Slashdot Mirror
What Actually Makes Up "Linux"?
David A. Wheeler sent in linkage to his extensive analysis of the true
size of Linux. There's an amazing amount of information in here, and although it focuses on Red Hat 7.1, it still has tons of interesting bits of information about the code that makes up the distribution. Break downs include languages, licenses, cost estimates, and stats that in no way clear up the legendary GNU/Linux debate that will undoubtably be engraved on tombstones somewhere.
Several hundred utilities.
And three hundred and fifty thousand annoying slashdot trolls.
Over half of which use windows.
How inappropriate to select vim, which is BSD licenced, not GNU.
In the glorious Digital Rights Management future, you will not be allowed to cut and paste in Windows without paying a fee to the owner of the web page...
If you want Microkernel then go and use Hurd. And when your head Hurds you can come back to Linux if you beg
Open a Microsoft Developer Studio project, run the debugger, and then try to open a Microsoft Internet Explorer 5.01 window simultaneously.
After you've rebooted, come back and tell me how far ahead of Linux microsoft is.
Most Linux users work without vi ever, let alone every day. They seem to have mistaken pico for an advanced editor.
--
Pico is easy to type in, it's not easy to use. If you actually want to do more than the equivalent of vi in insert mode, you're screwed. If that's all you want though, I agree it's a perfectly functional e-mail editor/basic text editor.
I agree that one should always use the right tool for the job. You don't say "I fancy Pizza" and then go buy tomato soup.
We'll take a common occurance for somebody who writes scripts, "parse error at line 1024". How to get to line 1024 in vi ':1024'. How to get to line 1024 in pico... page down page down page down page down, ^C (check to see what line i've paged down to so far)... page down page down page down..... 2 minutes later the pico user has managed to find the spot in the code that the vi user found in slightly less than a second.
Now let's say you finally find line 1024, it looks like the problem is that somehow you have mismatched braces. using vi, you check where your braces line up by putting the cursor over one of them, and hitting '%', instantly jumping to the partner brace. in pico.... well you go through the whole subroutine counting braces manually, because it can't help you. I'll gladly admit, the vi way is obfuscated, but at least it exists. With pico, you're just flat screwed.
As far as emacs being bloating... hellz yeah it is. If it had raw device i/o, and a boot loader, it'd be an operating system. But again, it all depends on what's the right tool for the job. If you get paid to handle large source trees, then time spent learning emacs (or any other IDE, be it visual slickedit or whatever) is time well spent. If your job is to troll slashdot, then that time is wasted.
Next time I suggest that you make your troll slightly less aggressive, and use better logic. While false logic is an excellent tool in the troll toolkit, it needs to be subtle to be effective. Additionally, you should use a user account, as your troll gets seen by fewer people when it's posted at 0. A user account increases readership, and it also makes the parent more likely to take your troll seriously, as it adds credibility. Older user accounts work best while trolling, as low UIDs increase believability. Preferably find an account with a UID 200, seeing as that's when the trolls started showing up.
--
Several hundred utilities.
And three hundred and fifty thousand annoying slashdot trolls.
--
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
If that's a community, it's the most dysfunctional one I've ever heard of. The "excellent free or shareware apps" are first and most importantly not Free at all in virtually all cases, and in my experience, of very poor quality. For every useful piece of software that functions as documented, there are 100 that bluescreen the system at startup - or never start up at all. It's ok not to be a professional hacker - I'm not one either. But there are professional amateurs and then there are amateurish amateurs. Almost without fail, the people who produce windows cruftware are in the second group. There are certainly enough of them writing Free software too, but I find far more from the first group in our community.
A small dot? I disagree. The point of this paper is that even if you define a community by sheer bulk, the Free Software community is a large one with serious resources. If you start to factor in things like quality and motivation, I would suggest that in fact the Microsoft community is a single trailer park full of dysfunctional trash next to our vibrant culture.
Windows is made up of the following:
You can moderate this down, but I challenge you to find proof that this situation is otherwise.
I'm sure Microsoft has spent that much over the years on Office+Win9x+WinNT+Backoffice+etc (basically the functionality provided by RH 7.1).
Microsoft has approximately 20,000 programmers from what I've heard, which at (say) $100,000/year gives you $2 billion/year.
Probably a better comparison would be lines of code created/changed, although really what we care about (and would find nigh-impossible to measure) is the increase in utility of the software.
Ooh, a sarcasm detector. Oh, that's a real useful invention.
Actually, two million lines of code for the Linux kernel is astounding considering what it does. I mean, what other OS can support as many architectures and devices in so few lines?
I mean, let's look at Linux 2.4.5. First, I see that in the arch directory, we have:
alpha, arm, cris, i386, ia64, m68k, mips, mips64, parisc, ppc, s390, s390x, sh, sparc, and sparc64.
Then we have the drivers directory which supports pretty much every bus I have ever even heard of and more devices than you can shake a handfull of sticks at. I'm not going to list them all here. But suffice it to say that everything from my video to my visor is supported.
All that in only 2.4 million lines of code.
Oh, and a microkernel isn't going to save us from a larger kernel tree, it may even increase the size of the code. We'll still need a TCP/IP stack, and a process scheduler, and drivers for all our nifty hardware. The only way to shrink the kernel tree is to fork the Linux kernel into platform-specific ports, but since that would cause massive redundencies in the drivers and a lot of the core code, there doesn't seem to be enough incentive for that to happen just yet.
Improvise, adapt, and overcome.
RMS can only name what he wrote; the fact of the matter may be that he (and many other volunteers at the FSF) wrote a large piece of a free operating system, which they are still working on, and have named "GNU".
Red Hat, Inc., gets to name their product, also. So does Caldera, and SUSE, and Mandrake, and Slackware, and Debian, and TurboLinux, and even LinuxOne as lame as they may be, among many many others, and not one of them has to provide a single acknowledgment in their advertising (which includes the name!) to Linus, RMS, or any other person or entity who wrote most of the code. This is a key freedom, according to the FSF (http://www.gnu.org/philosophy/bsd.html)
I believe that freedom is in fact the largest and most important advantage that the group of operating systems we are talking about has over the proprietary alternatives. Personally, I like the fact that my data is not being held hostage by a potentially hostile corporation, and that I will always be able to control my own destiny when using a Free system. Many of the people I support had no end of problems when one of their favorite software packages had no Y2K updates, because the publisher had gone out of business in 1998; other people I know are stuck because their favorite game was produced by a company which went out of business a while ago and never fixed the memory leaks before they collapsed. These are not problems when the software is Free.
However, it is rather disingenuous to insist that Freedom must include the freedom to fork the project and then insist that the old name must be kept as part of the new name, as is the case in insisting that it be called "GNU/Linux". That's just lame; the project was forked multiple times (once for each different distribution out there), as the people who created each of those forks has a perfect right to do, according to the freedom guaranteed to each of them, and they are not required to even give any credit in their advertising.
Now it may be a fact that completely ripping off someone else's project without any attribution and without providing any added value is in very bad taste and probably means that the individual or group in question is just a free-loader, but it also happens to be completely legal according to most Free software licenses.
If the FSF ever "fixes" this problem, they also ought to simultaneously issue the world's largest apology to the BSD project for the implied insult which has been up on the GNU webpage for years. Otherwise, the noise in "Linux" vs. "GNU/Linux" just sounds like petty jealousy.
Two minor points. First of all the article (which is very good BTW) did a very scientific, highly detailed accounting of the software in a typical (RedHat) distribution. Over half of it (55%) was GPLed according to Mr. Wheeler. The article actually dealt with the GNU/Linux versus Linux debate and stated that while the Linux kernel + drivers is the largest single contribution, it is far overshadowed by the code that whose copyright belongs to the FSF (not too mention projects that are officially part of the GNU project but whose authors retain copyright).
Of course, I still usually call it Linux. But if I was talking to RMS, I would call it GNU/Linux, and I would prepend it with "Thank you sir." Without GNU software none of the free Unixes would even have a compiler, much less a useful set of tools.
Another small nitpick is that Perl is actually licensed under the GPL. If you don't believe me check it out. Perl is also dual-licensed under the Artistic license (which supposedly is supposedly not terrible well-written from a legal standpoint), but that doesn't mean that it isn't GPLed.
Python, on the other hand, is not GPLed. It also doesn't look like line noise, but that is another debate.
The overwhelming popularity of the GPL is why there is so much interest in making sure that open source licenses are GPL compatible. Even those Open Source developers who take issue with the GPL and the FSF have gone to great lengths to insure that their software is GPL compatible. This isn't because RMS has some sort of mystical mind ray that makes people submit to his wishes. Instead it is because GPLed code makes up the lion's share of Free Software. If the software isn't GPL compatible then it is cut off from being integrated with a very large pool of software.
Uh... percisely! :)
LOL. Time to sleep now.
What I find amazing is that the Mozilla source code is larger than X (which isn't all that amazing when you look at the size of their respective tarballs). On the one hand, hats off to the Mozilla developers who've managed that monstrosity to almost-1.0. On the other hand, why is a browser bigger than X?! That's insane! That shows there's a lot a feature bloat in Mozilla.
What I failed to find in that article, though, is how much of that "operating system code" is actually beta code. Mozilla and other programs like it can hardly be considered OS code as they haven't even reached any level of maturity.
And that's why the system is GNU/Linux, and not 'Linux', which merely refers to the kernel.
A lot of the code they're listing as "Linux" code isn't GNU code at all. It's released under the BSD license (e.g. Apache). It's released under the Artistic license (e.g. Perl). Calling the system GNU/Linux simply because it has some GNU tools on it is like me calling my Windows box Netscape Windows because I have an old version of Navigator on it or GNU/Windows because I have GNU apps on it.
I think the reason people are more apt to further describe Linux as GNU/Linux is not because it uses GNU apps, but because it is released under the GNU Public License.
Basically the same thing. It's used as part of the Remote Installation Services.
Win2k has a capability of net booting a Win2k desktop image off the network. Well, more net installing rather than booting.
It makes it very easy to setup and configure new workstations. Turn computer on, hit F12 or whatever it is to boot off the network, enter username/password and a Win2k image customized for that username is downloaded and installed.
What do you mean there is no "Windows community"?
I go to local Microsoft events periodically, and I run into people I've worked with in the industry or know personally.
I belong to several email lists which have on them people from all over the country interested in various aspects of Windows programming.
I've been involved in open source development projects using windows, and even utilize sourceforge for our project to some extent.
There are even somewhat famous people that we will go out of our way to see speak. Some work for Microsoft, some don't. Don Box, Roger Sessions, Bill Gates, Steve Ballmer, etc. etc.
How is this not a community? What is it that you see as different about Linux that doesn't exist with Windows?
Perhaps it is just that you are unaware of the Windows development community, and the Windows user community is so large that you ignore it?
I'd actually like to see someone name one part which is ahead of Microsoft. Just one.
I've yet to see it, I see lot's of areas where Linux has copied Microsoft, Sun, etc. I see many ares where Linux is years behind Microsoft, Sun, etc.
But I'm not seeing the "innovation" or "advances".
Please help me! Where are these advances? I really want to know!
There is more to Microsoft Windows than it's ability to have a richly integrated cut -n- paste functionality. Besides that advance came back in '91 with Windows 3.1.
Microsoft has said they are committed to providing IPv6 support when there is a need for it. As of today, it still has not been clearly defined and not in high use. But you can go to research.microsoft.com and find an implementation.
I guess having many windows managers might be important to someone. I modify and recompile software all the time to change it... that's my job and I can do it as well, if not better, on windows than on Linux.
Microsoft initially started with NT by making it portable. The original development was done on MIPS, and when it was released it supported at least 4 different popular CPU architectures. But there was no market for anything other than Intel architectures.
Again, I would not consider that an "advance" of Linux, since Microsoft was clearly doing it first.
If I'm going to run Web/file/ftp/DNS/mail I'm going to use Win2k because it's clear that Linux is not the way to go. I don't understand why you jump to such a conclusion?
Ok, it's clear from your response that you obviously are unfamiliar with Windows.
I could go through point by point, but... come on. You're aware Win2k server comes with TFTP server as part of the OS?
You're aware that all of those free office suites for Linux are also free for Windows? Staroffice most notably.
Come on! Give me something challenging. I want an advance that has some meat to it!
I'm sorry, but...
Are you on crack?
How could you possibly claim that the development environment of Linux is light-years ahead of Microsoft? It is the development tools which are the killer apps. How do you think things like Office are possible?
By the way... Turbo Pascal had the first IDE and was released about 3 years before Emacs. (1982 versus 1985)
I think you need to get out more.
P.S. I still have my copy of Turbo Pascal v1.0 for CP/M-80 if you want to debate that point further.
I guess it depends on your needs. Windows 2k certainly contains a very rich scripting environment that is easily as powerful as the BASIC which came with MS-DOS.
.Net is how the compiler will be distributed. Right now .Net distribution is in two pieces... The .Net SDK which includes the runtime and compiler, and Visual Studio.Net which includes the IDE, debuggers, etc.
One of the pieces that remains somewhat unclear about
It almost appears as though Microsoft is intending to redistribute the compiler for free, and offer the Visual Studio.Net environment as a value-add product.
I think this makes some sense. It allows development for amateur home users at low cost, but provides the enhanced productivity pieces to corporate users who can afford it.
I did check my facts.
.Net.
http://www.jwz.org/doc/emacs-timeline.html
Your referring to a set of macros written for the TECO editor. The first release of Emacs as a standalone editor occured in 1985.
Interesting, I actually like Hungarian notation when working with C/C++. But you'll be happy to know that it's being replaced in favor of Pascal style notation with
Anyway, you're right. KDE has come a long ways, it's now up to the stage Windows was in 1994. In another 12 years it just might put Microsoft out of business!
Honestly. I owned an Amiga, and while I remember the clipboard I do not recall it ever being as extensible as the one which currently exists in Windows. I had my Amiga from Workbench 1.2 on through 2.04.
As far as the stack overflow goes... It's my understanding that the Sparc processor already disallows this. That is memory defined as stack space cannot be used to execute code. This is something down at the processor level.
It's a good idea, but obviously it is better implemented in the processor. Trying to do it within software does not guarantee what you think it does.
It's curious though, the greatest feature I recall from AmigaDOS was the device labeling. Being able to alias 'CYGNUSED:' to some location on my harddrive 'DH0:Editor/Cygnus' or whatever and run it from that new aliased device was quite cool. Moving the files to DH1: meant only having to rename the device alias in my system startup folder.
That alone was probably the greatest innovation I've seen that has been completely lost.
Unfortunately the Amiga became outdated back in '91 or so. That was when I finally sold mine, but it was very cool when I first bought it in '87.
"Their security model is based on security-through-obscurity and a daddy-knows-best approach. "
Sorry, wrong.
"It is worthwhile when you are a beginner, "
Sorry, been in this industry since '82. Was a Unix admin for a time from '92-'96, and would hardly call myself a beginner.
"You can't tell me all your Windows software is legal with a straight face."
Actually it is. Microsoft provides many ways to purchase their software at a low cost, if your in the know. The most expensive piece I've purchased was Office XP upgrade recently, but only because there was a signifigant bundle/rebate.
"Microsoft clearly aims to control. "
Oh dear.
"The obvious answer is, don't use Microsoft products or services. "
I suppose if you are a beginner and can't see value.
"You obviously enjoy being microsoft's bitch. "
Oh my.
"It costs me nothing but time. "
Which is a worthwhile proposition if time is worth nothing.
"This explains the poor quality of Windows software, especially shareware. "
Which is unfortunately(for Linux) far greater quality than the bulk of the stuff out there released Open Source.
Anyway, been there done that.
Nice attempts at insults, but I'm afraid they don't deflect your obviously poor argument.
No, but personally, I chose to just get used to it. There are some things that people just won't learn, no matter how much you try to teach them. =)
Um, kernel modules can be effortlessly separated from main source tree and compiled separately (assuming, of course, they depend only on the kernel headers)...
For example, right at the moment I have one device (QuickCam Express) that is built completely separately from main kernel tree. A couple of weeks ago, I had two such modules (Aureal Vortex2 needed a separate driver).
(Disclaimer: I'm not a kernel hacker.)
From the site:
"Finally, any software released into the public domain can be re-licensed under any other license, so there's nothing that keeps public domain software in the public domain - any of the other licenses here can ``dominate'' a public domain license."
IANAL
As I understood it, once a piece of software entered the public domain, it didn't matter what else you tried to do to it, it was still in the public domain. Even if you applied a proprietary license to it, anyone could use the software outside of that license because it had been in the public domain.
So why is he saying that it could get licensed and therefor be removed from the public domain?
neo
What's being asked here seems to me to be simply: "We know that a kernel isn't an operating system. So what is 'linux'?"
The difference is the GNU System and the utilities that were built up beside the linux kernel and supporting it. The difference between linux the kernel and linux the system that we all know and love is the GNU System.
And that's why the system is GNU/Linux, and not 'Linux', which merely refers to the kernel.
-- Truth goes out the door when rumor comes innuendo. -- Groucho Marx
You forgot:
2 camps of widget bigots
385 different versions of Solitaire for each widget set
1,675,394 would-be amatuer sysadmins trying to figure out why their laptop's soundcard won't work
3 tribes of BSD-fanatics jeering at the Linux proletariat
Potato chips are a by-yourself food.
I really don't feel that just b/c Linux cannot do this that MS is *years* ahead. This is something that MS wants to do, this is apparently not something somoene else has been motivated to do in Linux. Thus, it isn't really something that the majority wants (in theory, which I know is flawed)
Honestly, if the little gadget is necessary it is usually started fast.
Linux has its parts where it is ahead, MS has its sections where it is. I still think that the advancements mad by Linux in the time frame it has been available are more than what MS has done since its inception.
the more devices that are supported the greater number of lines. But think about it, re-compile the kernel with only support for devices you have installed, drop the alternate desktop packages, and you have a LEAN MEAN FIGHTING MACHINE. :)
This means that in order to draw in the 'uninitiated' some distro's have opted for a bloat of support. Next thing you know RH will be PnP
errr....umm...*whooosh* *whoosh* Is this thing on ?
tools, but you have a point. If this keeps up it could be GNU/BSD/GPL/LGPL/YADA/ETC/Linux, with you favorite distro's name and splashy release name added in to boot.
errr....umm...*whooosh* *whoosh* Is this thing on ?
between a Penguin and a YAK :)
errr....umm...*whooosh* *whoosh* Is this thing on ?
This was a quite thorough, well-written document all until the point he mentioned Bill Gates. Well, actually not Bill Gates himself but the immortalised words from his "Open Letter To Hobbyists".
In particular, the bit about documentation. The thing that Linux lacks these days is decent documentation in alot of areas, in particular things like devfs (which the author even admits is now poorly documents (the instructions that are available are now out of date)).
Coming from a BSD background (no, this isn't an excuse for a platform war - just hear me out), documentation is just as important as the code itself. This sometimes means that availability of certain features in BSD are a generation behind that of Linux, but when they arrive, the documentation is top notch, containing correct spelling and grammar, notes what bugs are present, provides examples of correct usage (this is especially relevent in documenting programming functions whose incorrect usage may have a security impact) and so on. Overall, it's an issue of documentation quality.
The author of the paper may scoff in the direction of Bill Gates, noting the ability of the Linux community to create and maintain an operating system, but what he's done in the process is brought the whole paper down by exposing the single thing that Linux as a "disparate sources, one distribution" model operating system can never have as what Microsoft products and, from my perspective, the BSD operating systems have - documentation that exists in a single form and written in a style that is consistent across the entire operating system. (This is not the case with Linux. Some things use manpages, others use "info", others use textfiles, others use html documentation. Heaven knows how a new user on Linux (advocacy is about attracting new users, right?) is supposed to navigate this mess without a considerable level of pain and/or persistence).
And before you let the flames begin, have a poke around on say, the NetBSD/OpenBSD/FreeBSD sites' manual page listings on their website and compare them to the ones you see on RedHat and so on.
Linux is what is keeping me from meeting women. : (
...i know this, and still, I find myself compulsively rebuilding my kernel.
Know what I like about atheists? I've yet to meet one that believes God is on their side.
2437470 source lines of code for the Linux kernel. Doesn't that worry some people out there? We have a monolithic kernel almost two and a half million lines long. I think that by 2.6 the kernel is going to collapse under its own weight unless the designers decide to reorganize it in a fundamental way. Maybe it's time for a Linux-Hurd fusion project that will turn Linux into a true microkernel.
Qu'on me donne six lignes écrites de la main du plus honnête homme, j'y trouverai de quoi le faire pendre.
Y'know, let's look at some other statistics. What has MS killed?
Dr DOS--yeah, there was a settlement, so the real story will never surface. But we all know what happened. So you can cross off the DOS achaeivement--what they acheived was getting a technically inferior solution rammed down people's throats.
OS/2--i know, I know, IBM killed it themselves by not giving it what they could (or, insert other theory). But, by accounts from all over, it was better. In the free software world, while there are casualties due to popularity/ego/etc, it's not nearly as bad as in proprietary models, where it's "to hell with the user, this is our revenue".
In the same vein, Lotus or WordPerfect may very well have been technically superior, but were simply deep-sixed by MS' deft use of control of the OS.
Windows 2000 I don't have much experience about--I'll give them that one.
I think MS bought one heck of a lot more of their starting stuff than Linux did, but it's a tenuous point anyway.
I think you have to take what they add and then remove what they subtract. Free software doesn't kill. (Ok, it might kill free time and overpricing, but I'm talking about technology.)
--
Liberty uber alles.
But you don't get all of those things out-of-the-box with a MS OS.
Try doing a point-to-point comparison of what you get with RedHat vs. NT. NT does not come even close to giving you a comparable value.
domc
Seriously, Windows can't really say that because there is no real "Windows community
Wanna know why there's no "Windows Community"? -- I live in the US, where most (but not all people) speak English. Even so, I notice there are very few clubs where people can hang out and talk about what a great language English is.
I also notice that there is no "HTML community". Is this because there's something wrong with HTML, or is just that common standards aren't worth forming a fanclub over?
However, anyone who thinks there aren't subcommunities in MS-space isn't looking hard enough. The local rags have adverts for Access programmer user groups, MS Office users, and so on. Not to mention a million small companies which do support and development for MS platforms and are usually completely loyal and perform their community service around the water cooler. And thousands of "partner" events and trainings and seminars that Microsoft puts on for these folks to network. Why hangout at a clubhouse when you can get paid good money for community activities?
Furthermore, the fact that there is a "Linux community" indicates that the OS does not have a mature user base. I can see valid communities forming around, say, Postgres or Perl. But the guys at the LUG are more interested in looking for like-minded pals than they are interested in discussing their configuration files. There ain't enough there there to form an authentic community around, and I think you could already find that the real Linux users aren't the guys down at the clubhouse.
--
Business. Numbers. Money. People. Computer World.
See, the joke is the jwz quote "Linux is only free if your time has no value." In both cases you have to put in the time to get the reward of supposedly free stuff.
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
On the other hand, software distributors and systems integrators have the benefit of using the code without paying too much for it, driving down their costs and either improving their profits or decreasing consumer costs. I bet in the end the economy is still better than it would have been.
Also, you'd have to spread part of the whole amount over many years, since GNU tools have been under development for a long time.
I work on stuff that I find interesting. It lets me get away from the sometimes boring stuff I do at work, it's a nice release valve and learning experience to experiment with some different kinds of code than I would normally see during the day, and in the end I get software that works more like I want it to. I don't think there are really that many people in free software "working for free" - I think most people get more out of it than they put in, and that's why they do it.
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
It might be more productive - I post on /. a lot more from work than I do when hacking at home :)
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
Easy - just run Sendmail on an underpowered box so it's as slow as Exchange, and every once in a while go back into the computer room and whack the box with a big axe to simulate the Microsoft showstopper bug-o-the-moment. Then, after a week of downtime, announce that your internal IT team has received great support from Microsoft to solve your problem (note: not their problem!), and slowly crank up the sendmail box again.
Lather, rinse, repeat.
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
Man, you had to ask.
We'll both sit here. You with your Windows(9x,NT,2k,Me, your choice) Box. Me with my Linux box.
Who can do the following easier and faster?
1. Find out what TCP ports the other is running, identify the other's operating system.
2. Run a reverse-arp server. Or TFTP. Or a number of other servers.
3. Find out how many files that start with the letter 'T' there are on the system, then sort by size. Ascending.
4. Stop the other's machine from running.
GUI
1. More configurabilty for look and feel. Don't buy it? Visit e.themes.org, kde.themes.org, gtk.themes.org... www.themes.org See for yourself.
Price: $0. That's quite a bit ahead. On top of that, major bug fixes happen in hours or days. rarely do they take longer. Plus, if you're a company depending on this bug being fixed, you can do it yourself. That is, if you have a business case for it. If you are dependant on it, the money should be there to fix it. If not, you have other alternatives.
You can't get any of the following for free for Windows, as far as I know. A decent port scanner, MP3 Ripper, office suite. SNMP-enabled network management, monitoring...
On top of that, you can get involved in merely requesting features. More of your needs will be met that way.
Take it to the next level. You need to set up a trouble ticketing system? Easily done. Choose from several. You need something "Web Enabled", you can.
Next level. You want to implement a terminal server solution? There are several ways to do that.
Today, I might want to deny packets at my linux firewall based on the TOS field in the packet header. Easily done.
I've only just begun.
Come on, man. If you can't answer the question yourself, you're haven't even looked for the answer.
rhadc
Yes, you could go through point by point but...
In the event that you don't have the money to spend, your solution is much easier had than under Windows.
Ok, it's clear from your response that you obviously are unfamiliar with Windows.
Been using Windows since 91. Regularly. It used to be an opportunity. Now it is a burden. If I'm not completely schooled in the products of Microsoft beyond 1999, it's because that's when I stopped considering their solutions.
Their security model is based on security-through-obscurity and a daddy-knows-best approach.
Their marketing department goes beyond the normal stretching of the truth. It works on FUD, false charity, lies, and propaganda to maintain the monopoly.
How can using this software be worthwhile? It is worthwhile when you are a beginner, or are unwilling to think about what you want to do with your computer. Windows is pretty good at being a single-user, several-purpose operating system. Load it down with lots of applications and you have yourself the slowest 1GHz machine around.
Not only does the operating system cost money, but the easily-attainable software still writhes in a pit of shareware, crippleware, etc. When designing or implementing a solution, you have to consider a few more factors that would otherwise not exist.
I prefer to NOT pirate software. I don't have to. You can't tell me all your Windows software is legal with a straight face. I've been there.
If you can't use Unix, don't. But if you can't acknowledge that Windows is limited in many ways compared with linux, or any Unix for that matter, you are uninformed.
Windows can do plenty of things Linux can't. Like force the user to make choices he doesn't want to. Example: MSN Messenger: "There is a new version of MSN Messenger available. Upgrade?" Choice A: Yes. Choice B: No(ask me next time I run MSN Messenger) Choice C: No - ask me next week.
I'm looking for the option that says: No, don't ask me again. Ever.
That's not available. I can try to hunt down a registry key to set "next week" to 2030, but besides that, I am trapped. Microsoft clearly aims to control.
One great response is: It's free, don't complain.
Well, my other software is free, and I can do what I want.
The obvious answer is, don't use Microsoft products or services.
So let me make this really easy for you. You obviously enjoy being microsoft's bitch. So go ahead. But don't tell me that being Microsoft's bitch is preferable to making decisions and having control.
One other thing: If Linux can't do it right now, all I have to do is write it. It costs me nothing but time.
You can't do that with Windows unless you use those same development tools that I have under linux. Gnu stuff. But they aren't exactly design for Windows, so they will be cumbersome. If you want to write software for Windows in an environment, you have to buy the software.
I've done this for a living: code for Microsoft OS's in Visual C++ and Visual Basic. I know what you can and can't do. The bottom line is, in Linux you CAN, FREE, and quicker than Windows. In Windows, you have plenty of CAN'Ts, for Money, and really really slow. Coding for windows is far more difficult than for Linux. This explains the poor quality of Windows software, especially shareware.
rhadc
That advance certainly didn't come in 1991, because the Amiga's clipboard.device already could do that earlier (at the latest, 1990 when 2.0 was released, and probably much earlier in the 1.x days of the 1980s but I'm not 100% sure). And this sort of thing wasn't really what the Amiga was famous for, so (I am speculating) that idea may have been stolen from the Mac.
Linux has faster filesystems. But Linux and NT both still suck at that, so I guess I should mention something more substantial:
An area where Linux is way ahead of Windows would be extensibility.
For example, Linux and Windows, when running on x86, both have a severe problem where code can be executed on the stack. If you run a network service and it has a buffer overflow bug, then bad people on the Internet can write their own code and execute it on your machine. So some guys decided it wasn't such a good idea for that to be possible, and they released some kernel patches to make it so that this infiltration technique doesn't work.
This actually reveals two ways that Linux is further ahead than Windows.
That doesn't put Linux just a few years ahead of Windows. It puts Linux a whole generation ahead of Windows, and even my beloved (but no longer maintained) AmigaOS. Freeness itself is a huge feature. (Alas, it's about all that Linux really has. But it's a biggie!)
---
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
But fanatical speeches about how "free software" based operating systems must be named according to one group's dictates makes me want to go out of my way not to do so.
I'm sure they're all nice guys and girls, really, and we'd get along fine otherwise, but pushing this name thing is doing about as much harm as good.
But then, maybe I just think other batters are more important to be fighting. :shrug:
I like elvis better :-)
-- Cheers!
All those people spending late nights chasing down obscure bugs
and adding new features, for no reason except to further the OS and the community.
half right. the reason anybody spends long hours or late nights these days chasing down obscure bugs and adding in new features is this:
so i can get this @#$%ing sound card to play my @#$%ing MP3s...
- Entertaining Bits from the Ancient Kernel Tree
Linux is in its newest incarnation ~25mg of tared and g/b2/zip'ed source code
How do you get the mass of a bzipped tarball? Is there some way to relate file size, or lines of code to milligrams (mg)? I have to say, though, ~25mg sounds like a pretty small mass to me.
--
--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.
Wheeler estimates 6.2 corresponds to a project that would take 4500 person-years to develop, and 7.1, 8000 person-years. This for two versions released within 13 months of each other.
Do the math. This represents the effort of over three thousand people working full time on free software in that period. More likely, it means the free and open source software in this study was written a heck of a lot faster than COCOMO suggests!
(OTOH, 7.1 included Mozilla and LINPACK; 6.2 didn't. These projects were started before 6.2 was released.)
Stupid job ads, weird spam, occasional insight at
There is a difference between ownership and license type. GPL is a license: anyone who writes code can release it under GPL and still retain copyright over his. It would not be accurate to call this code GNU code. So the case for calling Linux "GNU/Linux" is not that strong. Last time I saw an accounting of this sort, stuff from the Free Software Foundation -- software that can properly be called GNU -- made up under 10% of a Linux distribution. Lots of software is released under GPL these days, even Qt. To paint over them with the broad GNU brush would be a disservice to non-GNU contributors to Linux.
Also, KDE would rank quite a bit higher if it did not get listed as separate components: kdebase, kdelibs, koffice, kdemultimedia etc. It looks like a sort of statistical gerrymandering. After all, GNU projects like emacs and gcc did not get separated into smaller components, even though they are packaged in pieces. Why is KDE treated differently?
Where did the argument of who is farther ahead come into this? Microsoft has been around for 15 more years then linux. They've had plenty of time to "get ahead." The point is that Linux and the aplications that run on it are moving forward faster then microsoft and their applications are. This, in theory, means that linux will pass microsoft eventually.
Not only that, but Linux has pased microsoft in several key areas already.
From section 2.2 of the paper (my emphasis):
The ``physical source lines of code'' (physical SLOC) measure was used as the primary measure of SLOC in this paper. Less formally, a physical SLOC in this paper is a line with something other than comments and whitespace (tabs and spaces). More specifically, physical SLOC is defined as follows: ``a physical source line of code is a line ending in a newline or end-of-file marker, and which contains at least one non-whitespace non-comment character.'' Comment delimiters (characters other than newlines starting and ending a comment) were considered comment characters. Data lines only including whitespace (e.g., lines with only tabs and spaces in multiline strings) were not included.
Since the copyright statements are comment, I infer that none of their lines have been counted. If you want to check this statement later, you're supposed to be able to download David Wheeler's sloccount code here, but the .tar.gz file seem to be accidentally read-protected at the moment.
Come to think of it though, I would be even more interested in counts that included comments and documentation but somehow removed duplication, since comments and documentation also take an investment of time and add value (such as usability and maintainability) to the product.
soylent green is people too. coincidence? i think not.
complex
UltraEdit32 comes with Windows, does it? Under a Free Software license? I don't see it on my Windows2000 Professional CD-ROM...
BTW, I agree with the poster who points out that sed is more lightweight. The "all-in-one" nature of the perl trick was more important to me. YMMV.
Okay, now take a set of 15,000 web pages under a web server on Windows. Replace all "A" tags that refer to "url1" with "A" tags that refer to "url2" in all 15,000 pages.
How easy is this in Windows? I can do this with one command line in Linux (and any other *nix for that matter). Yes, I have to know REs to do it. Yes, it took me several hours to learn the RE expression syntax and several weeks of using them to make them second nature to me, but now I can do tasks like these in a matter of minutes. With ANY Windows system this would take several weeks.
"Useability" is a slippery term. Also, while Microsoft products do meet a certain level of minimum useability, there is a equal amount of crappy software from third-parties out there that are every bit as "unusable" as the hobbyist stuff for Linux.
And just how "useable" is, for example, MS Office? Sure enough, retarted monkeys can do the basics, but I would bet you 2:1 that 90% of Word users only acheieve 40% code coverage of Word -- in other words, if you start digging into everything Word is every bit as obtuse and difficult as state of the art 1970 glass teletype editors. More difficult, I would argue, because you could learn everything there was to know about those "unusable" editors in about two hours. Of course, you couldn't make a marketing brochure with those editors (unless you wanted to go out of business), but my point is that "useability" is pretty danged meaningless. "Suitability" is more to the point. Word is lousy if you want to do accounting.
Microsoft has actually substantially held back the increasing useability of systems by kepping the PC the dominant platform. Most people do not need general purpose computing devices. Home users need an "appliance" that does Web, e-mail, instant messaging, personal finance, word processing and maybe a spreadsheet. Business users need that plus presentation software, calendar/scheduling etc. These devices could be "embedded" type devices (think the Palm metaphor) that are much easier to use than PCs. Why should ANYONE but the very few who need more need to know about clock speeds, RAM size, ISA/EISA/PCI, irqs, USB, etc.?
The claim that Microsoft has advanced useability is absurd. They have been struggling against their own monopoly platform for over a decade, not because of their own failure, but because of the inappropriate design of the platform for its present use.
I will certainly grant that one must know a lot more to make good use of Linux on PC than to make good use of Windows on a PC. But which is easier to use, a Palm Pilot or a Windows PC? A TiVo or a windows PC? A Nintendo or a Windows PC?
Useability my rotund fundament!
Oh yes, that one-line command to edit 15,000 web pages?
/server/base/dir -type d -exec perl -e 's/url1/url2/gi' -p -i.bak *.html \;
$ find
I used to do it with a find within a find and a sed command, but the perl trick is a very nice shortcut, esp. since it edits the files in place and leaves backups behind!
gvim is part of vim
You're showing your ignorance. Go build a working Linux system sometime. I don't mean using Redsplat's gooey installer, I mean build everything from source code, and write your own init scripts.
I did that once. I came away with a much better appreciation of just what Linux is. It ain't merely a kernel. There's an entire infrastructure that it has to reside in. And that's before you even get to the GNU stuff. It's a hell of a lot more than merely the GNU System with the kernel swapped out. Much more.
A Government Is a Body of People, Usually Notably Ungoverned
Development tools. Linux (and the other unices) are light-years ahead of Microsoft with their development tools. Start with gcc, which is probably the most flexible compiler ever made - it handles C, C++, Objective C, Java, etc., and compiles code for dozens of platforms. Then you have emacs, which I consider to be the worlds first IDE, here long before Borland C++ or Turbo Pascal. There's Perl, Python, Tcl, the Unix shells, sed, awk, lex, yacc, gdb. Then you have all the libraries, too many for me to list, which are a tremendous help in enabling users to develop applications quickly. More recently, we have GUI toolkits - Qt, GTK+, Tk. We have source management tools such as CVS. All of these tools come with source code, so the programmer can get under the hood at will. The very design of Unix/Linux has evolved to be programmer friendly. Most Linux distributions come with hundreds of development tools.
While Microsoft's killer app is Office, Linux's killer app is its development tools. No other OS (except other Unices) even come close. That is why Microsoft sees such a threat from Linux.
Meldroc, Waster of Electrons
The first Emacs was developed in 1976 for the ITS platform by Richard Stallman, six years before Turbo Pascal. See the Emacs FAQ for yourself. Before you flame me, you might want to check your facts.
As far as development tools are concerned don't get me started with the monstrosity that Windows development has become. Between the Hungarian notation, the mixing of 16 & 32 bit code, the badly designed APIs, the not-quite-protected memory model and the maze of not-quite-compatible versions of the same DLLs, programming on Windows platforms is a nightmare. Comparatively speaking, Linux programming is much easier. Look how far KDE has come in just two years. They might put all the astroturfers posting on Slashdot out of a job at this pace.
Meldroc, Waster of Electrons
ALthough i agree with the solaris rant a bit...
:)
It is not the same kernel, there are different kernels for each sparc arch that get installed at setup... They are based ont he same come for the most part, but they have different code and are compiled with different opts..
BR just FYI
Yeah, sure. Can you help me to find the rpm's (or deb's, or tarballs whatever) for equivalents for:
And I didn't even start comparing features (I prefer Excel over the bloated gnome/kde variants).
Sorry lad, but basically Microsoft is still ahead of RH, Debian, and the whole Linux community.
----
... sometimes I fly with the white swan to my Liffey home.
I think you hit the nail on its head. One exception, which should become the rule IMAO is the GnuCash project. I was really impressed that it included its own help browsers(well, it is in fact an integrated version of the Gnome html browser), with high quality documentation. It is definitely on par with a lot of proprietary software, or even better. Another good example, although slightly less easy to navigate is the documentation included in LyX. I wished somone could come up with a scheme that includes the technical density of the man pages and the accessability of the documentation of the aforementioned projects for all the GNU/Linux/Hurd applications.
-- Spelling and grammar errors tend to be a sign of erroneous thinking.
You are correct.
I'll admit I don't know much about TFTP server other than that you can use it to load Linux kernels and do some remote booting stuff. Given that that is all I know about it, what does Win2K do with a TFTP server?
(I'll even pre-mod myself down for asking a dumb question)
Jesus was all right but his disciples were thick and ordinary. -John Lennon
I'll bite...
Linus still doesn't work for a Linux company. Linux is his side job (okay, yeah, he does that during the day, but it's not as if he works for RedHat)
No. Linux users are not a bunch of elitist pricks. That would be the users of the Macintosh (at least as portrayer by the media). My love affair with the Mac ended about 18 months before the iMac came out. I was tired of getting left with products that were not upgrade-worthy. I didn't do high end desktop publishing, so the mags didn't really give a shit. Even TidBITs no longer held any relevancy for my use. Quite frankly, if you didn't have a high end rig, nobody had time for you.
Then the iMac came out, and all the yuppie shits could buy them for their kids. But not somebody on a limited post-graduation income (tried to finance. Wasn't good enough for Apple, but was good enough for Kawasaki Finance. Whatever.) So that was that.
So yes, there is a single user experience for the Mac community. Since that didn't describe me, I finally gave up on it.
There is also the fact that Apple gave up the fight. You imply in your AC rant that Apple users were getting somewhere against M$. You were. Your buddy Steve (him leaving was the best thing to happen to Apple) was getting lined up to take it in the ass from Bill. Guess what happens if M$ stops making either IE or Excel for the Mac? Goodbye Apple. That's not fighting, and that's not winning. Perhaps you think the Vichy fought the Germans as well...
Yes, choice is a damned good thing. And yes, it can overwhelm new users. Guess what? I've never seen an install that upon initial boot said "which of these seventy WM's do you want to use?" No. You got AfterStep, Gnome, KDE, or whatever RH, Debian, SuSE, or whomever chose. The beginner does not have to make the choice. They also don't have to choose between any distros. Plunk down your money, pray it ain't Slack, and you'll get along fine.
It's interesting that you seem to be so fearful of choice. The only person I can think of who is so fearful (and also under the guise of the 'new user') is Steve himself.
Steve, is that you? Does Bill at least give you a reacharound? Why didn't you jump on the CHRP bandwagon? You're halfway their with your IDE and video.
What MS (and you) don't realize is that as a single user, I have won. For me, the war is over. I have more software to play with and use than I could in an entire lifetime. I have development environments to play with. I have databases at my disposal.
The chickenshit is not the moderator. It's the troll who hides behind the "Anonymous Coward".
Jesus was all right but his disciples were thick and ordinary. -John Lennon
If that is the case, why has Linux progressed farther in 10 years than Microsoft during that same time frame? Money is a motivator. Not the motivator.
Jesus was all right but his disciples were thick and ordinary. -John Lennon
Your points are not incorrect, but your debating ability is somewhat flawed. You have chosen to examine state variables, rather then the flow variables I mentioned.
Let us say that both Linux and Windows ** are on a 1000 mile trip. Windows started in 1975, Linux in 1991. By 1991, Windows was already at mile marker 500. Since then, they've reached mile 750. Linux started at mile 0 in 1991, but by today is at mile 532. In ten years, Linux has gone farther than Microsoft has gone in ten years.
Yes, certainly on average end-user friendliness and usability, Microsoft is ahead of Linux. But Linux is now moving at a faster pace. That was my point.
Jesus was all right but his disciples were thick and ordinary. -John Lennon
What I like about Linux is that I get the devel tools for free (as in beer). I liked the basic that came with MS-DOS. Where did it go? Why did it go?
I could care less about IE coming installed and 'part of the OS'. But IMNSHO, some basic development environment IS part of the OS.
(No, I didn't address your point about timing. I'm insufficiently armed for that battle. IOW, I don't know nuthin' bout that)
Jesus was all right but his disciples were thick and ordinary. -John Lennon
The further ahead thing came up as a probably deliberate obfuscation of a comment I made. I claimed that Linux has come farther in ten years than MS. Someone chose to interpret that as Linux has come farther since its birth than Win2k since its birth (counted since the founding of Micro-Soft)
Jesus was all right but his disciples were thick and ordinary. -John Lennon
What about Teco?
www.timcoleman.com is a total waste of your time. Never go there.
What Actually Makes Up "Linux"?
Posted by CmdrTaco on 4:14 Thursday 21 June 2001
from the somebody-did-actual-research-for-a-change dept.
David A. Wheeler sent in linkage to his extensive analysis of the true size of Linux....
I don't get it. The title of the article is "More Than a Gigabuck: Estimating GNU/Linux's Size" and in the first line it is made clear that he's talking about GNU/Linux, but in the heading and the abstract for this story, Taco refers to it as Linux. He's not talking about the size of Linux, he's talking about the size of GNU/Linux.
www.timcoleman.com is a total waste of your time. Never go there.
So my question is, what's the size of the non-development/non-application stuff? What's the size of the kernel plus the essential utilities (most of which are GNU, as RMS points out ad nauseum)?
The question is meaningless. If you want just a bare-bones system that boots and does nothing interesting other than run a shell with a few of our GNU friends (ls, rm, cp, etc.), you'll have a (compiled) size of about a single floppy disk.
If you want a rudimentary firewall, add a bit more. If you want a web server as part of your OS, more still. If you want some form of GUI for a desktop machine, you're including much more-- X, various other libraries such as qt, gtk, etc.
It all depends on what functionalities you consider to be a part of your operating system. This in term, depends on what you're actually wanting to do with the system.
"Intelligence is the ability to avoid doing work, yet getting the work done".
It's only software!
It's not the quantity of documentation that's the problem, it's the *quality* of it-- at least from the perspecive of a newbie.
/usr/share/doc/? Chances are you'll follow up pretty quickly with 'rm -rf /usr/share/doc'. Man pages, the HOWTOs et al are quite useful, but chances are that the average newbie will either not know where to look for them (or that they even exist), or will have trouble understanding a single word.
Chances are that you already have more than enough documentation installed than you'll ever care to read. Ever try running du on
Compare this with the documentation included with Windows. This is the complete opposite; there is a fairly consistent interface for it (althought the intuitivity of it is debatable), and it addresses almost all problems a newbie could have. Unfortunately it is completely useless if you ever try and look up something remotely technical in it.
What the Linux community (perhaps through the Linux Documentation Project) needs to do is compile the various documents that are out there into a consistent format and style. If the various distribution vendors would only then make their own distro-specific docs compatilble with this, Linux documentation would become just a bit more cohesive...
"Intelligence is the ability to avoid doing work, yet getting the work done".
It's only software!
I don't think the fact that Netscape uses Motif is the issue here. Take a look at the about: url to see what proprietary stuff is included in Netscape.
"Intelligence is the ability to avoid doing work, yet getting the work done".
It's only software!
Windows analysis - SHLOCK Count Method
--- Hot Shot City is particularly good.
But Linus can only name what he wrote.
The Tlog - a technology blog
No. :-)
Yet I found it strange that in GNU TeXInfo documentation the X Window System is usually referred to as "X Windows", even if that documentation passed several reviews (i.e. the GNU Emacs manual).
Perhaps there is more to it that a simple mistake.
What does this have to do with the kernel?
The increase in size is about what (usually preexisting) packages have been added, not about kernel size or programmer productivity.
I'm not really up on this stuff, but didn't Linux have IPv6 before Windows? I think it came out as a service pack for NT or something like that.
Here's one where Linux is way ahead of Windows - flexibility. How many window managers do you want to try? Don't like the way that program behaves? Edit, recompile, et voila.
Here's another way - portability. Linux has been ported to just about every CPU architecture out there. Regular Windows (NT/95) only currently works on x86, as I think MS abandoned Alpha. Only CE, which is a poor cousin to NT/95 in that it doesn't run the same applications, is being supported on multiple architectures. Then there's clustering, in which Linux leads.
The point is that for some people, these advances are big advantages, while for other people, Windows' strengths are more important. If you like playing the latest games, for example, you have to have Windows. As another poster pointed out, Linux is more stable. If you want to run a [Web/file/ftp/DNS/mail] server and not have to think about it very often, Linux (or a BSD) is your way to go.
An OS's name is not the right place for the FSF to try to advertise itself. Adding a GNU just makes things harder to say and write without adding any real benefit.
Let's follow the K.I.S.S. rule and leave well-enough alone -- Linux is a great name for the OS we all know and love.
..wayne..
This is like asking what part of a car is the engine.
Linux is the kernel. X is X. Utilities are utilities. Why is this seen as so bloody hard?
--
You're only jealous cos the little penguins are talking to me.
I checked the other week, and I have slightly
less than 2 MB of GNU General Public Licenses
on my hard drive. I have RH 6.0, and a 1.? GB disk.
I think I found them with
find / -name COPYING | du
or something. that and LICENSE.
Correction #1: the Mach micro-kernel
Correction #2: Eros
And the endless variations of the above, to name a few.
-- "So, what's the deal with Auntie Gerschwitz et all?"
everything seems to be remake of some existing ? usually commercial ? application.
I provided you with a few examples which weren't a remake of some commercial application. Mach (to a lesser extent) and Eros are representative of branches of OS design that commercial vendors have widely avoided for use in a commercial product.
Sure, Eros wasn't the first to use orthogonal persistence. But show me a widely used commercial (or free) OS which uses these concepts.. oh, that's right, there aren't any.
Here's the definition of "revolution" that gdict came up with: "6. A total or radical change; as, a revolution in one's circumstances or way of living."
True, true, MacOSX uses Mach (gee, it must be poorly designed, it's being used by a commercial vendor) but let's see: Most OS's are monolithic, so microkernal OS's are revolutionary; they're a completely different kind of kernel. Orthogonal persistence is likewise revolutionary, how many non-research OS's use it?
Here's a clue: Researchers come up with new ideas, corporations whore the ideas off to everyone else.
-- "So, what's the deal with Auntie Gerschwitz et all?"
Ok, I don't call the whole shebang "GNU/Linux", but you are still guilty of accusing all the authors of software other than the kernel, Linux "zealots". Many of these zealots work on projects that are platform independent.
There is a fine line between being a cultivated citizen and being someone else's crop. - A. J. Patrick Liszkie
Yes, I'm a Windows user. (No, it isn't because I'm too lazy to install linux. Rather, I work for a company which requires that I be fairly up-to-date on the latest 3d shooter games. Linux simply doesn't have the titles fast enough.)
I'm saddened to see that some moderator thought the parent post was flaimbait -- on the contrary, what really makes up linux -IS- the people, the community. I, for one, support krmt. Hear hear!
---
He he. Funny you should say that. What about all the new engines that are coming out? Is the monitoring system part of the engine? Is the turbo-charger part of the engine? What about the fuel monitoring electronics? What about the intercooler? There's lots of grey areas where engines are concerned. Again, is Linux Linux without GNU? Theoretically, Linux is just the kernel, true. BUT, if you take away X and you take away the GNU userspace, can the OS rightly be called "Linux" or is it simply an OS based on the Linux kernel?
A deep unwavering belief is a sure sign you're missing something...
The Linux Kernel (or UNIX in general) is just a bootloader for Emacs.
Isn't it?
That was truly the sickest thing I've seen this week. I actually screamed when I saw it! (I'm refering to parents .sig...)
Dive Gear
--- Think of it as evolution in action ---
Large open source projects are more likely than proprietary closed source projects to involve developers who are not physically co-located. This means that communication between developers is a bit more of a pain (e.g., you can't walk down the hall to discuss a problem). As a result of this and other factors, it's conceivable that physically co-located programmers may be more productive. As you may recall, there's evidence that a "war room" can result in 2x performance. I don't know much about the COCOMO model used in this paper, but I could imagine that it could be greatly affected by issues such as these.
Of course this doesn't take into account other benefits of distributed teams (e.g., more varied perspectives) or of open source programmers (intrinsic vs extrinsic motivation). But it's something to consider.
It would be interesting to see how many MB of space those "This is GPL" disclaimers take up.
On the plus side, you'd probably get a great compression ratio from that...
--
mod it down.... shit this is perfect!!!!
My only bone to pick with you is that there are few users/supporters of linux that dont fall into the zealot category....hmmm am i one cause i said this.... damn...
Non-Deterministic Finite Automata
No windows community? Who the fuck needs a wishy-washy, warm-all-over, community when you're the dominant player? Maybe the people who use windows at work and at home don't "talk" to each other, but they're there, and they number a hell of a lot more than linux.
Mac zealots' dependence on Apple? Aren't we dependent on the big projects' (kernel, X, KDE/GNOME) developers and leaders? How many of "linux community" are competent or interested or have enough time to fix/modify large projects, or write a kernel driver?
As far as "we make this," ya, coding is fun, but if I've spent years honing my skills, I'd rather get paid handsomely to work on a commercial product, thank you. And usually, things progress more quickly (and correctly) when there's the pressures of a capitalist economy driving production... I think that's evidenced by all the shit poor, half-baked, useability-retarded crap that floods freshmeat on a daily basis.
---
python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
The problem is, I think, that you can only copy/paste text in between applications under X. There's no concept of different mime-type copying, such as img/* or ole/table. Am I right in assuming this?
---
python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
:help copying
"SUMMARY
Vim is Charityware."
... not GNU/GPLed.
---
python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
Nice! I'm saving that one. I was doing it with find and a script:
#!/bin/sh
cat $1 | sed "s/$2/$3/g" >> $1.1
mv -f "$1.1" $1
---
python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
Linux progressed farther in 10 years than Microsoft during that same time frame
I don't see how that's true at all. In both technology, and the bottom line, Microsoft is *years* ahead. Technology: let me offer one example: go to a web page (IE) with some kind of table with data in it. Copy the table. Paste it into Word. It actually becomes a Word table! Paste it into Excel. It actually places the data, and the formatting, into the cells! How far is linux from that level of ease of use, that level of "object linking and embedding" across apps? Do you think the multiple desktop standards helps or hinders this task?
And in terms of bottom line, linux companies are still trying to figure out how to make a buck. Redhat just now moved into the positive column, VA and others layoff people seemingly every week.
I'm a fan of Linux because I'm a hacker. I like the shell, I like the flexibility and customisability that come with having dozens of "glue" tools. But the fact is, hackers are the minority of computer users, and this is only going to be more and more true in the future. For the masses, ease of use is priority 1, and it seems, at least to me, that the "other" platform has a great lead in that arena.
---
python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
Given that the author is David Wheeler, it seems pretty likely that you're right.
Happy Premise #3: Even though I feel like I might ignite, I probably won't.
After reading the analysis, two things sprang out at me. The first is that a lot of the stuff on a Linux system is meant for development, rather than just using the system. The second is that lots of the stuff on the list clearly is "application" and not anyone's idea of an "operating system".
Specifically, in the top ten, we have:
Development Tools
Applications
(Also in the top 20 are libgcj, teTeX, postgresql, and xemacs. And we won't get into the issue of whether Mozilla (#2) should be considered part of the operating system.)
So my question is, what's the size of the non-development/non-application stuff? What's the size of the kernel plus the essential utilities (most of which are GNU, as RMS points out ad nauseum)?
Happy Premise #3: Even though I feel like I might ignite, I probably won't.
much like a Japanese car built with 87% United States parts
You must be a Honda Accord owner..
.sig: Now legally binding!
He said vi, not vim. He may very well have meant Jvvi, JVI, nvim, gvim, etc.. Pick one, it's 90% sure to be GPL..
.sig: Now legally binding!
It would be interesting to see how many MB of space those "This is GPL" disclaimers take up.
Well, if you read the report, you'll realize that it isn't about Redhat. It is about Linux, with Redhat standing in as a representative distribution.
As for what useful features have been invented by the open source community:
And I think I'm only scratching the surface. The open source community is very diverse, including both individuals working for the sheer joy of programming and companies who have developed code for specific internal IT projects and have given out the source in a way that it is sharable (socks for example).
I would guess that on average OS does less reinventing of problems that have already been solved and more original work than any company could do. Even GTK which is arguably a distorted clone of Motif has some very interesting and unique inventions that are very useful for porting graphical toolkits to different language bindings.
LibBT: BitTorrent for C - small - fast - clean (Now Versio
Thank you kindly :-)
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
I don't know why I'm feeding this troll, but I'm going to. For me, "winning" against MS isn't about having linux everywhere (although that would be heaven). Winning is about having the choice not to use MS. I don't like Windows. Never have. Never will. I was a Mac guy, and just like gmhowell described so well, I was pretty much sick of Apple and jumped ship. I had the greatest, most unified UI on the planet (Win still can't touch it), and I jumped off.
And I, like gmhowell, jumped because I wanted to be free.
I don't like the fact that Apple and MS are able to dictate to users. Winning for me is the freedom, the choice, and the power to decide what I want to do. Don't like a new feature? I can disable it all the way to the point of deleting the code within the source itself. Don't want the new version of X? Fine. I don't have to upgrade until I'm ready, not when Apple or MS says from above "It's time krmt". I have the choice to decide how I want my computing experience to be, and winning for me is not allowing MS or anyone else to take that away from me. That's what Free software is all about.
It's not about having the most unified interface. It's not about our differences. It's not about the hubris or the third parties or the candy coating or the snappy wiz bang features. It's about Freedom. And all your complaints can never take that away.
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
... is the people. Seriously, Windows can't really say that because there is no real "Windows community". Mac people can talk about it, but they are still dependant on Apple for all wants and needs. On the other hand, Linux is written, used, and supported by the people themselves. Those figures, all of it from the the lines of code to the language percentages, just illustrate who and what we are as a community.
It's something I could go on and on forever about because it really is something special in a world dominated by the shadow of Gates and Jobs. "Those people" who work "over there" don't make this. We do! While all those numbers can start to quantify this, you can't really put a dollar value on it the same way you can't put a dollar value on freedom. Funny thing to be able to say that about a bunch of software...
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
That's incorrect, because BSD (at least FreeBSD, don't know about the others) will run Linux binaries. So that statically linked Linux binary will work on BSD. It's also not a fair comparison, because each BSD is really a different operating system (different kernel) while each Linux distribution uses the same kernel (though the original poster did not make that distinction).
Or from a workstation user's (i.e. Me) perspective,
one has to ask what good a free kernel would be without a free desktop environment. Or a free file/web browser. Or free office/multimedia/* applications.
For me, software like KDE is far more important to me, and gets used far more often than compilers, debuggers and shells are (of course I know they're essential, just not for me). Should I be calling it KDE/Linux?
Thell this to all people agruing aboud GPL versus BSD Licence
what separates linux from the older *NIXes?
Apparently, nothing big, because Linux designers decided to go for a well-known and working architecture (also because much OSS has been developed on ethereogenus *NIX platforms, look at what the configure scripts have to do ).
However :
- the old Unices are all platform dependent, GNU/Linux is multi-platform
- Old unices needed expansive hardware; GNU/Linux run on cheap PCs
- Old unices were sold on niche markets, often to be operated by a an elite of lab techs in white coats; GNU/Linux bring unix power to everybody wants to mess with it (still an elite, but with a totally different attitude).
And so onWhat I like of GNU/Linux is that, while being essentialy Unix, improved the original model in many little ways. Two examples:
Least but not last, it is somehow because of GNU/Linux if programmers now can have modern GUI toolkits like QT and GTK+.
Ciao
----
FB
Tom's emergency root-boot floppy distrib is moving over to Lua script for as many utilities as possible. it's impressive, and Lua is a very cool language to say the least.
If you just want to edit text, it is bloated, horribly so.
The illegal we do immediately. The unconstitutional takes a little longer.
--Henry Kissinger
Congratulations, you're the dumbest person ever to mention my sig.
The illegal we do immediately. The unconstitutional takes a little longer.
--Henry Kissinger
I suspect that very soon the Motif stuff in netscape that's still proprietary (the 606 files that are proprietary it says are entirely contained in netscape) will soon dissappear in favor of strict mozilla. Not that this would be a monumental happening (i myself don't use netscape), but it would mean that RedHat could finally say "0% proprietary code in our distro"
I mean, since most people assume redhat==linux, it could be a marketing buzz phrase.
on a sidenote, mandrake 8.0 is selling better than RH at best buy.
And walmart sells linux.
sig?
Having said all that, I just went and calculated the effect of this. By this formula, one 30M line program is 60% harder to write than 30 million 1 line programs. The difference between one 30M line program and 60 500K line programs is only 22% - so the overestimation is not large.
Quattuor res in hoc mundo sanctae sunt: libri, liberi, libertas et liberalitas.
(please feel free to correct me here, but I have never seen truly innovative OS project ? everything seems to be remake of some existing ? usually commercial ? application.)
Ok, here are a few:
- emacs (self explanitory)
- gcc (yes, it was 'just' a c compiler, but nobody had thought of attempting anything that flexible/portible before)
- TeX/LaTeX (still kicks any wordprocessor esthetically, and for highly technical publications is the standard)
- X11 (say what you like about it, but both the design and implementation were innovative)
- mozilla (the innovation isn't in the front end)
etc..
also a large number of softwares that will never see general use, being domain specific, but have benifitted from the OS approach.
You must be a Honda Accord owner..
I don't know about him, but I am. In fact, damn-near every part of the Accord is made in the USA. There are a few exceptions (seatbelts, some sensors, etc.), but most Honda vehicles (except the CR-V, S2000, and NSX, all of which are made in Japan) are comprised of such a high percentage of domestic parts that the US EPA classifies them as domestic vehicles!
--
That was truly the sickest thing I've seen this week. I actually screamed when I saw it! (I'm refering to parents .sig...)
Yeah, pretty fruity, huh? Aren't prolapsed rectums something else?!
If that sort of stuff turns your crank (you don't want to look, but you can't tear your eyes away!), you might be interested in these three links.
--
The kernel shouldn't be two million lines of code. How much of that is drivers? And how much of the drivers are duplicated from one driver to another?
ah comeone you can do better than that... anyways I'lll let you troll me.
Linux == kernel
GNU/Linux == all the nice applications (like vim) that are needed for a usable system.
"`Ford, you're turning into a penguin. Stop it.'" -THHGTTG
For ease of use (which, let's face it, is what most people want), DOS Edit beats every Linux command line editor hands down. I but I can add DNS records, crontab entries or hack up a perl script faster in vi (vim is my favorite take on vi). You can even have more than one person working on the same file at the same time in vi. I think it's the 'killer app'.
chris@xanadu:~$ whatis /.
/.: nothing appropriate.
Not 13 million Microsoft-quality lines of proprietary junk charged at $100 per hour, but 13 million lines of reasonably good code, sculpted for the joy of coding!
Explain how you justify that comment? Personnally, I have never seen any of Microsoft's source code and I doubt you have either (unless you're an M$ employee). However, I have seen a fair amount of Open Source code and guess what? It ranges (in my experience) from very good to very bad. My money is on M$'s code being similar.
I think the good reliability of Linux in relation to Windows is caused by other factors e.g. the the Windows GUI is too tightly integrated with the kernel, there is too little control over the quality of third party device drivers, the Open Source development model lends itself to fast turn around of bug fixes.
All I want is a secure system where it's easy to do anything I want. Is that too much to ask ~~ Randall Munroe
Depends on your definition of "operating system." GNU has created plenty Utilities, Applications, and Compilers. In addition, GNU has created a kernel, the HURD. One of the most significant contributions GNU has made is their C Library and their linker.
I would argue that GNU's goal is exactly that: to create an operating system that is free software.
Don't take my word for it, read The GNU Project for a complete description of the project.
This is essentially commentary on the health of the OS/FREE community; in the last year, the community added 13 MILLION lines of code to RH's distro. Part of the included code is bloat, but much of it is new projects, code mature enough for RH to distibute it to Ma + Pa Linux users. RH didn't do it, and the author doesn't say that RH did it. The programming community did it.
Look at that paragraph again:
That is incredible! Not 13 million Microsoft-quality lines of proprietary junk charged at $100 per hour, but 13 million lines of reasonably good code, sculpted for the joy of coding!I'm getting misty-eyed.
Or maybe 'Delta' code, still changing, and stuck in another quadrant of the galaxy.
All this talk about GNU/Linux is only valid if we all agree it is, or if some particular distrubution chooses to call itself that. That is the nature of the GPL as far as I see it -- you can call your product (distribution) whatever you want, as long as you give credits to the original author of the source code that you use. The product reffered to in this paper, for example, is not called GNU/Linux -- it is called Red Hat Linux 7.1, which is perfectly valid.
Sure, RMS and the FSF deserves credit for creating GNU etc., but, since they did not include a clause for that in the GPL, nobody has to call something which includes parts of the GNU project GNU/whatever. Of course, Linux was origanally only the kernel, but since the distributors tend to call their systems "Linux" that is what they should be referred to (in accordance to the GPL, anyway). For example, a product called Red Hat GNU/Linux wont exist until you create one -- if only by writing it on the cover of a Redhat CD-rom.
And poke her, with the soft cushions!!!
---
Linux is in its newest incarnation ~25mg of tared and g/b2/zip'ed source code written in C and covered by the GPL. Without gcc or some other compiler you can't even compile it. Without a shell you can't do much with it. All of those things come from the GNU or other sources.
Linux is in its simplest form much like a Japanese car built with 87% United States parts.
On a personal note:
In the beginning there was Linus and the word was with Linus. Accept Linux into your hart and you shall have uptime eternal.
Kernal 3:16:
For Linus loved man so much that he gave his first begotten OS.
Ascii artist &
20 hardcore source contributers.
300 source contributers who want their name in the CREDITS file so they can add it to their resume.
2 gatekeepers.
20 distributions that do nothing but add an installer front-end and offer tech support for an obfuscated OS.
1 obfuscated lightweight editor
1 less-obfuscated bloated editor
1 standardized windowing system struggling to keep up with a certain competitor -- driver-wise and enhancement-wise (anti-aliased fonts came to mind at one point).
Kernel modules/drivers with the same struggles.. (again, USB compatibility came to mind at one point).
2,000,000 zealots
You can moderate this down, but I challenge you to find proof that this situation is otherwise.
0.0023%
/. population that actually read the entire whitepaper.
This is the percentage of the
"My mother never saw the irony in calling me a son-of-a-bitch." - Jack Nicholson
Great programmers and lousy bean counters...
If voting were effective, it would be illegal by now.
No, it just means that it's 76% (or some percent correct.) If you refer back to the article, the author states that often, some large bulky code does something difficulty that some smaller piece of code could do elegantly. The EFFORT that goes in to making that smaller, more eloquent code is proportional to the effort of putting in a bloated piece of crap that is many times larger. IIRC.
--You will rephrase your request for me to go to hell. Goto statements are not acceptable programming constructs
But according to the Microsoft world view, the true size of Linux is irrelevant, since it will be replaced soon enough anyway! Seriously though, I never knew that Linux was that big... Now Linus has even more to brag about: (secret phone transcript) Linus: Nyah Nyah! My OS is bigger than your OS! Bill: LOL! Prepare to be assimilated! L: AAARGH!! I chose YOU, RMS! B: I chose YOU, Mundie! L: ESR-Use your secret pedant-o-propaganda attack! B: AARGH! LOOK WHAT HE DID TO MUNDIE! *sobs* Poor guy... L: Aww, I'm sorry. Shake on it? *BZZZT* B: Muahahaha... Nice guys finish last...
This
Ok, so this guy claims that Linux would cost a little over $1 billion (US) to develop. I wonder what the big deal is. I'm sure Microsoft has spent that much over the years on Office+Win9x+WinNT+Backoffice+etc (basically the functionality provided by RH 7.1). Intel spends billions to develop new chip technologies (IA 64 anyone?). The only thing incredible about this number is that most of that billion was completely unpaid, or at least underpayed.
"The universe seems neither benign nor hostile, merely indifferent." --Carl Sagan
I was up until 2am trailing through the disassembly of a DOS device driver for an obsolete piece of hardware, attempting to find out what I needed to write to which IO ports to enable this device, so I could add them to the Linux driver.
just submitted a link to:
Hackers move too fast for CIA
Wait to see how long until it gets rejected.
---
Hammer of Truth
Help Linux achieve mass-market acceptance and legitimize Linux in the eyes of corporate IT managers.
The operating system kernel (Linux) is the largest single component, at over 2.4 million lines of code (mostly in C); that compares to 1.5 million lines of code in Red Hat 6.2. Over 1,400,000 lines (57% of the Linux kernel) are in the ``drivers'' subdirectory, thus, the primary reason the kernel is so large is that it supports so many different kinds of peripherals. No other subdirectory comes close to this size - the second largest is the ``arch'' directory (at over 446,000 SLOC, 18% of the kernel), which contains the architecture-unique code for each CPU architecture.
Red Hat Linux 7.1 includes over 30 million physical source lines of code (SLOC), compared to well over 17 million SLOC in version 6.2 (which had been released about one year earlier). Using the COCOMO cost model, this system is estimated to have required about 8,000 person-years of development time (as compared to 4,500 person-years to develop version 6.2).
Had this Linux distribution been developed by conventional proprietary means, it would have cost over $1.08 billion (1,000 million) to develop in the U.S. (in year 2000 dollars). Compare this to the $600 million estimate for version 6.2. Thus, Red Hat Linux 7.1 represents over a 60% increase in size, effort, and traditional development costs over Red Hat Linux 6.2. This is quite extraordinary, since this represents approximately one year.
This is interesting, since it makes the Linux effort competitive with Microsoft in hours and bucks alone. (note that the IBM contribution is not included here, and is likely spread over several years (?))
Now as far as the Talent goes ...
;-)
Check out the Vinny the Vampire comic strip
"It is a greater offense to steal men's labor, than their clothes"
I basically agree with your analysis but why this comment?
Our notions of ease of use are not the same as Joe Sixpack's.
As the original poster pointed out - Joe Sixpack is in the majority of computer users. If free software developers don't write software that caters to his needs he's not going to consider using it.
Is that important? Depends what you as a free software developer care about
If you care about Freedom - in the RMS/GPL/Free-as-in-speech-and-just-as-important way - then you'll want as many people to use Free Software as possible.
If what you really care about is the technology itself - then you'll focus on the things that matter to hackers/developers.
I suspect that most developers of free software care about the Free bit but are truly passionate about the Software bit.
A few things to keep in mind about the names "Linux", "GNU/Linux", etc.
1. A system can be named by just one portion of it. I.E. "Windows" (the component that draws the windows?). The same could be true for Linux. Naming a system after it's kernel doesn't seem like an idea that's strange.
2. In favor of "GNU/Linux", one has to ask what good a free kernel would be without a free compiler. Or a free shell. Or free command utilities.
As I see it, the primary disfunction in free, shared software is naming a system. I mean, you can have "Redhat Linux 7.1", but that's not some release the developers agreed to. You can have individual releases, but what do you call those? Things will be thrown into an even BIGGER turmoil if/when GNU/HURD takes off. Are all of those companies gonna change their names from "Linuxcare", "Linux-Mandrake", etc to "Linux/HURDcare", "HURD/Linux-Mandrake"?
2) GNU utilities and other free/open programs
3) Distribution/vendor-specific setup utilities, packaging and similar "glue"
4) Commercial software, either evaluation version or free for personal use
The kernel is the central part but you can't do much without part 2). The two remaining parts are the real frosting when you compare to "Linux From Scratch" or even Slackware.
What is the sound of one hand clapping?
cat
The article says that Windows 2000 has 20 million lines, and Linux has 30 million lines; Do these 20 million lines include IE 5 or Office? They definately do not include the "development envirenment for Windows".
But that's bsides the point. The main questions I want to put forward is this:
Why does the linux distribution have to include everything? I mean, a lot of the drivers are not needed, and many applications are duplicated (Emacs/XEmacs anyone?) What we probably need, is a radical review of the installation procedure, where the installer detects the hardware available, and then only installes the needed drivers only.
What utter crap you speak, oh anon-one.
All M$ are doing is saying "Don't use "viral" tools when developing our stuff, and don't distribute our stuff with "viral" software. By "viral" they mean "software which imposes other conditions upon M$".
I.e. ALL they are doing is asking you not to cause M$ code to be forced to become GPL'd.
Nothing more, nothing less.
Deal with it.
People should not be afraid of their governments - Governments should be afraid of their people.
Are there estimate how much money in form of salaries were ever paid to programmers for the code and how much was in effect done not only voluntarily, but aslo completely on an unpaid basis ?
In the assumption that the difference between the total sum of salaries paid to all programmers for all the code included in the RH7.1 distribution and the 1.8 billion dollars it would have cost to develop by proprietary means, this overwhelmingly large, isn't that the best proof that Free/Open Software is far more powerful in promoting the progress of USEFUL Science, just because its so much cheaper to develop and so much more easy to make better ?
I wonder why the GPL should not hold up in court as the better license to prove the that Free/Open Software is much more capable to comply with the ideas of Article 1, Section 8, Clause 8 gives congress the power:
"To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries;"
Great analysis to read.
I think it would be very interesting (and somewhat important too) to shed more light on who has done how much work for how little/much/no money for how long.
May be /. could help to make such a poll and posters to the poll would actually give their names and be honest about their contribution. It would certainly help to read once something very uplifting on /. and might be helpful for other purposes too.
It might also be a way to publicly acknowledge in a way each person who contributed if he chooses to reveal his contribution here on /. or whereever one could set up a poll like that.
Use the TROLL distribution.
Lesson #1: Never ask which Linux is best.
Lesson #2: refer to Lesson #1.
Don't blame me 14M3R. You are the one that trolls here.
IIRC RH 6.2 didn't have KDE/Qt, right?
:)
I imagine that would account for the increase of C++ code. Don't think it will rise the same during the next year then.
--
--
Free Software enthusiast; Debian GNU/Linux (powerpc) developer
Really though, it was a good point. Just what is so revolutionary about linux? I mean, something is definitely special about linux (as seen by its success), but is anything actually unique, new, or whatnot?
:), the two could practically be the same. Samba, Netatalk, Bind, Apache, they're all the same, yet they programs like them are considered to be part of linux more often than not.
I recently setup a FreeBSD server to see what it was like, and other than some different file formats (very slightly different), a few more command line switches for various commands, and the lack of rpms (which i don't mind at all
Other than price and openness (which doesn't separate linux from *BSD, etc which are equivalents), what separates linux from the older *NIXes?
Scott
I have a sneaking feeling that the author used SLOC Count by David Wheeler to determine the number of source lines of code--particulaly because of the cost-to-develop estimations that the article mentions. Even if he didn't, this is a nifty tool.
Got friends?
uuuhhh, so how would you value a "revolutionary" idea ? Anyhow, what's so revolutionary about Linux ? It's just an OS isn't it, the way in which it proliferates amongst it's users might be different (ie. not sold but freely distributed) and it's users might have access to the source code but how does that make the OS itself revolutionary ?
Really? I thought it was a pretty transparent troll myself.
Using RedHat as a distro for this project isn't that good of an idea. Aside from the requisite bitching that goes on about it (corporate design, beta software that can't even compile itself), it's just an unrepresentative mass of programs and code! I can safely say that most Redhat users will never use about one-quarter of the programs in their distribution, especially if they install programs with RPM.
:)
Instead of using a binary-based distro, why not make a base system from scratch? This would nix the whole problem of making sure that you have all the original source packages, and, at the same time, you wouldn't have to worry about taking off SLOC for the RPM utilities and whatever other GUI-interface stuff Redhat happens to throw into their distribution.
Anyhow, this is a cool concept, and I'd love to see, for example, how the Apache web server compares with IIS in this arena
If this is a troll, it's a damn good one.
where there's fish, there's cats
Red Hat is profitible. Do your research.
end communication
I just want to thank the author for taking the time doing this, it looks as though it was very tedious.
MY only nit picks are these:
Quoting statistics/data going back to '95 is way out of date by todays standards, even '99 is now very old.
Only minor nit picks, thanks for the article.
StarTux
Try as many as you can until you find the one that suits your style.
Your question was worded badly. I personally like SuSE, but I have friends who like Mandrake, Redhat, Debian and others whose names I do not even remember.
Guess what? They are all happy with what they have, if they ever get upset with a distro they move onto another one.
StarTux
>Which might help explain another number that keeps cropping up: 5% of the OS market.
Seriously, documentation can have nothing to do with marketshare. M$ documentation is terrible. Even MSDN is filled with marketing speak.
If you want an example of good documentation, look at FreeBSD, and what marketshare does it have ?
> should head on over to http://www.linuxdoc.org and if you tell me that documentation is poor quality
The documentation at linuxdoc is incredibly sub-par. It's better than nothing, of course, but it's really not that good.
This is the deal: The Linux documentation that does exist tends to be the kind of stuff that ether breathing geeks find useful to themselves and their ilk. So it dives straight into the details without any regard for the big picture. I cannot tell you how many times I've stumbled upon documentation that dives straight into obscure nonsense without telling me the basic functionality of the thing it's documenting! I can read pages and pages and not even know what the damn thing does! Absolutely amazing. It all stems from hackers and their egocentric view of the world. No sense or care for the mere mortals (or even geek-diety-wannabees) who just want to know what something does and why.
It's poor discipline and lack of professionalism. There are plenty of bright and inspired geeks who can write useful code. But it takes an even brighter and wiser individual to recognize the need to articulate that code's purpose in language that is comprehensible and meaningful to a general audience.
Here here. Couldn't agree more.
Although the Linux Documentation Project is a noble effort, it does nothing to insure or even measure the quality of the documentation it presents. Additionally, there needs to be documentation on how to write documentation (and no, the 'LDP Author Guide' doesn't cut it). There needs to be a guide written for geeks that explains the basics behind decent technical writing and presents clear cut ways to go about it. Wouldn't hurt to talk a little bit about sentence structure and outline form either. Perhaps get a few things documented extremely well and then refer to them as examples. Something. Anything. I'd be happy if the hacker community would simply recognize it as a shortcoming. That alone would be a minor miricle.
(Great idea whjwhj. Get right on it!)
What we need to measure is LOD: Lines of Documentation. We measure that against SLOC (Source Lines of Code) and we would learn that Linux is, by any rational account, very poorly documented. And, compared to (more-or-less) intuitive full GUI environments, Linux really needs documentation. GOOD documentation.
Which might help explain another number that keeps cropping up: 5% of the OS market.
What I find amazing is that I bought $1 billion worth of software (Windows 2000) at my college's bookstore for $5.
Hi, I'm a pretentious cock who will make some gay comment about ignoring AC posts here.
Overall, a very interesting article (did you read the whole thing?), now I have some questions for the more savvy of you out there in geek land.
- I'm curious if any of you know why the more "obscure" languages were used in the kernel instead of C, or * shell, or assembly language. Is it because, as the author implies, that these languages are better at doing specific things?
- I see one of the major reasons that the kernal is so large because it needs to support so many different kinds of peripherals. Please excuse my ignorence, but why can't the dirvers be kept in an online database, with a user friendly menu, so the user could just download the mininum code to do the job. Aside from the work involved to set that up, it seems like a reasonable way to keep down kernal bloat. Is anybody doing this? Am I correct in assuming a smaller kernal is faster?
- I would propose the same "solution" for the "arch" CPU architecture code, if those two things were implmented, wouldn't it shrink the kernel size by more than 70%? Am I missing something big here, as to why this wouldn't be good?
Answers please.
-------------------
Hey, Ho, Let's Go!
R.I.P. Joey Ramone
If it don't GO... chrome it. ~ Frank Banks
Ummm... they're NOT used in the kernel? Perhaps you missed the chart which showed all the languages (and lines of code for each) in the kernel? It's hardly written in "only" C and assembly.
Let me refresh your "memory"...
SLOC Directory
SLOC-by-Language (Sorted)
2437470 kernel-2.4.2
ansic=2285657
asm=144411
sh=3035
perl=2022
yacc=1147
tcl=576
lex=302
awk=248
sed=72
Perhaps sed, awk, lex, tcl, and yacc aren't "obscure" to you, but they sure are to me.
-------------------
Hey, Ho, Let's Go!
R.I.P. Joey Ramone
If it don't GO... chrome it. ~ Frank Banks
Since when is the number of lines of code proportional to the quality of the software? If Red Hat 7.1 has 30 million lines of code over 6.2's 17 million, does that mean the product is 76% better? Is the code getting more sloppy as more programmers get involved? I feel like counsel is leading the witness for the author to say 7.1 has "60% more effort" under the COCOMO model. Kernel programmers, weigh in!
It's really nice to see this story on /. looking through the links, it looks like this is a legitimate, fairly unbiased analysis of not only the software itself, but the issues around Linux. Congrats.
My Karma is so good, I'm the Dalai Lama...or something.
So does that mean if RedHat can sell just 13 million copies, we'll all break even?
I don't think it's misleading. The paper specifically says that I'm measuring "GNU/Linux" and "an entire Linux distribution". It should be clear that I'm measuring an entire distribution, not just the Linux kernel.
There's a lot of discussion about "what is contained in Linux". Trying to define various groupings could be interesting. I've posted all my data for anyone who wants to identify and study various groupings. For most of today's users, I think such discussions are irrelevant -- to them, the set of CD's labelled "Distribution-X Linux" defines what "Linux" means to them - and that's what I measured. Perhaps in the future the Linux Standard Base will set what a "minimal GNU/Linux system" is. Don't expect a study of SLOC to tell you whether to call it "GNU/Linux" or "Linux"; I do discuss the issue in section 3.2.
> the study assumes one company/group doing all the development, which further separates the study from reality.
Nowhere does the study assume that. We all know that's not true unless you define the group as "developers of open source and free software." Red Hat did develop some of the code, but only a small fraction of it. All I'm trying to do is to estimate how much effort went into making this set of software - not who put in the effort.
> Another small nitpick is that Perl is actually licensed under the GPL... Perl is also dual-licensed under the Artistic license... but that doesn't mean that it isn't GPLed.
I don't understand how that is a "nitpick". Look at the data -- you'll see that Perl's license is assigned the value "Artistic or GPL". The paper is already correct.
> Since he didn't count either .xul or .js source files, the figure for Mozilla is much too small.
That's true, I don't count XUL or Javascript, so Mozilla might indeed be underestimated. That would move the total development cost to a cost even greater than a Gigabuck.
Just in passing - it's important to separate "development cost" from value. Something may cost a gigabuck to develop, but have a much greater value to the world at large (from saved costs worldwide, reduced barrier to innovation, lifetime profit, or whatever).
- David A. Wheeler (see my Secure Programming HOWTO)
> Using RedHat as a distro for this project isn't that good of an idea.... it's just an unrepresentative mass of programs and code! I can safely say that most Redhat users will never use about one-quarter of the programs in their distribution...
That's true for any of today's operating systems. No user uses all the code in Windows, either. Even real-time OS's have more code developed for them than is used by any given user. As a measure of effort, though, examining all the code makes sense.
> Since when is the number of lines of code proportional to the quality of the software? If Red Hat 7.1 has 30 million lines of code over 6.2's 17 million, does that mean the product is 76% better? Is the code getting more sloppy as more programmers get involved? I feel like counsel is leading the witness for the author to say 7.1 has "60% more effort" under the COCOMO model."
I never said it was "better", I said it included "60% more effort." Better is a value judgement. Effort is measured in person-years.
> The kernel shouldn't be two million lines of code. How much of that is drivers? And how much of the drivers are duplicated from one driver to another?
Section 3.2 specifically discusses this; 57% of the lines of code are drivers. Duplicate files are only counted once, but "partly duplicated" files are much harder to detect (and to discount when they happen); they certainly happen in the Linux kernel. However, the COCOMO model is based on real project data, and many other projects include cut-and-pasted code (for good or ill).
> Ok, so this guy claims that Linux would cost a little over $1 billion (US) to develop. I wonder what the big deal is. I'm sure Microsoft has spent that much over the years on Office+Win9x+WinNT+Backoffice+etc ... The only thing incredible about this number is that most of that billion was completely unpaid, or at least underpayed.
But I believe that is a big deal. Gates' "Open Letter to Hobbyists" assumed that if people just shared code, no large project would be developed. GNU/Linux and other open source/free software systems show the assumption wrong, and this paper has the numbers to prove it. You can argue which is "better", of course, but the notion that it can't be done is no longer debatable.
> Are there estimate[s of] how much money in form of salaries were ever paid to programmers for the code and how much was in effect done not only voluntarily, but also completely on an unpaid basis?
Unfortunately not; it's not even clear how to find out. You would have to go back to individual patches submitted to every project, and few people identify in their patches "I was paid to do this."
> 2437470 source lines of code for the Linux kernel. Doesn't that worry some people out there? We have a monolithic kernel almost two and a half million lines long. I think that by 2.6 the kernel is going to collapse under its own weight unless the designers decide to reorganize it in a fundamental way.
It's the nature of a monolithic kernel, and in any case, most of that is in modules (which are individually much smaller and only loaded when needed). I see no evidence of a "collapse", though clearly there are competitors (like HURD) that might eventually replace it in the market.
> Quoting statistics/data going back to '95 is way out of date by todays standards, even '99 is now very old.
It may be old, but it helps give perspective. A simple SLOC number doesn't mean much to people, unless it's compared to something else.
> The cost formula includes a term (ksloc**1.05): i.e. thousands of source lines to the power of 1.05. This reflects the fact that the bigger a program becomes, the harder it is to add new lines, because the system you are adding too is more complex. He plugs the size of the entire code base of RH7.2 into this formula. This seems unreasonable to me - these are many almost independent packages.
No, I don't do that (for the reason you cite). Section 2.3 of the paper discusses this: "Each build directory had its effort estimation computed separately; the efforts of each were then totalled." Appendix A mentions that sloccount was given the "--multiproject" option, which implements this.
Anyway, I hope people found this study interesting. It sounds like several people did.
- David A. Wheeler (see my Secure Programming HOWTO)
To me, it's the "true Linux." For more reasons than my non-sober state can manage right now.
My IP is 192.168.1.100 Hack it if you want.
I notice RMS talking about the goal of creating 'GNU OS' but is not totally true. The goal also included GNU Utilites, Applications and Compilers. The FSF did succeed in creating Utilites, Applications and Compilers. But they didn't create an OS.
And three hundred and fifty thousand annoying slashdot trolls...
...either employed by Microsoft or harboring the dream that someday THE BILL will bend them over the holy WINDOWS altar and give it to them in person....
Max
My god carries a hammer. Your god died nailed to a tree. Any questions?
-- Dr. Mike
I know this is slightly offtopic...
/usr/src contains only the core OS source: the nearest equivalent to Redhat is FreeBSD plus a ports tree with 'make patch' or the like run on every port... It could get quite big then.
:)
He mentions that it might be worthwhile to compare the results obtained for Redhat Linux to other distros, plus OSes like FreeBSD.
This could be quite interesting, but wouldn't really be a meaningful comparison. FreeBSD's
Thus I think comparisons would only be meaningful between Linux distributions with similar amounts of bloat^H^H^H^H^H included software (ahem)
how linux users point at BSD for being so splintered, when they're really the more organized, concentrated bunch. Not advocating so much as observing. This article brings a great deal of overlooked (perhaps intentionally) details to the forefront.
----------What the Chiquita banana?
Of course, I also think that people should be using more high level languages than C/C++. C/C++ programs are not particularly efficient (mostly because programmers have to do so much nitty gritty work that they never get around to doing the good stuff, and because everybody reinvents basic facilities over and over again), and they are difficult to extend and modify compared to programs written in more high level languages.
Maybe what we need is a Java, Perl, or Python-based Linux distribution--a distribution that throws out most of the C/C++/shell code other than the kernel and replaces it with small, extensible utilities in those higher level languages.
No, I was talking about technological side.
...and you can't blame meteors for everything.
Emacs:
... just another browser.
No way, it is simply an overblown editor ( equivlent of MS Word for programmers )
GCC:
C compiler - enough said. It has been ported to many platforms but it is still C compiler.
Tex/LaText:
This one is close. There are typesetting commercial system available but , well, you might have a point here.
X11: MIT project , has nothing to do with OS
Mozilla: Eh
...and you can't blame meteors for everything.
", what separates Linux from the older *NIXes? "
... Sure people managed to almost completely duplicate standard Unix platform without central management and often without funds but I fail to see how does it change anything.
This is precisely the question that people should be asking every time they encounter somebody claiming how revolutionary Linux is.
From the technological point of view, there is nothing there that has not been done before (often better.)
Frankly, the only unique part is HOW it was developed but it has not much to do with the technology itself. We are not talking here about revolutionary new stuff
One cannot escape conclusion that OS model is good enough for duplicatinh stuff that has been done before (no need for research) but completely fails at introducing innovative solutions and ideas (please feel free to correct me here, but I have never seen truly innovative OS project - everything seems to be remake of some existing - usually commercial - application.)
...and you can't blame meteors for everything.
There is nothing revolutionary there.
Frankly, show me one usefull feature on RH distribution that hasn't been done before ?
...and you can't blame meteors for everything.
In particular, it would cost over $1 billion ($1,000 million - a Gigabuck) to develop this Linux distribution[RedHat] by conventional proprietary means in the U.S
This is just devoloupement, it doesn't include the fact that some ideas are revolutionary and there value would be much more.
-------------------
Insert Witty Remark Here ===>____________________________
...is caffeine. Lots and lots of caffeine. I don't care if you're a programmer, a system administrator, or a homebrew hacker (in the old and true sense of the word). Without the readily available supply of that wonderful drug called caffeine, who would say that Linux would be even 1/4 the phenomenon that it is today? Hmm?
Am I the only person who cringes every time I read "x-windows?"
Or have they officially changed the name? (might as well...)
--
--
"You know you want me, baby." --Crow
The author was referring to (as you asserted) development COSTS; in other words, how much it might cost to pay programmers to duplicate all of the components of RH 7.1. However, it seems as if the author is speaking in terms of the kernel PLUS all of the userland components (including X,) so the gigabuck figure (and corresponding man-years) is really misleading. Plus, the study assumes one company/group doing all the development, which further separates the study from reality. Don't get me wrong, it was an interesting read, but I think the overall conclusion is a fallacy.
-- Never hit a man with glasses. Hit him with a baseball bat.
I began reading the essay and as I started thinking about the numbers presented (even if there is a 25% margin of error) and I was amazed. Think about all the effort that went into all the components that make up Linux, GNU/Linux, whatever... All those people spending late nights chasing down obscure bugs and adding new features, for no reason except to further the OS and the community. It's an astounding effort, and all of us who have ever written a single line of code, posted a bug report, or helped someone out in a news group should be proud.
Alright, enough with the sentimental crap...you may continue your bickering now...
"Karma can only be portioned out by the cosmos." -Homer Simpson
Of course, everyone realizes that theh 30 million lines of code are not all new lines of code when compared to the 6.x release. Many of those lines of code have been around for quite a while. The $1 billion number must also include some of the development cost for 6.x
<<the BSD operating systems have - documentation that exists in a single form and written in a style that is consistent across the entire operating system. >>
Are the BSD manpages as sucky as the Linux versions? Last I checked, there was no cross-reference capability, very small granularity (do people really enjoy reading through 200 screenfulls of information just to get to what they missed last time?), and myriad other problems. Yes, Linux uses sucky documentation formats too--but not all Linux is sucky.
There are reasons why democracy does not work nearly as well as capitalism.
-- David D. Friedman