development code ended up in the stable releases. Also, conversely, the vendors fixed problems in the stable versions, and sometimes the fixes were missed or weren't easy to then forward-port to the development series, because the two were just very far apart.
Basically, a multi-year development cycle simply doesn't work. It was reasonable and required for a while (we did some pretty radical changes there too), but with 2.6, the base kernel is in good shape, and we've improved our development process enough that we just don
Personally, my favorite aspect of this interview is that Linus himself basically considers the core OS now stable enough that anything new is likely to be incremental instead of ground-shakingly different, at least on the x86 platforms.
Which would imply a stability that leads to dependability which leads to usability which leads to widespread use. At least that is my hope in the enterprise, that the combination of commodity hardware with a commodity, high powered and stable OS can be coupled with increasingly powerful database engines such as mySQL, Veritas, etc. Oracle on Linux is now considered stable as well.
At home? stability leading to dependability leading to integration leading to crossover applications that will no longer depend on a proprietary OS stack to function. The only thing missing from my desired tool set on Linux right now is basically an easy to use, high powered MIDI to music recording and notation system -- and the pieces for all of that is already there -- it's my time to research and integrate the pieces that is in short supply.
I guess my point is that stability and upgradeability cause me to buy (several Linuxes and Win2K). Give me yet a large bulkier OS that doesn't really do much but add coolness (Vista or even XP) and I yawn.
...Open Source isn't the only answer -- but it's almost always a better value than the alternatives...
here http://apcmag.com.nyud.net/7012/linus_torvalds_tal ks_about
p.
If stability means it's dying. As I see it, the current Linux kernel does all that it needs to, and does it quite well. There is no need to upgrade it, because to do so would be an upgrade for upgrade's sake. Anyone in IT will tell you that to upgrade simply for the sake of upgrading is stupid, and will lead to a multitude of problems. The only reason that the Linux kernel would need a version 3.x is because of a fundamentally new hardware technology. Currently, software is driving hardware development; games are written requiring advanced graphics cards. In the 90's, hardware drove software development; chip makers like Intel put out a new processor, and then software was written to take advantage of the advancements of the new chip.
Even advancements in multi-core technology would not require a 3.x series kernel (unless I'm mistaken in my belief that the 2.6.x series supports multi-core CPU's), simply because once you cam make a dual-core CPU functional with the kernel, expanding that functionality to 4, 8, or even 64 cores is simply an expanding of the current code. And even if the current kernel does not support multi-core CPU's, that would be more of a 2.8.x series, rather than an entirely new kernel version.
Anybody want my mod points?
Because games don't work on them.
Maybe a hybrid os with a dual boot with reactos?
Linus' comments on Debian seem old. Calling Debian (and especially Ubuntu) "technical" in nature doesn't jive.
I share the idea that some systems I don't want to have to do a bunch of work to setup/maintain and my current choice for servers is Debian. desktop would be Ubuntu for most people. Personally I use Gentoo on my desktop, as I like the way it works.
The key advice remains "Try it all and use what you Like".
Anything is possible given time and money.
You know, he is pretty down to earth for a God.
You're a fucking idiot.
I only knew how to play make install and apt-get update!
...Windows is dying.
There fixed it for you.
>>Which would imply a stability that leads to dependability which leads to usability which leads to widespread use
while I agree with your Sentiment I must say that windows 95 was anything but stable, yet it still had widespread use.
i thought once I was found, but it was only a dream.
All right, so they'll duel first.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Same amount of work?? No way. If everything works, then it's only however much time it takes to compile and install, which is, what, 10 times as long as installing a precompiled binary?
If compiling was done once, fine. But whenever an update comes along, it's compile time all over again. How many packages might a typical user have? Maybe 1000? And how often is a typical package updated? Maybe 4 times a year? So, looking at perhaps 4000 updates per year, which comes to more than 10 per day. And how long does installation take? 1 minute per package? (For comparison, anyone have any idea how many updates Ubuntu has per day? Something like 2 or 3? Takes about 10 minutes to do?) So if time for compiling and installing is 10 minutes, that's over an hour per day spent staying up to date with Gentoo. That's too much. Gentoo doesn't lend itself to staying current.
And that's only if you don't encounter problems. What really irritated me was the installation of a package that had 20 dependencies that all had to be compiled and installed first, and which, after hours of work discovered that a suitable version of the 15th dependency wasn't available on any of the Gentoo sites, and gave up.
Gentoo might be pretty nice when networks and computers are so fast that a complete OS with a profusion of apps can be downloaded, compiled, installed, and running in seconds. Even then, there could be snags trying to get the extra configuration options right. Not saying that other distros don't have some of the same problems, like the lag between a Firefox update going live from Mozilla and the time a distro can get, tweak, compile, test, approve, package, and push out that update. But it seems simple practicality to go with a distro that avoids excessive work. Firefox is a little unusual in being one of the few applications that has a built in updater. For that reason, I usually rip out the Firefox that comes with a distro and install a binary straight from Mozilla, with appropriate permissions so that Firefox's automatic update will work. That's a bit of extra work, but I do it anyway. Gentoo however is too much.
Intellectual Property is a monopolistic, selfish, and defective concept. It is "tyranny over the mind of man"
I hope he was told the truth about Drop Bears and that he was kidding...
Thanks God he's not into distribution creation. His complaint simply makes no sense nowadays, nor has it made sense for quite some years now.
And by the way, it is kind of contradictory with his stance on Gnome.
Leandro Guimarães Faria Corcete DUTRA
DA, DBA, SysAdmin, Data Modeller
GNU Project, Debian GNU/Lin
On my FreeBSD system, I have around 500 packages, for a fully functional workstation.
I might need (read: want) to recompile 20 or so packages total (counting multiple recompiles of one package separately) a year. This takes a very limited amount of compile time, and with the exception of KDE and X, the compile time is minimal and you can keep working while the application is compiling. And you shouldn't have to recompile KDE or X more than once a year, or even that often, unless you are someone who absolutely has to be at the bleeding edge.
34486853790
Connection too slow for X forwarding? Try "ssh -CX user@host"
Enough with the Vista bashing, we're sorry.
Quack, quack.
Even advancements in multi-core technology would not require a 3.x series kernel (unless I'm mistaken in my belief that the 2.6.x series supports multi-core CPU's), simply because once you cam make a dual-core CPU functional with the kernel, expanding that functionality to 4, 8, or even 64 cores is simply an expanding of the current code. And even if the current kernel does not support multi-core CPU's, that would be more of a 2.8.x series, rather than an entirely new kernel version
I guess that a 3.0 version would be a suitable "label" to a conversion of the Linux kernel into a Microkernel architecture, I am not saying that it is going to be done, but I think that with the development of multicore technology and the overall new technology a microkernel architecture seems plausible.
In any case, I assume that just a modification of such size would make the version major worth of being changed.
Ubuntu is an African word meaning 'I can't configure Debian'
The kernel has supported n-core architectures for quite some time now.
Don't believe a word of it! Look at his flaky source - Wikipedia. Would you put your life in Wikipedias hands, or would you trust a genius such as Linus? You'll notice the interviewer never corrected him - because there was no mistake! Don't you make the same mistake the parent has!! The reason such rubbish gets in is all the city slickers that never go out back. In the wild bush, crikey they're everywhere, and they'll take you when you least suspect it. You can try toothpaste, vegemite, forks and other chicanery but the only true defense is a well developed sense of terror!
If you're looking for MIDI recording that generates notation for you in linux, check out Rosegarden. If you know of this program already and don't think it's easy enough to use, my apologies for assuming.
You think they'll bring those back? And you'd think NVidia'd have the good sense to allow the nvidia kernel module to supply an fb block device for dicking around with virtual consoles and non-X guis. Ah well.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
bork bork bork!
From TFA:
"LT: I spend a lot of time at the computer. But I'm writing this one-handed, because our puppy is sitting in my lap right now."
That's sick! You sick, sick, SICKO!
Actually, besides the fact that multi-core has been supported for ages, I can't currently imagine what kind of changes exactly would warrant the 2.8 or 3.0 re-labeling.
Even an improved graphics scheduler that's allegedly so amazing in Vista (read about it here a day or two ago) is just another switch, just like the kernel scheduler.
Ignore this signature. By order.
Linux ran just fine on 64 CPU systems back at the start of 2003 when SGI announced Altix. See, for example: http://linuxplanet.com/linuxplanet/reports/4612/1
If I recall correctly, that would have been running on a 2.4 kernel. Cores in a single socket vs. distinct CPU's is a really minor distinction, so all the work done to make Altix work will basically run just fine on a hypothetical Intel Core 4 Sixtyfourdro. So, 2.8 really shouldn't need to much work in terms of being able to handle the "new" multiprocessing.
Seriously, 1.0 was considered "feature complete" at the time of release, and there are some major architectural changes which will be required in order to improve scalability across multi-core as well as SMP systems, not to mention some fairly major pieces of work that are still under development which will need to be merged fully at some point (DAPL being one of the bigger). With the growth in the cluster market, I would also expect some meta-structure to go in to support the basic concepts. Even PCI-e 2.1 support is going to have a serious impact, due to the changes introduced in it.
If I was in Linus' shoes, I'd be pushing for these big infrastructure components to be readied and maybe placed in the -mm tree at this point. Once they're ready for the big time - which might take a while - I'd migrate them into the main tree and wait three or four cycles for last-minute bugs to settle down, then flip to 3.0 to mark the first of a generation of kernels that are keeping pace with the curve. I'd reserve 4.0 for when Linux is not only stable for mainstream use but defines the curve for OS development. I think everyone on Slashdot is at least aware of the research into new hardware technologies, new OS technologies and so on, so I don't think anyone seriously believes that Linux won't undergo more fundamental changes in its life.
Obviously, I am not Linus and he gets to do what he wants, whether I - or anyone else - would agree with his beliefs or not. However, his reluctance to flip digits is not new - I remember when kernels had a letter at the end to mark the sub-sub-version and it had to go into 2 letters because Linus ran out of alphabet. I also remember the first time sub-sub-version numbers ran into the hundreds. On both occasions, there was gigantic frustration with the absurdity. I guess he's forgotten the problems caused, or something.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
I for one welcome our delcious "kentucky fried penguin" overlords!
I only look human.
My mother is a halfling and my dad is an ogre, so that makes me an Ogreling
Did you first compile your compiler for your specific hardware?
Not sure if this actually makes a difference, but I've heard third hand that it does.
Granted, this is not a reliable source, but it made sense to me at the time. This person first compiled their compiler for their hardware, and then compiled all the source for their distro with this compiler. Any thoughts on this?
If I mod you up, it doesn't necessarily mean I agree with what you've said, sorry.
Even advancements in multi-core technology would not require a 3.x series kernel (unless I'm mistaken in my belief that the 2.6.x series supports multi-core CPU's), simply because once you cam make a dual-core CPU functional with the kernel, expanding that functionality to 4, 8, or even 64 cores is simply an expanding of the current code.
I'm sorry, but that's just not true. As the number of cores goes up, a great number of things changes and there's been tons of radical changes to support non-uniform memory, fix scaling problems with many cores etc. However, most of this has been implemented much earlier with SMP - basicly multiple processors on a motherboard. Not common in PCs, but common in servers which is where Linux has spent a lot of its time. Not as elegant but basicly to the OS the problem is the same - manage multiple cores. So no, you're wrong as it's far from trivial, but yes, most of the hard work is already done.
Live today, because you never know what tomorrow brings
Which would imply a stability that leads to dependability which leads to usability which leads to widespread use. At least that is my hope in the enterprise, that the combination of commodity hardware with a commodity, high powered and stable OS can be coupled with increasingly powerful database engines such as mySQL, Veritas, etc. Oracle on Linux is now considered stable as well.
Depending on what kind of infrastructure you needed, six or seven years ago, you were fairly likely to get funny looks if you announced you were running a significant chunk of your servers on Linux. If you were running a significant chunk of your desktop infrastructure on Linux, the funny looks were a dead cert.
Three or four years ago, the funny looks regarding Linux servers were long gone, replaced with genuine interest.
Today, nobody bats an eyelid about server infrastructure, and you'd be just as likely (if not more so) to get genuine interest as funny looks if you are seriously migrating desktops to Linux. About the only thing you can't easily replace is Exchange and the centralised configuration UI that Active Directory gives you (no, LDAP user authentication doesn't count). The centralised configuration isn't too difficult to work around, the full integration of Exchange is.
"LT: I think we have tons of areas where we're just better than anybody else. We handle portability better, we handle the development process better, and yes, we also end up having better memory management and a better filesystem layer than anybody else."
Let's see.
Portability:
NetBSD knocks the everliving shit out of Linux as far as supported platforms.
Development process:
It's his own process, of course he's going to think it's best. So how about another subjective opinion represented as fact:
OpenBSD's constant audits and focus on correctness.
Memory management:
Again, opinion, and again I have to lean (run?) towards OpenBSD.
Filesystem:
This is the same guy that's bitching about atime ffs.
It's also apparent that he's never heard of apt-get or the linux clones of BSD ports several distros like Gentoo use considering the way he's bitching about wanting things to be easy to install. Which itself never made much sense to me as NetBSD's pkgsrc is portable.
If I ever had a reason to bother with Linux again, I'd probably just go with Slack + NetBSD's pkgsrc.
I would be most interested to know what OS is on his wife's computer. If Linux, what distro.
The real "Libtards" are the Libertarians!
OK, a few days or even a week... but a month and a half on a piece that was covered everywhere is a bit of a stretch.
http://teasphere.wordpress.com - A little spot of tea
I wonder if your "name" is luser....
One of the most interesting things (to me, anyway) that Linus talked about in the interview as how proud he was of the technical merits of the kernel and of Linux as an OS in general. I thought that was fairly interesting.
I really don't want to try and turn this into a Linux vs BSD vs [something else] flamewar here, but since I'm not really qualified to comment on things like memory-management algorithms, I wondered if anyone wanted to weigh in on exactly what areas they think Linux really excels at -- from a purely technical perspective. I really like the idea that Linus is getting at, namely that the real confirmation of open source is technical excellence, but I'm curious exactly what areas the Linux kernel is "The Best."
In particular I've always been interested in how some of the different open-source OSes handle different technical problems. Is how Linux handles (just for an example) memory management quantifiably better than how BSD does it? And if so (or not) why?
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Linux's terrible support for SATA cards is a pretty good reason to upgrade. If the drivers ever actually materialize. /me looks over at 3 partially-supported SATA cards and weeps
A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
He likes to tend.
I'm not a ricer. I don't run the amd64 version. I like my OS to run what I want, and work well on my system. I've tried numerous distros (specially Fedora and Ubuntu, but I started with Debian), and they just didn't deliver. In my experience, the whole ubuntu and fedora style of doing things their way to make it 'less of a hassle' for the user has still not gone far enough. Linux distros are just simply not ready for full 'install-and-run-perfectly' functionality. That's why I use Gentoo. Maybe it's because I'm more comfortable using it than any other distro at this point, but I find that since I spent so much time configuring so many things by hand, I'm better able to troubleshoot the really dumb stuff when it starts happening. As opposed to ubuntu, which I'm running on my laptop. For example, one day, it suddenly decided that holding shift and pressing space bar would mean I get no space. This is a complete mystery to me, and I don't even know where to start, if it's something wrong with X11's dvorak keyboard layout, or if some app within the WM is using shift+space as a shortcut. Ubuntu isn't exactly troubleshoot-friendly... I find I get fewer mystery errors with gentoo than I do with more pre-built and packaged distros, and that's why I use it in my main box. People criticize Vista for smaller problems than Ubuntu has all over the place, but because it's linux, it's ok? And Fedora... you're telling me I should use a distro that doesn't even play mp3 files? Honestly, fuck that. When distros like Fedora and Ubuntu are actually solid, and don't take several weeks to package stuff, I'll switch to them. For the time being, since I'm a complete dork, I'll stick to my souped-up Gentoo install that gets me a whopping 1 second of accumulated extra performance per month.
These are really just novel stories. Linus is irrelevant at this point. Sort of like after Einstein gave us E=mc2 he ceased to be of any more use.
I disagree. We don't need a dictator for the kernel. But there are several candidates who could replace Linus, should he step down - Andrew Morton is the first name to pop into my head, but there are several other excellent choices. I don't think a committee would necessarily be bad, but would prefer there to be one lead developer, to whom all members of a committee would answer, but not necessarily a dictator. However the committee would have to be small and it would have to have certain requirements, like no person who is or has ever been a committee member of Debian is eligible, etc.
Just because A leads to B, that doesn't mean that every B was preceded by an A.
Linux is soooo Y2K. I run RedHat now.
I think there is a world market for maybe five personal web logs.
Your numbers are way off. The number of packages that take more than a minute at most to compile and install on modern hardware is, at most, about 20-30% of the ~600 packages I have installed (and 600 is a lot of packages, most people have less). Furthermore, the entire process is automated and you're free to use your computer exactly as if you weren't updating, so it really doesn't matter how long the compile takes. Thus, you can sync and update at your leisure. I usually do so once or twice a week, and it takes well under a minute of my time (per week!), which I think is quite reasonable given that that's all that's needed to maintain a completely up to date system tailored exactly to my needs.
For me, any distro other than Gentoo would be a lot more work, because I like to stay on the bleeding edge of software, and in my experience, Portage has always had the largest and most up to date selection of software that works with the least amount of fuss. Even if only one or two packages are missing or too old from a given distribution's repository, that's already going to cause me substantially more work than a minute a week to maintain it manually, and I ran into that situation frequently with other distributions. I've yet to find any software that I wanted that wasn't already available in Portage in several years I've been using it.
Another point to consider is that Gentoo is always current and up to date, you never need to upgrade to Gentoo n+1 as you do with basically every other distribution. Big upgrades are nearly impossible to get done perfectly (and software breaking is something that is statistically infeasible to avoid), so with a non-Gentoo distro, when something breaks, you have no idea what caused the problem, because everything was changed at the same time. With Gentoo, if something breaks after an update, (I or someone else) have a handful of packages to blame at most, making the problem exponentially easier to find and fix.
I won't say that Gentoo isn't daunting at first, because it is. I spent several hours doing my first Gentoo install (which is still running, over 3 years later), but now a new install of Gentoo takes me under half an hour, and a very trivial amount of time to maintain. The package management in Gentoo is second to none, and when it really comes down to it, that's the only part of a distro that actually matters. The choice is obvious to me.
Game! - Where the stick is mightier than the sword!
I used Gentoo for over a year, but it was sucking up too much time. The final straw is that Gentoo is often out of date (ironic) so I switched away from it.
Gentoo fans often claim Gentoo compiling gets you a faster machine. That was not my experience. My Gentoo box started up and shut down more quickly, but that was because it didn't have a bunch of unnecessary services starting up. The individual application speeds were no better than with any binary distro--which is what one would expect in this age of superfast processors.
If compiling was done once, fine. But whenever an update comes along, it's compile time all over again. How many packages might a typical user have? Maybe 1000? And how often is a typical package updated? Maybe 4 times a year? So, looking at perhaps 4000 updates per year, which comes to more than 10 per day. And how long does installation take? 1 minute per package?
Heh, that's optimistic. The problem with Gentoo is that there is no incremental patch mechanism. You have to reinstall the whole package any time there is a security update. The real stinger is that the packages often do not work. I'd say roughly 1 in 50 emerges quits due to some compile error. That's a lot of breakage on a typical system, which may have nearly a thousand packages--all of which often must be reemerged for security updates and for upgrades. Fixing broken emerges takes some time--"hmm, let me try unmasking the testing version to see if that works. Hmm, let me try changing the USE flags. Hmm..."
Gentoo would perhaps be acceptable if it worked a greater percentage of the time. Unfortunately it often breaks, and the response from other users is "oh, you shouldn't have upgraded your system" or "oh, any user should be able to fix that." True enough, so I have left Gentoo so that I don't need to bother fixing these things. No problem here.
Penny - plain text accounting
Someone dissing Linus? Protect the Queen!
http://www.faqs.org/docs/artu/
If you want a better understanding of the Unix philosophy then read eric steven raymond's The Art of Unix Programming. If you come from the windows GUI / Apple GUI world, it's a bit overwhelming / subtle to appreciate the elegance of the Unix principles. Besides, the pics of Ken/Dennis and a PDP show who's the Ubergeek.
Linux has been a powerful force, to be sure, but Ken and Dennis came before (they were mavericks too) and others will pick up the banner later.
Distros will rise and fall, applications will wax and wane, but the kernel is the engine that makes it all possible, thanks to the Linux team for making such a difference!
"Back when I was on RedHat, I'd see "package X" that wasn't part of the official distribution. So I'd find an rpm and try to install it. Then I'd find that I needed another library, and go searching for that rpm. Sometimes then things would work. But sometimes I'd find that some package was looking for things in SuSE layout instead of RedHat, or I was grabbing an rpm from somewhere that didn't play well with RedHat for some other reason. There was a non-trivial set of packages that I never could get installed and running."
That's why I prefer Debian based 'flavor's, all I have to do to install a package is (sometimes edit the sources.list and) type "apt-get install X". Same for update and removal.
mplayer is 35% faster? Presumably you mean mencoder, the encoding tool that comes with mplayer is 35% faster?
:-o I call bullshit on your bullshit on his bullshit! The only way you would see any benefit from this extra 35% faster mplayer, is if the generic package was juddering/skipping frames etc.
Sorry, but saying a media player is 35% faster is just retarded, what, you can watch a video in 35% less time than I can?
Yet if you can compare a generic packaged mplayer compiled for i386/i586 with something compiled especially for your processor on the same machine, you still must have some x86 compatible processor. What x86 processor do you have that has trouble playing mpeg files without skipping when using only i386 optimisations?
Fried in its own fat, you mean ?
Linux 2.6.x = UNIX System V
Windows 3.1x calc: 3.11 - 3.10 = 0.00
And that's that pretty much every Debian derivative (and Debian itself) has a vastly superior packaging system. Comparing Gentoo and RedHat is like comparing Windows 98 (Gentoo) and Windows ME (Redhat). Sure one is clearly better, but they are obviously both inferior to Windows 2000 (Debian), or even Windows XP (Ubuntu).
Much Love.
I hold very few opinions. I hold information based on observation and fact. If you wish to disagree, please use facts.
I agree that lots of people call the entire OS "Linux", but they are simply wrong (or maybe just being a little lazy or sloppy). It is one thing to use "Linux" as a sort of nickname for the whole OS. You are the first one I have encountered who insists that "Linux" is the full, proper name for the system.
Linux may have had the best filesystems some years ago, but I don't think it can make that claim now, with Sun's ZFS. (Incidentally, ZFS is open-source, but is unlikely to make it into Linux, save for an inefficient user-space version, because it's not licensed under the GPL. So it looks like OpenSolaris has the initiative in that area for the time being.)
That, and I could do without the preachiness. Maybe it's just me (well, and some friends), but the whole "it takes a village" ambience makes me actively avoid it. The African "unity" name, the touchy-feely logo, and the photos with every race laying side by side seem very contrived and fake.
;)
Having localized versions is very important and should be trumpeted, but in the grand scheme of things it is only one feature. It is not reassuring to me that this one feature is so obviously important to the distro maintainers that everything from the weird name to all the marketing revolves around this one feature.
In fact, if you think about it, the premise is exactly backwards. If everyone wanted the multicultural feel, they'd all be happy using the English version -- it would be "cool" and exotic. But the point is they don't want that.
(Also, the Human theme would be better named the "Human By-Product" theme due to the colors, that's easily remedied after installing
I mean, real collapsable threads. Because one asshole says "gentoo" in a FP and 99% of what was supposed to be about a Linus interview goes to Hell in a handbasket. So I wish it was possible to click on a "[-]" button the second I saw "gentoo" and be spared of all this.
Nothing against Gentoo, but this was a horrible example when this would've been a really useful Slashdot feature.
i ate crayons when i was a kid and now i have two braincells and the blue ones taste nicer
This is the first I have heard of alpine, I'll definitely have to check that out. I still use pine as my email client, and will have to check this out. I do have thunderbird installed, just in case I run into something that is cumbersome in pine, such as attaching multiple files to an email or viewing one with multiple image attachments.
Does anyone else here use alpine?
My beliefs do not require that you agree with them.
It's clear you don't fully understand compiler technology because you chose the wrong argument to backup your claim. I agree that recompiling isn't always necessary or the right thing to do. The whole idea behind compiling is that you compile once and from that point on you run "fast."
The main reason your argument is flawed is because there is no magic combination of compiler optimizations that will work optimally for every program. For example program A might run faster if compiler optimizations N was followed by M but program B might run faster if optimizations M was followed by N (note there are no such optimizations 'M' and 'N', they're just variables). In fact, some optimizations have a potential to degrade performance under certain conditions. Correctly compiling for optimization is a very hard problem because you need to determine what optimizations will even give you any benefit (this varies depending on the source code and hardware) and in what order to apply those optimizations.
To make the problem harder, certain optimizations may make overall performance faster (less time to execute) but system resource usage will increase dramatically. I took an undergraduate course on compiler optimizations and one of our projects was to benchmark a program based on different choices of optimizations. One of the optimizations called trace scheduling tended to generate some really nice performance gains. However, because it caused more parallelism more of the cpu's units where used and many paths in the code were taken that did not need to be taken. This cause the cpu to use significantly more power than the non-optimized version. So while this optimization increased performance, it could hurt under conditions where power is not freely available.
Finally, you may simply not know the best order of optimizations for a certain program. In this case you could have the compiler spit out all the possible combinations but that will take an insanely long time to not only compile but test. So you have to ask yourself, "how long do I want to be compiling?" One solution is to go purchase some time on a super computer to get your results quicker, but that is only a trade off of time for money. The end result may still be the same as you might find your optimizations were already optimal for your set of optimizations.
So I'm not convinced that we need to totally give up on compiling. At the same time I'm not convinced that everyone should be recompiling their sources just because they don't want a dependency on this and that. The real situation is there are a set of common platforms most people run, therefore we just need to target those common sets and compile for those different flavors. Right now most distributions just compile to be generic. I don't blame them either as compiler optimization selection is a very time consuming problem.
Ultimately, we don't know if our binaries could be running faster, but just because we aren't, doesn't mean everyone should stop trying.
dpkg-query -l | grep ii | wc -l
2306
is a typical number of packages for my debian workstations (some have more). I use FreeBSD and Gentoo, but I don't have enough workstations to justify the effort that they would require to duplicate the functionality I get with debian. (150 ports is typical of one of my FreeBSD servers that uses php mysql and apache)
I could see using gentoo on the desktop if I was developing desktop apps, or had enough workstations to justify a build server and quality testing of my workstations.
Work bio at MMWD
Linux has had very *fast* file systems, but it hasn't really had very *reliable* file systems. The traditional BSD file system still wins there, and since the introduction of Softupdates it's not been far behind the bleeding edge in performance for typical use.
Linux will move to 3.0 when it switches to C++.... ...
or C# :-)
There is no right to feel safe thru security vaudeville at the expense of everyone's freedom, privacy and tax money.
You're comparing apples to oranges there, Debian (and many other distros) split single packages into multiple packages where possible to somewhat emulate the modularity of USE flags and causes the number of packages to be larger than the number of pieces of software available/installed.
Fedora for example (used to) package Wine into something like a dozen separate packages, one for each sound driver (Alsa, OSS, NAS, ESD, and JACK, at least), along with a couple other package splits that I don't recall the details of. In comparison, Gentoo has exactly one package for Wine with a slew of USE flags and versions. You can see the same pattern repeat for lots of other software.
Game! - Where the stick is mightier than the sword!
Linus's famed aversion to politics shows here. He needs to be more open minded. I think he realizes that the free software crowd is more on his side than the commercial crowd, but he's uncritically accepted a lot of bullshit from them.
This view shows that he has not thought enough about the issues. Linux views people who have a different view as "pushing". If I were to use the same mindset, I'd say he was pushing a commercial agenda that threatens real software freedom. Because many more people look up to and will listen to Linus, whatever he advocates gets much more "push" than anything I say. If he paid more attention, he would "froth" more than me because he has much more of his life and energy invested in free software than a regular user like me ever will. A few questions down he hopes hardware vendors who don't cooperate, "die a painful death." People who go with "technically better" at the expense of their freedom soon have neither.
His aversion to politics has cost him - and that's the sign of a real idealog. Debian is not hard to use, even for a non technical user like myself. I'd say it was easier than Fedora in all things but adding non free software. Debian derivatives are easier still. Mepis, Ubuntu, Xandros and others come with all the non-free toys you could ask for. Mepis and Ubuntu have excellent compatibilty with the rest of the Debian tree, so you don't lose much more than a little stability and trust for the non free inclusions. My use of Debian has exposed me to the real beauty and stability of free software and I'm glad I made the jump from Red Hat years ago.
The thing that he should realize is that technical excellence happens when you have software freedom. In a world of bad patent laws and "trusted computing", there will be no new Linux kernels, or non you can run on commodity hardware. Future students in Linus's position will be thwarted if we don't guard the freedom that made his work possible.
Friends don't help friends install M$ junk.
I think you got Microsoft confused with Apple.
The way people and organizations select version numbers has always annoyed me and Linus is spot on on this topic. To me, version numbers are stupid. The only number that really matters is the revision number, all other numbers either encapsulate too much non-sense or require too much "thinking" and "creativity" based on the developer's part. At the end of the day it's just a label to say that this version was created after all of those other versions, nothing more nothing less.
In these days, versions and releases are becoming more and more of a marketing strategy or even a get out of jail card (see Google and "beta"). The real answer is, from the first day you decide to even write a document, a version number exists and should keep ticking with every change contributed to the project. That's probably too much information that the user doesn't care about so let's simplify it to only new builds of a binary. But the build of the binary could be from various sources with various optimizations and features. Blah blah blah. So the end result is someone gets a grand idea of "NOW let's give it a version number." BS, finish it and release it or don't. End of story.
At the end of the day, the user probably only cares about a couple of things: is the newer version better than the older version (what are the new features, fixes), is it compatible with my current platform software and hardware, and finally will it break anything or do something to make me very frustrated (dependencies, deprecated features). Version numbers should only be made to address these issues, anymore than that is just marketing. I don't care if it is major or minor in the developer's minds. That tells me nothing. Make the version number useful to the users, not another confusing marketing term.
"Linus sucks because he hasn't pissed away his entire life and soul for hating M$ as much as I do."
The Tao of math: The numbers you can count are not the real numbers.
Thanks, I'll be here all week. Be sure to tip your waiter.
that these days you can install Debian by popping in the disc, booting, and hitting Enter about ten times. Hell, they even brought back tasksel and created a useful set of tasks, so you can select "desktop machine" or "email server" :)
Your kernel does all you need until you get a new motherboard. I had to use an unstable kernel the last time I upgraded, just to get IDE functionality. I sometimes wish the motherboard companies would do their best to stick to a 2.x line.
"At least that is my hope in the enterprise, that the combination of commodity hardware with a commodity, high powered and stable OS can be coupled with increasingly powerful database engines such as mySQL"
Or could you use an already powerful and free database engine like Postgresql.
If I have seen further it is by stealing the Intellectual Property of giants.
I know a lot of people who use gentoo (me included) and performance is such a poor reason for using gentoo. Unfortunately whenever there's a thread mentioning gentoo on /. someone jumps right in with a fat target on his rice bowl.
The best reason for using gentoo, and the reason I've been using it since 1.4 is maintenance. I've had the box I'm writing this on for over three years and it still has the same root file system on the same drive. When the drive dies I'll probably copy right over to the new one.
So, yeah, I see the reviews on how gentoo isn't really so much faster than a binary distro, or how it takes so long to install. I wouldn't know anymore because it's been years since I had to install Linux. If I were using suse I'd have to wipe everything and start over every 18 months, and yeah you really have to because pretty soon it becomes impossible to find anything built against your tool chain. When I first abandoned suse I thought they were doing it on purpose to make me buy a new box set, now I know better but still, why should I have to reinstall my desktop over cause [distro]-10.3 RELEASED!
Anyway please ignore any gentoo users that start bragging about "performance and optimization".
Samsung took back my unlocked bootloader because Google wants me to rent movies. They're both evil.
"Compile everything by hand" ones simply weren't interesting to me,' ...I guess he's getting old too.
I totally understand his point of view...
It used to be cool to compile everything, there used to be time to compile everything. But then something happened - a decade or something...
We got older and if not wiser, at least found out that there is more to life than compiling kernel. While compiling kernel is still fun, it's just not always convenient. You have other things on your mind now, and you've actually come to realise that the other things are just as fun and important. - I guess that's what they call growing up?
But hey, you younger nerds: Keep on compiling! Keep on hacking those all nighters! We older DO need your enthusiasm. You have all the time in the world to grow old and growly - and you will, like it or not.
It took a while, but I got there. - I'm actually older now!
If all else fails, pull the plug and get out...
The Life is out there...
(Run-on sentence alert...)
Additionally, Linux audio is a failure. Even ALSA. There simply should not be any blocking of audio I/O at this stage. Relying on a happy coincidence between the app developers' sound server choice, Distro-X's favorite sound server, and/or the presence of premium multi-channel (instead of single-channel) soundcard HW just to keep sound apps working, accessibility aids speaking, softphones and calendar alarms ringing when they should be (instead of sitting mute because a Flash animation was left running in a minimized web page somewhere) is just not cutting it. Or how about a call not going through because an email or calendar alert went off or watching a video/audio stream at the wrong time?
Here are some wrong answers that are often repeated to me by the clueless:
* User was supposed to buy the premium sound hardware (which doesn't even claim Linux compatibility)
* User should pay attention to what apps are using which sound servers, and juggle the apps accordingly.
* User only has to visit ALSA development site, open a CLI and read Howtos.
* User can install and configure the mixing OSS driver any time they want (its free as in beer until the trial expires).
* User clearly did something wrong if Skype for Linux is either silent or echoes like a tin canyon on every single PC + distro they've tried.
* User deserves it because they haven't given up their real life to get in touch with their inner penguin.
* Blocking sound output is appropriate default behavior.
* User is accessing the wrong mixer panel: Run alsamixer from the CLI instead.
* User lazily ran their app from the supplied icon instead of from the CLI using an esddsp, artsdsp or similar wrapper.
Clearly this mountain and a half of BS and misdirected blame stems largely from Linus' inability to tackle this crucial problem after all these years. Neither he nor his delegates are capable in this area, although somehow I wouldn't be surprised if they rationalized it by thinking the issue of smooth audio output (icky desktop feature) was beneath them.
The future of Linux as a environment -- either desktop or server -- is out of the hands of the kernel developers and has been for some time. They can make things slightly better, or make things worse (which means we'll stay with an older version), but...
In all seriousness, how many here think that changes in the kernel are really going to be the big news for "the future of Linux" for the next decade? The revolution is going to happen elsewhere.
Linus is an interesting guy, but as long as he's focused on git and the kernel he's not going to be the guy who takes us somewhere new.
This really caught my eye in the article:
>> So instead of having two or three years between stable releases, we now have two or three months. Which means that the vendor kernels are much closer to the development kernels, and avoids a lot of the problems we used to have. Everybody is happier.
Oh. Oh, my. Pleae, please tell me this is going to be true. I am so very tired of under-experienced kernel developers who get a fetish about a particular kernel, turn it into an entangled and never documented proprietary mess, then force other people to backport things from the next major kernel release. I'm also very tired of the profound pain of major kernel upgrades.
Keep the upgrades small and frequent: small changes mean small mistakes, even if it breaks major OS vendor models of "the kernel we released with version 2.0 is the kernel we still support for it seven years later".
Only to idiots, are orders laws.
-- Henning von Tresckow
Linus was asked to list areas where Linux has a technology lead. He came up with portability, memory management and filesystems. Now it is possible that at some time in the dim and distant past Linux provided the best filesystems technology and the state of the art memory management but neither of these assertions are true today nor have they been true in the recent past.
As for portability, well yes Linux has been ported to a lot of platforms it is however a bit of a stretch to conclude from this that Linux is inherently more portable than Solaris or BSD.
So his three claims for technology leadership are mostly plain wrong or arguable at best.
Linus did however help introduce the concept of good enough technology probably not the technology leadership he was asked about hence the reluctance to claim it on Linux's part.