Slashdot Mirror
Unix: Which One to Choose?
I just found this story on Sm@rt Reseller which talks about which Unix (or Linux) they're suggesting to use for various uses (web, applications, etc..) - Its a very long article, and it talks also about the Windows 2000. Worth a read IMHO.
Are you sure? The Resource Kit contains two vis. One's a Win32 port and one's POSIX; and with the POSIX one you have to set up termcaps and things before it'll work.
I use the Win32 one; it's quite good for horrifying people who see me using it.
Personally...I like OS X Server.
I use RedHat 6.1 at home on IA-32 hardware...but at work I run OS X Server...
It's my fav of the bunch...but for now at least it only runs on Apple Hardware. If I could run it on IA-32 stuff...I'd be all over that.
The evaluation has nothing to do with any technical meaning of "best," but merely the notion of "most sold by retailers."
In which situation Debian obviously disappears, regardless of whether it has any valuable qualities or not.
If you're not part of the solution, you're part of the precipitate.
Personally, I don't think Sparc Solaris is good as a graphics workstation. Intel Solaris has a wide range of graphic cards available, but not the software (especially compared with BeOS!).
You can also see my Solaris Intel FAQ at http://sun.drydog.com/faq/
Yes, but on the very first page it says:
(emphasis mine).
The emphasized sentence is certainly a true statement - but so would be "However, the next public edition of the operating system, Linux 2.4, due out this summer, will run on x86-based PC's."
I.e., it's a true statement - but is also true of previous versions of the Linux kernel.
Perhaps you meant to say
which would also be a true statement and would speak of a difference between 2.4 and previous kernel releases.
Are you asserting that none of the stuff under #ifdef __SMP__ is ever compiled in, e.g. because enabling SMP support doesn't cause it to be defined (I don't see any place in the 2.2.14 source where, say, defining CONFIG_SMP causes __SMP__ to be defined, but a kernel tree on a machine here has -D__SMP__ in the Makefile), or are you asserting that the code in question is "experimental code" and, even if __SMP__ is defined, "none of it is ever even called"?
I agree with you on the choice issue - I just wanted to point out that there's a reason Windows is so restricted.
Micros~1 started off selling operating systems to desktop users, and computer hobbyists. They eventually developed their software more and more, till the point where they broke into the server market with NT. (It may have happened before NT to a certain degree, but they probably weren't that popular as a server platform). So here you've got a company that has always built desktop OS's, (arguably "toy" OS's compared to what is required of a server). Their approach has always been to insulate the user from anything approaching a technical decision, and to swath the deficiencies of the system in a pretty GUI.
That type of design and implementation is arguably quite good for desktop users and hobbyists, it plain sucks for server configurations. And once they had dominated the market, what's the point of spreading out and porting to other CPUs? (Especially when they're in bed with Intel)
I think the lack of choice on the wintel side of things has a lot to do with the evolution of windows, whereas UNIX has always tried to run on everything. (remember how one of the original brag points of UNIX was that it was portable?)
Just my $0.02
-- Truth goes out the door when rumor comes innuendo. -- Groucho Marx
I like how the article states that we have to wait until 2.4 to get SMP support. Gee. Funny how I could have swore I am running SMP now.... I think they meant to say "better support in 2.4".
---
They said LINUX has 29% of all public web servers, not Apache.
Just be sure to wear the gold uniform when you beam down -- you know what happens when you wear the red one.
For us, Sm@rt Reseller, we pick the operating systems we cover in large part based on their reseller presence. After all, they're our readers. Hence, I talk about Corel, but not Debian per se.
Steven, Editor at Large, Sm@rt Reseller
In our embedded system coverage where it belongs. FWIW, though I like the embedded Linuxes so far if I had to build an application for an embedded box tomorrow I'd use QNX. Battled tested and tried, well documented, and you get really fast code out of it.
Steven, Editor at Large, Sm@rt Reseller
Darn. And here I thought I'd recommended FreeBSD as the best choice for Web serving.
9 ,2469255,00.html
As for the BSDI/Walnut Creek Merger (not FreeBSD), that happened long after the article was in. For a take on that, see my sister in writing Mary Jo Foley's latest column:
http://www.zdnet.com/zdnn/stories/comment/0,585
Steven
And I have a J-40 8 way SMP machine that cranks out some nice numbers for distributed.net quite well. I may not like AIX, but it does a good, reliable, effecient job.
Bad Mojo
Bad Mojo
"If you can't win by reason, go for volume." -- Calvin
For home use, that doesn't matter - I'm not going to pay $100/month for support - I could buy a lot of O'Reilly books for $100/month. For a medium-sized corporate system, that does matter. Real large corporations will keep a staff of Unix (or NT) wizards around for support, but a 100-person office will need someone to backstop their in-house support. Downloading Debian doesn't provide that.
If I were my company's IT person, I could take that article to my boss to say "here's a good, well-supported, Unix system that will outperform our NT servers, and you'll be able to support it after I'm gone". That would be far more convincing to the bean-counters than a simple exposition of the technical superiority of Unix.
It only took me about an hour to get it working with NT the first time I tried it. It may not exactly be a no-brainer, but it is easier to get working correctly than most of the NFS add-ons I've seen for NT.
As for SMP in general, if I had to have heavy-duty SMP Right Now on Unix. I wouldn't use Linux or Intel. Solaris on SPARC and AIX on PowerPC is where you can really hit the gas with Unix and multiple processors. I haven't tested AIX recently, but Solaris eats NT's, and everyone elses, lunch.
What about IRIX? I have run Solaris on single and dual processor machines, AIX on IBM SP2, and IRIX on Origin200, Octane, PowerChallange, and 64 processor Origin2000. I find IRIX to be superior for most tasks I do (compuational chemistry).
Note: My workstation is a 2x PIII-450 running Linux, and i love it. But when I need to hit dozens of processors at once, I like IRIX.
Just my $0.02
I expect AIX to go sometime around the end of 2001, sooner if Linux can develop the features (HA clustering, journaling FS, etc) that AIX has but Linux doesn't.
Sorry, I just don't see it. The changes Linux would need to support SP2 type systems isn't going to happen within a year or two, whether or not IBM helps.
BSD says steal all you want, please!
It is logically impossible to steal that which is free.
A Government Is a Body of People, Usually Notably Ungoverned
While infrastructure -- the invisible technological foundations -- should be standardized for interoperability's sake, homogenized interfaces are bad all-around. This is because the interfaces that an individual works most efficiently with vary from person to person. A novice will set up TCP/IP on one machine most efficiently with a hand-holding gui, while an expert will set TCP/IP up most efficiently (and possibly on a large number of machines, quickly) with text files and scripts. Same infrastructure, different interfaces.
"Whatever happened to fair use?"
-- Duff-Man
Didn't SCO come out of Xenix? I think after they migrated from 80286 protected mode to 80386 page mode, it just got renamed... I'm not sure on this, but i seem to remember something about that...
---
Play Six Pack Man. I
PC = Politically Correct Personal Computer.
/. and I attribute it to the lack of interesting flame wars. Now, no one wants to go so far as offend anyone. This behavior is called being "politically correct" and makes for boring useless debate and generlally bland discourse.(note: /. is just more PC than it used to be, still not 10000th as much as the nightly news)
There is a much more balanced, bland feel to
Of course, the trolls are still funny.
--
ba-bu-ba-ba-baaa, da-da-dum. Re-boot the ser-ver.
ba-bu-ba-ba-baaa, da-da-dum. Re-boot the ser-ver.
+&x
Yup, but this AC saying that 'enalbing SMP doesn't actually do anything' is also pretty bothersome...
;-) Oh wait... that would include me on some subjects, too 8^)
Who let in the uneducated masses, anyway
"It's tough to be bilingual when you get hit in the head."
>2.3 isn't shipping anywhere. This is sm@rt reseller. Their target audience isn't slashdot.
True, and I addressed that. It is just as 'public' if you can get it from any internet connected computer anywhere in the world for free and without special logins as it is sold on a shelf. I made that point and understand what he meant. That doesn't mean that it is accurate journalism.
"It's tough to be bilingual when you get hit in the head."
That's a pretty odd stance, considering I know of several friends with dual-proc boxes running 2.2, and you can easily check to see the processes running on each CPU (or a quick MT RC5 check, too)...
or check out http://www.phy.duke.edu/brahma/smp-faq/
If you can't get SMP properly working, there are many people who can help, I'm sure...
"It's tough to be bilingual when you get hit in the head."
I meant that Solaris and AIX scale much better than NT does (example RS/6k SP ). I certainly wasn't bashing solaris (I've used solaris x86 and liked it).
Sorry for the confusion.
"It's tough to be bilingual when you get hit in the head."
Yes, but they didn't say that 2.2 does SMP, but it really isn't up to par with AIX/Solaris, which would be the truth. What they essentially said was that it *doesn't* have it now, but it *will* have it soon. They didn't say 4+ procs (after which NT gained minimally, I haven't made new #s on W2k yet), they mentioned 'SMP' which means 2 or more to me and just about everybody else. It's not defensive, it's ust poor journalism. They can rip on anything that I use, as long as they do it truthfully.
"It's tough to be bilingual when you get hit in the head."
>n terms of performance, as Sm@rt Reseller has shown over the last year and a half, Linux actually outworks--on low-end Intel uniprocessor systems--all other network operating systems. It also does well on high-end uniprocessor boxes. But Linux takes a back seat to NT on symmetric-multiprocessing (SMP) systems. However, the next public edition of the operating system, Linux 2.4, due out this summer, will offer SMP support.
That is the paragraph I was quoting. It says that Linux is great on 1 proc, and 2.4 will offer SMP support. Not *improved* SMP support, just that it will (sounds almost like 'will finally') have SMP... That and the "next public edition" thing... I understand what is meant there, but all of he 2.3 series is just as "public" as the 2.2 or 2.4 stuff. Free, anonymous downloads... (but it's not on a shelf at CrimpUSA, so it's not public).
I'm not disagreeing with the general idea, it was just poorly presented.
"It's tough to be bilingual when you get hit in the head."
True, but the article was specifically about x86 Unices.
perl -e 'fork||print for split//,"hahahaha"'
Thats ridiculous. Nobody uses Linux at home. If a Linux running home computer is not offered at CompUSA/BestBuy, etc, Linux cannot have anywhere near 4%. Maybe 4% of business desktops, but even that I doubt. Right now, there are maybe a 10 to 20 thousand people who use Linux as their home desktop machine. Nothing compared to the dozens of millions of home users there are. I think the real number is around .01% or so.
A deep unwavering belief is a sure sign you're missing something...
) More and more of the old school UNIX vendors will stop producing their own UNIX and switch to Linux, especially on or around the time when they release new (incompatible) hardware.
>> Better hope not. I doubt a bunch of hackers can better tailor their software to their system then they can.
I expect AIX to go sometime around the end of 2001, sooner if Linux can develop the features (HA clustering, journaling FS, etc) that AIX has but Linux doesn't. Expect to see more and more IBM UNIX developer-type folks working full-time on the Linux kernel instead of on AIX.
>> Lets hope AIX survives. Do you really want to buy a $30,000 RS/600 and run some crappy unix clone on it, one that originated on i386 no less? (Not a troll comment, but compared to AIX on POWER, Linux on POWER is a crappy UNIX clone.)
SCO will not see 2002. Neither will HP-UX.
>> Better hope HP-UX survives. I don't think Linux runs on PA-RISC, and if it does, its probably a pretty alpha port.
Solaris will hold out the longest, but the Linux tide will overwhelm it by 2005.
>> Maybe, but when Linux can scale to 128 procs (it can't really do 2 that well now)
2) We'll see a rise in the use of the *BSD family for the next couple of years, but as more and more people are exposed to the GPL through Linux, more and more people will _expect_ the GPL on their software. Sometime around 2003 the upwards trend will reverse, and *BSD will slowly slide to obscurity and historical curiosity - made worse as it loses developer mindshare to Linux.
>> Again, better hope BSD doesn't die. What would happen to all the servers that need the unity and rock solidity that BSD brings?
3) I expect that the governments of the US, Canada, Mexico, the UK, and Germany will require the GPL on all software in use in government institutions by 2004.
>> Doubt it. GPL is a license. The government would never do something that stupid, (like they did when they tried to make ADA the official language.) Say I come up with a critical piece of weather software that I'm selling for $50000. No way in hell I'm GPLing that. And you're telling me a bunch of hackers will be able to come up with something that good in a timely manner? When you show me a GPL product that can seriously stand up to its propriotory competition at something, I'll listen, but until then, its just talk.
4) Microsoft will declare bankruptcy by the end of 2005. The stock value will peak in early 2002.
>> Uh no, thats only 2 years away. Linux will be luck if it has 5% of home users by then.
5) We will see one more Windows release post-Win2k. So there will be Win2K, a stopgap WinFoo, and then that's all she wrote.
6) By 2010, the concept of "selling software" will seem as alien as selling air, or sunlight. >> They'd better be selling software, or else people wouldn't be making money.
You speak as if UNIX will take over the world. I for one will never use UNIX as for my media stuff until it can equal what BeOS is today. (But BeOS will be a lot more by then. The OS might die, but the concept of a light media OS won't.) Linux might take over the UNIX market, but believe it or not, there are many people who don't really think UNIX is the OS used by god.
A deep unwavering belief is a sure sign you're missing something...
BSD for server. beos for workstation
A deep unwavering belief is a sure sign you're missing something...
Hey, is anyone where running Solaris 8 on a single proc. Pentium II box? I want to put a UNIX on my desktop, and am thinking of Solaris because I don't really love Linux and its hodgepodge style (personal preference) and I would use FreeBSD but people bitch about its sound code. I do a lot of media stuff, so thats very important. Anyway, I'm used to using BeOS, so I'm a little pampered on the interactive performance front. Anyone using it who could tell me how it performs compared to Linux for 3D/2D graphics, media, and general use. (Clicking on files, word processing, moving around the system with the GUI. etc?) I heard that 7.0 used to be called Slowaris on Intel, how is 8.0?
A deep unwavering belief is a sure sign you're missing something...
(A vi fanatic)
The cake is a pie
The key is to grab all the other libaries and move them over.
And, the process *IS* documented in the freebsd mailing lists. How I did it is I looked at the old 1995 posts on how to get Solaris for X86 binaries running on FreeBSD and applied that methodology.
I was working on formal documantation on how do do it, when Steve Jobs took over and junked x86 Rhapsody. Given few would ever be able to use the work, it was scrapped.
If it was said on slashdot, it MUST be true!
Well, here goes, please hear me out before you start flaming...
1. The author makes a whole lot of claims, but fails to back up many of them with any real data. Netcraft is mentioned a couple of times, but other than that, there is no hard data.
2. The author is clearly biased towards UNIX over M$ products (you can basically discount his comments about SMP support, as everyone has pointed out that it is a bogus argument). Not only that, but nearly the entire article is about how great Linux is. I will be the first to say that Linux is a high quality OS, but there are also alot of things that Windows does well. I believe that his shots against W2K are unjustified and misplaced. Has any company/group released a bug-free/best performing OS on their first version. W2K had many problems to overcome, and considering the amount of code they had to merge in and make work on NT, I am actually surprised at the performance.
By the same token as he says to look at Linux when 2.4 comes out, the same can be said for Windows (or any other OS). Wait until SP2/3 comes out and you'll see marked performance increases. Anyone who builds a sufficiently complex system knows that it takes a while to iron out the wrinkles.
3. I use FreeBSD and Linux, and to say that you need to be a FreeBSD expert to run it as compared to Linux, is just plain wrong. FreeBSD takes no more monkeying around than Linux does to get it to perform well. As a matter of fact, I find the layout of some Linux dists (read RedHat) to be significantly harder to configure than FreeBSD.
The way that he dismissed the BSDs is IMHO shortsighted to say the least (with the exception of his comments on webserving). What he fails to mention is that FreeBSD can run any binary that runs on Linux. And there is a strong push to get a native port of Java (much like there was for Linux).
There are many other things I could say, but I'll leave it at that. The article was an interesting editorial, but it is just that, an editorial. The author clearly has biases and it should be taken as every other editorial is, with a grain of salt. It was nice of the author to promote interoperability, but it should have been done in a less editorial fashion.
My 4c (all little too long for just 2c)
The "Top 10" Reasons to procrastinate:
The "Top 10" Reasons to procrastinate:
10.
Well, it can't. There is the option of enabling SMP support when you do a make xconfig, but enabling it doesn't actually do anything.
Compile times for a Dual PII 450 w/448MB RAM. Resulting kernel was 650K. 'jx' is the number of make's running, s means single CPU.
egcs-2.91.66:2.2.13:j4 261.180u 20.280s 2:27.88 190.3% 0+0k 0+0io 334324pf+0w
gcc-2.7.2.3:2.2.13:j3 232.540u 20.200s 2:11.95 191.5% 0+0k 0+0io 265191pf+0w
gcc-2.7.2.3:2.2.13:s 261.76user 16.23system 4:38.48elapsed 99%CPU
What's really interesting is how people claim that their multi-CPU machine is "so much faster" when they enable SMP. Isn't that what psychiatrists call a "self-serving fantasy?"
Damm! That is some fantasy I have been having. I must have been dreaming when my kernel compile times were cut by more than 50%.
Think before you speak.Okay Rant time! I dont do this often and most often only people who read to the bottom of the stack get to see my rants but im gonna
RANT ON
Okay every time an article compares linux to some other opearting system or operating systems EVERY time I mean every freaking time the author decides to do what good journalists should and not print EVERY SINGLE little detail to made the damn article readable? Just maybe?
What happens? It is a freaking attack on poor Linux.
OH Linux can do that in Kernel X.Y.Z the author should have *KNOWN*. Its just so annoying to always see the group mindset of Everyone is out to knock Linux down. Welcome to Marketing 101!!! Get used to it and play harder dont whine granted Im glad to know the author made a possible oversight it just seems like its happening a little often here.
Someone makes a comment about FreeBSD? Flamebait! Heresey Get it out of here.
Gosh STOP this please. Linux is a Awesome operating system and it does a lot of things very good but that dont mean everyones always gotta look for someone to be knocking Linux down! They are not I did not really feel much of a Marketing ploy in here just an honest attempt at a little information spreading and *gasp* it just didnt happen to priase Linux right.
RANT OFF
One reason I love *NIX is that I appreciate choice.
With Win2K I have one choice, Intel and Microsoft.
With UNIX, and can choose a multitude of *NIX for INTEL, PPC, Sparc, Alpha, etc. Each hardare/software platform offers strengths and weaknesses, especially considering the level of support that is offered.
Choice is a strength to an IT organization. Choose the tool that most closely matches the problem you are attempting to solve. With the Microsoft / Intel choice you are very, very limited. If it doesn't work the way you like, wit and hope that Microsoft or Intel will fix things the way you like someday.
I like to be able to choose.
Magazines would prefer to cover one platform, with one OS. It is much easier to be an expert in one thing.
timbu
If I give you beer and include the recipe for the beer, does that mean that the beer is free as in software or free as in beer.
>Solaris on SPARC
Personally I'd add Solaris on x86 to that list as well. As I said earlier on in this thread, I'm currently running Solaris on a five year old dual P90 machine. Naturally this is running the x86 version of Solaris. The Solaris HCL lists x86 based machines with up to 6 CPUs and has done since Solaris 2.5.1 (which was when I started to use Solaris on x86 SMP machines).
There aren't that many differences between Solaris on x86 and Sparc when it comes down to it. Apart from the lowest level things like device drivers, certain parts of the kernel and certain commands which are very much architecture dependant, Solaris has the same code for both Sparc and x86.
That dual P90 machine I've got is also handling mail, news, DNS, web, ftp, SMB (file and printing), NFS serving, CD burning and the normal work I throw at it (ie HTML generation and some image processing) without too many problems. Whilst I'm thinking of upgrading this old dual P90 with just 64Mb of memory and a single Adaptec HBA is happily handling this load. Even after I add a couple of new machines I'll still be keeping the machine around as it'll have lots of now empty disk space for me to fill up :)
Moo
Ignore the "p2p is theft" trolls, they're just uninformed
"the top five Linux distributors--Caldera, Corel, Red Hat, SuSE and TurboLinux"
it makes me sick to see comments like this. when people talk about "market share" and think a distribution of linux is the most popular just because it has the most SALES.
that is total crap! most people i know DOWNLOAD linux. debian and slackware are very popular, but these writers think they aren't popular because they don't have as many SALES. i am going to go throw up.
A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
Once people, especially in the Open Source community, realize "one OS everywhere" is bad regardless of which OS it is we will make some real progress toward truly great computing systems. Our emphasis should be on interoperability and using the right tool for the right job -- articles like this are extremely valuable tools in this way.
--
: remove whitespace to e-mail me
We may not imagine how our lives could be more frustrating and complex—but Congress can. – Cullen Hightower
I ran Xenix on a 4.77MHz 8086. Yes, it was sllloooowwwwww. I shudder to think I actually used to program on that platform. The scars are still with me to this day!
Steven, Editor at Large, Sm@rt Reseller
My company gets the print version of this magazine. When that article came out, I took into my boss' office, and read the part about all unices moving toward linux compatibility.
/proc for tcp info? How many make sure that their utilities work under linux? How about tar, ls, or bash? Gzip? Make? Patch?
:P
We had a good laugh at such a completely stupid and meaningless statement, and then went back to work.
What the HELL is 'linux compatibility' supposed to mean? Linux hasn't unified anything, and I'm damned sure that the real (read as commercial) unix vendors aren't going out of their way to make their OS's linux compatible in any more than a marketing sense.
For instance, how many commercial unices include term type 'linux' in their termcap database? How many use
All of these programs exist on most unices, and they certainly don't behave in the way the GNU versions do. This is incompatibility, and for someone who write system software across a buttload of unices every day (DYNIX/ptx, solaris, AIX, and linux), these things all suck.
A handful have started porting some utilities or subsystems to linux, but the truth is that unix is based, loosely, on standards, and that linux adheres, loosely, to those standards.
It's a stupid article that makes a lot of false assertions written for people who won't ever know the difference.
Enjoy.
--
blue, bleeding karma from the eyeballs and loving it.
i browse at -1 because they're funnier than you are.
I just fired up a utility from the Win NT resource kit the other day that recommended that I remove the POSIX compliant portion of Windows NT Workstation, as it was a security risk.
That's not the trick. The trick is retarding your computer so that it will run. Race conditions render Xenix unstable past 486/75. I can't laugh too hard, though...it runs on a 286!
--The basis of all love is respect
They sort of insinuate that the 2.2 Linux kernel can't do SMP - the 2.4 kernel is SMP enabled or some garbage like that... I know that Linux SMP isn't the highest performing SMP implementation out there, and that 2.2 doesn't scale to 128 procs, but they said it 'takes a backseat to NT'... last time I check NT SMP above 4 procs wasn't getting you very much... If they said AIX or Solaris, I'd have to agree, or if they at least acknowledged that people *do* run lots of SMP linux boxen. Not very informed.........
"It's tough to be bilingual when you get hit in the head."
From what I was able to scrounge up from various sources, I've determined the following.
A) No-one uses Solaris for multimedia.
B) Solaris 8 is pretty fast, but for single proc. performance it is still slower than FreeBSD.
C) Even the x86 version is really fast for dual procs.
D) The sound system isn't anything special. Its straight OSS.
E) They do have some media apps, especially speech recognition and multimedia authoring.
F) Doesn't really support graphics cards that well. (In SunX)
So I guess Solaris is out. Anyone know if IRIX is coming to x86?
A deep unwavering belief is a sure sign you're missing something...
Getting the moderators blessing on /. is less much about being correct/insightful and more about fitting the adjenda of the moderators.
/., a large number of the moderators are not Pro-OpenSource, but are Pro-Linux. And, if you are not a Pro-Linux, Micro$oft bashing, GPL at all costs poster, you don't get positive moderation.
Think of it this way: At school, you got your best grades when your answer fit what the teacher wanted. And here at
I don't see it as humorous, just hypocritcal. And a very human response.
If it was said on slashdot, it MUST be true!
Linux doesn't inherit the 25 years of UNIX history. Other UNIXes are either SVR chidlren or BSD children. Linux "does its own thing" in a couple of key areas.
The linux IP stack was home grown. Note the plethora of of DOS attacks on linux's stack. These basically did not affect any other UNIX more often than not. Traditionally there have also been scalability/reliability concerns amongst ircd operators for instance, where last I heard freeBSD was the platform of choice.
The VFS/vnode layer in linux is quite different from the 4.4BSD implementation (or the SVR4 one for that matter). For more information, consider reading the websites/papers on GFS (the Global Filesystem at University of Minnesota). To paraphrase, the linux VFS layer is very local-file-system oriented, thus making it tricky to implement distributed/remote file-systems.
UNIX has been a moving target for years and years. Most modern unix deriviates are just that -- branches of the original. Linux is different. It does not contain any code retaining any of the original UNIX licenses. It was developed to work similarly to UNIX, and in many ways it does and sometimes can even work better. But it is not derived from the original codebase, nor from any of the subsequent branches. This has its good and bad points. Two "bad points" are mentioned above. The good points of starting with a clean slate (as far as implementation goes - functionality must be mimicked) are obvious to anyone doing software development.
My opinions are my own, and do not necessarily represent those of my employer.
I'm curious how they decided which five were top. Maybe my personal experience isn't indicitave of the general population, but I've personally met many more people running Debian than TurboLinux. Are they selecting based on the size of the corporation that owns that distro? Number of copies in use (and if so, how do they measure)? File size? It seems to me that most of their choices for "top distro" are the ones that have been in the mainstream news lately (Caldera, Corel) rather than the ones that are most popular or best suited for the jobs they selected.
Visualize whirled peas.
Firstly, the article implied that Linux pre 2.4 cannot do SMP. That is false. Linux has been SMP since 2.0. FreeBSD also does SMP, although it is rather weak. This was never mentioned.
Secondly, I found conflicting bias regarding Linux vs. FreeBSD. While they did recommend FreeBSD for web services, they made BSD sound like it was on its death bed. There was no mention of the BSDIFreeBSD merger, nor was there mention of the whole slew of companies embedding FreeBSD in Thin Servers/Server appliances. They cited a lack of development tools, but lets be honest here. I honestly dont think that Delphi for FreeBSD will help, as FreeBSD is primarily a server platform, not a desktop platform. GCC Anyone?
I don't know. I had a bad taste in my mouth when I read that article. It sounded like it was written by someone who based the article on the advice on others, and not experience.
the positive reporting of Linux is a good thing, but at what cost?
Jailbrekr.
Feed The Need[goatse.cx]
Ones that fly in the face of what we've been told.
A while ago, Unix==Big Iron hardware. Today, Unix can equal a 486/33. The only 'Big Iron' I'm gonna need is the hardware it'll take to make W2K fly.
I also smiled when he said 'Unix will have more software, enjoy it while it lasts.' All in response to the incompatibilities between MS'es own new OS and their own software.
On the flip side, Linux wasn't without growing pains. The lurch from libc5 to glibc wasn't too pretty a while back, but nothing was really stopping you from installing compatibility libraries. Even going 2.0 -> 2.2 was amazingly compatible. In Debian slink, I think there was about a half a dozen packages, out of thousands, that needed upgrading with the kernel.
It's amazing just how badly MS is handling the growth of Unix. You think they'd borrow a page from the Book of Good OS'es.. But instead they go on doing their thing. Scalability? Portability? Unheard of. POSIX seems to be the future of OS'es.. Who'da thunk it?
I've used a lot of UNIX flavors over the past 25 years, and I'd have to say that I recommend Microsoft's XENIX for any task. The NFS support really blazes, and it supports "ksh" or Korn shell. Give it a shot.
------------
a funny comment: 1 karma
an insightful comment: 1 karma
a good old-fashioned flame: priceless
this sig limit is too small to put anything good h
You know the joke about how "I am firm, you are stubborn, and he's a flaming *******"?
Linux/Unix is the same thing. To most people, Linux and Unix are synonymous because they have the same architectural structure, same POSIX libraries, same POSIX tools, etc. Does it run X? Does it run vi? Do you have a command shell somewhere that takes lots of cryptic commands? Then it's Unix.
Even the vast majority of developers will not see a significant difference in the way they develop code for a Linux vs. "Unix" system. A few files are in a different place, a few commands have "odd" flags, but overall it's about as much difference as between Dallas and New York, vs.
New York and New Dehli. With the common use of GNU tools, there's much less perceived difference between Linux and *BSD than Solaris and HP/UX (or AIX!)
But in the same way that many French Canadians can't forgive the British Canadian majority for a defeat hundreds of years ago (going as far as putting "I Remember!" on their the car license plates) we have a few tormented souls who want to make sure that we never, ever, forget the fact that the Linux source tree can't list three pages of "begats" that lead back to King Davi... sorry, back to the original AT&T source.
Is there a real difference? Yes, but the number of people who actually have to worry about them will probably fit into a small room. For the rest of us, the only real difference is a group that's coming across as increasingly bitter that they have finally achieved the Holy Grail of "Unix" Integration only as they band together to fight the Linux intruder... and they *really* hate to be told that this constant "Linux isn't Unix, nah nah nah" harping is exactly the childish mindset that lead to Unix fragmentation in the first place. This is how they snatched defeat from the jaws of victory a decade ago, and many of us really don't want to see a repeat of it.
I know one of the defining characteristics of geeks is great precision in speech, but it's time for everyone to remember the big picture. We're in the game show of life and one side has the MS family (daddy W2K, Mom Win98, insane child WinCE) and the other side has the Unix family (daddy AT&T, brother *BSD and adopted brother Linux), and it only helps MS when the Unix family's first response to a question is to whip out a gun and commit fraticide.
For every complex problem there is an answer that is clear, simple, and wrong. -- H L Mencken
I'm amazed this has got to be the first "battle of the OS's" type article I've read in a long time that not only presents the strengths and weakness's of the various OS's in a balanced and honest fashion but also presents a conclusion at the end of the article that makes logical sense based on the reviews given during the body of the article. I've gotten used to stories (usually involving NT or w2k) that slam a certain OS for performance, stability, and price at every turn but in the end give it their "Editors Choice" because it has prettier widgets than it's competitors. What I really liked best about this story is that not only did the conclusion mesh with the rest of the story but that the author didn't name one OS King of All Unix (on Intel) but gave a nice little chart with very logical recomendations. A nice use the right tool for the right job approach. To add yet another random and disconnected thought to this little ramble of a post, I noticed that the author seemed to put a very high weight on stability, on par with any *nix user I've met, which is nice because I personally value stability above all else for both my workstations and my servers. Ok I'm done with my little pre-coffee post ;->
"Listen: We are here on Earth to fart around. Don't let anybody tell you any different!" - Kurt Vonnegut
After testing extensively the different brands of flamewars available on the market, our SmashDot(tm) test center has decided to extensively analyze the performance of the following flame wars on geeks, nerds and system administrators. The results are as follow:
Of course, your mileage may vary. Honorable mentions include newbie questions, news for nerds that were posted before, the classic "Emacs vs vi" flame burst (unfortunately losing steam these days) and the "This does not belong on Slashdot!" flame war, which may well be a strong contender as long as
Since we certainly want the best, hottest and brightest flame wars for our own site, we'll stick with the tried-and-proven favourite: the Jon Katz flame war.
Jon Katz -- a good roasting guaranteed every time! Get yours today!
Another public-service testing from the SmashDot(tm) team!
The right to offend is far more important than the right not to be offended. (Rowan Atkinson)
The question: What OS do you run.... FreeBSD of course.
:-)
FreeBSD is a OpenSource OS with a licence that promotes the software to be used ANYWHERE, is stable and robust, and has a strong UNIX(tm) tradition.
Oh, and it can run Linux binaries, SCO binaries, Xenix binaries, and I've gotten mine to run Solaris X86 and Rhapsody DR1 binaries, with some tweaking.
(And on a more humorous note: Linux script kiddies come knocking but don't get in. It takes a BSD script kiddie to get in
If it was said on slashdot, it MUST be true!