Windows 7 will be out tomorrow (Oct 22). I can't wait.

Well, I am looking forward to it, though the RC works well enough that I'm loathe to reinstall. Here's hoping there's a way around that.

The only thing I care about is how this will affect WINE and how it will run in an emulator.

LOL at those benchmarks:

â¢How long does each operating system take to install?
â¢How much disk space was used in the standard install?
â¢How long does boot up and shutdown take?
â¢How long does it take to copy files from USB to HD, and from HD to HD?

I omit Richards Benchmark because Windows 7 beat Linux on that one.

Seriously though. Installation is a one-time thing. If it takes longer for Windows 7, big deal. Startup and shutdown are a non-issue if you just put the OS to sleep. Someone so concerned about 15 seconds of extra startup or shutdown time could just put the system to sleep.

That leaves large file transfer to and from USB. I can't imagine why Windows would take twice as long as Linux but it would be interesting to know if this is still a problem considering that this test was done on a 9 month old beta.

Just sayin'

Dave, LOL that Windows can't beat Linux on a single benchmark. These benchmarks use all of the basic functionality of the system (well, OK, not the install, but all the others).

Regarding the age of the OS being tested, that age-thing applies as well to Ubuntu, for instance, which has improved in all of those benchmark areas. Has windows improved in all those benchmark areas? I doubt it very much. LOL.

You're kidding right? Those are meaningless benchmarks for most people who do anything other than install their OS, startup or shutdown or copy porn back and forth to their thumbdrives. My point about the age of the OS is that this could simply be a driver issue. If Windows can beat Linux copying a file from HD to HD then there's little reason why it shouldn't be able to do the same to a USB drive. That's why I'd like to see this test done with the latest version. This could simply be a driver issue.

There are dozens of other benchmarks that would be more useful than these. How about graphics performance? Application performance? How about code compilation? How about any of the zillions of other things that people *actually* do with their computers.

I'm sorry but seeing an article like this at Greg's page seems almost like satire. Granted, his post could have been sarcastic. If so, I apologize to Greg.

One thing I did find amusing was a statistic that Windows 7, even before release, had nearly the same market share as all versions of Linux. I'll try to find the source of that.

I don't dislike Linux, FWIW. I respect it and use it. I just think that Linux and Windows users should be united against the common enemy; Mac users! ;-)

As a non-guru, I appreciate the fact that my Linux machines almost never strand me on the wrong side of an hourglass for a half hour while some mysterious thing happens in the background. Downloading updates? Keep working. Running three desktops at once? Keep working. Whole bunch of windows open at once? No stress, it just keeps working.

And yes, installation time matters in Windows, because you have to do it so often.

Personally, I'm waiting for the Google Chrome OS next year.

When they port Call of Duty Modern Warfare 2 to Linux, give me a call.

David, no I am not kidding you. It really is true, for starers, that people use their computers for reading writing and modifying files. That is not an unimportant bench mark. It really is true that people care about how long it takes to turn their computer on and off. And so on.

No, it is not a driver issue, that is absurd. There is a cluster of benchmark values for numerous flavors of Linux and a cluster of benchmark values for various flavors of Windows. Windows numbers are all clustered together in the "inadequate OS" range. LOL.

Code compilation? You want to use a benchmark for code compilation instead? Name one average user who has ever compiled a single line of code.

George good point.

Les: Version 1 is working fine in Linux. Version 1.7 is being worked on.

http://appdb.winehq.org/objectManager.php?sClass=version&iId=9699

"average" users don't use Linux.

People who do use Linux are much more likely to compile source code than Windows users.

Again, if someone needs fast shutdown or startup time, they are more likely to just put their system to sleep.

I want benchmarks for things that I'm doing in between startup and shutdown. We can multi-task while our computers are starting up or shutting down. It's not like we have to just sit there wasting our time. I don't see this being, even remotely, something that someone would care about enough to choose Linux over Windows.

"It really is true, for starers, that people use their computers for reading writing and modifying files."

Don't think I addressed this properly. Reading, writing and modifying files is a very important benchmark. Benchmarks for copying files to and from USB drives or from HD to HD are not valid benchmarks for reading, writing and modifying files. Again, Windows won for copying HD to HD. Why it was slower to a USB drive is something I'd like to find out about. If I find an answer, I'll post it.

Dave, average users do use Linux. Friends of mine are giving their old PC to someone who refurbs them as email boxes for grandmas in North Dakota and runs Linux on them. It's just that Linux users are more likely to be expert users.

I appreciate your input, Stephanie. However, average users are, essentially by definition, not experts.

My livelihood is software engineering. When I started using computers, I was storing my programs on cassette tapes using an Atari 400 back in 1981. I don't know a single person who uses Linux the same way that someone would use Windows. That is, the people I know who use Linux do software development or IT. These are not average users.

The average Linux user is indeed more of an expert than the average Windows user. This means that the majority of people using Windows are probably not going to be able to use Linux very well. Things have improved for Linux a lot but it seems that the way Linux is improving is to become more like Windows. Now, I'm all for that. Especially since it will allow people who can't afford Windows to do useful work on their computer. It also lights a fire under Microsoft to do better.

I still stand by my points. These benchmarks are really weak arguments for Linux. The things that Linux beats Windows on are things that average users don't need to do. That is, as web servers or database servers and that sort of thing.

OK let's take this poll. How many of you reading this

a) use Linux and compile code on it

b) use Linux and don't compile code on it

c) don't use Linux

Me: B ... I use Linux but I have not ever compiled code on it.

And I am an expert in many areas of computer operations. I am also not average and I am also not Dave it would seem.

All that I know is that Karmic Koala release is next week. I am an average user. I am not a programmer, an IT person nor a software developer and I am a nearly exclusive Ubuntu user (excluding one machine I bought for a poorly designed app that I have to run and which only runs on Windows,)

I am quite comfortable with Linux, Dave, especially in installing great apps without paying for them unless I choose to donate.

a) use Linux and compile code on it

In addition to a Linux autobuild server that builds my code and is managed by our IT staff, I have Linux running on an embedded system that I can compile on and I have an Ubuntu Linux system that I also compile on (different code). I guess that being in the software industry, I would of course know more people using Linux for building software. However, considering the fact that I can't think of a single non-developer, non-IT friend that uses Linux says something too.

Mike and Boris, have you ever installed software from source which required it to be compiled? I know that many apps are now distributed as binaries but it doesn't seem that long ago that most Linux software was distributed as source so it could be built on any platform.

The average Linux user is indeed more of an expert than the average Windows user.

Ex dad-I/L had XP on his PC, no idea about computers, bit senile, had to reinstall for him every month because he got it broken by viruses, downloading crap etc.
Installed him Kubuntu 2 years ago,showed him how to use Firefox, it's been running since then without any problems.

Just another datapoint against your argument...

d)still doesn't understand even what linux is. ;^}

> "average" users don't use Linux.

Talk to my wife, who converted to Linux some years back, and now installs Fedora Linux on her own, thank you. I'm the technical one in our house, I installed the first few versions. But after that, she realized the Linux install process was "just clicking buttons" and does it on her own now.

Or talk to my mom. She's been a happy Linux user for some time, too. Again, I installed the first one for her ... until she moved out of state, and she just figured it out on her own from there.

There's a university student I know who is getting a degree in history & economics, and he switched over to Linux. "Because it runs better." I didn't convince him, he just decided to do it over the summer break, as an experiment, and loves it. Not a geek, just an "average" user.

So there's three examples against that argument.

Different Dave here, b) uses linux(Ubuntu) compiled once, because I had no choice, not done it for 3 or 4 years.

Having just acquired a laptop with Vista pre-installed, and set up a dual boot with Ubuntu, I can say that the review of installation times is meaningless. It took about 10 minutes (maybe less) to install Ubuntu. I then spent 10-20 minutes finding 64 bit drivers for the video card, and for flashplayer, After say, an hour, the system was up and running and doing useful work.

I then spent *hours* with Vista, installing a free anti-virus program, fighting Norton utilities, removing miscellaneous crap like spyware and AOL searches, kid's stuff, spurious 'helper' software, and all those 'trial offers' of software I can get for free with OpenOffice.

Vista works better for non techies, becauuse it hides what's going on. But when it goes wrong, and grandma can't find where she stored those photos, who's she gonna call?

OCT 22 IS THE INTERNATIONAL CAPS LOCK DAY! CELEBRATE!!!

I've had idiotic coworkers swooning over Windows 7. I mean really, isn't it just basically a bugfixed Vista with a new coat of paint, and a horrible Control Panel?

As long as XP still works fine for my gaming, I'm still dualbooting Linux/XP.

b said:

> "As long as XP still works fine for my gaming ..."

Earlier in the thread, Les said something similar about COD4 not being ported to Linux. Why insist on Windows to do your gaming? I used to do a lot of Windows gaming up until about 1998. Then I realized it was a lot of hassle to keep my gaming rig up to date with the latest hardware, SLI cards, etc. It got too expensive when all I wanted to do was just sit and play games.

So I bought a game console instead, and haven't looked back. Seriously, game consoles are where the great games are. If it's available for Windows, it's available for consoles (PS3, XBox360). And you spend a lot less, because you aren't upgrading the damn thing just to play a game. More time gaming, less time doing "system stuff."

b.) Use Linux and do not compile code (with exception)

The few compilation programs I use are EDA tools for work that only exist for *nix systems, so there would be no comparison anyway.

Another mostly-b. Used Linux (Ubuntu, Linux for mortals) until I accidentally blew up my partition.

Compiled code once, following strict instructions. Found the process of downloading all the libraries and parts and compiling them and getting missing dependencies and forcing install heinous and unfriendly when compared to the Windows version of the same software. Installation on Windows took one day, Ubuntu took a month when I was unemployed and could work on it full-time.

This is why I dismiss some assertions made by open source evangelists, as in the Collective Imagination post Greg linked to a while back (http://scienceblogs.com/collectiveimagination/2009/10/james_hall_open_s…) "Users should be developers?" Seriously? Users may contribute, but should not need to be able to write the software to be good users. The people who figured out that the average user is now Grandma, like the people who put out Ubuntu and Firefox, are doing more to spread OSS than any number of experts who won't leave the command line.

[I know, the command line crack is a big strawman, but when you've been told that it's ok to say software works on Windows because "nothing prevents them from changing the sourcecode to make it work"...the real world has slipped away.]

My god, it's like arguing with creationists here. Claims that Linux won all the specified benchmarks when it clearly didn't. Claims that because some relative or friend heard a voice from God, I mean, is an non-expert Linux user and that proves that the average Linux user is as unsavvy as the average Windows user. Seriously? Given the miniscule percentage of Linux users out there compared to Windows and you're going to argue *against* the idea that they are more of an expert at it? It smacks of desperation to promote Linux as a viable alternative to Windows. For some users, it may be. I'd agree with that. I mean, if all you want to do is edit a document or spreadsheet now and then and browse the web, I'll admit that Linux is just fine.

It may be 100% true that Linux can be as easy to use for the average user as Windows. It doesn't take much web searching to figure out that Linux is still struggling to win over the average Windows user and that they're doing it by trying to be more like Windows.

There may be many reasons why someone should choose Linux over Windows. The idea that these benchmarks cover those reasons is ludicrous. Startup and shutdown time as a determining factor for choosing an OS? I can't believe that this argument is even being put out there. Installation time as a factor? Really?!?! How about a benchmark for how many applications are available to the user? An operating system is nothing without applications and Windows is the clear winner in that arena.

Let's face it. The real reasons to choose Linux over Windows are: they ain't Micro$oft (gotta hate the evil empire, right?) and it's free (or, you're running a web server or something).

It's obvious that Greg posted this entry to get a rise out of people. I can't believe that he would think that these benchmarks support the claim: "Sorry, Windows, you're just not that good and probably never will be."

It's okay, Dave. You're still hardcore even if regular people can use your operating system too.

Seriously, creationists? You were the person who said, '"average" users don't use Linux.' You were wrong, and people told you so, with examples. Deal with it gracefully instead of assuming that your anecdotes must be more correct than everyone else's.

As for apps, most of those average users will be using an email client, a browser, and maybe OpenOffice. That's what average users do.

Um, examples of average users that use Linux are not counterexamples to the statement that "the average Linux user is more of an expert than the average Windows user". You can't have counterexamples to an average; those cases are already included in the average.

I am in category A: I use Linux and compile code. I also use Mac OS and compile code. I prefer Linux for code-compiley tasks, but Macs for general use. I used to use Windows, but I have been completely off it for almost a year now.

No, lylebot, the examples are responses to comment 12--and in response to Dave's general attitude of "Oh, Greg, how could you post an item like this?!? You're not laying out the complete story in your one, singular blog post that's really a link to a couple of other posts!!!"

Here's a question: there's basically one LaTeX GUI editor for Mac OS: TeXshop. For Linux there seem to be a lot of choices: Kile, TeXlipse for Eclipse, KDevelop, TeXmaker, ... There's nothing stopping anybody in either case from writing their own GUI, but Linux users have obviously taken it up with far more gusto than Mac users. Why is that?

The same question could be applied to a lot of different tools. One choice for Windows, one for Mac, lots for Linux. I have to admit, I prefer just having one (or two) choices for most applications, assuming those choices are good enough to do what I need to do. Windows choices tend not to be good enough, but Mac choices tend to be pretty good. Linux ones tend to be all over the map in my experience.

So Stephanie, why are people still replying to comment 12 even after the refinement in comment 15?

Never mind, I see that's really your refinement. Sorry! "Average" users do indeed use Linux sometimes.

It is exactly like talking to creationists. Windows users have no really good reason to prefer Windows over other systems like Mac or Linux. They've bought into a bad system, drank the kool aid as it were, and are quick on their feet justifying their choice. If those benchmarks were a different set of tests, then THOSE tests would be the ones that are not relevant to the decision. "You've gotta love the evil empire" is roughly equivalent to "Evolution is just a religion." And, the standards of rhetoric, modalities of conversation, types of humor, and criteria allowable for personal choice are controlled by the side making the creationist-like argument (the Windows symps, in this case).

Is that what you meant, Dave?!?!? :)

Linux is clearly the superior system and has been for years. No one seriously doubts that. It is true that Linux is not for everyone. Many people can't use Linux because they are punished if they do. Others because they don't have the balls. Others because they prefer Macs and that's a good system too.

And then there are all these others who use Windows but don't have to. They're just dumb-asses. But Linux is not for dumb-asses. Not because dumb-asses can't use it ... it is quite possible. Rather, we just don't want them to. We prefer to be uncontaminated, pure, free of such things. Life's better that way.

Reposted from a comment on slashdot:

Imagine seeing an old girlfriend suddenly show up on your doorstep wanting to get back together. She's pretty, of course, but too pretty. She wears too much makeup and carries that desperate look in her eyes. The fragrant haze around her is the perfume she overuses to mask the scent of failure.

But standing there in that low-cut top, you'd almost forget for a moment what a psycho she was- how she used to shut down in the middle of a date and forget everything you were talking about and how she was only happy when you were buying her things. You'd almost forget about carrying around her legacy baggage or those nights when, for seemingly no reason at all, she would simply stop speaking to you and when you asked what was wrong she'd just expect you to figure it out.

You complained about her for years before finally deciding to get rid of her. But here she is again. Though, somehow she seems like a completely different person now.

Tempted though you may be, you know that over time she'll get bored and slow down on you just like she always does. And then you'll be right back where you started: trapped. She keeps you by convincing you that you don't have a choice. You're just not smart enough for one option or rich enough to afford the other.

"But I'm different now," she says, batting her eyes innocently. "I've changed."

Indeed she has. Apparently, she's really into Cabala now or something like that. It's helped her discover loads of untapped potential in herself. But it also means that you'll have to buy all new furniture to fit with her understanding of feng shui. That's not the only change she has in store for you. The minute you let her move in, she'll have a new alarm system put in that succeeds only in preventing your friends from coming over on poker night, but your TV gets stolen anyway.

She doesn't love you, but she doesn't hate you, either. The truth is that she couldn't care less one way or the other. She's here because she doesn't want to be alone. Like all human beings, especially those well past their prime, she wants to feel wanted and - after a string of lost jobs and bad investments - she needs a place to stay.

But all in all, she's OK. She's a 7. She'll do, I guess.

That's Windows.

Meanwhile, as you stare in disbelief at the girl at the front door, your thoughts wander to the love of your life, who is at this moment in the kitchen dishing out some ice cream for you to share. And checking her phone messages and writing a list of thing to do tomorrow and reading the newspaper all at the same time. She's a true multi-tasker. And always has been.

Today she's dressed in subtle earth tones and has the comforting, home-spun look that you usually associate with her, but memories of yesterday, when she seemed to be living in a totally different skin, where it was a little harder for you to find what buttons to press, but wow, what a result when you finally found them... And you think how fun it is that her whole personality can change across such a wide variety. Other girlfriends can't do that.

And you think how your new girlfriend wakes up bright and cheery in only a few seconds every morning, and goes gently and quickly to sleep without asking a lot of dumb questions every night, how she makes no demands on your time or pocketbook when she moves easily and cleanly from one task to another. So, you havn't quite figured out how to watch all the possible movies together, but you never have a total breakdown just because one thing or another goes a little wrong.

And you understand her. It's like she's made of basic, reliable, simple components with a handful of straight forward but powerful abilities that can be put together to do almost anything. You never worry that if you are doing something important that some other task will make her forget what is going on and have a total breakdown. You know that if you find something going wrong, you can sit down and work out, with her, how to manage and how to make it all nice again by looking as deeply inside her as you need to using simple methods.

And most important of all, you know that she will never, ever lie to you.

@flynn (#27):
This is why I dismiss some assertions made by open source evangelists, as in the Collective Imagination post Greg linked to a while back "Users should be developers?" Seriously? Users may contribute, but should not need to be able to write the software to be good users. The people who figured out that the average user is now Grandma, like the people who put out Ubuntu and Firefox, are doing more to spread OSS than any number of experts who won't leave the command line.

Hi, you're referring to the Open source software in the real world guest post I wrote. (Part 3 should be up soon.)

If all you took away was "Users should be developers" then I think you missed what that section was saying, and probably the whole point of the article. This guest post discussed the features a free/ open source software project needs to have to get off the ground. (Part 3, should get posted soon, will talk about things a F/OSS project needs to do to be successful over the long run.)

The "Users should be developers" section you mention starts with:

The basic definition of open source software is that the source code must be made available for others to see it. A necessary side-effect of this condition is that anyone who uses the program has an opportunity to make improvements. A well-managed open source software project will accept any improvements in the form of patches, which modifies the program to solve someone else's slightly different (but similar) problem. Releasing new versions of the software with the new features ensures that everyone benefits from these changes.

To emphasize: source code must be made available for others to see it, so that anyone who uses the program has an opportunity to make improvements.

That post goes on to use the example of writing the FreeDOS Kernel. In the beginning, everything was written by one developer (Pat). When we released the source code under the GNU GPL, everyone could see it. Those users who were also developers looked at the code, and began to contribute bug fixes, and patches to add other functionality. FreeDOS quickly improved performance, added network and CD-ROM support, LBA, etc.

All this was possible because we enabled our users to be developers.

However, this does not mean that your users must be developers. We have lots of people contributing in FreeDOS who are not developers, who know nothing about "compilers" and "debuggers" and "source code". They just want to run a DOS system. Some use FreeDOS as a dedicated system (for example: building an arcade cabinet using the MAME software), and others run FreeDOS so they can play old DOS games. Nothing there about compiling code to get things to work.

This whole "you need to compile stuff to use Linux" is a myth. I run Linux on my laptop, but I haven't compiled anything on it. My wife runs Linux, and has never compiled a thing. It's not that I don't know how (I think I've demonstrated my developer cred ... see also GNU Robots, GNU Emacs, Atomic Tanks, Freemacs, Cats/Kitten, etc.) but that I just don't need to compile anything to use a Linux system.

Not sure what you were trying to compile on your Ubuntu system, but my guess is that you could have just installed a pre-compiled package for whatever you were trying to do, rather than compile it. If you're a developer - sure, I can see you trying to compile the code from scratch so you can see how things go together, then contribute a fix for something. But if you're not a developer, or just not interested in compiling - don't compile, man. That's why there are pre-compiled packages that are easy to install.

Oops. I had a typo there: "We have lots of people contributing in FreeDOS who are not developers ..." should read "We have lots of people using FreeDOS who are not developers ..."

OMG - they have converted Linus! http://picasaweb.google.com/cschlaeger/JapanLinuxSymposium#539540000045…

Jim,
Thanks for the response. Soon after I posted my comment, I realized that my interpretation was from the perspective of the user, when you were writing from the perspective of the developer in a role as project manager. That does change things, and I'd believe that a project that never allowed or attracted developer/users would suffer from that lack of participation.

I'm pleased to agree that, as all here have testified, we don't need to compile software at all in most cases. This is strikingly different from the time just a few years ago when a friend installing BSD had to go find a keyboard driver and recompile. How does someone without a working keyboard get a keyboard driver? The leap forward in usability from that to my Ubuntu machine in such a short time is a good sign.

The time I mentioned that I did need to compile was while installing a program used only by my discipline. There was no pre-compiled package at the time and may not be now. Maybe not so rare in that kind of situation, but hardly representative of daily, non-work computer use.

James @ 25:

I'm really disappointed to read your "if you want to play games, buy a console" comment. I keep reading that one of the biggest reasons to use Linux is to get away from having somebody else (usually Microsoft) tell you what you can and cannot do with your computer.

I LIKE to play games on my computer. I've been trying Linux intermittently for the past couple of months and have had almost zero success in getting my games to run in it. I've pretty much given up on Linux gaming for now. Sure, it's much easier when I'm at work using multiple RDP sessions in Linux than in Windows, but when I get home, I have little reason so far to fire up the Kubuntu part of my dual-boot with Windows.

When I finally get around to firing up a mail or webserver at home, I expect to do it with Linux instead of forking out a thousand dollars for Windows Server 2003/2008. But for home computing, Linux doesn't really suit my needs right now.

Dave: How do you wish to compare compile times in different operating systems? In a bench mark test? Different compilers are different software, and correct compiles use different parameters for different systems. There is no possibility of a fair comparison.

It's sad that an interesting discussion of operating systems should degenerate into insulting genderist comparison. Of _course_ women are only any good when they're simple, fulfill every need for you, and get you ice cream. It's kinda like being barefoot, pregnant, and in the kitchen, isn't it, boys?