Linux Desktop Stigma Draws Ire and Fire

I think a great many more people beyond the many that already do so today will find Linux' freedom very attractive and refreshing. It will take time, but we'll get there. Of course, software patents can put an end to freedom if we don't eventually eliminate them ( http://www.linuxtoday.com/news_story.php3?ltsn=2010-06-17-015-39-OP-MS-SW "Watch Microsoft megatrolls monopolize Linux" )

Some vocal Windows enthusiasts seem to be deeply disturbed by Linux' growing success. Over 95% of supercomputers, 44% of servers (equal to Windows), and a third of netbooks run Linux. Even the most pessimistic of claims shows Linux desktop usage still growing, if slowly, somewhere between 1.13% and 5% of the market. Linux is the fastest growing smartphone OS today between Android, webOS, and the upcoming Meego, and a wide variety of tablets were recently announced to ship with these Linux distros in the coming months.

Windows 7 has been reasonably well received, and Windows Phone 7 has a least received some positive press (it should be on actual devices by this Christmas). It's not necessary for these Windows enthusiasts to be so defensive. But since so many Linux fans were back when we were struggling, I suppose tolerance and patience is the best response today.

I'll always advocate for freedom. But respectful, fact-based encouragement is certainly the best approach.

The dogs bark and the caravan passes on. I, for one, strongly believe that the time spent on defending Linux (and FLOSS in general) as a superior IT solution could be better invested in focusing on how to further the superiority of the solution. I think wisdom would agree with the assertion that most men cannot be contented. But most men can be impressed. I've noticed that even among Apple users (excluding those brushed-aluminum worshipers and adepts of the secret cult of the Church of Stevus the free world use to call "fanboys") even Apple users as I was saying can't help but admire my not-so-fancy Ubuntu desktop (I say ubuntu but it could have been any popular distro) running state-of-the-art software in four degrees of freedom (thing GNU), leave alone my using a bare USB drive to start up a full-fledged operating system after spending less than 5 minutes to install it, distro and tools download time included, on _their_ computer and without writing a single command.

A couple of years ago I used to say "Linux rocks, yeah just give it a try, etc." and everybody said I was a zealot. Now I just open my laptop or boot a live-USB on others' PCs, use my favorite OS and when they say "Wow, what's this ?" I answer "Oh, that's nothing, that's just Linux." Usually, it takes less than a couple of minutes before they ask back "How can I get this on my computer ?". Passive marketing, it works !
Of course, if they want Linux I have to hold their hands in the first couple of weeks but I like it (since I can't contribute code I'd better help code be used.)

It is good thing to share out point of view on global society issues like freedom as well as on technical ones, but in no case should we take part to flame wars and heated debates over trifles like Linux_sux_Windows_rox. After all, it we do our duty in showing the way to software freedom but others nonetheless disagree, then we should respect their freedom not to agree with us and even not to agree with facts and logic. We should even pity them, without condescension, because they are not smart enough to see what's good for them. Stupidity is probably a flaw but definitely not a crime.

Interesting. I just so happen to be a proud member of Linsux. I use Linux on my desktop, I publish Linux how-tos, and I write open source software for Linux freely available to all.

"It's also awesome that these folks have so much free time on their hands that they aren't contributing to society by volunteering or creating something for the good of all. But, they sure love to criticize those who do."

What is awesome is that you wrote an article completely devoid of facts.

These inaccurate statements just make you yet another fanboi, just the uninformed zealot kind.

If you had actually taken the trouble to actually read LHB or linsux, you would see that the vast majority of the posters on both sites are Linux users who are sick of the fanboys and the harm that they do to Linux.

Most of the "Linux hatred" on the internet comes from Linux users who are sick and tired of the fanboys who only use Linux because it has become fashionable in certain circles to hate Microsoft and see their use of Linux as "sticking it to the man", these people generally do not know a whole lot about Linux and are so determined to preach about how much better Linux is than Windows they studiously ignore all faults in Linux and assume that anyone who does criticise Linux is either too stupid to use Linux, is a Microsoft employee spreading FUD or is simply lying. I have been accused of all three of these when I have pointed out obvious flaws in Linux.

Linux on the desktop is a disaster (the fact that so many servers, supercomputers and phones use Linux cannot take away from this) and it will never stop being a disaster until the Linux community wakes up to the fact that it is a disaster and starts addressing the problems instead of trying to shout "WorksForMe(TM)" so load that they drown out any criticism at all. When you have a desktop OS which has millions of fanboys screaming about how wonderful it is then you end up with developers like the Ubuntu devs who believe that they can do no wrong and that it really doesn't matter if they release a system which breaks hardware functionality for a significant number of users on a regular basis.

Instead of responding to sites like LHB and Linsux with comments like
"* By awesome, I mean that they suck.

** By intellectual, I mean that they're not intellectual."
and dismissing everything bad said about Linux as FUD, the Linux community would do better to sit up and listen to what is wrong with Linux.

The article is a good read. Years ago when I began using a "computer" I purchased a Radio Shack Trash 80. It was good learning tool, but I wanted more and that led me to looking at Apple, but the cost was and still is prohibitive and so I opted for the PC. I purchased a NOrthgate 286 PC with 1 meg of RAM, but I was not pleased with the OS. To replace the OS I bought DR-DOS and stuck with that until I upgraded my system to a 486. At this time I was really interested in UNIX and began seriously at purchasing UNIX. It was about this time that I also heard of a new OS that was FREE and I researched it. My curiosity about this new OS was overwhelming and I proceeded to download the OS to nearly 100 diskettes.

I dumped DOS and installed this new OS and found it lacking many things. However, I saw an opportunity to learn more and see where this new OS would go, and of course the new OS is LINUX. There was much promise and I continued to observe the progress of Linux in the computing landscape. It wasn't long before Red Hat appeared on the scene and I was quick to download it. Today I have Fedora 13 and I am very impressed with its ability to discover and install most of my peripherals. My wife now uses Linux and her computing experience is better and a far cry of calling me and asking what the error message in windows was telling her. I can't wait for the new Fedora to come out!

That's like Michael Jordan being a member of a Basketball Haters Group or Simon Wiesenthal being a member of the Nazi party. It doesn't make sense to me. So, perhaps it isn't my facts that are flawed but the logic of being fan of something and a member of a hate group of that thing. Odd. Very odd.

Interesting. I just so happen to be a proud member of Linsux. I use Linux on my desktop, I publish Linux how-tos, and I write open source software for Linux freely available to all.

"It's also awesome that these folks have so much free time on their hands that they aren't contributing to society by volunteering or creating something for the good of all. But, they sure love to criticize those who do."

What is awesome is that you wrote an article completely devoid of facts.

These inaccurate statements just make you yet another fanboi, just the uninformed zealot kind.

That's like Michael Jordan being a member of a Basketball Haters Group or Simon Wiesenthal being a member of the Nazi party. It doesn't make sense to me. So, perhaps it isn't my facts that are flawed but the logic of being fan of something and a member of a hate group of that thing. Odd. Very odd.

No, the problem here is that your logic is just flawed. You should have done some research before blindly assuming it is a Linux hate group.

Lovely analogies, both, and elegantly skating near to the Godwin fallacy without quite falling through the ice.

As I understand it, Michael Jordan enjoyed playing basketball and was happy to spend his working life doing so. If he'd hated basketball, this analogy would make more sense.

From memory, I'm pretty sure that Simon Wiesenthal was Jewish, and known to be Jewish. As such, his chances of joining the Nazi party were slim to none. This analogy would be a closer approximation if you transposed it a little and made it "Joseph Goebbels insisting on all his speeches being broadcast in Yiddish." At which point your logical fallacy is, I would hope, apparent.

Don't miss out on The TM Repository, incidentally. We have three main weapons: Linsux, LHB, TMR, and a fanatical devotion to Ballmer. No, wait, we have four main weapons ... I'll start again.

We need Linux (in all its flavors), we need Windows, we need Apple, and we could use still other choices. When ATT had a monopoly on the telephone you had a black, dial phone with lousy acoustics and lousy service. You would still have it today if the monopoly was not busted.
There are Chevy fans, Toyota fans, Ford fans, ad infinitum. They foam at the mouth about their preferences and why they have them. In theory (and I believe in reality) having many technology-sources drives innovation and quality. Multiple, independent sources are forced to ask: “How can we make our product better? How can we distinguish ourselves? What are the consequences of our mistakes?”
Linux developers chafe at Windows developers. Something that rhymes with Megaloft worries about at the Linux threat. They say, “Let’s make our product better. Let’s try to make the choice a no-brainer.” Ideally, the computer-using community benefits - no matter what individual choice is exercised.

"Schmindows?" "Megaloft?"

OK, I'll join in. "Snapple?" No, that's taken. "AssHat?" No, I like RHEL. "GoodForYux?" No, it's more tragic than comic. "Supercalifragilisticexpialolaris?"

Jeez. There are school-children who would consider this stuff beneath their dignity.

Thanks Doc,

I guess my main point was to say that anything named with a suffix of "sux" typical means that you aren't particularly in favor of it. For example, if I start a website named say, www.fewtsux.com, would you think I was a fan or an opponent of Fewt?
Unless of course, "sux" doesn't mean what I think it means.

Megaloft? I've never heard that one. I can't figure out what it means. And, for the record, I said it rhymes with Shmindows. I have school children. They just ignore me.

"Schmindows?" "Megaloft?"

OK, I'll join in. "Snapple?" No, that's taken. "AssHat?" No, I like RHEL. "GoodForYux?" No, it's more tragic than comic. "Supercalifragilisticexpialolaris?"

Jeez. There are school-children who would consider this stuff beneath their dignity.

sucks != hate

I know that you know this already, you seem like a smart guy. Just admit your error, and move on.

I bet you probably haven't spent 2 seconds thinking about how it may mean Linux zealots suck, or Linux quality sucks. What am I saying here? I am implying that you are being short sighted.

You are stuck in zealot mode. People can think something sucks without hating it. Sucks means needs improvement, or doesn't meet my standards. It doesn't mean "hate" as you imply.

Shmindows, schmindows. The former is a better fit, I guess, but we're in louse/flea territory here. The Megaloft thing comes from a post directly above.

You can't really start with a paragraph that uses less-than-clever references like "iCrap" and "Shmindows;" follow up with a paragraph that ends with the resounding statement "We, I'm sorry to say, are equally guilty;" complain in the third paragraph about virtual tagging; and then describe sites that you don't agree with as "awesome [not] and intellectual [not]" without exposing yourself to ridicule.

If you're going to be ironic, then be ironic. Don't trample over your own irony.

If you're going to criticise other sites, then making the absurd implication that they consider themselves "awesome" or "intellectual" is fairly self-defeating.

Intellectual? Next you'll be calling "us" Elitist. Actually, we're just plain folk who are often forced to use Linux on the desktop, and object to that fact. We're not as intellectual as http://lambda-the-ultimate.org/, say, but then the subject matter is hardly intellectually inspiring. (Feel free to substitute the OSX desktop or the Win7 desktop, and this assertion holds equally well.)

Essentially, it looks like you're trying to make a reasoned and balanced argument about playing nice, and then you drop a ten ton weight on your foot before jamming it straight down your throat.

Thanks Doc,

I guess my main point was to say that anything named with a suffix of "sux" typical means that you aren't particularly in favor of it. For example, if I start a website named say, www.fewtsux.com, would you think I was a fan or an opponent of Fewt?
Unless of course, "sux" doesn't mean what I think it means.

A website named say, www.fewtsux.com? I'd say that you were an opponent of Fewt, or possibly that you're addressing an extraordinarily narrow niche in the internet pornography market. In what way is this "main point" relevant?

I'd also assume that (unless it's pornography) you'd get very little traffic indeed, unless there are thousands and thousands of people out there who know and loathe Mr/Mrs/Ms Fewt. Which is unlikely. There is, however, a market for sites that criticise Linux. I know Mrs Finkle in fourth grade would like us all to be nice to each other and remember that, after all, we are all the same under the desktop skin; but, really, that doesn't work once you're grown up. Only a really naughty nine year old would call their site www.happy-penguin-feet.com and then trash Linux. Honesty ... yes, honesty. That reminds me. It's another commodity that self-proclaimed Linux evangelists sadly lack.

I tend to be strongly critical of things I don't think have too much hope in terms of being fixed to my liking. This is why I am highly critical of mono, of Microsoft technology, and of Microsoft. If I had hope in the monopolist or proprietary model, I would not talk up the negatives nearly as much as I do.

I would think most people most of the time when criticizing Linux strongly or in a way that would tend to have an observer think lowly of the product or to ignore it would be more interested in that effect (or have little hope for improvement) rather than in helping to improve the product.

I don't know if any of those websites falls into this category, but I am not a fan of the bits I have read from some of them because I have gotten the impression there was a greater interested in creating damage than in helping.

Also, I am suspicious of sites that don't post comments (or who take them down) critical of strong closed source supporters and monopolies (eg, like Microsoft) as has happened to some of my comments in various places (I don't remember much about those websites mentioned above, btw).

Is it not possible this is a response to all the windows hate sites on the internet? Not just slashdot (although that could certainly be counted), there's a huge pro-linux bias on sites like digg, reddit, etc, not to mention blatant anti-ms hate sites like windows7sins.org and defectivebydesign.org. Those 2 are actually funded by the FSF, a non profit that takes donations and is dedicated to furthering the progress of FOSS. So, you donate money to the FSF and they hire webdesigners to slander MS, they apparently decided that the best way to advance linux advocacy is to start windows hate websites. Not to mention those sites are highly misleading, lying by both omission and by twisting the facts.

It's easy to hate hate hate on everyone, nobody bats an eye when people criticize every single aspect of windows design but as soon as you touch the linux sacred cow, you get zealots coming out of the woodwork to harp at you for being 'uninformed' or a 'noooob'. Linux hate sites are a response to this kind of attitude. I am an informed, security conscious IT professional with 10 years of sysadminning under my belt, and I like the features and ease of use windows gives me. To me, making things easier to administrate should be seen as a GOOD thing, a sign of progress. The linux philosophy seems to be to make everything as convoluted as possible, because if you dont need to read documentation for 5 hours to configure a single system service, clearly the software is inferior.

Is it not possible this is a response to all the windows hate sites on the internet? ... So, you donate money to the FSF and they hire webdesigners to slander MS, they apparently decided that the best way to advance linux advocacy ...

It's easy to hate hate hate on everyone, nobody bats an eye when people criticize every single aspect of windows design but as soon as you touch the linux sacred cow, you get zealots coming out of the woodwork to harp at you for being 'uninformed' or a 'noooob'. Linux hate sites are a response to this kind of attitude. I am an informed, security conscious IT professional with 10 years of sysadminning under my belt, and I like the features and ease of use windows gives me. ... The linux philosophy seems to be to make everything as convoluted as possible ....

A Linux hate site by people that don't like Linux. That makes sense.

What you have to remember is that Linux let's you have all the blueprints. You are free to change it. You can know what the software is doing.

What is the value of source code and of rights to modify and rebuild?

Can you understand why people feel they are being short-changed by Microsoft in relation to Linux?

Now, some people don't care about what might be going on with the software or having control. That's their business, but understand where the FSF is coming from (or at least to the extent many support them) when they work to raise awareness of these short-comings in Microsoft products as well as awareness of tactics used by Microsoft to suppress competition and keep people in a position where it is inconvenient or very difficult for them to move to a freer better-valued alternative.

To some people it is an insult and threatening to be forced to use Microsoft software in order to accomplish certain things. In some cases, we are just talking about conveniences and it's not that serious, especially since Linux comes with it's own set of advantages and simplifications (and low cost of course), but in other cases, it affects or could affect things at a more serious level.

It's natural that people point out that if open source software were used by the majority, regardless of vendor, then everyone would have access, but, because of closed source software and the network effect on lock-in products and documents (with Microsoft's central monopolies being the greatest supporter of these) most people are then forced into difficult compromises of their convenience, market options, security, or privacy.

As for why so many complain about Windows, there are a lot more people that have been burned by Windows than Linux for obvious reasons (Windows has been everyone for many many years), so, even without considering that Microsoft keeps secrets from you and works to encumber your documents and applications with dependencies on their platform, you simply have a lot more people that have built up bad experiences and gripes against Windows and Microsoft. [People also don't like to be gouged and more so when the monopolist keeping key blueprints from the public actually breaks the law.]

As for the FSF having nothing better to do, then you clearly have not seen their participation and contributions in many other areas that, if they affect Microsoft, it's because Microsoft simply champions so strongly tools that disable consumers and their competition. [Eg, DRM, software patents, Microsoft's anti-GPL rhetoric and policies, and core open platform technologies that enable competition to Microsoft's closed platforms.] Among the numerous things the FSF has done is to help maintain and preserve the integrity of a large number of core open source products (and licenses). Check out their line-up of key infrastructure software.

Finally, as for Linux being hard to administer, well, the focus has been on accuracy. Sometimes you can't turn precise control into a GUI or at least it takes time. There are drawbacks to pre-mature attempts to simplify administration. I hear many shops that move from Windows to Linux find a significant gain in productivity because of correctness and fine control that they were missing. People don't have to use fine control, but those that have access (and have learned how to use it) tend to like it over alternatives. You've heard about not judging a book by its cover, I am sure. Some book covers go to far, while others don't go far enough. [Not coincidentally, Windows is a highly marketed high cost product in relation to community owned and created Linux.]

PS: I expect the savvy individuals that are aggregating software patents, if not neutralized by good fair law or court decisions, will end up with so much leverage that they will decide who will own which market. This will hurt many VCs and business managers and software developers, all of whom will lose significant leverage based on their skillset or any other assets they have. You can't be too wealthy if you share too much! That is why some people love monopolies and keep trying to grow them into new areas. Of course, consumers will have to deal with much higher costs (many fewer freebies) and lower quality due to diminished opportunities for competition.

There is one very important point I forgot to make as concerns the quality and value of open source software.

Microsoft themselves have admitted directly that they want all of the open source world to code for Windows and support their platforms, API, and protocols. They know how far behind they will fall (and how costly it will get for them) if developers keep focusing on Linux over Windows.

And to throw in a few more items:

Of course, because of the interlocking way Windows probably works, it's very possibly against the GPL for GPL software to be run on Windows. We know Microsoft has gotten many fits over the large amount of GPL software that exists.

With the open source licenses they released not long ago (to try to neutralize the spread of existing licenses), they can even try to play all sorts of games on developers' code rights by adjusting the licenses themselves over time and the terms of their EULAs. They lead you into a situation difficult for you to back out of, so that surrendering to their new much less fair terms is the simplest way out at that point in time. Of course, they will argue that the terms are "fair" in the sense of being driven by market realities and negotiations. But this is exactly why monopolies and cartels need to be shun and dismantled whenever possible. They create significant pressures in the market place in their favor -- yet another reason to dislike Microsoft or at least their position in the market place.

"A Linux hate site by people that don't like Linux. That makes sense."

Let me ask you some very, very simple questions. Have you ever wondered why some people simply don't like Linux? Has it ever occurred to you that to some people, your so-called "software freedom" is nothing more than a misguided concept? Think, buddy. Think HARD!

"What you have to remember is that Linux let's you have all the blueprints. You are free to change it. You can know what the software is doing."

This is one of my biggest gripes with the FOSS culture, that, at this day and age, its adherents still honestly think that source code on its own is good enough to serve as a "blueprint". Are they even bothered enough to stop and think what people mean when they say "top-down documentions"? No, they will just give you a large swathe of spaghetti code, thinking that you have unlimited man-hours to waste on it, and smugly tell you that you CAN change it.

And you can't. Well, at least not without reverse-engineering the code first, finding out which part of the code needs to be changed, making the changes and then hoping that said changes will not result in some emergent issues that you need even more man-hours to deal with. Honestly, I feel sorry for those who had the misfortune to work on Android. Or any other similar projects, for that matter.

"What is the value of source code and of rights to modify and rebuild?"

Source code on its own has exactly ZERO value. No, given the resoures you need to expend on it in order to make it worth, I am even willing to say that source code, in fact, has pretty much nothing except negative values measured in dollars. "Free software" is nothing more than a pathetic excuse to shift the responsibility of quality assurance to the users themselves, and I dare you to say otherwise.

"Now, some people don't care about what might be going on with the software or having control. That's their business, but understand where the FSF is coming from (or at least to the extent many support them) when they work to raise awareness of these short-comings in Microsoft products as well as awareness of tactics used by Microsoft to suppress competition and keep people in a position where it is inconvenient or very difficult for them to move to a freer better-valued alternative."

Right... When you eat out at a resturant, do you know exactly what they have actually put in your food, how they have cooked it or even just who the cook is? When you brush your teeth with a toothbrush, do you know exactly what polymer the manufacturer use for the bristles or the techniques the use to put the bristles on the handle? When you use a hammer or a screwdriver, do know you exactly what proportions of metals the manufacturer use for the alloy? I am sure, at one point or another, you have used something that you don't know how exactly it is made or how presicely it works on the inside. In fact, you are even standing on this very planet, breathing its atmosphere inside this very universe without knowing EXACTLY how each of these things works or what they are made out of or how they are supposed to hold themselves together. You are talking about a futile pursuit that humankind have long given up since the pre-historic era, and unless you can prove that free software advocates are really practicing what they preach, then I think it is pretty safe for me to say that they are all in fact nothing but hypocrites.

"As for why so many complain about Windows, there are a lot more people that have been burned by Windows than Linux for obvious reasons"

The first question that comes to my mind is, "what do you mean by 'burned'?" The second question that comes to my mind is, "have you ever listened to complaints about Linux without just throwing the bog-standard you-can-fix-it rhetorics in return?" The third question that comes to my mind is, "have you ever wonder why your pre-historic ancestors opted for the barter system, or common currencies, or eventually what would become our modern economy?" And I am directing these questions at you.

"To some people it is an insult and threatening to be forced to use Microsoft software in order to accomplish certain things."

Oh, yeah? Do you know that, to some people, it is threatening and unsafe to be forced to fix things on their own? Or do you know that, to some people, it is an insult to be duped into using something that is unsuitable for the purpose? When a FOSS advocate gets a door slammed in his face, will he know that its just time for him to get lost? Or will he just insist on spouting the "virtues" of software "freedom" for even longer? Blaming the competition for your failure to deliver a VIABLE alternative will not make your product look better - it will just make you look silly.

"It's natural that people point out that if open source software were used by the majority, regardless of vendor, then everyone would have access"

No, it's neither natural nor reasonable to assume that "open source" means access to anything at all (except the decompiled binary). A large part of this reality has something to do with - as I have mentioned - "top-down documentations". The rest is pretty much just the individual ability to read code - be that the ability to read machine code, or the inability to read a higher-level language.

"Sometimes you can't turn precise control into a GUI or at least it takes time."

Ah... So implementing a decent GUI actually TAKES TIME! I wonder how much effort has there been recently, within the FOSS circle, in creating any GUI that functionally matches its .conf counterpart, or just a GUI that delivers what it is supposed to deliver, for that matter. Excuses are easy to find but not always easy to be made convincing.

"I expect the savvy individuals that are aggregating software patents, if not neutralized by good fair law or court decisions, will end up with so much leverage that they will decide who will own which market."

People work AROUND patents, you know. If your business strategy happens to revolve around following what has been done by others, then it is only appropriate that you get knocked out of the market anyway.

"This will hurt many VCs and business managers and software developers, all of whom will lose significant leverage based on their skillset or any other assets they have. You can't be too wealthy if you share too much! That is why some people love monopolies and keep trying to grow them into new areas."

This is like saying that you can't make soft drinks if you don't know the recepie for Coke. At the end of the day, it is just the same ol' excuse used by people who refuse to innovate or simply change their market focus.

"Of course, consumers will have to deal with much higher costs (many fewer freebies) and lower quality due to diminished opportunities for competition."

This is just as ridiculous as saying that computers lock your documents into binary digits and make things difficult for people to switch to file cabinats. Honestly, if people decide that an alternative is more suitable for their goals than their current solutions, they WILL switch no manner what you say about it or what price tag you have put on it. Consumers' reluctance to switch to your alternative means one thing and one thing only - that they have generally decided that this alternative is by no means more VIABLE to what they are using - and you can bet with your head that no amount of pseudo-religious rhetorics will convince them otherwise.

>> Have you ever wondered why some people simply don't like Linux?

I'm sure many people will have many reasons. There are an awful lot of people on this planet.

>> so-called "software freedom" is nothing more than a misguided concept?

It would be have been interesting if you would have followed up on this. I'm not going to try to read you mind.

>> its adherents still honestly think that source code on its own is good enough to serve as a "blueprint". .. spaghetti code .. top down ....

Well, I can tell you that I have gotten a WHOLE lot more from the worst open source I have ever looked at than from Microsoft, in terms of being able to reasonably verify what the product is doing.

>> Source code on its own has exactly ZERO value... in fact, has pretty much nothing except negative values

You are a comedian?

OK, well, ask Microsoft to throw their source code liabilities into the public trash can and watch their billions in current and future earnings evaporate.

Ha.

Funny man.

Joe is right. Source code without practical purpose and users has no value.

As for Microsoft, you can analyze what Windows is doing it's really not difficult. It only requires a different set of tools.

If you want to see their source code, become a partner and you'll gain access to it.

fewt, I'll repeat the message I left on the other forum a few days back (though they may not have posted it):

>> #2: Speculation

When contrasting Linux to Windows, I think most would agree that Linux is much easier to fork (to change the behavior in significant ways without Microsoft authorization). Such promotes diversity.

I think data suggests that Microsoft software malware affect a significantly larger percentage of their customers than is the case for Linux.

>> #1: Much of the Windows source code has been published to partners. These partners include security companies who's responsibility is to provide security protection to Windows.

These same security companies plus the rest of world+dog (including tomorrow's all-star security companies) have access to ALL of Linux, without restrictions, at all times.

I highly doubt Microsoft releases things that Microsoft might want in there but that is likely to raise serious objections. A back door can be but a few small changes added to an otherwise perfect piece of work. [And large body of software and "perfect" don't go well together.]

And, in addition, you can't judge the security of the whole forest when you can only see a small part of it and/or must ask for permission or pay and deal with delays to perhaps see a little more. On Linux anyone can follow any and all hunches without limits.

http://en.wikipedia.org/wiki/Blind_men_and_an_elephant

Just looking at their ability to contract out work (likely through deals very favorable to them overall), I'd wager Microsoft has a lot to worry about should they lose their high market share. Losing Windows asset value like that would lower their leverage quite a bit. The flow of dollars starts to reverse and that only exacerbates the effect.

By the way, what fraction of code coming from Microsoft and running on a person's Windows PC would you say can be verifiably known to have been vetted by even a single entity outside Microsoft?

I even expect that most people within Microsoft also have to face significant access issues if they wanted to look at arbitrary parts of the code. [Never mind that they don't know what the build system is doing].

On the other hand, how many black hats might have unauthorized access that might reveal certain problems that most others cannot ever easily vent? The more secret something is, the higher a price that can fetch in the black market, leading to increasing temptations for those able to grab access. [Note that you can find 1 or 10 holes perhaps (or buy that information at the right price), but it's different to audit the entire thing to find all exploitable weaknesses or compartmentalize as you can with Linux where the entire stack is not integrated.]

Google dumping Windows as a standard development platform (?) is long overdue. They were sitting ducks.

>> #3: If you are aware of it. Isn't the point here though that it was 7 months before it was discovered?

Do these 7 months translate into 7 years for Windows?

Can you give me any guarantees or reason to trust Microsoft "assurances" when independent verification is not possible?

What about that Microsoft gains in keeping such information quiet and few would be in a legal position to make most allegations to the contrary?

What about back doors created by the primary vendor? Contrast Microsoft with these open source developers. [One builds a very limited piece of software and others build the software using openly available tools. The other builds the whole stack and builds the system, producing only the binary.]

>> Both platforms give you the tools, but why wasn't this found for 7 months if Linux tools are so good?

Windows does not give me or most people any source code.

What do you mean "the tools" exist on both platforms?

[Let me add quickly that you cannot compare tools that reverse engineer binary with source code tools. In particular, efforts to create code that creates code and looks things up within odd data (apparently as a bug), which can be prodded through the right triggers after a run-time update.. and code that effectively gets recompiled after significant blob updates frustrate reverse engineering, more so when it's something as large as is Windows+integrated apps. I wasn't done with my reply to Joe, but to use his spaghetti example, binary reverse engineering can't do more than what a source code editor can do.. taking spaghetti into account. Don't think the build system that Microsoft sells is what they use on Windows (not the people who use their tools know how what it's actually doing). Here is another example, if I trust code at point in time X and you change it in source, you can be caught much more easily through analysis of diffs. We know where we were and exactly everything that changed. If you are studying a binary OTOH, a *few* key changes and backdoors can be inserted by the vendor and transferred to your system as a 200MB "update".]

Continuing from the last long paragraph (in reply to fewt), let me ask this. How many developers do you know that write in binary (casting aside all mnemonics) and share updates with each other in binary.. and do any significant volume of serious work (vs. a fun hobby of few lines of code) in binary?

Again, this doesn't even factor in all the things the compilation/build processes can do to code to obscure its nature terrifically if that is the intent of the developer prior to distributing it to others.

Imagine having a peer change **one** semantic item within the application but then "recompiling" with their own set of libraries and integration so that you get a product that agrees with what you had in almost nothing (extraneous register and memory manipulations added and others recoded) except of course that over millions of instructions it does appear to achieve essentially the same software at the human observation level? Are you going to re-analyze 20 megabytes of data almost as if it was the first time you were looking at it? And note that 20 megabytes is tiny for a modern system (ie, if that application was integrated with a full software stack).

Note as well that our measure of time is in large fractions of seconds and on past seconds, minutes, etc. Computers can achieve a truckload in even 1/100th of a second (so that we cannot even identify the lapse in time, much less appreciate all the possible changes that took place).

Truly, there is a reason developers don't write directly as the CPU understands things. There is a reason software advanced significantly over the decades as people made efforts not to obfuscate (at least not intra-company or within the recent FOSS phenomenon) and used sophisticated tools to allow writing at a much higher level than 1s and 0s.

[Note that while anyone can write code and open it to the world, if it is crap, the world won't adopt it, generally. Or they may rewrite it to improve upon it.]

There is a reason we are taught to build at the modular level (it lowers the level of complexity for analysis tremendously). Modularity is what allows 2 raised to the power of say 10,000 (which exceeds the estimated number of "atoms" in the universe, I think) as the total number of ways to achieve something across say a few megabytes (think spaghetti code of the worst kind) to be reduced to a much smaller number because of a much smaller number of large subcomponents interacting with limited interfaces to the rest of the code/components. If you don't modularize, a team of humans can't easily divide the work amongst themselves and a single human is not going to deal with the much larger number of possibilities. So now take a relatively modularly built system (at the developer pov) and shuffle the internals of each together, adding "safe" inter-relationships, etc, through a mega obfuscating compilation process.

And further, keep important components out of the system except through a mandatory set of updates that are necessary periodically to keep the system working. In other words, get off the Internet and the system loses functionality and maybe even, by design, starts to fail more and more.

As an example of the latter point, consider a system where key code and data are hidden among many picture files that are part of a required GUI theme update.

As another example, consider a system with code set to crash the system (eg, crucial components not there) unless an update is acquired before the system goes through the point in time where the damage is done -- this is a great way to design systems where you want to force users to upgrade -- just stop feeding the system its required dose of online updates and it will degrade over time perhaps as to become unusable for anything. In fact, build a system with many security vulnerabilities that take a while for others to discover them, and you will get this effect naturally, assuming there are people out there willing to write malware.

[And you can even contract a business entity to "take care" of the situation, eg, to give others enough information and a potential business model so that they will be tempted to write the malware. As I stated above, it's about time Google consider banning Windows from their developers' desks. Though instead of grinding down, I would be worried about how easy it might be to leak trade secrets and sabotage projects.]

There is a reason higher level tools for building code keep capturing the imagination of developers. We need to work at a higher set of semantic blocks in order to build large systems (where we can assume the details still meet the high level contract). Remember, software, as we experience it, for us does a relatively small number of things we can identify (with our very limited minds) across many seconds and minutes worth of running. We assume all of that time was simply spend achieving purpose X. As developers we have to code the internal components that together achieve the large X. However, we only go so far down and allow build tools to map to the CPU. This means that of all the possible things that could be happening at the level of the CPU, we build so that we constrain this to a very very limited subset.

OTOH, if we try to reverse engineer from the low level CPU detail, there are too many possible paths upward that could be taken in order to achieve the few observable things X, Y, etc. AND along the way, maybe X', Y' and a great many other things the user does not like and can't sense, also get done. It's too hard to identify the precise forest you are in if you mostly deal with trees. So we can write tools to constrained a top-down approach, but we cannot generally manage the level of detail to identify **everything that might be done** when we use a bottom up analysis. Put differently, it's one thing to want a forest A and build it. It's another to be given a different set of trees of some forest A' and then prove that A equals A'. If it's true that an "almost" infinite number of distinct set of trees can define the same A, it's also true that a much larger number of combinations can define any of many A' forests that appear like A in various ways but are not [eg, our private data is analyzed and leaked out in obfuscated fashion onto the Internet over time without us identifying it]

As Joe Monco stated, there is a reason we build "top-down" in many cases. [Eg, the operating system proper separated from user level process (and even Windowing and many other parts) is a great top-down organizational structure since the main top-down design contract is set in stone and people can them build out the lower layers in parallel (and in competition) independently from each other.] "Top-down" suggests we focus on the forest and then work on filling in some set of trees that we know will produce our forest. If we have to analyze from trees up, we can potentially be faced with more human years of analysis to see if the forest we get is a safe forest in the sense of including the features we wanted and omitting all the features we despise.

I called Joe Monco a funny man because it almost seems he(?) doesn't write any serious code to suggest source code doesn't have value, or, as he inferred, that the open source code used today by many doesn't have value. Trust me, many people have gone out of their way *not* to reveal their source code and to obfuscate their binaries. This is a trade secret for them, and Microsoft has shown how effectively it can be used to "manage" competition. Source code written by people to solve a problem has lots and lots of value.

On the other hand, Joe recognized that an obfuscated mess is worth effectively zero. For example, if the "source code" is an obfuscated binary blob (as we might get from a vendor trying to keep the details of the code difficult to decipher).

Joe was addressing fewt's (at the time) unwritten comment head on if not intentionally. The real funny man(?) is fewt I think.

[Joe, can you point to the FOSS source code that is giving you problems? I'm sure you can find some spaghetti source in the FOSS world (not to mention that Microsoft is now "contributing").]

Let me summarize (and slightly augment) some of the main points of the last couple of comments I wrote.

As concerns software development:

-- Developers that want to share with each other and build large things and manage complexity use good practices and use high level source code. This source code has a lot of value and so it is reused over and over with careful tracking of which parts have changed.
-- It can be virtually impossible to figure out the consequences of a large (real-time updated) obfuscated binary blob. In human time, you may be able to approximately see that desired features X, Y, and Z are present, but you can't define the entire set of unwanted features X', Y', and Z' that could potentially be present. The complexity for analysis bottom-up can be overwhelming if enough obfuscation is sought by the vendor. Also, remember that every online update of many megabytes scrambles a significant amount of information you may have collected from the system in its earlier state and potentially introduces or changes many aspects of the present set of "unwanted features".
-- A single introduced semantic change can be obfuscated (dissolved) into a large binary blob (that itself changes form significantly) so as to frustrate analysis of this change (contrasting the old and new blobs).
-- Many undesired features (from the point of view of the customer) can be introduced through a relatively small number (perhaps thousands) of semantic changes at the source code level (to be hidden within the binary blob and multiplied in effect in many places).

And as concerns Microsoft:

-- I think I would wager billions of USD (if I had them) that "Microsoft" does not release to anyone nearly all the information they would need or that would be helpful in order to resolve many of the "unwanted features" in their gigantic blobs.
-- I'd wager my hypothetical billions.. that they obfuscate tremendously (so that even most of their developers have no clue), thus, very significantly raising the level of complexity for reverse engineering the whole blob.
-- I'd wager ... that a proper analysis of any of their systems intended to be connected online would involve many key bits that are not distributed with the system proper at time of purchase and whose interplay within the system would not be easy to decipher (in some cases the bits might be moved around and transformed many times (over days and weeks even) before they are finally leveraged).

Note that hooks and flaws can be either intentional or accidental. And we are talking about thwarting third-party interoperability as well as any other number of consumer undesired features.

I'm not just talking about theoretical maximums and such. For example, even software you craft yourself carefully can end up many times with undesired effects. I am talking about identifying a significant number of significantly undesired effects.

Let me describe one particular type of trap some might be falling into.

We have a binary blob. We have a Reverse Engineering Tool (TM).

The Reverse Engineering Tool (TM) comes as a binary blob itself, and you have not tracked down RET(tm)'s behavior through your own binary reverse engineering tool whose behavior you do understand really well or have reason to trust really well.

RET(tm) can look at a blob and unscramble parts of it to more easily reveal good behavior X; however, RET(tm) does not reveal anything about certain pieces of the blob which rewrite good behavior into mischievous behavior. In other words, RET(tm) ignores pieces of the blob by design, and these pieces are key to defining the overall behavior of the blob.

An analogy would be HTML comment lines with embedded scripting (perhaps even proprietary scripts) not showing up in an analysis of that webpage but certainly changing the actual behavior of that webpage into something undesirable or unexpected or more simply unknown.

Another analogy would be undocumented (or buggy versions) of CPU instructions that RET(tm) skips over or interprets incorrectly but which do have real effect when run on the CPU.

Another analogy would be that RET(tm) skips over or adjusts some parts of the blob byte stream after it sees a "secret knock" or identification sequence in the stream. This would allow the blob's creator to hide code behavior by using the secret knock to reveal nice behavior and not the full behavior as adjusted by the parts not shown. [And note, as mentioned earlier, that the secret knock could be an intentional feature or an accidental feature (eg, in the case of undocumented CPU behavior).]

A fourth analogy would be that RET(tm) was built as a binary using compilation tools from Untrustworthy Vendor (TM). Untrustworthy Vendor (TM) also makes the large blob you are trying to analyze using RET(tm). UV(tm) compilation tools add the "skip over this type of code" feature silently into RET(tm) when it was built with their tools. In fact, UV(tm) might even have created (or bought out and modified) RET(tm). UV might promote RET(tm) over other tools by putting lots of glitz into it, giving it away, creating obstacles to other reverse engineering tools, etc. Thus, when you use RET(tm) on UV(tm)'s large blob, you have many key parts that never show up anywhere.

Analogy five: UV(tm) might build a set of system libraries used by most ret vendors, and it is these libraries which ultimately cue in the "skip over (or modify) this type of code" feature.

>> In fact, you are even standing on this very planet, breathing its atmosphere inside this very universe without knowing EXACTLY how each of these things works or what they are made out of or how they are supposed to hold themselves together.

You are correct, but I see the software problem as a solvable problem. I can collaborate to build my own $0 "toothpaste" and as many variations of it as I want and while being much more able to trust it. I can have this for life.

Our current understanding of nature (eg, chemistry) is more limited than the full scope of what we can carefully and predictably design through (idealized/math/digital) software. Thus, software is placed in charge of much more than is the case for any other class of products. From the wide uses of software, we know it attempts to accomplish a whole lot more than what toothpaste or a car does. Heck we can write software cars that do a whole lot more than real cars facing the limits of the physical world. We even improve upon physical cars using digitalization (software). Our stomachs digest most things we eat and turn it into a much smaller number of components and that is what ends up affecting us. Software's effect, OTOH, is considerably more complex in terms of what we can potentially identify and control.

[Plus, bad food is simply rejected. We have learned not to eat rocks, viper venom, lead paint, etc. Bad software can be created in a flash and it's effect not identified despite being used (eg, through binary blobs). OTOH, we don't create new foods with potentially serious deficiencies that readily. Like I said, new combinations, all go down the hatch and get broken down into constituent parts. Software works only in its constituted and much more complex form.]

We even have many regulations for most things around us that don't exist for software. Source code is a great thing to have in the absence of regulations and with EULAs that are horrific to read. Third parties are constantly finding in proprietary software security holes and bugs, some of which are being exploited to a large degree. If they revealed source code, many of these problems would be anticipated and solved once and for all faster.

It's your choice to avoid open source software. I don't see the point in trying to convince everyone. As long as enough people switch to open source, the network effect will mean we can all buy from our preferred vendors rather than be locked in as a society into the proprietary binary blobs of monopolists or near monopolists.

>> "what do you mean by 'burned'?"

A ton of people have had gripes over the years about Microsoft products. As a starting point, google "BSOD" and then "Windows malware". Never mind the many that have experienced other data integrity or data lose problems, program crashes, unexpected behavior, etc.

**All** software has problems. My point there was to recognize that when you have a clearly imperfect product (and so is Linux) on the market for many years longer (and used by many millions more users) than Linux, you will build up a larger number of consumer gripes.

>> "have you ever wonder why your pre-historic ancestors opted for the barter system, or common currencies, or eventually what would become our modern economy?"

Ah, you don't know how to service, extend, and generally create original things or solve customer problems on Linux for money? That explains why you think Linux has no value to vendors. Of course, even if you were correct, it certainly has value to many end users, but we know you can solve essentially at least the same set of business problems with Linux as you can with Windows as many in-house developers have been doing for decades. Linux' transparency enables a greater number of good solutions. This is particularly useful with DIY customers (eg, this is why Linux is used almost 100 times more frequently to create high powered supercomputing solutions, at least as measured by these folks: http://www.top500.org/stats/list/35/osfam ).

[I'll cover the DIY aspect more below as a rebuttal to your criticism of it.]

Open source is built in a distributed fashion (as science is done) rather than in a secretive method as was the case for alchemy. Note that there are a great many people employed today that practice science. They share and leverage each others' workload, make a living, and they frequently do other things besides 100% science (eg, they might teach, own a related business, etc).

>> to some people, it is threatening and unsafe to be forced to fix things on their own?

Most users of Linux (even Linus Torvalds) do not "fix" the majority of the software they use.

You are missing the point of DIY possibilities.

>> Or do you know that, to some people, it is an insult to be duped into using something that is unsuitable for the purpose?

That's why I don't use Windows.

Over time (though not sure how long), I'm confident there will be a vast number of superior solutions in open source in most categories, but, as you likely can tell, open source adoption has been slow for various reasons (eg, lock-in effects and marketing by many proprietary vendors, with Microsoft leading the pack). Open source has definitely been a disruptive paradigm change, but we still have much legacy lock-in and vendors not on board or with only one foot in.

>> When a FOSS advocate gets a door slammed in his face, will he know that its just time for him to get lost? Or will he just insist on spouting the "virtues" of software "freedom" for even longer?

We use and build our own software. The more people that join, the faster it grows and the faster we can overcome social lock-in to monopolist closed source software.

Every day, FOSS gets rejected coldly as well as adopted whole-heartedly (or somewhere in between).

From a selling pov, you should know that you don't stop after you encounter rejection.

Unlike Microsoft, we will still be here (and much larger in size) improving our software "when revenues dry up". Heck, revenues started at $0. If we didn't stop then, do you really think we will stop as revenues improve and as there is more useful open source software to be leveraged? There is value to having source code access. This is why FOSS hasn't been killed off (just look at the fits Microsoft has gone through).

>> Blaming the competition for your failure to deliver a VIABLE alternative will not make your product look better

Most people complaining about Microsoft point to the anti-competitive moves they make (eg, giving away their software only after Linux made inroads in some segment). We complain about closed source lock-in (it's tough to figure out all the dependencies being created by the large binary blobs). We complain about back room deals and misleading marketing.

This is not to say open source is better in all cases. Not at all. Unfortunately, we do get attributed as a flaw the failure to fully interoperate with closed source products. These complaints go down once people start creating new stuff using open source and more open standards and learn to avoid proprietary lock-in.

When measuring the value of open source, you (well, I) factor in the low cost and the high degree of flexibility (auditability, control, etc) that comes with having access to the source code. Of course, not all users find as much direct value in source code, but, as an industry, we improve faster the more code that is accessible.

>> nor reasonable to assume that "open source" means access to anything at all

You are trying to pretend that the individual is the society.

You may not know how some source works, eg, Linus doesn't know how all parts of Linux work, but society as a whole covers your back.

Open source means letting world+dog have access rather than have access be controlled by a very small group. You will end up with faster growth and lower prices for a given functionality when you break monopoly grip.

Also, while you may not know all the code, with open source you can pick and choose. You might come to the conclusion that learning and modifying some aspect here or there can give you large bang for buck (you extend it, reuse the code, or fix an obscure bug that was affecting you significantly).. or you might pay someone to do something, and because of greater access (and competition), they might solve your problem faster and for less than if you had to write Microsoft a letter (yeah, try that one!).

As one example, an artist might learn a little programming in order to modify one of his/her applications to create their own stylistic effects a little easier (or to reorganize the GUI). In many cases, they can share this code back and gain overall as others will help maintain it and improve it. The artist's edge is in knowing how to use the technique, not in hiding the details of the code that helps him/her more easily use the technique.

So the artist was able to gain all of that simply by learning some and making a few changes (maybe doing so with help from many others). Perhaps the code additions weren't that great but still helped. Others could then improve it. With closed source, access would be more limited and cost more money, making growth happen slower.

Like I said, a failure of your imagination (IMO) is not going to discourage those that are growing faster daily through sharing information. Stay in your island or join. It's your choice. There is no one to reject you because all the key parts are open, available for $0, and with many there to help you out (this is not to say it is easy in all cases). You really can DIY, but feel free to ask a question on any forum you want. You will frequently get a decent reply or a tip (sometimes the tip will not be in an easily identifiable form). Google.

>> So implementing a decent GUI actually TAKES TIME!

I think you are unaware of all the GUI available for FOSS (the "list" is updated all the time and not always easy to track down because FOSS is a distributed phenomenon (but many work at centralizing the information)).

I also hope you didn't miss the point that there is such a thing as over-simplifying so that key details are obscured. This can lead to problems and an inability to really know what is happening or happened.

I'm not criticizing Microsoft or Apple (or any of the oftentimes more creative folks these two giants tend to buy out) left and right on the quality of their interfaces. Everything can be better but also could have been much worse. My biggest criticism is based on a failure of them to reveal source code and to give rights. Why bother with them? I have never thought about begging them because I know the "community" is solving our own problems. We bypass the monopolist. With community code, everyone has access (if not necessary the desired access path). It feels good to have independence, control, knowledge, access....

>> People work AROUND patents, you know.

Patents are too broad, specially for software where the greater work is in building not in general ideas. Even if we very wrongly assume patents aren't too broad, you can't use the "work around excuse" all the time. How do you work around the round wheel, for example. How do you work around mathematically optimal truths?

Of course, patents being given with such broad claims and so easily (eg, **unobvious** to a Jack Developer) means that many times it really is very difficult or impossible to work around a patent. And if you work around it, it might only be with an inferior solution. And you will have lots of independent re-invention and hence at least the stifling of free speech.

Eg, the shortest distance between two points is a line. This is a mathematical fact within a model that happens to coincide for all intents and purposes with the real world. If you patent, "using the shortest distance between two points when you want to minimize length within context X", then you are simply trying to grab exclusive control for 20 years over a mathematical reality. The precise wording you use (and your skill and timing in doing so) will define the scope over which you will have control over this mathematical fact.

Patents are stifling. Software patents are very very stifling.

Some useful links:
http://www.techdirt.com/article.php?sid=20100616%2F1037129853&threaded=true&sp=1
http://www.linuxtoday.com/developer/2010062000135OPLLDV

>> If your business strategy happens to revolve around following what has been done by others

That's a funny one.

Patent claims are being taken out on ideas we all have.

As an example, look at how easily it is to write up ideas: http://www.feld.com/wp/archives/2010/06/mailing-out-patent-absurdity.html#IDComment79749495

Giving someone a monopoly on it is very stifling when that affects many, as is the case for software: $0 to acquire, modify, and distribute; is based on an idealized model (eg, as is math and fiction) where the creation is in the mind without a need for physical product constraints or experimentation.

If software patents were implicit and $0, as is the case for copyright, a lot of current patent holders would find that related prior art would pre-empt their patents.

Social leeches study others' prior art and then write patents to block as many as the law will allow. Social contributors write open source that enable everyone.

Weren't patents to protect the "little guy"?

Here is another comment about how society has changed since the late 1700s but patent law has gone in the opposite direction, if anywhere: http://www.techdirt.com/articles/20071227/010830.shtml "Promote the progress in a different world".

And these effects are much worse for software and related patents than they are for "expensive" patents.

>> This is like saying that you can't make soft drinks if you don't know the recepie for Coke.

You totally did not understand my comment (likely because of how I wrote it).

What I said was that all of these patent monopolies being created and traded around are allowing tremendous (government supported with artificial scarcity broad monopoly) power to be concentrated in few hands. This means that most businessmen and engineers will lose leverage they would otherwise have based on their skills, work ethic, creativity, etc.

Leverage exists when you can compete directly or else withhold your asset in order to have significant negative impact (eg, opportunity costs) on the person over whom you are trying to exert leverage (eg, in negotiating a salary).

With monopoly creation in the hands of very few, you will not be able to compete or to withhold anything that will really impact their earnings. Ergo, you lose leverage. This can mean all sorts of things including having to deal with a less fun, interesting, flexible working environment and lower pay.