Slashdot Mirror
Microsoft Releases SP4 for Windows 2000
Snake_Plisken writes "I checked Windows Update today on a lark and found that Windows 2000 Service Pack 4 has been released." You can read a short CNet article discussing the media player patches as well as one more about
the fixes in SP4.
I wouldn't call a Service Pack for one of the most popular operating systems just "every patch". Service Packs come out about once a year.
And if you really want to talk about relevance, I'll guarantee you there's far more Win 2000 boxes out there than any of the Free OSes...
Since OS News had this yesterday (as did Arstechnica) I've seen a number of people installed without issue. I've done 2 servers in my lab (both were up to date on all hotfixes already) and they seem fine so far. I heard rumors (fud? who knows) that some systems that were only at SP2 had problems after upgrading, but I wouldn't think that is a widespread problem right now. Test, retest then deploy
..compared to how many millions of lines of code? 2 million? 200 million?
200 million lines of code and only 63000 bugs? Not too bad a ratio. That's what, 32:1?
Cruising the internet on my TI-99/4A @ a whopping 300 baud!
As much as I hate to say it, 2k is a really good operating system. XP uses the same kernel as 2k and it seems to run pretty well too. If it had a bit more finish to it then I probably would have stuck with it rather than going up to XP.
I hate to say it, but when I read changelogs for many Linux apps (or the kernel), they simply say "Fixed bug in foo.c". That doesn't tell me a whole lot as an end-user.
My beliefs do not require that you agree with them.
The part that worries me with how well 2000 works, how many defects do the 9x, XP, and NT versions contain?
It's not really that 2000 has less bugs, it's how well it handles them. Windows 98 might have more or less bugs, but it crashes way more often.
Not to mention how much more functionality is included in Windows 2000 than Windows 98.
Windows 2000 has finally given Microsoft the "five nines" it needs to compete in the business market. (99.999% uptime)
Let me start off by saying I am a Linux user and love it, but it always amuses me when Microsoft issue patches, the chuckling that comes from the Linux community...Like Linux never needs patching, in fact Linux is patched ALOT.
We all know the people exploiting security holes and writing virii for Windows computers are Linux users...What would happen if people as skilled as the Linux community turned on Linux...I'll tell you, patches would be released as quickly and with as much fanfare as for Windows.
Is Linux a better OS than Windows..YOU BET, but lets be realistic.....LET THE FLAMING BEGIN!
Except that Windows 2000 is still the latest version of their server OS and for most companies their client OS. Even Microsoft wouldn't make a product obsolete that they are still selling.
"The Scary part is, I've found Win2000 to be the most stable and reliable Windows ever released. 63,000 defects?"
I wouldn't read too deeply into the 63,000 number of defects figure without considering a few things:
- A defect does not always mean "Will cause Windows BSOD". Some defects are an interpretation of a problem. Fictional example: "Defect #24013: There's a post-it note icon on Internet Explorer 6 that is mileading. It looks like the notes icon in Outlook 2000." A lot of them are probably design considerations.
- 63,000 is a huge number, but you have to remember that Windows runs on a very broad range of machines. Not only that, but there are tons and tons of people running it who are supplying defect reports.
- We each only use a small part of Windows. You'll probably never know if there's a bug in the Win32 API unless you're a programmer.
I wouldn't these types of statistics too seriously. There'll be a day when Linux has that many defects, if it doesn't already. All it takes is complexity.
-Lucas
I allready migrated my domain server to Windows XP. I can recommend it.
Really? Was it XP Professional or XP Home? I wasn't aware that either version supported operation as a domain controller. Did someone port SAMBA to Windows by any chance?
To be fair, W2K is a decent operating system and W2K3 is better. That said, its vendor is sleasy, deliberately breaks standards and (in my view) will eventually bleed dry anyone who locks themselves into Microsoft. For business, I use Windows a lot. But, I recommend to everyone who will listen that they should position themselves to be able to move to alternatives.
I have 5 Win 2K servers and two Win XP boxes in my home lab, along with three FreeBSD systems. The XP and 2K systems have about the same uptime as the FreeBSD ones.
The only restarts for *any* of my systems in the last two years were for moving, installing new hardware, applying some update or another, or the occasional power supply failure- had two of those in the last year. I've had exactly zero crashes related to software in several years. I get a BSOD on my laptop every now and again, but that comes with the territory of running a debugger.
I routinely get 6 months or more uptime out of my desktop, and more than that for my servers. Any operating system can be made stable if you know what you're doing.
Unexpected Delay When You Log Off
They finally fixed that? Wow, that's probably going to be my number 1 reason to install SP4.
With Windows 2000 Service Pack 4, the software maker added various security and compatibility fixes and provided support for USB (universal serial bus) 2.0 and for wireless Internet connections. Im pretty sure my computer already supports USB 2.0. I want some Hi-Speed USB support instead.
If slashbot was filled with more software developers and less "power users", people would realize that the more "known defects", the better. If you know about them, you can priortize them, fix them, or defer them.
During the flamewars over this issue, people would post crap like "In all 30 Million LOC of Debian GNU/Linux, there's only 2000 open bugs!" Which only indicates a lack of systematic QA rather than some sort of inate superiority.
You're talking about processors, I'm talking about general hardware. Linux certainly deserves praise for being portable, however, it has not been run on near as much hardware as Windows has. Few companies make mass-market type PC products for anything but Windows.
Note: I'm not saying Linux won't run on as much stuff as Windows. I'm saying that Windows, because it's the de-facto standard out there, has the most hardware made for it. Sorry to bring it up.
Well, I guess another difference here is that Windows is one conceptual app does a lot of apps, compared to Linux being a lot of apps that do one thing.
Windows 2000 is, after all, equivalent to the linux kernel, glib, x server, window manager, web server, web browser, etc. ad nauseum. I wouldn't doubt that there are more than 63,000 conceptual functions of Win2k -- not even considering the obscure combinations of these, like opening a print dialog in IE vs opening a print dialog in Paint.
There are a countably infinite number of combinations of these as well. If testers, during their "what wierd shit can we make this OS do" phase, discovered 63,000 obscure bugs but 1,000,000 plus functions worked perfectly, I'd still ship the thing.
After all, all software over ten lines has bugs and implementation decisions. Some of those 63,000 may have never been found by consumers, while thousands more were no doubt discovered on the first day of release. That's how this industry works. Nothing is flawless or bulletproof...the benefit that Linux has is daily releases. Of course, that's if somebody cares enough about your bug to fix it...you might get stuck doing it your damn self.
Hey freaks: now you're ju
"Five nines" leaves you only 5 minutes 15 seconds of annual downtime (0.001% of a year).
I guess if you're running your entire infrastructure off of one measly little server then you only have a little more than 5 minutes a year of downtime for your enterprise.
On the other hand, if you're really in a situation that requires 5 nine's of reliability, you probably know about things like clustering and load balancing, redundant backups, and hot-swapping.
In the end, it's not the uptime of one machine that matters so much as the uptime of the service. Don't confuse the two.
I can't for the life of me understand some of the comments I read in response to this article about SysAdmins who are actually INSTALLING this thing right now to a bunch of users without testing!
This isn't a flame against Microsoft, it makes sense to fully test anything like this, be it OSX, Redhat, Windoze, whatever. Those that are deploying without testing are doing SysAdmin's in general a complete disservice-- it makes us all look bad when something goes wrong.
It just doesn't make any sense to me to even consider deploying before it has been out a while and tested. A service pack is a cumulative rolloup of security fixes and bug fixes and occasionally some enhanced features. Yes, there are additional fixes that haven't been distributed yet, but unless you HAVE to install it, you can wait a couple of weeks and test it in production before deploying it to everyone in your company.
Look at Winnt SP3 and SP3a. They released SP3a shortly after 3 because of some problems with the service pack. Frankly, I wouldn't want to be the sysadmin who installed it on all my clients to discover all the problems! Crazy!
"Look! There! Evil, pure and simple from the Eighth Dimension!" --Buckaroo Banzai
>In the end, it's not the uptime of one machine that matters so much as the uptime of the service. Don't confuse the two.
So, in that case, windows 95 has 100% availability, as long as you patch it fully before deploying it (there are no new patches, ever) and ensure you always have someone ready to hotswap a machine in every 49 days, 17 hours, 2 minutes and 47.296 seconds.
Wow! I guess overheating C64s can also acheive that, considering if you have 3 or 4 on at a time, one's always going to be available in a pinch.
W00T! I love stats!
If you could be told what you can see or read, then it follows that you could be told what to say or think - BoC
"Come on, how many icons and groupings can a default win2k have wrong? I doubt they have 10,000 help pages."
:)
Doesn't sound like ya know, does it? I don't know how you can leap from an "I doubt" comment to a "this is clearly..." statement.
"This is why Windoze 2000 can't run more than a few days in a row."
Funny, I had a home-made Tivo running Windows 2000 that had an average up-time of 3 months. I've got an NT4 Exchange Server that's been up for 80 days here. We used to have an IIS webserver running gon NT4. It was up for well over 6 months. We never needed to restart it, but we did have to physically move it a couple of times. My desktop machine has been running for about 13 days now. It'd have gone longer but my UPS is flaking out on me. Not bad for a machine that I do 3D animation and play games on.
" Trust your observations to tell you that software simply sucks."
Think I'd be defending 2K if I had 'observed' that the software 'simply sucks'?
"Balderdash! Windows2000 runs on intel 386. Wince runs on ARM. That's it. What do you think this is, free software that's compiled to specific x86 processor families, Motorola, ARM, Alpha, "
A machine is not a processor, it's a complete setup. Never heard of the old "Windows is on 90% of desktops" stastic before?
"you will always be at the mercy of the service patch that requires you to give up hope of privacy."
Actually, if you had read the EULA instead of going by the sensationalized Slashdot version of it, you'd know that the purpose isn't for MS to go sniffin around your machine. The reason it's there is to support a number of the features they added to Windows to deal with the virus problems that have been plaguing it. Go read it.
"Pull your head out of your closed source place please."
I would suggest you do a little thinking on your own instead of repeating all the stuff you've heard on Slashdot that gets modded +5 Insightful.
"Free software has fewer bugs and does more than any dinky windoze distro will ever. "
That's a myth. I'll give the Open Source community credit for responding to bugs in a timely manner, but you need to face facts that Open Source Software is rarely both well designed and bug free. Run a few commercial apps in front of an ordinary user and then run a few free apps in front of an ordinary user, most of the time he or she will be able to tell you which is which. "well, the commercial one seems to be friendlier to me while the free one is confusing to use."
" The complextiy you are thinking of is a legacy of all the dirty tricks M$ used over the years to kill of software rivals. That does not exist in free software and never will. "
Yeah, that's scientific. Heh.
"This is why free software PCs don't have to be turned off until the power fails."
Riiiiiiiight. We'll see how stable your Linux machine gets when games start becoming available. You'll find out just how 'rock-solid' it is then. Linux machines are not being used like Windows machines are, so drawing comparisons like that is not very informative.
Nice bit of Linux propoganda tho. Bucking for a +5 Insightful?
I found it funny because it sounds like he doesn't check it regularly...considering how often patches come out, his is a box just ripe for hacking in about 6 months...
People will pass up steak once a week, for crap every day.
I wouldn't know. I would say no because the key and ISO showed up in newsgroups at least 2 to 3 months before XP was released (There's even a picture of someone holding the rip in front of Microsoft's X days until XP launches sign). I myself had it one month before...Not that I kept it or anything ;P
I'm sure Microsoft got wind of the key long before anyone actually used it and I would guess whoever may have gotten the key was issued new a new one.
Snarf This.