Slashdot Mirror
Windows Server 2003 Is A Small Step Forward
b17bmbr writes "According to eWeek, 'The release of Windows Server 2003 is a small step forward for the platform -- an effort that really should be considered Windows 2000 Server Second Edition. With the exception of Internet Information Services 6.0, there aren't any far-reaching or fundamental changes in the product.' And from CNet Microsoft prepares Windows Server ads, 'The ads are geared toward IT managers on tight budgets.' This is probably Microsoft's last chance to turn the tide and take mindset and market share from FOSS."
This is probably Microsoft's last chance to turn the tide and take mindset and market share from FOSS.
please. they have $30 billion in cash. i think they'll be able to buy some other chances.
They took Windows 98 SE, dressed it up a bit, and called it Windows ME.
It's lousy from a consumer standpoint, but enough people thought it worthwhile to buy it and make it profitable for Microsoft.
It's not the most upstanding business strategy, but it still makes them money. And any business is not in it for the ethics, but about the cold hard cash.
The Pigloo
nothing that apache hasnt had forever!
Wouldnt it be cheaper for an IT manager on a tight budget to stick with 2000 Server rather than 2003 Server. I know I dont need it and I have a tight budget. We have most of our infrastructure already upgraded to win2k server at-least the stuff that will be migrated over. We will not be upgrading to 2003 server but rather get it as it comes preloaded on any new servers we buy.
This must be Thursday, I never could get the hang of Thursdays.
I see that tongue planting firmly in your cheek, there. Seriously, though, Microsoft has three main businesses: enterprise software development tools, office applications and the desktop itself. Linux is taking a major chunk out of their server market, and slowly, inexorably the rest of their business will follow. They will either adapt or die. My guess is that they'll adapt, and take their new place to the left and just below the Throne of the Great Penguin. They're certainly not going to go away, but in about another five years, they're not going to be calling the shots anymore.
It currently is a bitch and a half to get web sites setup exactly the same way across multiple web servers in a farm behind a load balancer without the use of third-party utilities (IIS Export is really nice and isn't too expensive... Google it for more information). Also, if you have web servers that are not in a domain and you want to restore the IIS metabase on a rebuilt system... good luck. Even with some help by Microsoft, the process is very painful and isn't perfect either. Instead, we had to use IIS Export to migrate all of the sites from one server to the rebuilt server. Not a fun task to do for over 50+ sites.
Having text-based configuration files would be a godsend for people in such a situation! It would also make creating an restore image of a server much easier since you only have to update the web content to the latest version in production.
No one can make such a predication, what if there is a new linux that gets developed that completely revolutionizes computer use. Linux is only 10 years old and look at the influence it has gained in that short period. I don't think that linux will be that last of the computer OS revolutions.
Checking out my form of escapism.
Being a small bit of a geek, I think myself qualified to say whether a term is esoteric or not, and I must say, I've never seen FOSS in my life before. My first thought was "How is my local tourist goods shop suddenly competing with Microsoft on a global scale?".
Free Open Source Software (FOSS). Thanks, that's what I want. More adjectives. And, once more, have them all thrown into an acronym I can't recognize. That's not going to encourage cliquishness or scare away people who might otherwise be interested.
I even thought to look at E2 to see if the obscure FOSS had been noded. If it had been, a little link could have at least been provided to make this more accessible. Nope. Then again, I remember reading something in the Slash CVS which mentioned the E2 linking (with those little question marks) was broken.
You like splinters in your crotch? -Jon Caldara
Depends what you mean by Linux. Kernel? Or GNU/Linux, operating system?
There hasn't been a technical revolution anyway The revolution has been in licensing and using the Internet as a development environment. The technical work is all evolutionary, small steps.
In proprietary software, you create the appearance of a revolution by giving something old a new name. You can't see the source so you can't see that it's nothing knew.
So I wouldn't be surprised to see Linux around in 20 or even 30 years, and I'm sure DOS and VMS will be with us too.
Yours Sincerely, Michael.
Back in Jan '00 it was at $60 - now it's about $25...
That would make it "a good time to buy".
I guess all my pro-Microsoft / Linux sucks posts got scrolled off the list. All of the information I listed is public knowledge, read the literature.
It's accurate that I'm anti-MS now. I always thought they were evil but hell I'm evil too, so you go where the winners go. I think Linux is a winner now, that's why I'm in this camp for a while. Later it will be something else...
The reason you see so many employment adds for these technologies is because (1) they keep having to fire the incompetents they hired the first time around and (2) the incompetents they fired are the ones who recommended that technology that they paid big $$$'s for and are now stuck with.
Address : www.dbcodegen.com only open port is 80 IIS6 Prize Money : $100 US winning entry - must replace the homepage with your contact so taht prize money can be paid. Ricky
$100 bucks huh? Either you are wanting people to work (hack) at slave labor prices, thus doing your dirty work on the cheap, or you only have $100 worth of faith in a product that costs alot more.
If someone just needs the $100, I could use some help this saturday spreading mulch and chainsawing several large trees here at the house. Pays cash. Bring gloves and a lunch. Beer provided afterward.
Tequila: It's not just for breakfast anymore!
Times are hard. All the good jobs are occupied and the lousy jobs go begging. That's why you see so many MS sysadmin jobs open.
"Only in their dreams can men truly be free 'twas always thus, and always thus will be."
--Tom Schulman
Yes, I know, there are always programs you can't do without and can't find in UNIX. There's also, of course, the problem of getting people to try something new. For many people, learning anything new is hard, and once they've done it, they don't want to do it again, so there's lots of resistance to change. However, all these new features look like Micro$lop's finally realized that they did it the wrong way in the first place, but don't want to admit it.
I think the biggest problem for me is that all these shiny new features are still piled on top of all the existing Windows cruft. The old vulnerabilities are all still there, still unpatched, still waiting to be exploited. As long as a glitch in a display driver is still allowed to crash your server without warning, I'll stay away from anything built on NT "technology," TYVM.
Good, inexpensive web hosting
MSFT P/E == 28.98. That's high for a producer of a commodity product. OSes and office suites aren't rocket science anymore.
I know this has been said before, but it seems almost everyday Windows become more Unix-like (cleaner, faster, more stable, better) while Linux becomes more Windows-like (less stable, slower, more bloated and less stable [why is is that the 2.2 kernels are generally considered more stable than the 2.4 series?]). With current predictions showing PDAs are going to overtake desktops in the next few years, the Linux community has to concede the desktop market to Microsoft and move on. Servers are is where Linux/Unix strength is. It just always seems to me Linux is playing catchup to Microsoft on the desktop while MS is learnig from their mistakes and trying to move forward.
And yet, you're apparently willing to play the Dependency Hell game, not to mention the hours required to make anything work the first time. And then there are the kernel upgrades every other Wednesday and the futile attempts to find any non-pre-pre-alpha products that actually work. Good luck getting that work done.
Sounds like linux astroturfing bullshit to me. Where do they find you people?
Windows is so dumb, it is like so dumb. What I mean to say is that if Oreck can lift a bowling ball with it's suckage, Windows 2003 could lift all of Russia, France, Germany's balls or maybe their bowling balls because they show no signs of testicularosity. It is a shame that it wasn't completed in time to be used on Iraq, who would have thought a 8x10.2x2.35 inch box would be more effective than a nuke.
\Windows is horrid
</burning_karma>
Who is this "Poster" guy and why does he own all of my comments?!?
Sorry to bust your (communist) bubble, but someone needs to pay for bandwidth and various streaming media stations. In case you haven't noticed, advertising provides a means to do this in the age of the internet. Sorry, bub, this isn't PBS -- no one's gonna call up and pledge. If you can't find a means to cover your operational costs you'll be out of business in no time.
Regardless, these are *features* that are optional. It's nice to know that they're there if you need them. So quit trolling with your GNU/Jihad about "information must be free!" That's sure a nice utopia to live in but come down to earth and enjoy something called the "real world."
Although someone complaining about optional available features on Slashdot doesn't surprise me in the least.
This is the real problem with MS these days and no amount of reform on the part of engineering is going to cure it. Win2k3 may be the best thing since sliced bread but pair it with MS legal and the MS corporate culture and it's not a partnership that I'm entirely comfortable recommending to anybody these days, even confirmed MS shops.
I'll probably renew my MCSE credentials in order to help out customers on migration and interoperability but without some forced reform like the Teamsters went through, I can't imagine how the public can trust MS with anything.
Don't forget about costs related to implementation and support. It is often quicker to get a Microsoft-based solution up and running than an equivalent Linux-based solution.
Then there are costs related to support. UNIX/Linux specialists often bill at higher rates than MCSEs (good if you have solid experience in that line of work, bad if you're on a tight IT budget). It can also take more planning and vendor support to design/implement a Linux-based solution. Every peripheral company is supporting Windows, but Linux support isn't quite universal.
The bottom line is often pretty much a wash, and then the decision-makers go with whatever they are most comfortable with...
That's the problem with a psuedo-database. The "right" tools suddenly become remarkably more complex and harder to debug. A "single interface" can easily be applied to configuration files if there is a unifying standard. The need for a "single interface" doesn't require a single datastore. This is one of the rather nice aspects of a "real database".
As far as "multiple users go". That too can be wrapped around standardized, human readable files. Although most metadata will be specific to a particular user and there's no compelling reason to burden a centralized system with data only relevant to a particular user.
Also "it's a bit complicated for the average user" is an absolute show stopper for an OS marketed to users expected to be morons.
A Pirate and a Puritan look the same on a balance sheet.
A better reason to unload em is that they have lots of downside and darned little upside. Think it through. Their stock has been stagnant for several years and their whole business model is predicated on ever increasing revenues leading to an ever increasing stock price. The rising stock price allowed them to fund the bulk of their payroll with ever increasing stock options. Those options ain't worth shit these days and I'd bet it would take a couple of years back on the gravy train for the jaded rank & filers there to believe in becoming a millionare through options.
Then there is the revenue problem. Once you are a monopoly there isn't much room to increase revenue through increasing market share and the natives are already restless after their first attempt to squeeze more money from their installed base with Licensing 6.0. So if revenue is going to be flat/down the only option left is greater efficency.
But not at Microsoft because they are already too good in that dept. Their cost to produce the next rev of Windows or Office is already close to nill per copy so half of almost nill wouldn't make a big difference. Cut marketing and they are toast since the bulk of what they write off as marketing is shady deals to enforce the monopoly.
That leaves their rightly feared cash horde. They can cause much mischief with that stash, but if you are expecting that to prop up their stock value you should think again. Exercise for the student: Go to your fav stock info site and find the number of shares of common stock outstanding and divide it into the approx $40 gigabux they claim in cash and short term investments. They have split their stock so many times that even 40 bil gets watered down real heavy.
Are they doomed? Not likely, they long ago passed the size where a company is too big to be allowed to fail. But as IBM was once the mighty titan that dictated terms to the entire technology industry, Microsoft will also pass into being one among many peers.
Democrat delenda est
Will Microsoft survive? Sure.
Will Microsoft lose their domination? Yes, they will, eventually.
As an investor, I see Microsoft as a company that has very much to lose and not much to gain. (On the desktop, the only place where things are going reasonable well for MS, there is no room to grow and everywhere else they are losing. But also on the desktop, especially OpenOffice is starting to eat their MS Office)
Linux is not there to slay any Goliaths.
Linux is there because many people believe it solves their computing problems, most importantly it solves the problem about who decides how to handle your own computer resources.
With MS you have to upgrade when they say you must, to what they say you must, under the conditions they dictate to you and there is absolutely nothing you can do about it if you have become completely reliant on MS based stuff for your crtitical computer work.
With Linux and other OS OSes the ones that make those decisions are the owners of the computer infrastructure. Red Hat does not support your version of RH Linux but you can't upgrade? No problem, all the code is open, you can pay somebody to fix any outstanding issues if that is what is required. Or you can move the data and programas to other Linux (and most probably any UNIX platform) when you are ready.
The variety of Linux distributions gurantees that you find the right solution to your problem and not a one size fits all approach that most probably will not please all the different types of needs out there.
Regarding Linux training, there is enough stuff around, but to be perfectly honest any good UNIX training is good enough and any particular caveats for a given distribution are easily covered by books or standard documentation.
I never learned Linux anywhere but had my first working machine up and running in no time at all. I had been working on Solaris before that, I felt at home on Linux.
IANAL but write like a drunk one.
While i was there I evangelized using I/O Completion ports, Overlappped Async I/O, Scatter/gather APIs, Interlocked Linked Lists (newly documented in XP/2K3), &c. the "MS-specific razor's edge" that IIS and SQL and any high-perf app on Win32 uses. It is a true statement that if you follow *exactly* that set of techniques you get world-class perf. It is also true that if you do anything the way the rest of the world does it you get shitty perf. It is also true that every ISV I worked with had huge perf problems because they didn't "walk the razor's edge."
Example: a guy in the kernel team was bitching that WaitForMultipleObjects has a 64-handle wait limit. Dave Cutler stepped in and gave historical reasons for why the limit was set, and argued eloquently (as you did) that the Unix way of select() on thousands of sockets was inferior to using I/O completion ports. His argument was compelling, as is yours. But the original guy said: "I cannnot go to a large ISV and tell them REPENT! and use I/O Completion ports! as does IIS!" The MS-attitude is IIS is a great socket-listener, and MS wants people to write web services using ASP.NET, so why reinvent the wheel?
So that's why I switched. It's not that MS technology is inferior. It's just that it's myopic. Linux is out there letting a thousand flowers bloom. It's more fun!
I see feminists are so desperate to point out their relevance to technology that they make their own site just for them. How cute! Are they called "witi" because they know how to laugh at themselves?
Since, as usual, a feminist is bad at supporting even her own case, I'll help you out: Lovelace and Hopper. These were two people who didn't make a big deal about having no dick, but just got on with the job, and as such were recognised as "practitioners" rather than "angry feminists".
Similarly, successful men just get on with the job rather than explaining their achievements in terms of having no tits. Content of character/mind, not colo(u)r of skin or shape of reproductive system, see?
Either problem alone would scare potential buyers, so it seems to be forbidden to discuss.
It would be convenient to skip the upcoming deluge of advertisements and astroturf and see trade magazines feature the F/OSS tools instead. Ads cost a fortune and MS could instead use the money to 1) hire developers to rewrite software in a secure, stable form, 2) hire lawyers for the upcoming willful negligence lawsuits.
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.
The thing is .NET doesn't run at the kernel level. There is no integration into the 'core os', its just another runtime making win32 API calls.
.NET so fast is:
What makes
1) MSIL is designed for optimization, and to be compiled not interpreted. Several features of the IL, inferring types from stacks, etc, allow you to make some great improvements in perf.
2) MS has a HUGE team of developers working on high speed, high quality JITs that are using some of the latest ideas in compsci. In-line optimizers also optimize code on the fly, so the longer your program runs the faster it gets. (You can see this if you make video processing loops, etc.)
3) This huge team of developers gets more bang for their buck than from, say, IBM, because MSIL is easier to optimize. IBM has made some INCREDIBLE advances to Java performance, but Java's stack assembly is so much harder to optimize that IBM's team has to work a lot harder for the same gains.
(IBM has some incredible scientists though, and these guys really made Java viable on the performance end.)
I think you switched because you were an intern and didn't get back a 'Will Hire' recommendation. :-)
You walk the razors edge on ANY performance tuning you do. Be it Linux, or Unix, or Windows.
Take a look at Oracle for Linux - it works best on specially tuned versions of linux, and requires all kinds of kernel performance changes and patches.
As a developer, I'd rather just 'get it right once' then say 'Ok now patch your kernel for optimal performance.' Once you optimize your server for Win2k, you don't have to worry that someone's kernel settings will be off and kill your optimizations, it will work on ALL of the Win2ks out there pretty much the same.
That is worth a lot more to me as a developer. You walk a razor's edge, but you do it once every few years. That's why Windows TCO is said to be cheaper than Linux's - Oracle's recommendations on how to tune your kernel for Oracle might go against another application's recommendations for tuning. And all that tuning still requires a sysadmin's time to go in and recompile kernels and make other changes.