Slashdot Mirror
Squaring the Open Source/Open Standards Circle
Andy Updegrove writes "Before there was Linux, before there was open source, there was of course (and still is) an operating system called Unix that was robust, stable and widely admired. It was also available under license to anyone that wanted to use it, and partly for that reason many variants grew up and lost interoperability - and the Unix wars began. Those wars helped Microsoft displace Unix with Windows NT, which steadily gained market share until Linux, a Unix clone, in turn began to supplant NT. Unfortunately, one of the very things that makes Linux powerful also makes it vulnerable to the same type of fragmentation that helped to doom Unix - the open source licenses under which Linux distributions are created and made available. Happily, there is a remedy to avoid the end that befell Unix, and that remedy is open standards - specifically, the Linux Standards Base (LSB). The LSB is now an ISO/IEC standard, and was created by the Free Standards Group. In a recent interview, the FSG's Executive Director, Jim Zemlin, and CTO, Ian Murdock, creator of Debian GNU/Linux, tell how the FSG works collaboratively with the open source community to support the continued progress of Linux and other key open source software, and ensure that end users do not suffer the same type of lock in that traps licensees of proprietary software products."
The REALLY nifty thing about UNIX is the userland. Without it and its tremendously clever and somewhat unique approach to solving problems in the way it does, it's REALLY not just "Linux" you should be talking about when referring to a modern UNIX-reimplementation. It's GNU/Linux.
:%s/Open Source/Free Software/g
YTARY!
This article is more about standards in Open Source development, specifically Linux
To me Open Standards are much more important than Open Source. Open Standards allow Open Source solutions to be created that are compatible with the other solutions.
For me, the most confusing thing about linux is the variance of licenses.
I can understand the variations of programs: window managers, desktop managers, mp3 players, video players, etc, etc. However, I can't even begin to grasp the variation in licenses associated with linux programs. Anyone else as clueless as me?
Unix was killed by the high price of licenses. Unix during the early 1990's was supposed to be for the big boys --- the enterprise customers willing to pay up to 10,000 USD per seat for a Unix license.
With the license for Windows NT starting at less than 1000USD, the enterprises which formed the majority of the paying Unix customer base soon found a way to make do with NT and delete their Unix installations.
It wasn't open standards and the fragmentation that did Unix in, it was plain hubris among the Unix vendors who cannot fathom a future where a cheaper Windows NT would replace the robust, stable and widely admired Unix they are selling.
All mainstream package formats have the full installation path hard-coded in the archive. LSB does not address this yet. The other problem of RPM, namely binary compatibility between different library versions, is already solved by compiling with apbuild. This works surprisingly easy, and allows my to provide one single package that can be installed everywhere [1].
[1] I can recommend to compile packages at Slackware because Slackware ships most packages without patches. Compiling an app at SuSE for example, made binaries depend on ABI changes caused by SuSE patches.
The best way to accelerate a windows server is by 9.81 m/s2
The article summary is a bit of a flamebait. In order for a product to fork there must be two forces in action.
1) Licensing that allows a fork.
2) Frustrated users who feel like they can't shape the future of the product via existing channels.
This is why there are at least three forks of java and none of perl. I suppose one could argue that the forks of Java are not true forks but attempts at re-engineering but the end result is the same.
Will linux fork like Unix? Well in a way it already has, there is real time kernel, different kernels for devices etc but not in the way the article talks about it. The article isn't talking about forks per se it's talking about distros. The author seems to have missed the point that the Unix forks were actual forks in the kernel not "just" distros.
Weird article really. Kind of pointless too.
evil is as evil does
Those wars helped Microsoft displace Unix with Windows NT, which steadily gained market share until Linux, a Unix clone, in turn began to supplant NT.
When did this happen? I must have missed it.
Just because they (UNIX and Linux) have samelike POSIX architecture, that don't mean they are same entity. Linux is a kernel while UNIX (and all of its descendant like *BSD, Solaris etc.) is a whole system (userland).
I'd like to support the nonfragmentation of Linux - as I guess many would. But looking at the LSB 3.0 certified list http://freestandards.org/en/Products, just shows Red Hat, SUSE and Asianux. Are these all the choices I have?
Could someone please explain me?
Mostly because different applications were emphasized by the needs of the individual companies liscencing it. SGI Empahized engineering simulation and graphic design and later on server applications, Sun emphasized server and database applications and IBM emphasized mainframe operations and engineering applications. Companies centralized different forms suited to their own needs and those of their customers based on the development of their specific hardware. The hardware was the thing back in the early 1980's rather than the OS, it was mostly about selling their own products with the development of the OS being somewhat secondary to the hardware. Unix was already well establishedas an OS by 1980.
Unix at the time back in the late 80's and all through the 90's was better suited to this level of customization than DOS or anything else. It had already garnered the majority of its market share in research institutions and universities by the time NT came into the market. NT is just an imitation of Unix based on BASIC rather than C. The two operate very similarly but unix is much more robust and securable even to this day.
UNIX is still going strong, but linux, being open source is gaining against this because everyone can access the source code and customize it to their own specific needs, putting their own empahisis on their own particular applications.
You can do anything with Unix and Linux, moreso than NT. Most market share for any operating system can be attributed to timing in the market and luck, AT&T and IBM had already established Unix as the main OS in their own systems and had that market share already established when M$ started as a company. AT&T really did most of the initial development work and handed it over to the universities because at the time they weren't allowed to enter the computer market by law.
Timing is everything in any market.
Stupid Humans.....
Sure this leads to some incompatabilities and duplication of work but there are several ways for developers to mitigate this. Open standards are essential as they allow code be ported between distros rapidly. Another good idea is for devs to be involved (in some way) with using multiple distros. Different projects could work together more closely to achieve better interoperability.
Its an essential aspect of forking to accept that many forks are dead ends and should be allowed to die or merge back into the tree where desirable. There are many good projects out there and it isn't really in everyones interest to reinvent the wheel continuously.
the more they over-think the plumbing the easier it is to stop up the pipe
All your (Linux Standards) Base are belong to us.
My father is a blogger.
Similarly, Linux isn't displacing NT, it's displacing commercial UNIX.
The overlap of functionality between NT and Linux is, really, quite small. There aren't many cases for which Linux is a good solution, where NT could also be (and vice versa).
The thing I do not like about the LSB is that it seems to be pampering too much to the desires of the closed source application suppliers who only ship binaries not source. When the application is available in source form, many of the issued addressed by the LSB become either irrelevant or much less important. The binary distributions can build binaries for RPMs, debs etc for their own distribution and for the most part, users can install from source using the 'standard' "./configure; make; sudo make install" without having to worry about having the exact layout and library versions mandated by the LSB.
NT? Plan9? BeOS? VMS?
You don't seem to like Linux.
http://michaelsmith.id.au
The funny thing about the LSB is that it concerns APIs for use by userland programs -it has _absolutlely nothing_ to do with the kernel. All of the requirements for LSB compliance concern calling conventions, executable formats, libc, POSIX facilities, filesystem layout an other extra-kernel configuration, most of which any UNIXoid system could support.
There are no obstacles to Darwin, *BSD and Solaris systems meeting LSB compliance, because it has nothing to do with kernels and everything to do with the specific details of a UNIX userland environment.
Generally I don't get into 'Linux' vs 'GNU' discussions but the LSB is once case where I feel the name 'Linux' is used completely inappropriately.
By the time Unix died, it had already given birth to Minix, which had already given birth to Linux. Linux is very much not dead.
stillborn
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
"Not only is UNIX dead, it's starting to smell really bad." - Rob Pike 1991
Plan 9 from Bell Labs.
You sounded like a pretty technical person until you came out with this "snobbish" comment which, to me, makes absoultely no sense whatsoever.
I'm in a similar situation to you except that I use Linux about 80% of the time and keep a Windows XP machine for a little desktop stuff and for gaming. But I have no qualms about putting XP on my network because I keep it updated and don't use Outlook for email on it - likewise, I virus/spyware check it regularly and don't go to those sorts of places on the Internet where I might download a trojan. As a result, the XP box is secure enough for what I need it for and, to my knowledge, has never had a virus on it.
Sure, as a predominantly Linux user with a lot of UNIX/Linux expertise under my belt, I prefer Linux and my computing time on Linux writing scripts, compiling software and generally tweaking & learning is far more productive than on my Windows machine where I spend so much time running virus checkers and installing upgrades. But I don't view the XP machine as being a "threat" to my Linux ones, it's a "necessary evil" to me, nothing more.
I just find it strange that because I read on Slashdot constantly about how secure the BSD Unixes are (I've no reason to doubt those statements either), you would consider a Windows 2000 machine as a threat to a network of them....
Gentoo Linux - another day, another USE flag.
The kernel, however, leaves a little to be desired. There is no support for aio_* or pthreads, for example. The kernel still doesn't support any kind of fine-grained locking, so it will not scale beyond about 4 CPUs.
I use OpenBSD on a couple of machines, and it's a real joy to use. Hopefully it will catch up in terms of kernel features soon, although not at the expense of doing things the right way.
I am TheRaven on Soylent News
turn up the baud rate baby !! make my terminal go faster
% stty
speed 38400 baud;
lflags: echoe echok echoke echoctl
oflags: -oxtabs onocr
cflags: cs8 -parenb
erase intr
^H ^?
I want to SSH in to the server and set up my new wave NFS
Do you think that is the pinnacle of OS design ?
if you s/pinnacle/nadir/ and you might be on to something
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
If a standard is as simple as possible, while being reasonably
efficient, then it is best.
This is also known as KISS, Ockhams Razor, etc.
Or as Einstein said: It should be as simple as possible, but not simpler than that.
Kim0
"Before there was Linux, before there was open source, there was of course (and still is) an operating system called Unix that was robust, stable and widely admired.
.jpg to verify you're not a bot. hmm... i'll be off, doing... stuff...
please cue the "and before linux we went barefoot over snowy hills to our servers, and liked it" jokes. thx.
before spammers, you didn't have to decipher words like "condom" on a
The pinnacle of OS design is when I can throw my computer out of the window and reflash my brain and connect it to the brain internet at will.
The brain internet is coming! You probably heard it here first!
Oh dear, This means I have a room full of Zombies at my office
The only things certain in war are Propaganda and Death. You can never be sure which is which though
I wasn't being snobbish. I don't let it on the network, because it is the only Windows install I have here and I don't want to set up rules to allow this machine on. I do OS fingerprinting and by not willingly allowing Windows on the LAN it helps to keep things simple.
The closed source nature of Windows means that I cannot be sure of what a Windows machine is saying over the LAN, not that I am overly sure of what my Unix/Linux might be 'getting up to'. I also resent having to check a Windows box for trojans and such like. It is an informed and reasoned decision not to allow the w2k install to have net access. It keeps things simple and that is how I like it.
Apologies if that offends anybody, it wasn't my intention and I think I have been misinterpreted slightly. The Windows 2000 install does what I need, in playing a few games, without the need to keep it updated with patches and other updates, that I am not permitted to look inside of.
The biggest challenge in what we do is probably no different than in any other standardization effort: Balancing the need for standards with the need for vendors to differentiate from each other.
I think the biggest achievement isn't so much the software (OS / OSS) but the whole concept behind it. While the software has ofcourse helped big time when it came to 'swaying' the masses the really impressive part in my book is that it gained so much momentum that it was enough to draw some serious attention and show the big companies (Microsoft, Sun, HP) how it can be done. And even though some claim to follow the market with close interest you can't convince me that some of these companies didn't startle when noticing the growing success all of a sudden.
So... Open standards, open source, e.a. are very nice and important, but I think the biggest achievement has been made in the overal field.
What's this about "various types of licenses" under which Linux is supposed to be available? Linux is GPL, so forking is possible, but there is no risk of UNIX-style fragmentation because the source is open and copyleft. For somebody to create a "closed Linux" they would have to start from scratch. You can't add closed bits to GPL software and keep them hidden, so any incompatible Linuxes ("fragments") could always be re-connected by users irritated about the differences.
The nonsense about UNIX displaced by NT and NT in turn displaced by Linux already set off my alarm, but the above really is FUD designed to further somebody's personal agenda.
It is not possible for UNIX-style fragmentation to happen to Linux, because of the GPL.
From the submission:
Unfortunately, one of the very things that makes Linux powerful also makes it vulnerable to the same type of fragmentation that helped to doom Unix - the open source licenses under which Linux distributions are created and made available.
I believe fragmentation has very little to do with the issue concerning the doom of UNIX. My three top reasons are:
1) Price of purchase
2) Expensive/hard to administer
3) Stagnation in development
Users want the cheapest, easiest and most feature-filled solution. It's pretty straightforward actually, and a Personal Computer with Windows was the first to fill the niche, if you leave out Apple.
Apple lost because they wanted monopoly on _both_ hardware and software, while Microsoft only wanted to control the OS (in the beginning). More importantly, Microsoft was better at hyping/marketing their next generation, something that Apple has learned to do better in the recent years.
UNIX and IBM lost because they failed to scale down to personal PCs, which is where the commodization of computing happened in the 90's. IBM and other mainframe dealers refused to understand the Personal Computer (too much vested in big contracts), thus the clones took over along with Microsoft Windows while the dinosaurs waited it out.
Without the IBM PC Clone, the computing world would probably look very different today. In those days it was very attractive to be able to upgrade the PC, exchange parts and use commodized hardware for the whole rig. Many tasks which rented expensive CPU-time on UNIX mainframes, were moved over to PCs during the 90's.
Fragmentation, no doubt, can be very bad for development, but it is also a boon since it leaves developers free to explore different avenues regardless of politics and limitations. I think once a system becomes popular enough like "Linux", the demand for standardization will pull it together. Hey, even the BSDs keeps compatibility with "Linux".
What killed UNIX was lack of creativity, focus, commodization, too much control and maybe most importantly: arbitrary high prices just to milk customers.
Linux may have killed off UNIX (oh what irony), but NT have been beating the crap out of it for many years. Linux and UNIX never actually competed on even terms, because UNIX has already been pretty much abandonded for a long time - it's owners only keeping it for milking the last drops.
My pet peevee with bash and the GNU utilities is the lack of standards, and lack of further development of the command-line. In that regard, I hope "Linux" can progress without having to be beat by Microsoft releasing a better command-line.
POSIX is really an antique joke compared to what could be possible via the command-line. So the trap "Linux" might fall into, is the same as for UNIX: stagnation, because most users drool at eye-candy and not the actual implementation in the back-end. However, maybe the cost of switching command-line is not worth the gain, time will tell.
http://www.debunkingskeptics.com/
It might be that the LSB makes life easier for distribution but does it also have an effect for developers and users? I don't remember I ever have looked into the LSB when designing and coding an application nor when distributing source files. And I'm quite sure most users don't even know that the LSB exist. While the LSB is very important for the binary distribution, it's influence on a Linux system is rather limited. Yet the FSG only cares for the LSB and therefore it's importance is also rather limited.
A useful Linux system needs some user's GUI guidelines, more specific a single set of guidelines. A set of guidelines which are usable anywhere not just on a single desktop, best if usable cross-platform, something like wyoGuide (http://wyoguide.sf.net/). IMO the FSG should not only standardize the binary interface but also delve into standardizing the GUI interface. This would IMO give the FSG much more importance.
O. Wyss
See http://wyoguide.sf.net/papers/Cross-platform.html
Eunuchs can't give birth. They don't have any... oh, you said "Unix," not "Eunuchs" Nevermind.
You missed the point of not having the win2k box on the network. It has
nothing to do with protecting the BSD machines and everything to do with
protecting the win2k box.
I do something similiar with an XP box at work. If it's on the network,
then I'm required to let the network security goons run their software
on it that lets them monitor and make changes without my permission. By
not having it on the network, it remains a stable development platform
that I have full control over.
*sigh* back to work...
The kernel is fairly much irrelevant for portability. Userland headers, C library compatibility, file locations, compiler options, linker options, Bourne shell incompatibilities, C compiler incompatibilities, C compiler and library bugs, word size, etc/passwd handling, ... - those mattered. The differences in the kernel itself are fairly irrelevant.
The kernel has generally communicated with the userland through a fairly narrow range of paths: Mostly the syscalls and the ioctls, and making available data through filesystems. The narrowness of these paths has meant that it has been fairly easy to deal with the differences, at least compared to the userland differences.
Eivind.
Doubting the existence of evolution is like doubting the existence of China: It just shows that you're uninformed.
Unfortunately, one of the very things that makes Linux powerful also makes it vulnerable - the open source licenses under which Linux distributions are created and made available. Happily, there is a remedy, and that remedy is BSD-style licenses.
There, fixed that for you.
xterms have a baud rate even though they are not serial consoles
NFS - puke
SSH - band aid
etc. etc.
it *could* be so much better
But you have to get people to even see it is a problem, like being alcoholic.
Unixaholism is a disease, DrSkwid is the cure !!
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
if you don't modify the source code, you don't have to go out of your way to make sure it is present if you redistribute the software. the original source code is already freely available.
the OP's point is that you don't have to worry at all about the legality of burning/uploading free software to which you have access.
my password really is 'stinkypants'
Unix has and/or had lots of standards. POSIX, CDE, Openboot, and OSF are some of the big names that leap to mind.
It is somewhat simplistic to assume that the relative lack of success for Unix is due to any single factor. Unix has lots of things working for and against it -- price, maturity, perception, marketing, legal history, fragmentation, ISV support, etc. However, it's worth noting that having standards really does not prevent fragmentation. Most vendors don't ship a minimal product with only standard features. Instead, vendors ship "value-added" tools and APIs along with the standard. Any ISVs (independent software vendors) that target a platform have to worry about any ways in which they drift from the standard. The Linux LSB does not solve this particular problem -- despite the LSB, ISVs continue to target specific Linux distributions.
NT did not broadcast its serial number. You could buy a single copy of NT and install it a thousand times. If you needed a new file server or a temporary file server, it was so much easier to setup another NT box.Yes you do. But they're still in the organizations that had them before.
What has changed is that Windows servers swept through the smaller companies. Those companies never had a *nix box. They might have had LANtastic or NetWare or nothing, but they did not have *nix.Okay, I can agree with you on that.I guess that depends upon what business segment you're talking about.
Linux has been showing double digit growth for the past 5 years (maybe longer). Businesses are deploying it. At the server level.Now you're talking about the desktop segment.
The corporate desktop segment is different than the corporate server segment.
And the biggest problem with the corporate desktop segment is all the Access databases that have been built over the years.
The 2nd problem is all the not-supported-or-sold-anymore Windows apps that users "absolutely must have to do my job" that they've acquired over the years.
Changing 10 servers is easier than changing 10 workstations for users who've spent 10 years with the company.You might want to take a look at Google before you talk about "hobbyist market".I'll have to disagree with you on that.
While that would be nice, it is far more likely that one distribution will become dominant and that distribution's structure will become the de facto "standard".
And it seems we're already on that path with Red Hat and Ubuntu.
O wise master! I have emptied my cup! Please teach me your mysterious ways!
I thought forking and merging (thanks to the license) are strengths of Linux. It's an evolutionary development process where the best or most popular ideas rise to the top. Basic standards arise out of the most widespread/adopted projects.
It's hard for me to see in this chaotic (but necessary) environment how much external control developers are willing to have "imposed" on them by such standards - unless of course from a development/technical standoint it makes sense.
My understanding is that Linux really isn't in the game of competing with anybody (Unix, Windows or otherwise) anyway. it's just about the code, love of things computers and a new way of doing things.
E.
Offtopic: I'm I the only one who believes that the various color schemes provided with those desktops must have been designed by blind monkeys?
is that you can sound like you are saying something even when you say nothing. The parent comment could be applied to almost any article. Each statement in the comment may be true, but without even an anology between the truisms in the comment and the story what good does a comment like this do? Do you see a solution that is needlessly complex, or something that is only a complex as needed by the problem?
Think global, act loco
Where's the script that installs symlinks on any distro's filesystem that makes every "filesystem API" available on any Linux distro?
--
make install -not war
IMHO the reason most "Forks" of Linux have happened, is just because it is allowed. Someone that is a real geek, and I mean that in the nicest way, decides it would be really cool to take the kernel, and do something that suits them. There is nothing wrong with that, as it is what the heart of OSS is about (again, IMHO). I think as long as we have some standard distros, that folks should still inovate, as without inovation, Linux would have never happened in the first place.
I personally have tried over 10 different distros, one works for certain things REALLY WELL, and another works for OTHER things really well, it would be great if they all worked for everything, but that will NEVER happen, not with UNIX, Linux, or Windows.
----- I have bad karma for a reason! -----
Those wars helped Microsoft displace Unix with Windows NT, which steadily gained market share until Linux, a Unix clone, in turn began to supplant NT.
;-)) Linux would have begun to supplant Windows 2000.
Does anyone have figures for this - that Linux has supplanted NT.
Not that I don't believe it - Since it's many years since NT was sold, it may
be possible that Linux has begun to supplanted NT. Hopefully by 2010 or so (when
Vista will be released
That's because of shared libraries, a hold over from the time of extremely expensive hard disk space and RAM. There's little to no need for it now, apps should contain the files needed to run them, so they can just be contained inside their own directory. No breakages with newer releases then, no wondering which version lib is needed by this or that, as the package releaser would include it. Download app, stick it in a programs directory or wherever you want, it will run then.
I know this is heretical in the old school unix way, but really, it just isn't needed any longer. They went to shared libraries because of hardware cost issues and because machines were still slow, it was a work around compromise that is now solved, so it is not needed any longer. The success on the hardware side is being *resisted* on the software side by the "save the dinosaurs!" mindset. Normal cheap consumer desktops can now easily hold two or more gigs of RAM and have fast processors orders of magnitude faster than processors from 15 or 20 years ago, and hard drive space is down to what, a quarter or 50 cents a gig or something wicked cheap like that? This is buggywhip mindset *inertia* more than anything else keeping us stuck in incompatable library hell, which is the main reason you see menus not being functional and why stuff breaks with every dot point release cycle of distro xyz. It's nuts to keep doing that particular thing all the time when it is *not needed* anymore.
Linux has one thing Unix never did - if someone forks it and does something innovative, then ther forks/branches can use it too, thanks to the GPL. The varous 'Unix' flavors didn't allow that.
So the LSB is a blanket organisation - like the 3rd International? Or is it more like the 4th? Don't these things inevitable schism?
There was an arrogant attitude toward PC hardware in the mainframe and workstation market. If you wanted to do real computing, you wouldn't use a PC -- those were just toys! Drop 15 grand on our workstation and then we'll talk. Well PC's WERE toys for a few years, but you had to have blinders on to see that they weren't going to make progress. That arrogant attitude persisted while the 386 and then the 486 came out, while all the while Windows NT and to a smaller extent OS/2 started stealing more and more business from the traditional UNIX vendors.
And while the UNIX vendors arrogantly believed they had a better product, not a single one of them ever made an effort to push the GUI portion of UNIX beyond CDE (Well... except NeXT and SCO, but SCO's offering was a step back from CDE.) Gnome, KDE and Enlightenment were all efforts of the Open Source community and to my knowledge Sun's really the only one of the old guard to even consider using one of them. Hell, even Afterstep is a step up from the commercial vendors' offerings.
In the end it was cheap Intel hardware and cheap Intel operating systems that did the old guard in. Windows on a pentium made a server that worked well enough that it was impossible to justify the price jump of an order of magnititude to get just a little bit more. And I doubt there are more than a handlful of companies that would even consider putting UNIX on an employee's desk. Had the old guard of UNIX vendors played their cards right and embraced PCs as a natural extension of their high-end UNIX systems, things might have gone differently.
The current situation is rather interesting. The cost of Windows licenses is significantly more than the cost of Linux licenses. Microsoft can't really compete with free, so they have to find other avenues of attack. That, more than fragmentation, is the biggest danger to Linux. Most commercial companies only deal with RedHat or SUSE anyway. I don't know what the future will bring, but we most definitely live in interesting times.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
It was also available under license to anyone that wanted to use it, and partly for that reason many variants grew up and lost interoperability - and the Unix wars began
That's a nice, often-repeated story, but it doesn't correspond to reality. In reality, UNIX variants were no worse than the different versions of Windows or MacOS we have had to live with. In fact, once the ABIs were decided upon, arguably, things were somewhat better. And today, at least on x86, the Linux ABI is a de-facto standard that's supported by several UNIX systems. For a brief time, Windows and MacOS had their act together a little bit better in terms of software packaging and installation (driven by necessit), but UNIX and Linux are now lightyears ahead of both Windows and OS X in that area, too.
The primary source of binary incompatibility between machines was, in the end, simply the use of different processors by different vendors; as long as that existed, software vendors had to recompile and ship different binaries anyway. Note that Microsoft briefly tried to handle multiple processors with Windows NT and the effort flopped completely.
What could be possible on the command line that isn't currently?
You'd have to get into some pretty contrived piping examples I'm sure, and how many people actually want to do such things? Even then, the pipe shell should be your shell of choice rather than bash. I think we're all agreed that csh and tcsh are broken, but bash and a few others are pretty powerful.
There is power, I grant you that, but the standards are antique. What you have is:
stdin, stdout, stderr, pipes, some loose POSIX-convention for flags
Take ls and find, two of the most used UNIX-commands and I'm sure POSIX mentions them as well.
man ls
man find
One problem here is that find uses one minus-sign (-name) for long options, while ls uses double minus signs (--all) for long options.
Even worse, many programs allow a single minus for having multiple flags. Eg. In grep this is legal: grep -ive
It is highly ambigous in so many ways. The semantic difference between "find -name XXX" and "grep -ive XXX" is not obvious. The find-example is one option, while in the grep-example it is three flags.
Sometimes flags can be sent in after the parameters, other times not.
Then there's the parameters to each option, which can be either separated with space or not, depending on the utility.
Strings may be sent in, but maybe the program only accepts the first character. Other datatypes are impossible (I'm not saying it is needed on the most basic level - which would break a fundamental UNIX-philosophy).
Errorhandling is not consistent either.
Input, output and error must be parsed and can break for language-specific reasons, or differences between versions. Let's say you make a utility to parse the output of ls -all, which is pretty standard for most ftp-programs, but then the dates get parsed wrong because they might be in a different language!
The whole concept of the command-line of UNIX/"Linux" is antique and should be replaced with a more modern system which is consistent and builds on the same UNIX-philosophy, but do things right in a modern way. I would suggest making input and output available in XML, as an option. This will help for programs that can be automated, like KDE-programs via "dcop" etc.
The problem with the command-line are many:
1) It is hostile to new users
2) It promotes bad conventions and inconsistencies for those familiar with the systems
3) Only humans can understand the interfaces, which makes it bad for automation
Improving the command-line is really a starting-point of a thesis or a big book. Lots of work can be done here, but the problem will be maintaining some sort of backward-compatibility, or making people jump on the wagon.. *cringe*
http://www.debunkingskeptics.com/
Drink from the furry fountain of life and your thirst will be quenched.
One more vessel must you seek : http://www.maht0x0r.net/the_mug.jpg
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
The evaluation of an action as 'practical' . . . depends on what it is that one wishes to practice.
You shouldn't see different binaries for different distros. A Linux app should be an Linux app, period.
/usr/local, including the config files. Most linux distrubtions seem throw everything including the kitchen sink into /usr/bin and then put the config files in /etc. My own personal feelings on the matter are that nothing but core os components should go into /usr/bin and /etc (and no, GNOME and AFTERSTEP are not core OS componets). Everything else (including config files) into /usr/local or /opt.
Amen! Not only is it frustrating figuring out where all the config files are, but having an app fail to install or work because of dependancy or lib versions is also frustrating. I remember having fits trying to install Oracle 8i (circa 1999-2000) and having the install fail because the linker was choking over libc version incompatabilities and LD_ASSUME_KERNEL settings. Ofcourse, all the problems could be resolved by tinkering and patching, but that turns a 30 minute install into a 2 hour install. Installs should be a clean process, not a tinkerthon.
I saw another post here that mentioned apgcc. This was the first i've heard of it, but from the description at its website, it looks like a good idea. Basically, it looks to enforce lowest common denominator libraries and static linking. I like the idea of a fat binary. I also like the idea of self contained app directories (I've never owned a Mac, but I've been lead to believe that is the way it works). Diskspace is cheap nowdays. I don't see why everything needs to be dynamically linked.
Now let me digress into the config file issue. This seems to be a favorite flame topic, so let me don my asbestos suit and jump in. IIRC, the cannonical UNIX practice is to install everything not part of the core OS into
If Apple can do universal binaries across architectures, you'd think all us linux whiz kids could get a cross distrubtion (and cross version) system of binaries working. Ofcourse, Apple has unitary leadership and direction, instead of "hearding piss-ants".
I'm not out to start a flame war, "just evangelising what I love to use." (quotations added for emphasis)
/.
Yes, but sometimes that's all you have to do to get flamed at
If you lend someone $20 and never see that person again, it was probably worth it.
Unix died not because of its downfall but because of MS's rise.
And for Unix vendors, Unix software was not the problem. Unix hardware was the problem.
They gambled that Intel would always build 'toy' processors. But Intel had the volume and competitive security to allow it to become the processor $/performance (and eventually, just plain performance) champ.
MS gained critical mass from being funded by being un-pirateable and ubiquitous (OEMs tend to pay for those copies of MS/DOS and Windows). Awful as their software was, eventually everyone found that they had to learn it.
Eventually even the 'old guard' IT guys couldn't resist the price/performance of a MS/Intel NT box.
In other words, the extra "BS" costs incurred in dealing with a Wintel PC due to inflexibility and quality - though substantial - were less costly than the price/performance and universality aspects of the PC.
MS/Intel proved that indeed you *can* drive a 'toy' car on the freeway given time, money, and enough crashes and money to refine it.
Where Linux wins, it too has a better value proposition. It has just as cheap & fast of a hardware 'home base' as Windows, and its software cost savings out weigh its extra costs.
Does it hurt to hear them lying? Was this the only world you had?
I dare anyone to run a standard complexity metric on OpenBSD and see if you still think the code is correct. OpenBSD is a bloody mess when analyzed by a formal metric such as McCabe or Halstead. It isn't even close to calling itself "clean" or "elegant" or "correct". It is an ugly mess when analyzed by formal metrics. I wouldn't trust it to be correct. Not if my life depended on it.
Unix is great at providing source level compatibility. You can (usually) easily create source code that will compile binaries for a wide variety of operating systems and versions. What Windows does beautifully is take this one step further and provide binary level compatibilty between different versons. I can take something created for Windows 95 and it installs and runs well today on Windows XP. You can distribute a version of a program that works with pretty much all currently used Windows versions as one single file, and that's what vendors typically do. Linux doesn't have that. LSB is a step in that direction but we have a lot further to go still.
.net 2.0, vbrun6.dll, glibc 2.3 etc.) to fill in the gaps so that application vendors don't have to bundle them with the application. It should also take care of installing menu entries in appropriate places and the user should be able to configure where they go and if they install silently or require prompting.
One thing that I think is lacking in all distributions is a good method for installing OS independant third party applications. Most Linux distributions like to rely on the OS's native package management system for this task but the requirements are somewhat different and the two systems should have different interfaces for third-party packages. Third party applications often have options as to what components get installed, are less trustworthy and therefor should have more restrictions on what they can do, and should have extra functionality allowing the package to be non-OS specific. Installing packages today using the OS's package management system requires root privledges and allows a package to pretty much do whatever it wants. This isn't secure at all when dealing with third-party packages.
What would be great is if someone created an open source third-party-program package management system that would allow the creation of packages that would install on lots of OSs (Windows, Mac, Linuxs, etc.) and architectures (x86, x86_64, ppc, etc). A publisher could pick a whole set of OSs and platforms and bundle it all up in one file or separate them however they wanted. Only the OS/archetecture specific parts would have to be duplicated for each bundle. A third party package manager should be configurable to run from non-administrator/root but allow root/administrator things to be done as long as the user is prompted and agrees. Big warnings should be given for anything that would run by default or alter the system significantly. It should also be configurable to prompt for incoming/outgoing IP ports the application may use and other security related things. The installation manager could prompt for downloads of common depenancies that weren't already installed (java,
Putting the reqirement on the OS to make the program work instead of vice versa just makes sense. The current situation makes Linux very ugly and dangerous when dealing with software that's not already in the distribution. A little midlleware glue software that would figure out how make a vendor supplied distribution work properly on this OS would make Linux much more attractive to end users and if the same file they installed on their Windows box also installed on their Linux box, it can only help interoperability between the two.
set softtabstop=4 shiftwidth=4 expandtab nocp worlddomination
Bad luck try again ;)
Ah yes, this is only because the number of cases where NT is a "good solution" is tiny and vanishes in importance when compared to both Unix and GNU? Next thing you will tell me that NT and it's derivatives are a useful desktop or widely deployed server platform.
Microsoft kills what it buys and owns and Vista shows they can't develop their way out of it.
Friends don't help friends install M$ junk.
There have been more than enough windows exploits that didn't require outlook or anything, just RPC holes or similar. Sure, you can turn off services and run antivirus, but why go to the effort when there's no particular need to put it on the network?
I am trolling
"Linux is very much not dead."
I guess that must be true if you say so.
Interestingly, the tide is turning again. Despite the ongoing anti-competitive activities, people realize that they've been burnt by MS, even if only as a result shelling out for software assurance. Though many have a longer more serious list of grievances and disappointments. With all other options gone, that basically leaves only 'Linux'. As a result we are now seeing that sales of Linux servers have shown 15 consecutive quarters of growth. That's sales not general market share which would also include Linux servers installed over other operating systems.
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.
This is one thing I've been trying to figure out.
/lib/UbuntuLib/, and Slackware will put it in /var/log/opt/etc/usr/lib/. So why can't we just massively symlink the bloody directories together? Someone create a script file with two hundred and ninety-six lines of:
/var/log/opt/etc/usr/lib /lib/UbuntuLib /var/log/opt/etc/usr/lib /usr/lib/ObscureSuSEdirectory /var/log/opt/etc/usr/lib /some/other/RedHat/lib/location ... [etc, and vice-versa]
So different distros will put their files in different places. (Actually, I can't believe programs will actually have the library locations hard-coded in, but whatever; I'll accept that the alternatives have some disadvantages.) So Ubuntu will store its WonderfulLibrary.so in
ln -s
ln -s
ln -s
or whatever, and just make sure every possible library directory is symlinked to every other library directory, and we'll be done! It sounds like this way, a distro can meet the (file location requirements of) Linux Standards Base and still be backward-compatible. And we can actually have packages from one distro installing on another! Wouldn't that be great?
It seems so simple and so logical that I must be missing something. Someone please tell me why we're not taking advantage of that epitome of what makes the POSIX filesystems better than the Microsoft filesystems, the symlink?
404555974007725459910684486621289147856453481154 in hex is "You sank my Battleship?"
[GPG key in journal]
>I just find it strange that because I read on Slashdot
>constantly about how secure the BSD Unixes are (I've no reason
>to doubt those statements either), you would consider a
>Windows 2000 machine as a threat to a network of them....
You misunderstand. The problem is not that
Windows 2000 is a security threat to BSD or linux for that
matter. The problem is that any Windows OS is a
threat to the User's data since you cannot see
what it is doing or what it is talking to etc.
Can you guarantee your XP box is NOT phoning
home to the mothership in Redmond? Just what do you
think it is sending back? How do you know?
Clearly you either lack caution or knowledge of even the
most rudimentary experience with loss of personal data
in the MILLIONS of users due to the exploit of
the week, etc. Please, as a public service do a little
googling on how much personal data has been lost because
some company chose to use Microsoft products.
--Johnny hates stupid
However, I can't even begin to grasp the variation in licenses associated with linux programs. Anyone else as clueless as me?
Pretty much everyone is clueless to some degree except for the author of the license. However it sounds like you are even more clueless regarding propretary licensing, especially for the MS Windows platform. If you DID have a clue about licensing in the proprietary world and actually *read* the EULAs (end-user LICENSE agreements) that come with your Windows OS and software then you wouldn't be complaining about licensing under Linux--you'd be heaping praise on it for SIMPLIFYING licensing.
Let me explain: A typical Linux distribution (sorry--GNU/Linux) is comprised mostly by GPLed software, and most often entirely out of OSI-approved licenses in the case of non-commercial distros. This means that there are common characteristics to all the licenses and a user can (relatively) easily determin what their rights are--and generally the end user has a lot of rights (re-distribution and copying, view and modify source code, etc). Furthermore, as new releases are issued of most open source software the license terms remain the same.
Contrast this with Windows and proprietary Windows-based applications: Microsoft alone--ONE VENDOR--has DOZENS of different licenses. Microsoft often changes its licenses with each new software release too! Furthermore, while open software can sometimes be distributed under "dual licenses" like mySQL, Microsoft has MULTIPLE licenses for many of its products simultaneously. Often these are variants of a standard EULA but there are different details for each one. Win95, Win98, Win98SE, WinMe, WinNT 3.51, NT 4, 2Kpro, 2KServer, XP, 2K3 Server ALL have slightly different EULAs, and in fact the EULAs are different for the SAME EXACT PRODUCT depending on if you are a retail customer, OEM customer or volume licesne customer. You know how that EULA box sometimes pops up during Windows Update? Guess what? Sometimes when you click OK you are letting MS add, change or take away some of the rights you were originally granted when you purchased the software--without compensation! Throw in other vendors that all pull the same EULA crap and you've got an impossible nightmare in terms of licensing.
Given the restrictive nature of typical proprietary EULAs the associated software also contains enforcement mechanisms. I'm sure all but the very casual PC user is painfully familiar with "Product Activation"...
Contrast this with Linux, where the vast majority of open software is licensed under a dozen or less OSI-approved licenses that all share a farily straightforward set of common characteristics. If you think about it, this isn't a disadvantage for Linux at all--it is an ADVANTAGE. In a way, what LSB is trying to do with Linux software distribution is something along the lines of what the FSF and OSI have had some success doing with licensing. We have a baseline for what constitutes open source software licensing that is well established, and if we could have a similar baseline for software distribution so a person can say "this package is an LSB package so I can easily install it on my LSB-compatible OS" then it would be a great step towards catching up to Microsoft's market share.
Something like Autopackage already provides this :-) The reason I was silent about it, is because it's asking for flames most of the time. Autopackage files install in the right location for every system. They also offer tools to compile your software in such way it can actually run on older systems and different distributions.
Autopackage doesn't intent to replace RPM/DEB/TGZ, since these packages are quite good for managing core stuff; a base system [1]. On top of this, their format allows endusers to install the latest version of their favorite end-user/desktop applications: Gaim, Firefox, etc.. I think this separation of core/end-user is a Good Thing(TM) and would improve the landscape a lot.
[1]: it's 2 months since I've released a new version of my OSS project, and none of the major distro's have updated their repositories yet. In the meantime, I'd rather allow endusers to install the latest version through Autopackage or from source. The central-repository model doesn't really scale here for me. ;-)
The best way to accelerate a windows server is by 9.81 m/s2
UNIX: "The reports of my demise have been greatly exadurated"
Lets go on with a list of some the Unixes out there.
FreeBSD OpenBSD HP-UX Solaris AIX (HUGE user base)
Unix is hardly dead. The problem is most people don't know what UNIX acctually IS. GNU stands for GNU is NOT Unix. Unix is a standard that is owned by The Open Group http://www.unix.org/. If your OS meets the standards they set out, you can apply to call your product Unix. Linux, doesn't meet these standards and so can't be called Unix. The BSDs on the other hand DO, hense why we call them BSD Unix and not BSD Linux or some other moniker. To say "Unix is dead" is stupid. Large clustered servers using virtualization is a huge and growing compoent of IT infrastructure. So while you may think you are accessing a Windows NT server, that "Server" could very well being running as a virtualization in a large cluster that is running AIX.
People tend to only see what is in front of their eyes. Users think that Microsoft is the only way to go, or what their offices use exclusivly because that is what they see every day. Their entire backend could be Unix or Linux and they will still say they are a "Windows Shop". As for administrators, how many of THEM consider what OS their routers are running when asked "Do you run only Windows?"
Unix isn't going anywhere. The Open Group is doing just fine, and as Unix is a standard, not a company it means that ANYONE can make a new Unix. Stick that in your pipe and smoke it. Once MS is gone.. you will be able to say "Windows is Dead". Unix, like Rock and Roll, will never die, becuase someone will always insist on digging up the corpse.
Peace