Is Windows 7 Faster Or Just Smarter? 619
Barence writes "The Windows 7 unveiling garnered largely positive coverage, with many hands-on testers praising it for being faster than Vista. But is it actually? To find out, this blogger ran a suite of benchmarks to see just how much quicker Windows 7 really is — and the results weren't quite what he expected. 'The actual performance gap between Vista and Windows 7 is ... nada. Absolutely nothing. Our Office benchmarks and video encoding tests complete in precisely the same time regardless of which OS is installed. [...] It's tempting to see this as a bit of a con. They've sped up the front end so it feels like you're getting more done, but in terms of real productivity it's no better than Vista."
Re:A bit smarter would be welcome (Score:1, Informative)
Shlemiel algorithm [joelonsoftware.com]
Joke? (Score:3, Informative)
Of course video encoding and the same old office build won't be affected by the OS.
What people want to be faster is booting up, logging in, connecting to networks, detecting hardware and installing drivers, and running those damn .msi installers.
Re:Trick Question (Score:3, Informative)
Wrong. The correct answer is mu [catb.org], and you are a moron.
Re:It looks just like Vista now because... (Score:4, Informative)
Since 7 is still a year or so away at this point they're just showing you mostly user interface changes with little or no changes to the core underlying os.
Excuse me?
A year or so away from release, they should be done with fundamental changes to the core OS. They should be working on the details and polish, before they send it of to testing and QA.
Of course, they'll probably pull a main component two months before release when it turns out it'll never work anyways, re-write some stuff and not have enough time for testing left.
Re:Where was the real improvement (Score:3, Informative)
Although I don't think that was his intent, all too many people DO think that way.
Hence they end up having a subpar system because of subpar hardware (subjectively).
It would be like trying to install Win98 or ME on a 386DX40 with 8 meg of RAM. 5 to 7 years before 98 was released, that was a decent system. BUT, we where all running pentium class machines by then (even the 486DX2 systems had spec's similiar to the pentium 66+ at LEAST.. And when we figured out that bus speed made the BIGGEST difference, our 5x86 systems with an 86 mhz bus speed where as fast or faster than a typical Intel Pentium, and we didn't have math problems :) ).
So his point is valid, and one I've been making for about a year. Vista works fine, with vista qualified hardware. The minimum MS guidelines have always been a little lean on performance, but if you go out and purchase a laptop with Vista on it, then format it WITHOUT all the bloatware, it's a fast, lean, working system.
I know, I did it.
--Toll_Free
Running X64 and X32 Vista, and liking it. Just as fast as my Ubuntu system
Re:Smarter not harder (Score:3, Informative)
I guess it could be painful if you don't know how to reinstall grub or lilo from a downed box and have no way of looking up a document on how to do it.
IME, you just need to boot off your linux install media, go to the recovery mode, enter the command for your bootloader that tells it to reinstall itself to your hard disk then edit the boot list to include windows and your set. Its really nothing too complicated for someone who will be dual booting...
Its always been easier to install windows first because the linux boot loaders make setting up dual boot an automatic process (for the most part). The process to fix windows is just the same if you mess up its boot loader as well. Boot from media, recovery console, and fixboot and/or fixmbr.
Re:Productivity originates from the users percepti (Score:5, Informative)
Unsurprisingly, the German magazine C't has just compared XP, Vista and the new one.
They came to a different conclusion - W7 is noticeably faster than Vista and roughly the same as XP.
They found no difference on laptop Battery life between Vista and W7 though.
When it comes down to C't and some blogger, sorry - I'll take C't. Those guys take independence very seriously.
Re:Mp3 Locking? (Score:1, Informative)
http://kottke.org/98/11/my-mac-sucks [kottke.org]
Re:Perhaps this alpha releases uses Vistas kernel? (Score:4, Informative)
I hate to contradict you sir as your points _should_ be entirely true.
Sadly the number one complaint from my users (around 10000, at least 400 complaints) about the new locked-down desktops we installed 18 months ago is that they cannot change their wallpaper (screensavers as they call them...) And to top this, one of they key reasons we chose to force this on them was the sheer amount of work we had to do removing clunky screensaver managers or custom cursors and icons (not to mention the virii accompanying many of these).
I often wonder if my father (an engineer) had the same problem with people wanting to customise their wrench so it had blue handles?
Re:Faster interface = improved productivity (Score:3, Informative)
Re:Worse than that. (Score:4, Informative)
man screen, it does wonders on ssh especially on a slow line.
Re:Productivity ... Really? (Score:3, Informative)
The file copying slowness was actually related to something new in their networking stack which I believe was fixed in SP1. If I wasn't lasy I would try to look it up and find links for you. I use a Vista box for copying large numbers of files between servers because the copy dialog actually provides useful information, doesn't cancel on simple errors and the speed seems the same to me. (If it's a LOT of information I use a backup utility).
Ah, wait 5 minutes before posting a comment....so I got a link for you:
http://mytechweblog.blogspot.com/2007/04/slow-file-copymove-in-vista-here-is_05.html [blogspot.com]
It mainly talks about the fix and not the cause, and I remeber reading somewhere that SP1 flips that off by default.
Re:Worse than that. (Score:3, Informative)
Oh you're right it was rxvt not xterm that was indeed the fastest. Here's some simple performance numbers I got:
yes | dd of=/tmp/y.txt bs=1024 count=10240
10485760 bytes (10 MB) copied, 0.271632 seconds, 38.6 MB/s
(for calibration purposes)
dd if=/tmp/y.txt of=/dev/null
10485760 bytes (10 MB) copied, 0.0366388 seconds, 286 MB/s
gnome-terminal: dd if=/tmp/y.txt
10485760 bytes (10 MB) copied, 22.3385 seconds, 469 kB/s
xterm (jump): dd if=/tmp/y.txt
10485760 bytes (10 MB) copied, 57.5998 seconds, 182 kB/s
konsole: dd if=/tmp/y.txt
10485760 bytes (10 MB) copied, 32.9735 seconds, 318 kB/s
linux text console: dd if=/tmp/y.txt
3971584 bytes (4.0 MB) copied, 288.027 seconds, 13.8 kB/s
(I aborted it cause it took too long)
rxvt: dd if=/tmp/y.txt
10485760 bytes (10 MB) copied, 7.40936 seconds, 1.4 MB/s
This is on FC6 with a cheapy Quadro FX 550 and an intel core2 6300@1.86GHz.
My initial statement that xterm was faster appears to be wrong at least on this hardware.
But clearly rxvt is doing something right that no one else is doing.
Re:Mp3 Locking? (Score:2, Informative)
Re:Tip for you: (Score:4, Informative)
Interesting. When I used to use the dev snapshots of Beryl they were lovely and quick (AMD64-3200+ with GeForce 6800). Unfortunately either the compositor or the window manager would crash at least twice a day, so it was a bit unnerving.
On a box at work (Old P4@3Ghz, crappy GeForce 5 series) I can notice windows redraw, tabs pause before disappearing, the works. It is very annoying, although I suspect that the driver is not installed properly and I don't have root on that box. I also don't have the time to explain to our sysadmin how to fix it :(
I used to use BlackBox a few years ago. I love that snappy feeling, and a few milliseconds in the wrong place completely destroy a user interface. But at work I put up with the slow annoying rendering - because metacity is rock solid and it doesn't crash on me ever.
At home, I ditched linux the day that I decided that I wanted a unix laptop with properly working hardware support (like hibernation that works the way it should do). So I bought a mac :) I agree about the perception issue, there are exactly two ways that it can be correct: instant snappy response, or smooth transitions in-between. Any noticeable redraw/overdraw lag is too much.
Re:Worse than that. (Score:5, Informative)
Re:Mp3 Locking? (Score:4, Informative)
I'm actually the opposite. I always heard comments like yours (maybe you're astroturf from MS marketing?) and thought it made sense. Other than the DRM quirks, I figured Vista would be at least as good as XP, with no real problems outside the removal of backwards compatibility.
My first experience with Vista was when my grandfather-in-law bought a new computer and couldn't run any of his small business' software. He had to downgrade to XP.
Then I got a new job where the boss required us to use Vista. My coworker (started same day I did) was totally against it and only grudgingly gave in. Personally I was excited to give Vista a try and shoot a hole through all the BS whines/complaints.
I found out that Vista sucks. REALLY sucks. It had crashed and bugs I had never even heard of. Because I was working, I only even used a small portion of Vista's software. Yet very basic things like FTP, Windows Explorer, and Remote Desktop had huge and glaring flaws that made working a painful experience. The UI in general is abysmal. The very first day I was installing software I ran into UAC popping up prompts behind other windows, for instance. "Hmm, why is this installer frozen?"
I'm sorry, but Vista really is *that* bad.
I'm now working at another software development company. My brand new Dell laptop they gave me has a Windows Vista sticker on it. It's running Windows XP. Same goes for my wife's new computer from Dell, except I had to install XP myself because it costs $100 extra (on a $400 laptop!) to get XP instead of Vista.
Re:Worse than that. (Score:5, Informative)
Yes. I doubt that the CPU or GPU are of differing quality. Except, of course, that the Mac has a much faster CPU than the eMachine, which is also where some of that cost comes in.
Uhh. NO. While Apple's case designs are *much* better than most third party cases that are available, they still use Seagate/Maxtor HDD's (just going by some G3/G4/G5's I've had my hands into), and standard off the shelf RAM, even off the shelf CD-ROM's.
Continuing...
I built my own PCs for a long time, but finally got fed up with low quality, buggy components. I don't have spare RAM, mobos, and PSUs lying around to troubleshoot crappy hardware to figure out which piece is bad or incompatible with some other random cheap crappy piece. I still buy PC desktops, but my last several computers have been barebones kits from Dell and Shuttle. When I buy RAM it's always from a reputable company, as I've had as many unusable sticks of Kingston and the like as have actually worked.
It is not the "PC's" fault that you purchased bottom-barrel crap components. Know what Apple uses for motherboards? Slightly modified Intel parts (again, going from what I've seen in some newer machines that went to surplus at my place of work because of how expensive they'd have been to repair). I've never purchased motherboards (even some bottom-barrel priced boards), memory sticks from various companies (though I've settled on G.Skill lately), or other parts that were "buggy". Yes, I've had an occasional DOA part, but that's what warranties are for. Just going by your own words, I'd say you are one of those "Yea, I can be a PC-Tech here for (insert company), because I built my own machine at home!" people I see all over the place. The kind of person that knows *just* enough to assemble a machine, but not enough to make sure all the parts your ordering/spec'ing for a machine will actually work together. This isn't the fault of the aftermarket parts producers -- it's yours.
And to suggest that they're the "same thing" as an $200-$400 eMachines or Dell is a complete farce.
Actually, they *are* basically the same thing. Like I said, you get an awesome Apple designed case, but the parts inside are basically off the shelf PC parts. Hell, I bought an Intel 975xbx2 mainboard that, with pcefi, worked perfectly under OSX 10.4 and 10.5 with vanilla kernel/kexts/etc (of course I did have to keep the couple modifications needed to bypass checks for the silly Apple ROM's, etc).
Having re-read what I typed, it sounds an awful lot like I'm harping on YOU instead of your argument, however, that's not my intention. I'm just really tired of seeing "Nu-uh! Apple uses much higher quality parts than you'll find in a PC, that's why they're better/more expensive!" arguments. If there were any truth to it, I'd be quiet on the subject (or even blast the "pc makers" for the same), however, it's just not true. Now, maybe when you get to the $200 Bargain Basement E-Machine, maybe (though even still, many of the parts *are* the same).
I think the major bits making Apple machines as expensive as they can be are the case designs and the modifications made to the mainboards (it doesn't help that Apple charges ridiculous prices for RAM/HDD upgrades). If Apple pushed the kinds of volume that, say, HP or Dell did, they might get some better volume discounts on the custom parts...
Re:Worse than that. (Score:3, Informative)
Looking at Vista vs XP, I've got a sense of deja vu from the old OS/2 v2 vs Windows days. In both cases, the newer OS (OS/2, Vista) was internally a better operating system than the predecessor/competition, but suffered a reputation of being slow and unfriendly to use, which generated a stigma against the new OS.
In both cases, the newer OS (OS/2, Vista) actually was a superior operating system, but lost points with users due to UI problems. Both were considered confusing and slow by a large part of the user community, including the tech press. But looking at the fundamentals, like kernel robustness, stability, security, etc., the newer OS actually was/is better. The achilles heel was a weaker user "experience": poor or nonexistant device drivers, annoying GUI features that couldn't be disabled, etc. And in both cases, the vendor produced a successor (Warp, Windows 7) that was basically the previous OS with the serial numbers filed off, but with UI enhancements/bug fixes, rather than base OS changes. Just as Warp was OS/2 v2.2 under the covers, Windows 7 is Vista 1.1, or however the hell you'd represent an incremental upgrade using Microsoft's version number of the week scheme.
I remember when OS/2 v2.0 came out, it was panned for its' horrid installation procedure. This was back in the days when everyone installed their own operating systems, so this was a big deal. For many, it was a deal breaker; people who failed to install OS/2 would try to install Windows 3.0 (and a week later, 3.1), and if successful, became Windows users instead of OS/2 users.
IBM spent a lot of time on this issue, as you can imagine. I installed OS/2 and Windows 3.x more times than I care to admit, logging the time and my reactions to it. And now, we come to the actual point of this post.
The most interesting statistic that I remember was that despite the most common complaint being that the OS/2 install was horribly slow compared to Windows, wall clock time spent was almost exactly the same. Depending on the hardware configuration, OS/2 install time was almost always within 5% of the Windows install time. And yet, Windows installation got glowing reviews as being speedy, while OS/2 installation was a chore.
The difference was the user experience. People installing Windows got coloured screens, filled with information about the new OS ("did you know that..."). Microsoft put enough text on the screen that users read while doing the install. In comparison, OS/2 just put up a dead, black screen, with a textual process bar at the bottom, saying "1%... 2%", etc.
Functionally, there was no difference. If it took OS/2 93 seconds to install drivers for an Adaptec SCSI card, it would take Windows 90-95 seconds to install drivers, as well. But during those 90+ seconds, the Windows user was scrolling through two pages of text explaining some new Windows feature, while the OS/2 user watched a dead black screen, while his watch went tick... tick... tick.
If I interrupted someone installing OS/2, they'd appreciate the distraction. If you interrupted someone installing Windows, they'd usually ask me to come back in half an hour; they didn't want to miss anything during the install. OS/2 users *did* want to miss the install.
I think that Microsoft is doing the same thing with Windows 7 that they did so long ago with Windows 3. They're not changing the underlying OS in any way. They're responding to the user reaction to the OS, and fixing the bad user experience. And that's a good thing.
When addressing the installation issue a lot of IBMers (read: execs) completely ignored what I've said above, instead pointing out that the wall clock time spent installing OS/2 was the same as Windows, and therefore complaints were due to whiners, people prejudiced against IBM, etc. Fortunately, Microsoft appears, at least on the surface, to be responding to user complaints and reacting to them.
Windows 7 will not be simply a new paint job and some detailing on Vista; there will be some new features added. But the base operating system will be the same, just as OS/2 Warp was an improvement over OS/2 2.x (especially in TCP/IP networking). But the end result will hopefully be a win for end users, and that will be a good thing.
Re:Trick Question (Score:2, Informative)
link [wikipedia.org]
Re:Trick Question (Score:3, Informative)
Does anybody else remember when the greatest thing in Windows was After Dark, with it's screensaver of flying toasters?
Yeah, that was developed on the Mac first, too.
Re:Trick Question (Score:2, Informative)
The funny thing there is that After Dark was, AFAIK and IIRC, originally a Mac product...
--- Mr. DOS