InfoWorld says WinXP much slower than Win2K

iforgotmyfirstlogon submitted an InfoWorld story that makes the shocking claim that XP is slower then 2k for business use. Pretty graphs, comparisons of SMP, and they even tested without the eye candy. My favorite comment is this one "it appears that for light-duty service on the newest hardware, Windows XP with Office XP is an acceptable choice -- if an 11 percent performance hit, or 53 minutes added to an 8-hour day, is acceptable." And thats the best case scenario.

Comments

not the only performance hit

[Lumpy]

You also have a huge amount of retraining because XP doesnt look or act anything like the NT4.0 or Win2K models.

You now need to re-train your users on how to use the Operating system...

Gotta love how they say how linux is too hard to switch users too but dont mention that Microsoft does the exact same thing every 2 years to their user interface.

Re:not the only performance hit

[ConceptJunkie]

The first thing I did with XP was turn off that ugly, stupid, round, bouncy, primary-color assault they call "Aqua", wait... whatever they call it. Like almost every visual UI change uSoft has done in the past 10 years, it only chews up more screen real estate with wasted pixels.

Once that was changed, I had a reasonably lean, nice-looking UI with the benefits of the enhancements to the task bar (very nice!) and the start panel (kinda cool, but pointless IMO, also that can be turned off too). And on a laptop, ClearType is worth the upgrade price alone.

Now, configuring is another matter. For instance, I still can't always get volumes to share on my home network the way I want them to on the first try, but then again Joe Worduser isn't going to be doing those kinds of things.

OTOH, XP boots radically faster than Win2k on my IBM i-series laptop (Celeron 433, 192 MB RAM) and shuts down faster too.

Now going from 9x to NT/2K/XP, that takes some re-training :-)

Mostly in unlearning that you have to reboot your computer every hour.

Re:not the only performance hit

[tmark]

Gotta love how they say how linux is too hard to switch users too but dont mention that Microsoft does the exact same thing every 2 years to their user interface.

I gotta love how Linux zealots downplay the difficulty in switching users to Linux but then jump on top of Microsoft when they change their UI - as if that made the abhorrent UIs currently available on Linux somehow even approximately as useable as (the still crappy) Windows UI.

Re:not the only performance hit

[Masem]

If only we told people how to *use* computers as opposed to teaching them to click on specific buttons and using specific menu items to do specific tasks...

Typically, as long as an OS has a consistent interface *and* plenty of help, I doubt a well-trained computer user will get lost. However, sheep that are trained on how to use Windows or Word or Excel will flounder when they have to deviate a bit from the path.

The XP interface isn't terrible. (I'm a sucker for AA fonts and alpha layer fun). All the buttons are in the usual places, but look just a bit different. But because we've "trained" people to look for a black X on a grey button in the top right corner of a blue-background window border in order to close a window, the new XP interface will give those sheep headaches. Microsoft's fault? No; I blame "..For Dummies" and the rest for the sheep mentality.

Re:not the only performance hit

[geomcbay]

For the record, I've been a computer programmer since the C64 days. But I really dislike this "people who don't know how to use *real* computers are sheep" attitude.

Computers are no different than anything else: Cars, VCRs, whatever. Yet hardcore tech guys tend to think people are stupid for not learning how to "use" computers, often snickering at those who have never used a CLI. The simple fact of the matter is many people don't WANT to learn the in-depth operation of a computer just like most don't want to learn the exact details of how a car, tv, vcr or whatever works. They just want an extremely simple, fairly standard UI. They have lives to lead that include many things other than computer use and aren't willing to spend 1000s of hours learning the ins and outs of computer use just to type letters and surf the web -- and they shouldn't have to.

Re:not the only performance hit

[kaladorn]

For the record, I don't consider people like this very smart. If you have no desire to learn the details of how something works, you shouldn't be using it. I know a guy who had no desire to learn how his car works

I will observe that there is a difference between having the interest in something and having the time to investigate it. For example, there was a time when a car was a fairly rudimentary thing to do most operations on. This was well before a dozen sensors controlled combustion, before vehicle engine control modules with arcane diagnostic codes that require a mechanic or a nasty piece of hardware to get at. This was before the level of complexity of those systems grew to the level it currently is.

I also don't know... you may have tremendous amounts of spare time. Most people don't. I'd love to know why my Doctor prescribes a certain medication, but at some level, I have to take his word for it as anything more than a beer n' pretzels explanation will exceed my university chemistry knowledge level. Similarly, I ask my mechanic why he's doing something, but once he moves off into bafflegab, I'm left with two choices: trust that he's a professional or don't.

I think I could probably walk around almost anyone's house or the environment they work in and identify at least some things about which they know little or nothing other than how to operate the item in question in a simple way. We humans now live in a complex world and not only is it infeasible for you to know something about everything you come in contact with, it is inefficient. Specialization is efficient. Knowing a little bit about a lot of things has some utility, but that isn't having real knowledge of those things. A veneer isn't in-depth knowledge. And if you spend your life trying to investigate all of the existing objects that you come in contact with, you won't be doing much else.

Reduction of complexity to usable levels is how humans cope with an increasingly complex world. Reducing formerly complex tasks which had to be understood in detail to black box technologies that anyone can get at least average utility out of is how we move on to dealing with higher level concerns. I for one am glad that I don't have to worry about what IRQ or DMA channel or I/O space address my various PC cards occupy now. There was a time when I did. I wasted hours friggin' around with these things because it was a necessity. It no longer is, and I can worry about new (and more worthwhile) concerns.

And, strictly as an aside, the snide down-the-nose look that most Geeks tend to give the untermensch that compose most of our world and who use M$ products doesn't exactly enhance the reputation of our caste nor encourage people to seek our help our to try our chosen solutions. In fact, it drives them deeper into the hands of those who offer them no pain and who pander to how the world is, not how we all wish it were.

Re:not the only performance hit

[vsync64]

I've said this before and I'll say it again. I don't in any way expect people to know every detail of how every component in their system works. I do, however, expect them to understand the basics before wreaking havoc upon the lives of others in the community they invade. I also expect them not to run away in disgust when knowledge is presented to them on a silver platter.

I don't know how everything in a car works, but I understand the basics of axles and gears. I don't know every line of telephone switch code by heart, but I understand the basics of country codes, area codes, and prefixes. And so on. For a computer, the equivalent is understanding the basic differences between short-term and long-term storage, how file systems are organized ("This is a folder/directory. You can put files or other folders in it."), and the like. For networks, the basics are host addressing schemes (TLDs, user@host, etc), simple protocol knowledge (no, your mail is not sent over the Web, even though there may be a Web front end, and there is more to the Internet than the WWW), and what "client", "server", "upload", and "download" mean. I have no sympathy whatsoever for anyone who refuses to learn these basics, any more than I would expect sympathy from a police officer if I attempted to drive without first learning what the pretty colors on traffic lights meant.

The second and more important issue I mentioned is that people tend to shun the acquisition of knowledge, especially when computers are involved. No, I don't know everything about how a VCR works. But I had a VHS tape I needed to watch the other day get crumpled, and what did I do? I broke out the screwdriver. And now I know exactly how the VCR releases the locks on the tape spools, how the tape feeds through the cartridge, and all the other knowledge that will let me more quickly diagnose a problem with my VCR if it ever occurs again, and I watched my video on time.

I don't expect every computer user to open the case every time something goes wrong, but I expect it to make at least a token effort at proper use and maintainence of the system it uses on a daily basis. If I say "your file system keeps getting corrupted because you keep turning off your computer when it is writing to disk", I expect you to stop cutting power in that fashion. If I say "the reason it keeps having to write to disk so much is that you need more than 32MB of RAM to run 5 applications at the same time", I expect you to get more RAM or stop whining to me, and certainly not to say "but it has a gigabyte" when I have explained the difference between short-term and long-term memory 5 times in the past 3 minutes.

If someone wants to be lazy and ignorant, that's his problem. But he shouldn't expect his life to be a seamless ocean of perfect technological ecstacy, and he shouldn't expect me to gladly mop up after him for free. I reserve the right to charge high prices for the accumulated technical knowledge of a lifetime, when he could have learned what he needs to know in 5 minutes of his own time.

Not much of a surprise

[Shimmer]

Does anyone doubt that, say, Word 95 installed on Windows 95 would run circles around Word XP on Windows XP (on the same hardware)? I hope not. You can call it bloat, but there's probably a reason why people (not just "lusers", but also "power users" who "know better") keep upgrading anyways.

This is the foundation of the Wintel monopoly: Harness ever-expanding software to Moore's law and reap the benefits. We don't have to like it, but at this point its not a surprise either. Maybe instead we should try to understand why it's been so successful.

-- Brian

Re:Not much of a surprise

[joshjs]

This is why I still have my typewriter: you don't even have to tell it to print, it's so damn fast.

even more shocking...

[Anonymous Coward]

"InforWorld story that makes the shocking claim that XP is slower then 2k for business use"

even more shocking is that InfoWorld has changed their name to InforWorld, and that the previously respectable publication has taken to printing such elementary mistakes as "slower then" rather than "slower than"

troll me, flamebait me, i don't care. you should check your spelling and learn the difference between the common homonyms! and here i thought this site was for geeks and nerds...

Re:even more shocking...

[sulli]

But without the typos, it wouldn't be salshdto!

Re:even more shocking...

[iforgotmyfirstlogon]

In my defense, Taco edited the piss out of what I posted. It was error-free when I sent it in.

- Freed

I'm suprised by this

[johnburton]

Ok, so I'm not exactly a microsoft fan but I got XP professional recently because as a professional software developer I need to at least be aware of how it works, and what it does...

But I'm suprised because I subjectivly find it works noticably faster than 2000 seemed to do. Programs seem to load quicker and ot just seems more responsive. Could be because I reformatted and defragmented my disk I suppose.

Have to agree with the comments about moveing things around. Not a problem for me, but it did take me ages to find a few things first time.

Hard to argue with statistics

[Pov]

It's hard to argue with statistics from an authoritative source, but I'm running XP right now and I have XPerienced no qualitative decrease in performance over my old Win2K install. I would say I fit into the power-user category since I usually multi-task through a couple of applications and run with about 8 windows of something or other open at a time. I can't benchmark with pretty graphs, but I don't think I'm losing 53 minutes a day or even five.

This is a benchmark of only Office XP

[gburgyan]

How often in the last couple years did you notice "Gee, this word processor can't keep up with my mad typing skillz." So what if an automated benchmark can't make a bazillion documents as fast.

In the end there are lies, damn lies and benchmarks.

As someone who's used XP, the time lost (microseconds per day) are more than made up for with the added reliability of the system and the much easier recovery process. (Personal experience -- I was evaluating the system for work and purposely installed some crappy drivers that I knew would blow up; the system recovered just fine)

I read this in the print version of InfoWorld a few days ago and got pissed off then too. If you're going to beat up on M$, do it better for crying out loud. This is just like the dumb VM debates for Linux.

Microsoft announces Windows XP Turbo!

[Refried Beans]

Now with compiler optimizations!

The age old programmers vs. engineers problem

[Uttles]

It's easy to point to certain features in a new OS as examples of progress, but end-users often find that a new OS performs like molasses compared to the version they were using.

So why does that happen? Well I'll tell you my educated guess: every year, electrical and computer engineers make amazing advances with comptuer hardware, making RAM more plentiful and less expensive, making hard drives larger and faster, implementing devices like L2 cache to decrease read/write times, and most popularly making Processors faster than ever (at least by clock speed.) You would think that these advances would make all software simply fly, be faster and more responsive than ever, and you'd have unlimited storage space for your files. However, that's not the way it is, and somehow, you still run out of disk space, don't have enough RAM, and have programs running slow (on a 2 GHZ Machine!!!) So what is it? Programmers. "Computer Scientists," rather than improving on software that ran well on old architectures, go by the thought process "well now that we have all this power, why don't we use it all" and so they end up writing applications and OS's that hog all the newly available extra resources. I'm not saying all Comp Sci's do this, I mean look at Linux, it's pretty damn efficient. When it comes to commercial apps though like Windoze, rather than make something extraordinarily efficient that runs on the newest machines, they say "well the hardware takes care of efficiency, let's just make something with a lot of bells and whistles." What you end up with is grossly large applications that sloth along on extremely powerful machines that have the capability to be so much more. This is yet another reason to use Linux.

Re:The age old programmers vs. engineers problem

[slow_flight]

I think what happens is that programmer's focus changes. When I first started as a professional programmer 12 years ago, we concentrated on code size. The distribution media of choice back then was the floppy disk, and we really, really wanted to keep our app small enough to distribute on a floppy. That, and user's hard drives were small. Priority one was code size (and remember that this priority ultimately resulted in the Y2K debacle).

Eventually disk and memory became cheap, so responsiveness became the focus. "Don't worry about spending a few more bytes, just make it faster." Then processor power became cheap too.

Now that memory and performance are more than adequate on the platform, the priority has become features. "Don't worry about how big it will be, memory is cheap. Forget about how slow it is, everyone that matters has at least a Pentium III to run it. Just get some new features in there asap."

My point is that code is slow and bloated because no one cares about that anymore, at least with regard to the 'unwashed masses'.

It's a lot like the way you grow into your income. I bought a house nine years ago when I was making one-third of what I make now, yet I still seem to go paycheck-to-paycheck even though I'm still in the same house.

Re:The age old programmers vs. engineers problem

[Christopher Thomas]

When it comes to commercial apps though like Windoze, rather than make something extraordinarily efficient that runs on the newest machines, they say "well the hardware takes care of efficiency, let's just make something with a lot of bells and whistles." What you end up with is grossly large applications that sloth along on extremely powerful machines that have the capability to be so much more. This is yet another reason to use Linux.

And of course, here on Linux, we never make the same mistake. We're all just chugging along with fvwm as our window manager and pico as our editor and all of our apps have a footprint of less than 4 megs in total...

And I have a bridge in Brooklyn to sell you.

Bloat and feature creep happen on all platforms. It's just easier to escape under Linux, because you aren't locked into a single toolset. Calling it a Windows-only problem is a gross misnomer, however.

The real problem with sluggishness under Windows is actually device probing (during boot) and hard drive seek time (when launching anything, due to the many configuration files it checks). Swapping isn't a concern if you use your system wisely, and applications are usually quite responsive (YMMV). Branding Windows application programmers as lazy when your system bogs down is grossly oversimplifying.

Oh please!

[mr.nobody]

Gotta love how they say how linux is too hard to switch users too but dont mention that Microsoft does the exact same thing every 2 years to their user interface.

What a load.

XP is the first time since Windows 95 that Microsoft has made a major change in the look and feel of the GUI. As 95 begat 98 which begat ME (and NT 4 begat 2000) minor things have changed such as the placement of the Windows Explorer icon and a Control Panel group or two, but the same basic grey bar at the bottom of the screen with the start button has always remained.

Every 2 years? Complete FUD.

Wow... ignorance is bliss huh guys?

[Anonymous Coward]

Slashdot is so predictable. The second I, and many other people who enjoy knowledge over ignorance, read the Infoworld benchmarks... I knew that Slashdot would post these, and ONLY these, and ignore the many other benchmarks that show WinXP performs as good if not better than 2k.

For instance, here is what Paul from WinInformant has to say:

"InfoWorld stood alone this week when it declared that Windows XP significantly underperformed Windows 2000 and Windows 9x in its tests. Not only do the controversial InfoWorld results fly in the face of Microsoft's published results and actual real-world use, they refute every independent XP performance test performed to date. One gets the idea that ... nah ... InfoWorld was trying to make XP lose. Don't believe me? Consider this: The following organizations have tested XP, independently of Microsoft: CNET/ZDNET, eTesting Labs, eWeek, PC Magazine, and PC World. All these independent labs came to the same conclusion: XP meets or exceeds the performance of Win2K and Win9x. The InfoWorld results are also at odds with real-world XP use, which already includes hundreds of thousands of beta testers, tens of thousands of IT professionals and developers, and hundreds of thousands of enterprise customers. "Microsoft has not received any indications that users are experiencing reduced performance compared with Windows 2000," a company spokesperson said. "We have had extensive feedback that Windows XP is better performing than Windows 9x." Go figure. And yet, you just know that every anti-Microsoft site on the planet is going to run with the InfoWorld story and not any of the positive stories. Ain't life grand?"

Looks like he hit that nail right on the head, huh?

Re:Wow... ignorance is bliss huh guys?

[borzwazie]

I have also tried XP. It has a lot of really nice features, and some not-so-nice things.

First of all, if you turn off all the GUI bloat (My Computer -> Properties -> Performance Settings) it is nicely quick indeed even on low-end machines PROVIDED that you have lots of RAM. A Celeron 400 with 192 megs of RAM boots into XP much faster than 2000 does, and performs at or above the level of 2000 with a similar amount of RAM.

It is true, with any newly-release software, that there are probably bugs and compatibility issues. But XP really shines in some areas, notable Firewire. A friend who had a 1394 card that he couldn't get to work in win98 lent me the card, so I tried it and a DV cam in XP. Not only did XP not need any drivers, it mounted the camera up as a drive instantly, and has a built-in image capture utility that works extremely well (though it could use some extra settings for images). Color me impressed as hell.

XP by default, installs about 1.2 gigs of stuff! It also ships with a number of security concerns like Remote Registry, Remote Desktop, and some other services turned on by default... XP also has a lot of services that are not necessarily security concerns, but, do we really need easy wireless configuration turned on by default if you don't have any wireless devices? Same for the camera image services. Those turn on if disabled as soon as you install a camera anyway. So, there's LOTS of room for memory bloat improvment. I got a significant speedup by disabling a lot of unneeded services (probably about 12-15 by default).

We in the linux community, instead of bitching, need to look at XP as the new target. Look how good OSX and XP are. If we want to remain a competitor, we need to make our desktops this good. No excuses.

Numerical FUD

[spellcheckur]

I have to question the validity of any simulation that thinks that 100% of the average user's 8 hour work day is spent doing things that a computer isn't orders of magnitude too fast for.

I write code for a living, and while I've got a well tuned linux box to do all my compiling, any (speed) advantages it has over an $800 low-end Windows box when I'm writing emails or posting to /. is lost when I pause at the end of a sentence to consider my next thought.

(voiceover indicating speed being wasted goes here)

Your "average" user, in all likelihood, isn't running 100% processor intensive tasks. They're composing emails, or preparing presentations, or IMing their coworkers about the wording of some useless document. An 11% slowdown is going to cost them seconds on a day, not minutes. Certainly not 53 minutes.

Now don't get me wrong, I hate Windows with a passion, but isn't this the same kind of FUD we've been laughing at for years, just going in the other direction?

(Besides, who works an 8 hour day?)

Well it *IS* slower in many cases.

[tcc]

It's not really supprising, I've installed XP on my dual celeron 400 system and a week later, I've trashed it and reinstalled win2k.

There are 2 issues here. The HOME version and the (supposely) PRO version.

the PRO has WAY too much "take me by the hand I am a complete newbie" stuff in it, too much monitoring, too much popups for crap that I should do myself anyways. While I can understand this in the "home version", it's PERFECT for home, it's nice looking easy and made for newbies... but the professionnal version really is NOT aimed at professionnals that's for sure...

heck I can see myself if I need updates
I can see myself if my drives are full
I don't need any importer utilities to copy files from my previous version, even less converting my Mp3 to WMA
I don't need a stinkin popup window everytime I throw in a CD-R or RW.
Etc...

Oh and for those who will say "you can turn them off too you know?" Well, I don't need these options turned on by default and having to turn them all off just to be less annoyed right off the start, it should be the opposite, no? (again for the pro version, not home).

I don't want this to be a flame, even if it sounds like bitching, but if you look at the points mentionned, it's weird that a "pro" version has all of these little annoying things, people that will buy pro are used to NT/2K environment (usually) so why would he downgrade to the "clippy-age" when he upgrades?

As the old saying goes...

[ptomblin]

Grove giveth, and Gates taketh away.

XP isnt slower, Windows Networking is Faster/smart

[BrookHarty]

I installed XP on my work laptop, 650mhz p2. No slow down if you turn off the pretty gfx, its the same speed. Friends with older PC's have told me its slower and have stayed with win2k, ymmv.

But windows networking FLYS compared to my win2k. I can open network domains with 10000+ pc's and it only takes seconds now. Printers and shares now remember the passwords. I can log transparently into a domain for printer shares only. Network login is actually faster now. FTP transfers are the same speed thou. I dont like to log into the domain, but It authenticates me for printers and exchange.

Only crash I'ved had was the 3dfx driver I have in my docking station, disabled the onboard ati card and no problems. The reason I run a 3dfx pci voodoo3, its pci half-height, and does 1600x1200.

On my Home PC, dual 800, I left the gfx on, and turned off shadow menus, that was the main slow down. Only crashs are the geforce nvidia driver (28.88 with newest gf2mx bios, god love those russian unreleased driver/bios sites)

Configuration?

[yamla]

This document is meaningless without knowing more about their test setup. They list the CPUs of each computer but do not mention how much RAM each has. Consider the possibility that each machine has 64 megs of RAM. Unlikely, yes, but it would explain these results. Or perhaps the hard drives are set up with DMA enabled in Win2k and not enabled in WinXP.

Besides, as has been mentioned already, a system that performs 11% slower than another only means 53 minutes out of an 8 hour day if your CPU is 100% busy all the time.

Have you interviewed a brand-new BSCS lately?

[jcr]

If you had to get a engineering degree in to program, we would have a lot better code out there.

LOL! Do you have any idea how many people are walking around with BSCS degrees, who can't even tell you why qsort beats a bubble sort?

Code is the only cred that matters.

-jcr

Independence test

[wytcld]

"independently of Microsoft: CNET/ZDNET, eTesting Labs, eWeek, PC Magazine, and PC World"

Independently of the big three automakers, ExxonMobile says, "We objectively consider the newest crop of SUVs to be the most wonderful ever, and urge their immediate purchase by all."

ridiculous extrapolation

[macpeep]

I'm not going to argue over the 11%. It may or may not be true but let's assume it IS true for a minute.

A user spends most of his/her time writing emails and documents, surfing the net etc. where the computer more or less idles all of the time. Even if a certain operation is 11% slower, you can't extrapolate it over the course of the day and come up with a figure of 53 minutes. I'm sorry but that's just idiotic. Most users will lose a few seconds per day over this! If there's even a small increase in productivity due to other things like, let's say clearer and simplified dialogs, faster access to your documents due to thumbnails in the file manager or something similar, it MORE than makes up for the lost couple of seconds.

Articles like this are really pathetic. You know, you don't HAVE to post something negative about Microsoft EVERY day.

Malda Flamebaiting?

[Null_Packet]

I honestly have to wonder how many more of these stories Rob is going to continue submitting on this same line of articles. We have seen over the last year or so a steady increase in these kinds of articles by standard Slashdot Editors, and I have read a strong increase in support for Microsoft on Slashdot, strangely enough. I don't mean to imply that all or even most of Slashdot's readership is MS-biased, but I think Malda is letting his own bias show. Most Engineers who get frustrated with a particluar release of any software package vent by the water cooler, but I think Malda is venting via the articles he chooses, which shows a poor display of bias.

XP Launch

MS FrontPage

MS Loses Delay Appeal

Whether or not this post is modded up, I hope CmdrTaco takes notice that while he has founded and continues to heavily influence one of the best Tech-News Sites ever made, he needs to keep some kind of restraint. I'm not defending MS, but rather trying to promote the idea that you don't sit around all day and bash something you don't even use. I could understand if Malda was teased all day for running Linux in a Windows Shop, but I would guess that it's typically the other way around. When was the last time you even saw XP in person, Taco? or 2000? I don't post criticisms about the drivability of Ferarris and Saabs, or even Peugots- why? Because I have contact with them, and I don't consider myself to be anywhere near an authority on them. Maybe this kind of consideration should be taken to newer windows products with some of the Slashdot editors.