Servers with a Smile 141
9jack9 writes "Fortune Magazine has this article entitled Servers with a Smile. While they probably get almost as much wrong as they get right, it's still an interesting article, if for no other reason than it's in Fortune, with a readership undoubtedly consisting of people more focused on business than technology. To me the strange thing is that in portions of the computing world Linux and related phenomena (GNU/Linux, OSS, etc) does seem to be "the hottest thing", but in other parts of the computing world it is all but invisible. It reminds me of NT in the early days. There is also a related article Does Software Yearn to be Free?."
Re:I agree (Score:1)
".. Intel-based servers weren't fast enough, and the only operating system written for them, Microsoft Windows, wasn't reliable.
Today Intel machines are often faster than the proprietary systems from the three big vendors, and with Linux, the systems are just as reliable."
There is no word about AMD. It's saving a lot of money too.
After a bit of slashdotting (Score:5, Funny)
Re:After a bit of slashdotting (Score:1)
Journalism has never been a hard science. (Score:4, Interesting)
One draws the impression from this statement that these capitalists are using the word 'free' to describe the cost of Linux, rather than its nature. Hopefully the recent move by Linux's premier distributor to solidify the Linux desktop and put an end to the endless bickering over controls and widgets will do something to dispel this notion of Microsoft's invincibility in the user interface department.
Re:Journalism has never been a hard science. (Score:1)
Re:Journalism has never been a hard science. (Score:5, Insightful)
Lack of offical support,
Right. RedHat, SuSE, Mandrake and others will be happy to take your money for them to supply 24X7 support for thier distro. Oh, and support from newsgroups and Linux specific message boards are usually just as fast and thourough as any tech support dept. I've ever called.
Swing and a miss - Strike 1.
culpability,
Ha! Have you read any EULA included with "expensive software built by huge companies staffed by thousands of tech support folks"? EULAs exist to make sure you don't sue that same company if the software you buy from them blows your business to smithereens. "Sorry, Mr. Customer. Guess you shouldn't of bought the line from the Marketing Dept. Or (heh) our software package..."
Swung at a ball in the dirt, Strike 2!!
insuring[sic] that the software company will be there to lean on when things go wrong.
No, not lean on. Call and give more $$$ to so they can fix the problem with thier software, or tell you how to use thier crappy UI. They sold you a defective product, and you have to pay more for a correct product. IMHO, they're leaning on you, as in "Vinnie, I needs youse to lean on dis stooge 'til he gives us da dough..."
Strike 3!!! Batter out!
Software Libre _is_ good - read the article.
"It's a sea change," says Bridget O'Connor, a top technology executive at Lehman Brothers. "Now I can play all the vendors off against each other to get the price I want. I never had that negotiating power when all my machines came from Sun."
Power in the customers hands, not the vendors. Competition based on technical merit and value. That's what Free means, bud - freedom to choose. Spin that to the execs next time they ask.
Soko
Hello Free Software Zealot! (Score:2, Insightful)
Swing and a miss - Strike 1."
Its this kind of thinking that unfortunately keeps the adoption rate of Linux/Free Software artificially down.
Corporations, ones bigger then your local mom and pop that is, don't want to obtain their tech support from a bunch of geeks in a web forum or on a mailing list. They want teams of experienced consultants and tech support specialists whom they can call upon at any hour and not recieve answers like this "Stop being so pushy! We do this for free you know! WE DON'T OWE YOU ANYTHING!"
Re:Hello Free Software Zealot! (Score:5, Insightful)
Anyone who has ever done any real system/database/app support knows that the web and message boards are your best source of information. Even the best tech support organizations suck at the first tier. The best I've ever seen was Oracle, and it depends on the phase of the moon as to whether or not you get anybody who has even the slightest clue. Sun's support is the worst, until you get to the third tier. Microsoft is better, but not much. With them you have at least a passable chance of getting someone competent at tier 2.
Besides the parent post did not say that forums and message boards are the only way to go, he simply said they are often faster, and they are. He said you can purchase support for Linux and you can.
Just because he's a Free Software Zealot... (Score:2)
So, if you are getting your support for free by all means feel free to bitch somewhere else.
This is a non-issue (Score:2)
Frankly, professionally supporting an open operating system like Linux is easier than a closed one. As a Linux guru, when push comes to shove I can get my hands dirty and dig into the source myself to solve the problem. That simply isn't an option for 99% of the Microsoft support firms.
Re:This is a non-issue (Score:1)
There are some people who doubt that source code is ever searched through by people who want to find problems but aren't working on the project that produced the code. Believe it or not, this actually happens. As an example, I had a problem with the GNU MP library that I couldn't figure out. I looked through the source, and found what was going wrong. It involved division by zero, and it was my fault.
This is just an example. It does help illustrate, however, that having source code is useful to actual users, although perhaps not your normal "where's the any key?" user.
Re:This is a non-issue (Score:1)
cheers.
Re:This is a non-issue (Score:1)
Ultimately that's the rub. Open-source software will always have a hard time getting out of the back room until Joe and Jane User are comfortable and experienced with the UI. I have a Linux box or three in our server room and would be happy to have one on my desk too, but few of my users can even spell Linux. Their comfort level is with Windows and Windows applications and despite what some of the Microsoft bashers around here would crow the reality is that our workstations are quite stable and the work gets done. At the end of that day that's all the bosses with the big offices care about - did the work get done.
Me? I like the whole idea of open source software and enjoy the moments when I get to pull up a chair in front of a Linux box and play. I'm always interested to see the latest tool for this and that even if it's something I know I'm only going to tinker with for a few days then uninstall.
My users just want to do their jobs though; and their jobs have nothing to do with compiling the OS. Windows provides them with a level of comfort.
-CoachS-
Not in the same ballpark (Score:1)
Re:Journalism has never been a hard science. (Score:3, Informative)
One draws the impression from this statement that these capitalists are using the word 'free' to describe the cost of Linux, rather than its nature.
You have to remember, of course, that to much of the business world, it is the bottom line that counts. Most could care less about the ethical grounds behind free software.
Re:It's a time-honored capitalist tradition: (Score:2)
1. Yes, that is a quite common way for businesses to make money.
2. Yes, that is a quite common way for businesses to make money.
3. Yes, it does work in our (and by our, I mean consumers) best interests. What free software does is turn the whole thing topsy-turvy. Businesses can (a) cut costs and improve their bottom line by getting software gratis, and (b) not become morally bankrupt in the process.
I'm just pointing out the fact that many businesses really only care about (a). The fact that (b) occurs is a bonus for the rest of the community.
Re:Journalism has never been a hard science. (Score:1)
Re:Journalism has never been a hard science. (Score:3, Insightful)
But what about business desktops? In those situations, you often have an IT department to manage things, so that's not a problem. Employees shouldn't be playing a lot of games. The required software often amounts to email and a word processor, which open software has fairly well covered these days. Free is a huge plus, saving on up-front costs as well as the headaches of managing licenses. Seems to me there's a great case to be made for the business desktop in many situations (not all, some businesses [or portions thereof] are more tied to Windows-only applications). And if Linux can make inroads there, it'll increase mindshare and familiarity, speeding adoption with consumers....
Re:Journalism has never been a hard science. (Score:4, Informative)
Thanks to Codeweavers you can have Office, Notes, Quicken, IE, and many other windows apps running flawlessly in Linux. Thanks to Transgaming you can play many windows only games in Linux, such as Warcraft III, The Sims and many many more. Now, unfortunately, Loki is gone, but I'm glad I got my Quake III, Railroad Tycoon, and SimCity 3000 before they went under.
Last but not least, UT2003 will have a Linux version at the same release as the windows version. Return to Castle Wolfenstein runs in Linux and so will Neverwinter Nights, if Bioware ever gets it out the door anytime soon. If it wasn't for Battlefield 1942, I'd never boot into windows.
Plus there are many great apps on Linux already like Bluefish, GIMP, Grip, XMMS, Evolution, Ximian Connector, and Red Carpet to name a few. I think this is no longer a real issue, but that's just me.
Re:Journalism has never been a hard science. (Score:2, Insightful)
Re:Journalism has never been a hard science. (Score:2)
Re:Journalism has never been a hard science. (Score:2)
Re:Journalism has never been a hard science. (Score:1)
actually, what recently got me thinking about this is working in an office where up until last year, they only used typewriters. They actually are quite well organized with paper files, phones, and typewriters.
i fear for them, entering into the digital world.
Re:Journalism has never been a hard science. (Score:2)
For example:
Imagine a group within the company that uses email, IM, a ticketing system, a browser, very little office apps, and the rest internal n-tier applications.
Now imagine that the company decided to replace their PC's with SunRays, only they didn't use any Solaris versions of the applications they need, only windows apps. Instead of using Mozilla and its mail or some other native mail client, everyone used Outlook Express and IE (mostly because clueless developers couldn't make their internal web apps multi-browser). Instead of using StarOffice they use MS Office. Instead of running a Solaris version of Vantive, they use the windows version. Now imagine that they are running this all through Citrix. So in effect they are running a thin-OS from a remote server, running a thin-os from a remote server.
Now that sounds like someone read that article and completely missed the point. And sadly, I've seen this in two major enterprises.
An even better quote (Score:3, Insightful)
Does anyone here take seriously anything said by a publication that would actually print something like this? "Over-granolaed"? Is that even a word? Is that even a valid concept? How does one become "overgranolaed"? If you eat a lot of red meat, do you automatically become a good businessman? If this is the target audience for Linux adoption, it's obvious why it's been less than successful. These characters are congenital idiots.
Re:An even better quote (Score:2)
Granola has long since been discredited as a health food; anti-oxidant fruits like purple grapes, green algae, and tofu are the correct breakfast food for berkeley commune linux nuts these days.
Re:An even better quote (Score:1)
Re:An even better quote (Score:1)
The best quote: (Score:3, Insightful)
What, Fortune? Sure. The fact that they don't see the world the way you do is fine. Indeed, irrelevant. Just like the fact that we see Free (Speech) and they see Free (Beer). It's nothing to be shocked at. They're doing it in terms their readership will understand, just like Slashdot has spelling errors (at least, I _hope_ that's why) and the inevitable Microsloth slam (whoops, did I just contribute?).
As for the comment on the culture? If the shoe fits... I keep thinking back to the old Dilbert strip, with the bald, bearded Unix guru ("Here's a nickel, kid, buy yourself a real operating system"). Like it or not, Berkeley is viewed as a hippie commune, a view that many (in Berkeley) subscribe to and act in accordance with.
And finally - it's largely irrelevant WHY they choose Linux, isn't it? After all, even if they're just buying it because it's "free beer", (a) the techs that will be running it know the difference (b) they're happy, (c) the techs are happy, (d) Microsoft is unhappy, and (e) WE'RE STILL RUNNING LINUX.
Re:The best quote: (Score:2)
As for your proficiency at setting up straw men, I would strongly suggest that in the future you consider responding to what I actually said, not what you find it easy to refute. Consider the old joke about the drunk and the lamp post: A guy walks up to a drunk crawling around on the ground in the middle of the night apparently looking for something. Guy asks the drunk what he's looking for. "Dropped my keys, Mister," says the drunk. "Did you drop them under the lamp post?" the guy responds. "No," says the drunk, "I dropped them over there," and points towards the wall of a building. "Then why are you looking under the lamp post?" "Light's better over here," answers the drunk.
Quote... (Score:5, Funny)
Heh, 3,800 lines developed over a decade. More like a month of work.
Re:Quote... (Score:2)
Re:Quote... (Score:2)
Re:Quote... (Score:1)
Anyway, most of the big companies don't seem to crank out much more than 10 lines of code per programmer/day.
Most disturbing quote (Score:3, Interesting)
When we had Solaris machines, we needed professional Unix system administrators. Now that we have Linux instead, any geek with an undergrad degree can do the job for 80% of the pay!
Scary. Keep in mind he's talking about the sysadmins for production e-commerce server farms, and explains the cheaper labor by saying "Linux programmers are more plentiful and cheaper to hire than ever."
Re:Most disturbing quote (Score:5, Insightful)
For a time. Then as demand for Linux geeks increases the salary will increase as well.
Re:Most disturbing quote (Score:5, Insightful)
That's not the point. The misconception held by the pointy-haired boss (and reinforced by this article) is that junior programmers can do the job of senior system administrators if only you use Linux.
The end result of this will be high-profile Linux failures; not because of flaws in Linux, but because of bad decisions driven by drivel like this piece.
Re:Most disturbing quote (Score:2)
Failure or success, high or low profile, doesn't have much effect on these persons' decisions. Are we talking Code Red scale disasters? Remember, membership in the PHB group is a fuzzy thing; last year's script kiddie antics has only nipped at the tail edge of the PHB bell-curve.
What does affect these persons' decisions is the herd mentality. Once the herd starts moving towards Linux, demand for Linux sysadmins will rise, along with salaries.
And, in any case, while hiring of competent sys admins is important, in all probability junior Linux geeks are probably better than cargo-cult Windows drones (of course, expert senior Windows sysadmins would be better than junior Linux geeks).
Not even close (Score:2)
Re:Most disturbing quote (Score:1)
Re:Most disturbing quote (Score:1)
Re:Most disturbing quote (Score:2)
Nah, I'm not buying that one at all.... "Bob, the guy who knows computer stuff" is not the one developing those exploits and coding worms in his spare time.
The only reason you see increases in exploits, virii, and worms for a particular platform/OS is due to a rise in its popularity. In general, people writing destructive code do it for 2 reasons. #1 is to gain some notoriety. #2 is to do damage to a particular business (ex-employer?). In both cases, it only makes sense to aim at the platform/OS in widest use.
Sure, the news media may pick up on a virus or worm that goes around for a long time, due to people who don't do the security patches. Still, patches only come about *after* the fact, in most cases. Nobody writes 100% perfect code, and the more people have reason to try breaking a particular piece of code - the more times they'll succeed.
they would been killed faster (Score:1)
If real professionals had been hired, then I think that all the worms that are still circulating would not have gotten as far in the first place.
Re:they would been killed faster (Score:3, Insightful)
Still, I'm not so sure "better trained I.T. staff" would solve the worm circulation problems. For starters, many people run small web servers from their home, and they're not "professionals" at all. Every copy of NT includes the IIS server, installable with a single click of a checkbox during setup. How many people thought "Hey, I'd like to host my own web site. I think I'll try that!" and became worm distributors, without ever realizing it?
Also, I've worked for companies where it's like pulling teeth to get the I.T. manager to allow you to do upgrades and patches. They're so afraid of service packs that break expensive/complex applications loaded on the servers, they refuse to patch things until months go by and they know for sure it's "safe". After all, if you have a million dollar in-house app developed using an obscure manufacturing control development package, you're much more concerned that the "security fix" might GPF the thing and stop your factory from producing product. The idea that "it keeps your system from spreading the XYZ worm" seems pretty insignificant by comparison.
Re:they would been killed faster (Score:2)
Also, I've worked for companies where it's like pulling teeth to get the I.T. manager to allow you to do upgrades and patches. They're so afraid of service packs that break expensive/complex applications loaded on the servers, they refuse to patch things until months go by and they know for sure it's "safe".
True, and if there isnt active development going on for those in-house complex apps the company is going to be in a world of hurt sooner or later - when they face the choice between freezing all upgrades & fixes on the core servers or paying through the nose for the app to be reengineered to cope with changes in the IT landscape that naturally occur over the years.
Where I work, the in-house apps have a defined release cycle for regular updates - after all, patches aint just for the OS :) This means that when I have a patch for any OS component I put it on the integration test & QA setups first, if nothing breaks there it goes on development and then later into production concurrent with the first release of the app that was developed and run through an entire test & QA cycle running on the patched OS. Sure, theres some delays involved and I occasionally break this rule if theres a serious security hole to fix but these all get argued on their merits and I have not had a problem getting a management signoff for critical stuff. In general I find my systems are at least as current on their patches as most of my colleagues at other companies and often much better. The only thing that makes this difference is that the apps I must not break have developers working on them too and they are savvy enough that if an OS patch breaks it they dont care whether its their code or the patch that has the "bug", they just know they have to work around it before the next release.
Re:Most disturbing quote (Score:2)
You made me squirt cofee all over my screen, ill sue you!!
Here in sweden we have 90% Bob's. They all say that education and skills says nothing. I know plenty of sysadmins that scripts their servers to restart every night instead of trying to find the memory leaks that makes them die. I even saw one put in more memory so that he didnt have to reboot it so often. I dont think this is Microsofts fault. Someone emplyed theese goons and they are responsible.
Re:Most disturbing quote (Score:3, Insightful)
Additionally, when linux expertise was scarce, you might have been forced to hire someone who was overqualified, which, strictly speaking, is a waste of money. Now that the linux labor pool is larger, companies are more likely to find an employee with just the right qualifications without having to pay someone who is overqualified just because he/she is the only person with expertise.
The point in the article is about efficiency.
Re:Most disturbing quote (Score:2)
It could just as easily mean that overtime is down. Many places pay their Tier 1 administrators hourly.
Slashdot adopts new GNU standard naming (Score:1)
Re:Slashdot adopts new GNU standard naming (Score:2)
Canibalizing (Score:3, Interesting)
Re:Canibalizing (Score:2)
Right on the money. The example in the article is Amazon.com's migration from Solaris to GNU/Linux. Microsoft doesn't view this as a loss. If anything, they're delighted that Amazon.com has moved from Sun hardware to something that can run WinXP -- makes the business case that much easier for migration to an all-MS solution.
Re:Canibalizing (Score:3, Interesting)
Who in their right mind cares? Ofcourse the *nux guys pick it up first. They are "home" at first boot. I would be more surprised if the MCSE's would be the first to pick linux up when they dont know anything about unix. Its easier to go from someting you know to something almost the same.
After the unix guys have changed is when the Microsoft camps will start to think about it seriously.
Re:Canibalizing (Score:2)
UNIX was ripe for the picking. Microsoft has been going to businesses for years and telling them that they could replace their expensive UNIX servers with Intel based servers, and their message has been pretty well received. I know lots of places that were primarily UNIX shops that have been switching to Windows because Windows was "good enough" at a lower price.
Now, along comes Linux, which also runs on inexpensive commodity x86 hardware. Even worse, it is basically a drop in replacement for UNIX. Migrating from UNIX to Linux is easy, and your staff needs very little retraining. All of a sudden the UNIX to Windows migrations have stopped, and instead they have become a flood of UNIX to Linux migrations. While this is devastating to Sun, it is also bad news for Microsoft. All of a sudden Windows is having a much harder time getting server sales. UNIX shops, generally speaking, are looking at Linux first, and they are liking what they see. Once you have shifted to Linux there is very little chance that you will ever use Windows on the server. Microsoft's ace-in-the-hole has always been the fact that they ran on commodity x86 hardware, but Linux has stolen this advantage and is "free" as well.
Without the built-in growth market Microsoft has to find other ways to keep their revenues up. Their new licensing plan is just another example of how Linux is affecting their game plan. The harder Microsoft squeezes their customers the better Linux looks.
Re:Canibalizing (Score:2)
It's still early in the conversion cycle. IDC says servers running Linux represent only 5% of the servers in operation, compared with 27% for Windows and 43% for Unix, the umbrella term for the proprietary operating systems largely in use by IBM, HP, and Sun. And there are still big business applications like Siebel's that don't yet support Linux. But IDC predicts that by 2006, Linux and Windows will have all but replaced Unix as the dominant server operating systems. It forecasts that 26% of servers in operation will be running Linux, 56% will be running Windows, and only 12% will be running Unix.
Yes im sure MS is shaking in their boots about the prospect of more than doubling their server market share in a 3 year period according to these projections (which yes, could certainly be flawed but nonetheless it sure looks like good news for them to me)
Of course that projection of Linux more than quintipling their share ain't to shabby either
Re:Canibalizing (Score:2)
Microsoft has seen enough market shifts to know what happens if you allow a lower priced competitor to set up shop in your market niche. They have competed against Lotus 1-2-3, WordPerfect, Novell, and now commercial UNIX by being "good enough" and less expensive. If Linux is allowed to grow its marketshare then Microsoft is going to have serious long-term problems, and they know it. With every new version of their software their customers are going to compare the cost of upgrading against switching to Linux. And Linux is going to be more feature-competitive with every release. When I first started using Linux in 1995 it had almost no applications that weren't development tools available for it. Now there are Free Software equivalents for almost all server side software (and lots of commercial server software as well), and Linux is even a credible desktop.
Re: Canibalizing (Score:2)
In every unix-to-linux conversion, a potential sale for Microsoft is lost.
In each case, the "linux mindshare" is increased. Linux is legitimized by its wider deployment, and word-of-mouth communication. Linux's initial success in some parts of an organization pave the way for wider adoption, which only means more lost sales opportunities.
And worst of all for Microsoft, linux capturing a very substantial share of the market means that (almost) nobody will tollerate lack of operability between Microsoft's client/desktop software and linux servers. Had linux not come along (or had MS figured out how to write a true 32 bit protected memory operating system many years earlier), Microsoft would have been able to capture most of the server market with NT (as people were predicting in the late 90s)... and then they'd be in a strong position to dictate communication protocols that marginalize the small remaining fraction of installed competitive software.
It's not GNU/Linux (Score:1, Offtopic)
RMS is partly right (Score:2)
Software Yearns To Be Free! (Score:5, Funny)
Re:Software Yearns To Be Free! (Score:2)
"Lemme tell you a story. I knew a guy, a lot like you. Didn't have his rent check. So I kicked his ass out on the street, shoveled all his things into a furnace. Parents came by looking for him, told them he ran off with some guy. Right now, he's thinking about drinking his own pee."
"But you don't understand, I'm freely exchanging ideas in a competitive environment."
"Ain't that something? I'm doing a little EX-changing of my own! I'm exchanging your broke ass for someone that can pay the rent. Me and you, we're like kindred spirits!"
Re:Software Yearns To Be Free! (Score:1)
Re:Software Yearns To Be Free! (Score:1)
Re:Software Yearns To Be Free! (Score:1)
Re:Software Yearns To Be Free! (Score:1)
Re:Software Yearns To Be Free! (Score:1)
So I guess Ellison's message is.... (Score:2, Funny)
Oracle + Postgres (Score:1)
Ellison ought to consider an Oracle-supported distro of Postgres. Put in the work to make porting between Oracle and Postgres trivial and then position Postgres as the low-end Oracle offering.
If this pulls more MSSQL deployments into the Oracle camp than the number of Oracle deployments it pushes to Postgres, it could be a net win for Oracle and OSS.
And if it enables Oracle's own enterprise apps to scale down in price to appeal to a broader audience, it could be a big win for Oracle.
GNU/Linux a related phenomena to Linux ? (Score:1)
So GNU/Linux is a related phenomena to Linux ? Damn, I have to read that FAQ [slashdot.org] all over again !
Amazon's review (Score:4, Insightful)
At my college (Virginia Tech), I suggested for 2 years while I was there that we switch to Linux as the primary server software for the engineering department. I always got a 'No.' response...why? "Not enough people know how to administrate it, what happens when you leave?" I hated this reason, mostly because *I* knew that there were other people out there that could admin a freakin Linux box, but they didn't. This response from Amazon helps out more than all plugging us geeks can do.
--trb
Re:Amazon's review (Score:1)
Had the same thing here. When I came in, the intranet web server was IRIX. They wanted to move away because of the HW support contracts being too expensive. I suggested going to an Intel platform (we had extra machines around that had plenty of power) running Linux. Same excuse, if you leave, who's going to support it? Well in a systems group of about 20 people, 3 others ran Linux servers at home.
And when you're only talking about FTP access for a few web developers and a simple Apache install (our Intranet doesn't use any scripting, all static HTML) its not difficult to maintain. Instead, we spent several hundreds on a Windows NT license, and just recently another $1500 on a 2000 Server license. Meanwhile, there have been several applications that I could have built in a couple of days using PHP/Perl and a database like our room booking software (over $1000) and a telephone directory (still using an Excel spreadsheet).
Awww man (Score:1, Offtopic)
Change is inevitable (Score:2, Insightful)
Clearly the author never upgraded a company from Win98 to XP.
Ballmer's "Linux is cancer" pitch misfired! Great! (Score:3, Interesting)
"A year ago Microsoft CEO Steve Ballmer called Linux a 'cancer' that would cause the death of intellectual property as we know it. Microsofties were encouraged to make that saber-rattling pitch on their corporate sales calls. But that just made IT executives angry. Peter Houston, Microsoft's head of industry strategies, says, 'became clear that we were being seen as having a polarizing and myopic view.'"
All right! Maybe there's hope.
(Now let's just hope that the same IT executives are savvy enough to see what Palladium can do to them--and don't see it just as a way of stopping college students from trading music).
Quote at the end of the article (Score:2)
The same companies that balkanized Unix are now jumping on the Linux bandwagon. Some may try to introduce proprietary extensions. I don't think that's going to work, mostly because of the GPL, but also because any company that tried would face a backlash from developers that would impair the company's ability to get features added to the standard kernel.
The question of whether the suits own Linux or not boils down to whether or not the suits can get the features they want, at the expense of everyone else. Case in point: DRM. Suits love it. Linux users hate it. Linus may include it in the kernel, but Linux will be dead the day he includes it *without* a compiler switch to turn it off.
Re:Quote at the end of the article (Score:2)
Nobody in their right mind will be able to sell a proprietized Linux because its "Optimized for their hardware". Anyone with half a brain will immediately throw the word UNIX back in their face and the calamity caused by the fragmentation of UNIX.
hmmm, that sounds a lot like a never. OK, they shouldn't be able to pull it off, but its amazing what some people can sell.
either way (Score:1)
Workstation companies on way down and out (Score:1)
More and more it is looking like an Intel world. AMD mearly adds some price competition. I worry about where we'll see the competition from inovative hardware architectures?
Re:Workstation companies on way down and out (Score:1)
You mean PPC not MIPS, right?
That last comment... (Score:2)
What I really hated was that last comment made about "Suits are taking over, and are here to stay."
Linus is possibly the closest thing to a 'suit' that will have influence over the kernel. "Suits" are left with the GNU system to "Take Over." I will laugh my ass off at any businessman that tries to remove the freedom of GNU from GNU. RMS, and others will not stand for such stupidity, and fortunately the GPL protects the system from pointy - haired idiots.
There were references to changes made to the Kernel and OS that would make it run faster only on their machines. This is to be expected! Ever heard of Gentoo? My version of Gentoo runs fastest on my machine, an Athlon, and not on an Intel Pentium 4. What point is this guy trying to make? That the version of GNU/Linux that IBM/Sun/whateveh comes up with will be proprietary? Is he trying to say that the source code for this version will not be released? Is he trying to say that this version of GNU/Linux will be the only version?
Bad businessman, Bad! You are meddling in things that you cannot possibly understand.
Anyone renember SCO? (Score:3)
solaris is not dead (Score:1)
Solaris is rock solid, and will continue because of the proprietary systems. Damn, even their x86 Solaris is Rock solid. Oracle needs rock solid for mission critical. This is why Sun owns the mid to high end market......
Re:solaris is not dead (Score:1)
I agree. Linux makes a very nice "we-also-run" for Solaris/Oracle shops. Develop with Linux on the desktop, put Linux on continuous/automated build servers, use Linux for smoke and unit testing. Then deploy to test and production Solaris servers running Oracle.
This combination is effective because it allows a development team to explore, and even break stuff, on low-cost but highly capable systems. These systems don't need the "grooming-the-queen" approach of mainframe administration, or the "you-can't-handle-the-root" management of Solaris system administration. Most of the well paid Solaris sysadmins I've encountered have proven to be well worth their pay. That makes it even more important that we concentrate their effort on optimizing and maintaining production systems (storage management, fail-over, system security, OS patches... etc.).
Let the developers worry about bouncing an application server or web server to test a change. Let the developers worry about scripting and scheduling builds. Let the developers figure out how to configure the application and try several different ways of doing it. Let the developers do all of this on Linux.
Sorry for the rant... I'm out of breath now.
The opinions expressed in this post are my own, and not necessarily the opinions of my employer.
Re:solaris is not dead (Score:1)
backhanded compliment (Score:2, Interesting)
A Linux conversion is not for the faint-hearted. Companies must install new hardware and software on a large scale and prepare for bugs and hardware incompatibilities galore
What a crock, so as soon as you put GNU/Linux on a system it might explode?
This article is full of misinformation, the worst being the focus on "free"! How many times have we read here on Slashdot that PHB's don't want free (gratis)? You can't give anything away, you have to charge for it. Sun showed us that with Star Office. Am I happy Fortune did a write up on GNU/Linux? Sure, but I think articles like this do more injustice to us than anything. At least thanks to IBM (per the article) GNU/Linux didn't die a death of neglect!
By the way, is granolaed a word?
Re:backhanded compliment (Score:1)
Is this a contradiction? (Score:2, Interesting)
Then later the article reads, "Levanon says the hardware and software savings are nice, but what has really made the conversion compelling is that his labor costs are down 10% to 20%."
If there are so many bugs and incompatibilities then why does administration cost 10% 20% less?
Re:Is this a contradiction? (Score:1)
I think it makes the statement more impressive. The Linux hill is easier to camp on even when it rains.
Tying Linux to proprietary hardware? (Score:4, Insightful)
The article says a few stupid things (and a bunch of interesting things), but I think this part is just silly:
Companies like Sun, HP, and IBM could derail Linux if they co-opt it--that is, modify it enough so that their versions run well only on their own hardware systems. That is exactly what Sun executives plan to do. They believe the profit motive will prove too strong for IBM, HP, or anyone else to resist making such a play either. "The reality is that profit-making companies like to get paid," Sun's Schwartz says.
How can you possibly tie Linux to a proprietary platform? Sure, you can make weird hardware that requires specialized drivers and whatnot to run, but if you want the large volume of Linux software in the world to run on your machine you have to keep the kernel APIs the same, which means that your customers still have the option of swapping your hardware and customized Linux for someone else's at will. Unless, of course, your hardware is just vastly better, in which case you have no need to try to lock people in, 'cause your boxes will win on their own merits.
If you choose to change the kernel API, or wrap it in proprietary utilities and libraries that provide a substantially different application interface, which you then use to implement your applications, well, you've just managed to tie your applications to your version of Linux which is tied to your hardware, but you have to convince *other* application developers to do the same to have achieved any significant lock-in. Either that or you have to create some truly killer software that is so good people will buy your platform just to run it.
And even if you achieve that, as soon as someone else ports your kernel modifications (which you had to publish, Yay GPL!) to your competitors platforms, or as soon as someone else reimplements your killer app on a standard kernel/OS, your lock-in quietly evaporates. You can also try to lock people in with contracts, licenses and marketing approaches, but that pisses customers off and, if you're very successful, it also pisses off the Department of Justice. Not to mention that all of this customization of Linux will cost you a crapload of money both to do and then to maintain.
Finally, as for Schwartz's assertion that Sun's competitors will opt to find ways to lock their customers in, I think he's on crack. Not only do the customers not want to be locked in (they've been there, done that and got the scars to prove it), many of the vendors have also learned that strategy doesn't work as well in the long term. IBM, in particular, found that you can make huge amounts of money for a while that way, but eventually it comes back to bite you.
Nope. Open Source operating systems make sense precisely because they allow hardware and applications to compete separately. Sun can't change that no matter how hard they wish they could. And IBM, at least, likes that fact, because IBM thinks the key to selling hardware is services, and they're much better at services than Sun, HP or Dell.
It reminds me of NT in the early days. (Score:2)
oh wait, I'm old (over thirty), nevermind.
Scary Thought (Score:1)
And you thought they were being nice...
Date (Score:1)
IBM & MCA (Score:2, Insightful)
Here we are now, and IBM is the biggest backer of open source. Seems like they have a corporate memory, at least.
Software is going through the same growing pains that hardware has already gone through. And the proprietary hardware vendors are choking because open standards have caused huge price drops along with huge performance gains.
I'll be damned if I know where it will all end up, but the landscape 5 years from now will be a lot different than it is today.
Last Post! (Score:1)
forest, hunting bear. They'd rented a cabin, and, when they got there, took
their backpacks off and put them inside. At which point the salesman turned
to his friend, and said, "You unpack while I go and find us a bear."
Puzzled, the analyst finished unpacking and then went and sat down
on the porch. Soon he could hear rustling noises in the forest. The noises
got nearer -- and louder -- and suddenly there was the salesman, running like
hell across the clearing toward the cabin, pursued by one of the largest and
most ferocious grizzly bears the analyst had ever seen.
"Open the door!", screamed the salesman.
The analyst whipped open the door, and the salesman ran to the door,
suddenly stopped, and stepped aside. The bear, unable to stop, continued
through the door and into the cabin. The salesman slammed the door closed
and grinned at his friend. "Got him!", he exclaimed, "now, you skin this
one and I'll go rustle us up another!"
- this post brought to you by the Automated Last Post Generator...
Re: Oracle (Score:2)
Namely, Oracle would like to see more installations of their database on stable platforms. Right now, you've got quite a few small to medium-size businesses running mission-critical Oracle databases on NT servers. Most of these installations happened because of budget limitations. (EG. We only have a few I.T. guys on staff and just paid out *big* money for Oracle. We can't afford to purchase expensive Unix servers to run this stuff on, too. Let's just recycle one of our existing NT boxes (or get a new one fairly cheap), and that way we'll be using the platform our staff is already trained on. That should keep costs to a minimum to implement this thing.)
Unfortunately, when NT blue-screens every so often, people simply think "the stupid Oracle database crashed again!" and it makes Oracle look bad.
Linux can be the more stable alternative for these people that can't/won't spend any more on their hardware or operating systems.
Re: Oracle (Score:2)
Most of these installations happened because of budget limitations. (EG. We only have a few I.T. guys on staff and just paid out *big* money for Oracle. We can't afford to purchase expensive Unix servers to run this stuff on, too
Considering how much Oracle costs - I don't think the price of the operating system should even be a factor.
Jason.
Re: Oracle (Score:2)
1) I wish I still had my rant about how Oracle should have never been ported to NT. It belongs on Unix -- the NT port is pretty damn crappy. If you're running NT, you should be using SQL Server (ideally) or DB2 (actually the FIRST database to get Microsoft's certification for Windows 2000 -- yes, even before their own).
2) If you're in a budget crunch, you shouldn't be buying Oracle anyway. Oracle is very powerful, indeed -- in fact, I designed a very complex application using Oracle this past summer. But what I found out is that the company I worked for spent somewhere in the neighbor hood of 20-40 million dollars on Oracle licenses (this probably included the uber Sun servers it runs on). (I don't remember if this is the exact amount -- I do remember that I almost shit myself when I saw the numbers.)
If you are running Unix, Sybase and DB2 are viable alternatives. Microsoft did a study once and it showed that Oracle on Unix cost about 10 times more per transaction than SQL Server on NT.
If NT blue screens, something is horribly wrong with your servers. I've seen NT boxes with uptimes in the hundreds of days.
But then again, installing Oracle on NT reminds me of preparing a Thanksgiving turkey.....
Re: Oracle (Score:2)
As for which database to use, the problem isn't simply constrained to Oracle. I would say 50% of the Oracle installations I have worked with (600+), Oracle was gross overkill both economically and complexity. In many of these situations, SQL Server or Sybase would have done the trick and saved them quite a bit of money. Now, the funny thing is about 30% of the SQL Server/Sybase installations I've worked with (about 100) could have been easily handled with MySQL. It's like deer hunting with a bazooka.
The same is true for systems. I've seen Sun E450's when a Dell would do the trick. My personal favorite however has got to be a Win2k box to share files for a small software company with no budget.
Maybe its the herd-mentality of the PHB's, but I think a large part goes to the system analysts who make recommendations as if they are getting a commission.
BTW, I'm usually happy with the turkey after its cooked. I'm rarely happy with Oracle on NT, possibly because it never seems cooked.
Re:Everyone in the media bashes GNU/Linux (Score:1)