I think the article makes a good point. Most comments complain about some technicalities, like using the word "open source software development" vs. "open source software", or mention that "no it isn't inferior!" (which is true, but not the point).
A few months after meeting me, my girlfriend wanted to try a Linux system and I was happy to help her install it. In the end she didn't see what I thought was better about it. I thought I never said it was better, but apparently I gave off that impression, leading to a false expectation.
And it's true, Linux desktops are still notorious despite being actually very stable these days. Open source programs like Firefox, Chromium and Telegram all work the same; and some others like the average file explorer is functionally equivalent too. It's just not better, at least not for most people. It might even eat more battery because of inferior drivers.
The inherent advantage of free software being free (as in freedom) is what the article is about, and that is a good point.
The problem with this argument is that it doesn't achieve the aim of winning more users, even if it is true.
The "open source" movement recognised that users don't care about freedom, but they do care about other aspects. In a choice between 'proprietary and easy' and 'free but slightly-harder' users go with the first option: it's actually worse that this because consumer users are actually willing to go beyond 'zero cost' and actively sell their privacy in return for ease/sociability etc.
I think the actual issue with the OSI's marketing is that it turns out these words work for professional users of software and not for general consumers. Professional users such as IT departments care about 'power', 'flexibility', 'maintainability' etc - general consumer users don't think that way.
End users (particularly consumer users) care about achieving aims in very straightforward, simple terms! Neither, the FSF or OSI philosophy translate into perceived or real benefit for them.
To a large degree I think it's something like this:
Developers (care about freedom) <---> professional users (care about attributes) <---> consumers (care about ease of outcome)
'care about attributes' is the wrong word, but professional users tends to care about aspects of the software, they ascribe value to the software itself.
> it doesn't achieve the aim of winning more users
Winning more users is not an aim in itself, not to me at least. It's not a cult that I want to win hearts for.
But if people see how wrong it is to have a single entity control your computer (and almost everybody's computer), they are definitely welcome to ask for help installing something better (in that regard).
I didn't respond to the rest of your post because it builds on that assumption.
That's fair enough, there are lots of FOSS users who don't care about there being more users. They care about their individual freedoms - it's a perfectly rational view.
Perhaps I mis-spoke because I personally do care about there being more users of FOSS.
Really, we're trying to discuss how the FSF approaches its aim to "preserve, protect and promote the freedom" ...
Do you think that aim is enhanced by spending energy on what I consider to be academic hair-splitting, rather than addressing and explaining the philosophy to wider audiences such as users / academics / policy makers. Or am I misunderstanding how Free Software is promoted by denigrating open source?
You're conflating the idea of external remote control of software with freedom to do what you want on a computer. They're completely different, unrelated concepts.
Of all the operating systems, only Windows does things that it really shouldn't, and resists attempts to keep it from doing those things.
macOS puts limits on non-expert users, but if you know what you're doing it lets you work around the limits - so far, at least.
Even iOS is easy to open if you know how to install fuse from the command line. You can't run as root, but you can access the file system without having to go through iTunes.
Ubuntu meanwhile added links to Amazon that no one really wanted. In theory you could get rid of them by recompiling or using a fork. But creating and maintaining a custom Linux fork is something that is completely beyond the skill of 99% of users, and probably beyond the skills of more than 10% of professional developers.
And if you're using someone else's fork, you have no idea what's in the code. We've seen libssl infiltrated, and it wasn't even hard. Who knows what else is in there?
In reality access to source code gives you absolutely no protection from bad actors. There are no reliable tests for overall security, and the ecosystem isn't large enough, focused enough, or professional enough to appraise security with any confidence.
So the idea that "You can control your computer" is just rhetoric. You can't. You will never be able to. You will always need to rely on an ecosystem of developers, and large parts of that ecosystem will be opaque and closed to everyone for practical reasons.
I think users choose what they see every day. FOSS OS distributions does not have the marketing budget to push their product into the face of the everyday user or bribe governments to put their product in use in primary schools to lock-in the next generation.
"You can change this (if you spend three years becoming an expert on software development)" is a very strange definition of free.
If FSF produced software that ordinary consumers could customise and improve with ease, there would be some merit to the argument. If the FSF had sponsored original research towards that goal and driven the industry towards it it would have been a game-changer.
But that isn't what happened. The reality is that FSF 'freedom" only applies to the tiny percentage of the population who can use the command line.
In practice the FSF has always been perfectly happy with hard-to-use, amateurishly designed, non-performant software. Somehow the idea that it's "free" is supposed to excuse its many faults.
Stallman et al have never shown the slightest interest in making their idea of software freedom accessible to the 99% of users who aren't technocrats who dream vi commands.
"You can change this (by hiring a developer to fix the issue you are having)." <- There: I removed your infuriating straw man. Also: remember that one way to "hire" a developer is for you and thousands of other people to each contribute only $1; this is something everyone does quite often (by buying software) and for years now has been even easier to pull off (by crowd funding "changes").
I think you went too far there and that's where I lost you.
I sure appreciate that FSF gives me access to a kitchen where I can cook my own dishes. I know how to cook, it would be shame, and extremely infuriating if all chefs were constrained to a shitty one size fits all "Mcdonalds". "Mcdonalds" is plenty good for many, and that's fine, its just that I don't want to eat there for ever after when I know I or someone else can cook up something better at times.
Or to give another example, if I know how to handle a slightly more complicated control I would happily buy a car that permits that, rather than some generic blasé Chevy Vega, may be a 911.
There are other attributes that some professional users care about more than just a simple functional matrix.
For example, I've been in situations where the ability for the professional user to fit the software to their specific requirements over time made FOSS the choice. Even though the initial solution was worse than the equivalent proprietary one. But, the idea that they could develop and improve it over time to meet their specific requirements was powerful.
The things that - to me, at least - are better about a Linux/Un*x-based desktop only became apparent after spending at lot of time using it (a year, maybe 18 months).
Which makes me pretty bad at convincing people to give it a try - "Yeah, it sucks at first, but trust me, in year you will love it" is not a good pitch.
(OTOH, having first installed SuSE Linux about sixteen years ago, the process of getting a Linux desktop setup is soooo much smoother these days.)
Believing you're more secure simply by using Linux instead of Windows is dangerous.
The Mirai botnet is the living proof of what happens when you use Linux and forget to take care about your security, not using Windows is not an excuse to lower your guard.
As an example, think about all the untrusted code you run, something so simple like `curl -o angular-react-vue-example.sh | sh`.
OTOH, not being familiar with troubleshooting low level issues means you'll never know whenever you get infected, remember kernel.org's hack[1]?.
That would be missing the point entirely due to the fact that most users care more about ease of use and apps than "advanced configuration".
Not sure why we have to "sit anybody down" either...that sounds a bit forceful and unnecessary. Last time I checked, "better" is completely subjective.
But for the record, most people think Windows is better since that's what they've voted for with their dollars. Windows won the desktop war. It's got more users, more apps, more consistency, more stability, more commercial support...more everything. People still fighting for desktop Linux to become popular with regular users are like Hiroo Onoda, the japanese soldier who kept fighting WW2 for 30 years. Good luck with that!
The biggest issue imho is that there just isn't a nice consistent desktop story for linux. Making an application with a nice consistent setup and UI experience across all desktop environments and all distributions without having to distribute to several different package repositories or test on these distributions or window managers?
Until that works there is no one Linux desktop - only several different Linux desktops, each with too little traction.
You could sit your girlfriend down with nothing but terminal, the man pages for a printer or video card driver, and the root user password, and she'll understand why Windows is better.
mmc, regedit, powershell, sure this could cascade much further, but I think maybe op had the default privacy options in mind, which enable windows to use the camera and microphone feeds as the os sees fit.
Yes, this exactly. If anyone cared to check out the advanced configuration it is pretty much exclusively turning off security vulnerabilities.
One of them is "Automatically trust WiFi connections my contacts trust", a bunch of others are sending data to Microsoft. It's been a few months since I turned on a new Windows 10 PC but I was completely turned off by the advanced configuration.
The #3 story on HN right now is "Windows 10 in-place upgrades are a severe security risk"
The hardware's less than a year old, but it's low-spec. Still, using it is so miserable that it makes me physically enraged. The login screen loses keystrokes; the browser constantly prompts for my smartcard; the system can't do partial OS updates; it tries to sell me things on the login screen and in the start menu.
Everything is painful, and nothing is pleasant, and when I use it life is dark and miserable and existence is cursed.
Seriously, that's not hyperbole: using Windows negatively impacts my life.
Running Debian with stumpwm, OTOH, is so comfortable it's like slipping on a glove.
>To me, the Windows machine I have to use is a Trabant with a missing front passenger seat and a nasty tendency to catch fire when it's cold out.
>Seriously, it's horribly painful to use.
Haha, I have the exact same experience with linux.
Windows is a self-driving Tesla that just works while Linux is some 1960's beater limping into the modern world, requiring the 1500pg Shop Manual in the glove compartment on a daily basis because, you just know that old carbureted air/fuel gets knocked off and requires hand adjustment, and the throttle gets sticky if push the pedal too hard, and it just doesn't like keeping oil in the engine when the rubber seals are too cold.
I guess it just comes down to what you expect. To me, Windows is "push install button, come back to internet connected desktop" while Linux is "put aside an afternoon for terminal usage and extensive googling and be prepared to make sacrifices/compromises-- temper your expectations"
When I get into a conversation about Free software being better than proprietary software, I focus on real problems with modern proprietary software that average users have these days, mainly with operating systems. For instance, lots of people don't like Windows 10's interface, but there's also the spyware and advertising issues, and various other annoyances with it, plus the forced-update problem and the fact that you have to separately purchase so much software for it to make it useful. With desktop Linux, you don't have these problems: lots of software is easily installed from the repos and you wind up with a fully-functional system very easily, updates are ridiculously simple and fast and don't prevent you from using your computer, updates aren't forced on you breaking your software, your OS won't be forcibly "upgraded" to a new version you don't want, performance-crippling anti-virus software isn't needed, etc. On the Apple side, the OS is tied to very expensive and now-crippled hardware (see the new MBP which lacks ports, requiring you to carry around a whole bunch of dongles) and still has the problem of not having useful software by default. For people who insist on using some kind of proprietary software like Photoshop, I just don't bother, but for people who just want something to surf the web, write simple documents, watch videos, etc., desktop Linux works great so this is how I pitch it.
"and the fact that you have to separately purchase so much software for it to make it useful."
That's not a negative, especially for anyone who makes their living selling software. And there's nothing special about Linux that saves it from that; if it became popular, commercial software would be written for it as well, and you'd have the same problem.
>That's not a negative, especially for anyone who makes their living selling software.
So you think free things should be banned so that leaches who want to make money selling things can do so? How idiotic.
If I'd like to make a living selling air for breathing, is it wrong for people to breathe the air that's freely available? I guess so, in your world.
>And there's nothing special about Linux that saves it from that; if it became popular, commercial software would be written for it as well, and you'd have the same problem.
Here you completely miss the point. I'm not talking about special niche software, just basic stuff. I can't even edit a fucking text file in Windows without installing some special third-party software, because Windows doesn't come with a usable text editor. Luckily there's free stuff available on the web for some things like that, but it's a big PITA to go find some website, find a download link, and install the program. On Linux, this kind of basic software is part of a standard install, and at the very worst, all I have to do is something like "sudo apt-get install vim".
>So you think free things should be banned so that leaches who want to make money selling things can do so? How idiotic.
Point out where I said it should be banned. And nice that you equate someone trying to feed their family with being a "leech". How idiotic.
> I can't even edit a fucking text file in Windows without installing some special third-party software, because Windows doesn't come with a usable text editor
>Bullshit. Notepad is a pathetic excuse for a text editor. It can't even handle different line endings. Wordpad isn't much better.
They're still text editors that work just fine to edit text files.
>And yes, someone who tries to leech off of society providing something that can be easily obtained for free (such as air) is just that: a leech. Obviously, you DO want to ban such things, or else you wouldn't advocate leeching. You're a pathetic excuse for a person, advocating for people to be forced to purchase breathable air.
Again, show where I want to ban free software. And if I write software, how is it easily obtained for free if I'm charging for it? And your air analogy doesn't work unless I'm the one creating the air as well.
> It's just not better, at least not for most people.
Certainly if you define "better" as "do everything I currently need it to do", Linux/BSD aren't any better. But with just a little effort, you can apply nearly any skills you learn in Linux in far more domains than you can for Windows skills. For instance, robotics and embedded systems.
"Enabling a path to future growth" certainly qualifies as "better" in my mind.
The main point of the article and you seem to be missing: freedom has intrinsic value.
Personally I don't care if you voted in the last election, or if you ever traveled abroad. But I do care very much about the rights that I have and allow me to vote or exercise free movement.
I also don't care if you are interested in modifying the source code of your operating system or the driver of your printer, but I do care very much about having a world where these things are possible.
The article is trying to exactly show that "Free Software" is like this, focusing on keeping the intrinsic value of having software that can be freely read, distributed and modified - even when it does not translate in more utility value, a.k.a better software.
>The main point of the article and you seem to be missing: freedom has intrinsic value.
And not everyone agrees with that, or that the value is enough to forgo the other things.
>I also don't care if you are interested in modifying the source code of your operating system or the driver of your printer, but I do care very much about having a world where these things are possible.
YOU do. Someone who just wants their printer to work would probably place much more value on that, than if the driver were open. Remember, one of the main themes of the article is whether software with "freedom" is worth more than software with usability or features.
>The article is trying to exactly show that "Free Software" is like this, focusing on keeping the intrinsic value of having software that can be freely read, distributed and modified - even when it does not translate in more utility value, a.k.a better software.
And many people disagree. If it doesn't do what it needs to do, then the software isn't of much value, even if it's libre.
Free software or even open source software is neither superior nor inferior by definition. It is just software and its quality will largely depend on how much effort the developer is putting into it.
Typically some form of commercialization is often a good incentive to improve any type of software but obviously there are plenty of examples that go against this rule.
It is common to see these days open source or free software that comes with some form of commercial options and that's very good in my opinion because you get the best of both worlds - still it doesn't mean it is better or worse.
Is Gitlab better than Github? It depends who you ask.
> Free software or even open source software is neither superior nor inferior by definition.
Strongly disagree with your implication here and the rest of your comment, though I agree with the literal text you wrote in this sentence.
FOSS is superior for the end user because it can help avoid vendor lock-in (someone else can always take the code and run their own service/app that the user can then switch to) and allows the user themselves to pay someone to fix issues/maintain the software if it becomes unmaintained.
You are quite correct in your point that FOSS is not necessarily more featureful, more secure, or easier to use than proprietary solutions though.
I understand where you are coming from but I disagree with some parts of your comment.
FOSS is not superior to the end user because by no means it helps with the vendor lock-in. Remember that the vendor in FOSS is the developer and when they decide to ditch the product it does not mean that someone automatically will step in to take the job. Therefore, the customer is locked in anyway. You are assuming that the customer will support the product but they may not have the technical know-how nor the interest to do it. What is the difference between a dead closed-source product and a dead open-source product?
In fact, more often than not you see the opposite happening - someone creates an open-source alternative of an already established closed-source products or services to piggyback on the established success.
I am dissing FOSS. :)I am only only trying to present a balanced view that at the end of the day all software is equal as long as it does the job for the time it is used.
> What is the difference between a dead closed-source product and a dead open-source product?
The difference is that with an open source product you can always choose to pay someone to maintain it, even if just for long enough to move away from it. With a closed source product you are SOL and may not even be allowed to use it anymore after the company goes under depending on the details.
I disagree with your disagreement. The usability and functionality of the software is the paramount concern. Who cares if you're not "locked into it" if the software sucks to begin with? You probably wouldn't start anyway.
I respect that ethical motivations of free software in an age of mass surveillance, retail tracking and the copyright war but I routinely use proprietary software.
In practice when a free clone lacks a feature or has a bug that stops me, I have no choice. I seldom have the skills and time to pop the hood and fiddle with the code.
If I need to read a document typed in MS Word that is unreadable in libre office, I'm going to download MS office viewer...
You "may not have a choice" at the exact moment that you need to do something that only a proprietary solution can provide, but those that really want to support free software can still adopt a strategy to avoid such moments: every time one "consumes" proprietary solutions, take a percentage of the value spent and donate to a free alternative.
Imagine if every graphic designer that "has no other choice than Photoshop" decided to donate $10/year to Gimp, and that took one hour week to use it, provide bug reports, feature requests, etc.
Imagine if every web developer that uses Apple "because it is a better desktop than Linux" decided to donate 5% of that to Canonical (or Fedora, or Arch) and took a few hours every month just to try to get their issues sorted out.
Imagine that every company that is so invested in AWS (or GCS or Azure) realized that there are cloud providers that are based on free software solutions and have no lock-in, and that they made a commitment to make quarterly donations to them.
So, if a consumer has the reasonable expectation to pay for something which fulfils an immediate need, they do not "really want to support free software"?
Surely you must understand how that is a meaningless phrase?
Software services are R&D _at some point in time_, which is also a meaningless phrase. Right now I can go buy Microsoft Office and fulfil all my document editing needs I'll ever have. However MSOffice to me is a product, not some distant R&D objective.
Maybe FOSS could be better guided by focusing on what a product really is.
If your focus is only on your editing needs and you buy MS-Office and stop there - then no, you do not really want to support free software.
If on the other hand you have a document edit need that no free software can provide, and you understand that you can buy MS-Office to fulfill a need and also donate/support to LibreOffice or any other project you might be consider worthy of support - then you are at least walking the walk and doing something in support of FOSS.
It seems to me you are making an either/or proposition: either you are a consumer or you are an contributor of software. It doesn't have to be like this.
To be honest, I think we both lost the point of the conversation.
All I was trying to say is that those that (1) respect/value FOSS but (2) end up buying/consuming a proprietary solution due to the current shortcomings of the free alternatives can still support FOSS, which might lead to (3) not needing proprietary solutions in the future.
I don't want us to be arguing past each other, so please help me understand what you mean by "this is not how it works". Does it mean that you think that such strategy can not work? If so, why not?
What cloud providers are based more completely on free software than AWS? Aside from running your own cloud on your own hardware, I'm not aware of anything that is free software so I'm very curious what the better options are.
Does Amazon provide source code so you can run your own S3, EC2, RDS, Cloudfront, IAM, etc, etc, etc on your own datacenter? They don't. They build their services on top of free software, but none of the services provided by AWS itself are free.
OpenStack, on the other hand, is exactly that - a free software platform which gives no single "cloud provider" any kind of vendor lock-in whatsoever.
"Imagine if every web developer that uses Apple "because it is a better desktop than Linux" decided to donate 5% of that to Canonical (or Fedora, or Arch) and took a few hours every month just to try to get their issues sorted out."
You just illustrated part of the problem: There are so many different desktop environments, that it would make trying to pick one to be supported, and trying to pick one for writing software a fool's errand.
One of the main reasons developers like MacOS is because it has the UNIX underpinnings and terminal power, AND it has a standard GUI toolkit and robust 3rd party software market providing high quality, pleasant to use tools.
> There are so many different desktop environments.
I described Linux distributions, not desktop environments, and the ones I listed are known exactly for having huge support from their communities and that can be easily picked up by anyone wanting to do so.
> trying to pick one to be supported, and trying to pick one for writing software
This makes no sense.
If you are writing software that is multi-platform, anything that runs on some kind of VM, you don't need to know the internals of your development OS.
Hell, even if you are developing an application that is specific to one OS, you still should not take your development distribution in consideration. A distribution is just a nicely packaged collection of software, there is no reason to tie your application to any of them.
Also, bear in mind that we are talking about ways to have more free software, not to have "one winner alternative of software". Any kind of contribution helps, no matter what distro most people use. Using the excessive number of distributions compared to one single "MacOS" as a justification to not support free software is akin to complaining about going to the supermarket and complaining about too many kinds of healthy fruits and vegetables and using that as an excuse to pig out on Nutella.
> developers like MacOS
Yours is one of the thousand variations of excuses I heard from people that simply do not bother to even try to adopt free software. I can take you up on a bet that you will find absolutely zero "high quality, pleasant to use" tools that a web developer can only find on MacOS.
From web browsers to server software, from compilers to text editors, from source code viewers to even file managers... I can bet that anyone claiming that a Linux-based distro falls short in providing tools has not touched a Linux-based workstation in at least a decade.
>If you are writing software that is multi-platform, anything that runs on some kind of VM, you don't need to know the internals of your development OS.
Ok, so do I write my software using GTK or QT? Perhaps whatever XFCE uses?
>I can bet that anyone claiming that a Linux-based distro falls short in providing tools has not touched a Linux-based workstation in at least a decade.
I never claimed that. I claimed that the MacOS tools were higher quality with better aesthetics and usability. Once again, because there's only the Cocoa environment to target. If I were to write a tool for Linux, what do I use? QT? GTK? Some other thing?
XFCE uses GTK, so it is one less thing for you to be concerned.
> Once again, because there's only the Cocoa environment to target.
You are equating the tools provided by one vendor to the whole ecosystem. It is entirely possible to write MacOS tools using free toolkits on MacOS.
> better aesthetics and usability.
First, this is entirely subjective. Just as examples, I'd rather use Banshee as my music media manager over iTunes or any of the music players I found for OSX. Firefox and Chromium/Chrome exist for both, and I am yet to see someone that claims Safari "has better aesthetics and usability". Developer tools that are not exclusive to Mac are mostly ported from Linux and require workarounds on Mac (brew as one such example), so I'd argue that the usability on Linux is better.
Second, the "better aesthetics" is not due to some virtue of being closed. It is just a matter of economics and resources that are poured into the projects. If you don't think that is the case, ask yourself why many people prefer Android designs over iOS?
> What do I use? QT? GTK? Some other thing?
You use whatever you want, and whatever toolkit better fits your requirements. It is entirely possible to have desktop application that use one toolkit while the main Desktop Environment uses another.
You seem to be stuck in this same mindset from Microsoft developers, where everything is only accepted if blessed by their vendor. This is as anti-freedom as it gets.
If you are okay with living in a golden cage, fine. But at least own up to it. All of your excuses so far have been baseless.
>XFCE uses GTK, so it is one less thing for you to be concerned.
Still doesn't resolve the question. Do I use GTK, Qt, or any of the other Linux GUI toolkits. Perhaps FLTK? FOX? Do I have to use raw XLIB? What do I use if I want my application to fit in with the rest of the system?
>It is entirely possible to have desktop application that use one toolkit while the main Desktop Environment uses another.
Which looks and uses like ass. You have things that conflict in keyboard shortcuts, you have things that don't respect the UI/UX of the rest of the system, which the user presumably chose because they liked that setup.
>You seem to be stuck in this same mindset from Microsoft developers, where everything is only accepted if blessed by their vendor. This is as anti-freedom as it gets.
No, I'm stuck in the mindset that I want things to work, and work well, and be pleasant to use. Right now Linux has significant barriers to this. And to hand wave it away as not a problem, or something derogatory about "microsoft developers" is not only arrogant, but means that it won't get addressed.
>If you are okay with living in a golden cage, fine. But at least own up to it. All of your excuses so far have been baseless.
They're only baseless to someone who doesn't care about usability. If you want to sacrifice being able to use software without pulling your hair out for "freedom", go ahead. But don't you dare disparage those of us who value usability of our software.
Look, I don't want to get into a screaming match with you, and I am afraid this is what is starting to become. I see someone is already doing some drive-by downvoting on you, and this is not cool.
But you need to realize that you are holding very strong opinions who are based on the state of Linux system from at least 10 years ago. If you try to take a look at any modern Linux system, you will see that there have been many integration efforts and developments around FreeDesktop.org, which basically removed all of these straw-men you are putting. Theme Engines can be toolkit-independent, so they can and often do perfectly "fit with the rest of the system".
> If you want to sacrifice being able to use software without pulling your hair out for "freedom", go ahead.
My whole point for OP was "use proprietary software if you have to, but hedge it by supporting free alternatives. In the long run it can be a good strategy, because you won't have to sacrifice the utilitarian value of software, but at the same time you will be standing for a good principle", yet and you are the one taking such a violent stance against it.
You want to paint me as someone taking some kind of extremist view which has no basis in reality. If you need examples:
- A few months ago I wrote a tiny review about Ubuntu Phone (https://news.ycombinator.com/item?id=12201061). If you read it you will see I am not "pulling my hair out for freedom".
- I reinstalled my windows partition on a laptop, simply because I wanted to play with my guitar in the computer, and I was tired of fighting with low latency kernels, jackd and pulseaudio. It was too much knob-switching and lever-pulling just for what was supposed to be a hobby.
- I made a Hackintosh for me last year, because I truly wanted to see what I would be missing software-wise. Turns out that it was only XCode and iOS programming. Everything else worked just like it worked on my Xubuntu, or it had an analogue.
So, I wouldn't tell anyone that has some specific use-case that they should sacrifice everything just to keep using FOSS. But I will tell them when their use-case could be covered by FOSS and that there is no excuse not to use it.
And this happens to be the case of 95% of "coffee shop web developers". There is no extra value in using MacOS compared to Linux for someone that uses basically text editors/terminal line/unix server software/web browser. If you take an honest attempt at using Linux for 2 weeks you will see that, and you will soon realize that all of this talk of "valuing usability of our software over freedom" is just a rationalization of your prejudices.
I once was on an elevator with Richard Stallman at MIT with my 14 year old son. He turned to my son and asked him what OS he used. My son replied 'Windows', because it's free.
Stallman might have lost his mind at this point, but instead smiled and said 'Windows isn't free and free software isn't free'.
I think a core issue here is that despite being a buzzword, 'free' just isn't all that meaningful or important to most developers, especially since the freedom that is granted to end users is explicitly not given to creators (see GPL). Open source however, especially with something like the MIT license, at the very least gives you access to read and hack at cool source. F is not the meaningful part of FOSS unless you happen to be a lawyer or really enjoy lawyering.
It's true that the definition of "free" is somewhat overloaded, and means different things to different people. But to say it's not all that meaningful or important is, I think, going way too far.
As for your point about creators' freedom with copyleft, that limitation is explicitly to ensure that end users' freedom remains unrestricted. And, by logical extension, that creators are not exploitive of the original authors' efforts. Think of it as "locked open." So yes, it may be less "free" in the strictest sense of "do whatever you want free" - but we as a society have clearly decided that level of freedom is unacceptable (see eg, murder, slavery). GPL has simply chosen to draw the line closer to rights for the people than rights for the ruling class (ie, end users vs corporations, respectively).
> F is not the meaningful part of FOSS unless you happen to be a lawyer or really enjoy lawyering.
I'd be careful about diminishing the importance of government and legal processes with respect to software (or anything else). It may not be as interesting to some as the technical aspects of software construction, but it absolutely has an impact on what software gets developed and how it's used.
A lack of interest in this fact does not insulate you from its effects.
> "Open source advocates must defend their thesis that freely developed software should, or will with time, be better than proprietary software."
IMO open source advocates should stop defending and start making better free software. (most) People use what's good not what's ideologically pure.
Example : you don't have to defend Linux as a server OS because it's an excellent OS for servers. Now if we're talking as a desktop OS ...
The "I'll use it when it's better" argument is self-defeating. If we can convince people why this software that may be practically inferior is better for our society and get them to invest in that mission, the missing features and bugs can be addressed more easily.
Suggesting a long-term investment instead of a short-term one is a self-defeating argument? Or the suggestion that convincing people about the value of long-term investments is feasible?
The 'argument' is an infinite regression. At the end of it, if we can convince everyone that FOSS is better for our society, we can make it better for our society. But how do we convince everyone? By demonstrating it is better?
>But how do we convince everyone? By demonstrating it is better?
That would be an infinite regession, but free software is not about being better in a practical sense, but being better socially. You also make an assumption that everyone needs convincing, which is also not true.
I think one has to be fairly pedantic to hear the open source argument and not realize that some open source projects are better than others. That said, the philosophy is a bit dated. Good, solid software often comes from having a large user base. With lots of people using a tool, issues get identified quickly and there is more motivation to fix those issues. What can make the open source method so powerful is that it opens the door for numerous people to improve the software. For important projects, like cpython or boost, this often means having significantly more resources than proprietary software. Its not an absolute rule, but it can often play out that way.
So if I'm reading this right, the freedom that in principle someone could make the software better is more valuable than actually having better software?
So the freedom isn't valuable because of what it achieves, it's just valuable intrinsically by existing in principle.
That's not completely bonkers. I've never been to Arbroath in Scotland, but I could if I wanted to. That freedom has value right? But only if I can actually exercise it in practice.
It feels as though the open source community (I'm talking about the Linux Desktop space here) is being pulled in two different directions at the same time.
On one hand, there's a strong desire to have a closer to the metal experience, allowing more veteran users to quickly adapt and modify the software to how they see fit.
On the other hand, there's a push to bring new users in and create a more beginner friendly environment.
Frequently these two objectives clash with each other, especially when trying to find solutions for common problems. It's pretty common to see senior members complaining that the user is at fault for not reading the appropriate doc/guide (not to say the user is totally absolved from reading them), or seeking an "easier" solution that doesn't involve applying advanced terminal commands to people who have rarely used one before.
After using Debian based distros for a while, I switched to Arch. Installing wasn't too bad except for a few weird hiccups. The amount of looping from the help forums to the Arch Wiki, which doesn't contain your issue, would make your head spin.
This piece looks to me like a major strawman argument example.
Let's get back to the sentence that's quoted at the beginning:
"Open source is a development method for software that harnesses the power of distributed peer review and transparency of process. The promise of open source is better quality, higher reliability, more flexibility, lower cost, and an end to predatory vendor lock-in."
1. This sentence is about "open source development method", not "open source software" (two different classes of objects). Software can be open source (if its licence is open source, i.e. acknowledged as such by the OSI) but not developed using the "open source method" (i.e. if, still quoting the initial sentence, it doesn't "harness the power of distributed peer review and transparency of process").
2. A "promise" is just a promise. Whether it's realised or not is still, to a large part, up to the talent and determination of the developers, or organisations, doing the work on the project.
3. Also, "better", "higher", "more" and "lower" should be defined in the context of which object we're talking about. In this case, this is implicit but I guess the comparison should be with similar software not developed using the open source method, i.e. with not distributed peer review and no transparency of process.
4. Last, I'm sure that both FSF and OSI agree that putting "an end to predatory vendor lock-in" is a good thing. Two bad the FSF doesn't mention it in this text.
You probably missed the part where they say how many sourceforge projects have only one developer. That clearly says something about the development method, and how often are these promises realized.
I don't think that proves anything at all. If you said to a free software advocate: 'Freedom is pointless because most of us don't exercise most of our freedoms' that would clearly be a nonsense argument. Enumerating open source projects and counting their contributors is barely more coherent than that.
This article is an example of what over ~20 years has steadily turned me away from the FSF. How does this article do any good?
At the bottom of the page it says:
"Our mission is to preserve, protect and promote the freedom ..."
And yet, they're endlessly stuck dividing the world into smaller and smaller cohorts of purity.
It's 15 years since the OSI. Open Source as a term has done more to get FOSS into business than anything else: Open Source is the Heineken of Free Software whose slogan was "reaching the parts that others can't"! And yet, they still don't get it!
There is value in being controversial as it makes headlines and spreads the word. But, they're too insular - arguably, at a time when the values of FOSS needs more positive explanation and engagement, given the direction of software and technology today.
>There is value in being controversial as it makes headlines and spreads the word. But, they're too insular - arguably, at a time when the values of FOSS needs more positive explanation and engagement, given the direction of software and technology today.
What direction are you referring to? As I see it, Open Source has won. The shift to cloud based solutions means more people are using at least some open source than ever before, whether it's portions of their browser to massive Linux or BSD server farms, they are engaging with open source. Companies that once sold proprietary systems are moving more and more to open source solutions (e.g. Microsoft).
The irony is, of course, that as open source software has won, freedom has actually been damaged.
I feel like FOSS is growing/winning for developers and shrinking/losing for users.
FOSS is used a lot by the major SaaS companies to deliver their solutions, but none of their solutions are themselves FOSS. For example:
Google search, Google email (I mean why not?)
Facebook
E-bay
Amazon
SalesForce
LinkedIN
All of these companies now believe in Open Source for developers, and not for users.
And I think that division has firmed up over time, a new generation of Web-centric developers have come through. Challenger app developers are highly in favour of FOSS for infrastructure, but I don't think they even consider it for the application [0]. For example, Asana, <and a bunch of stuff I don't use>
If anything, I think the steady move towards apps/closed-gardens is a big risk to FOSS as a vision for users, though the vision for developers continues to grow.
[0] what everyone is doing may well be rational from their perspective, we've learnt since the 90's that open source is a development methodology not a business model.
And this is because the experience for users is no longer in the software. Whether I use Firefox or IE matters little when what I actually want is Google Maps and the massive amount of information it provides me. The FSF needs to shift from free software to focusing on "free (as in freedom) service". Cloud services that respect your freedom and privacy.
I don't know how that world would look, but I suspect it will involve heavy government regulation of things like data protection processes and data usage restrictions.
I feel that the direction things are moving, those two categories of software will eventually be the same thing. FOSS was most prominent at the bottom of the stack, first kernels, then major libraries, then UI engines, toolkits, game engines and so on.
Once open source reaches the level of web-app, the users will be directly interacting with a program that's open source from top to bottom.
The things on top of that are data, content, support and other things businesses add. Maybe next those will be become open too.
Or maybe a new layer of proprietary solutions will emerge above the current stack (VR?), and open source will scramble to catch up.
This is why I find it important to use AGPL3+ for my code. We need more developers who are committed to user freedom and who work not only on free software but specifically software with a strong copyleft.
> The irony is, of course, that as open source software has won, freedom has actually been damaged.
Exactly. This shows one of the core differences between the Free Software movement and Open Source. Little is won when there are more users of Free Software (copyleft or not). The goal is not only to displace proprietary software but freedom-restricting practices involving software. This is one of the reasons why the GPL was updated to version 3.
The hugest blow open source received in recent times was Minecraft being bought by Microsoft.
Minecraft is/was what personal computers where for the open source community- a chance not to propagate abilities or rights, but a mindset.
You can do everything yourself- if you can create it - you can share it with everyone.
If they would have added a "Shrink-e-nator" that allowed for complex devices to be shrunken down to one block size and carried around and a GitHub like system to share constructs & turtle programs- it would have been a whole world-wide new generation, growing up with open source.
> hugest blow open source received in recent times was Minecraft being bought by Microsoft
Minecraft is and always was entirely proprietary and closed. People reverse engineered java bytecode to get what they wanted, but that doesn't make it any less proprietary.
Yes, but the whole community that spawned around it- is a open source community by mentality. People today dont have to setup a OS to get a computer going, they dont have to learn anything, thus many of them never get to know creation, and the joy it brings.
Minecraft reintroduced that - lego is programmable matter, and minecraft was programmable matter, and most servers where/are anarchic.
Its true it was proprietary, and it should have been wrestled with a real and better open source alternative from that fate.
Whoever at microsoft saw this danger growing, and voted for hugging it to death, i draw my hat.
I really think you are overthinking this. Minecraft is a fun game. Legos are fun toys. I don't think either was going to respawn a mass love of computer programming that probably never existed to begin with. People who want to create have always had outlets to do so. I would argue with the introduction of robot toys, raspberry pis and arduinos, that creation has never been more immediate than it currently is.
Im part of a open source game community (yes, the can even hold hostile and weird people), and we botched that- our tools got better and better - aka more powerful, and our new dev-base dwindled and declined until it is by now consisting only of devs dropping by from other projects.
Even the map-editor, which used to just take bitmaps, is now in need of several complicated picture formats and a complex configuration file,and requires some knowledge about comp-science to get going.
Others have worked against this trend- there are some neat tools in the making, but the community is not really behind pushing for easy access.
You are right, im over-interpreting that there was a microsoft plan for this. But the perfection of minecraft at open source, that holds true. It was the perfect mixture- easy to get into,hard to get out, instant showable result, a lot of inspiring achievements, which you could investigate yourself and find out how they where made. Depending on the scope-infinite time spend digging on your "repo".
It would have been a generation, for which openness was as natural as the openness of air and water, if the github of builds would have become real.
But correlation is not causality. Lets say in two years time, on some bigger repo, a survey - did you play minecraft before you did start to code. Just to be certain.
The risk of software not being maintained anymore now or some time in the future is important in most contexts, especially if data formats are specific to the software.
This risk is higher with open-source projects, especially if being developed in academia, or by few developers (not a community).
But what about companies shutting down, axing the project, mismanaging the project and stonewalling fixes, shutting down the APIs, revoking licenses, or any other mishap that can happen to commercial software?
I feel like the risk is significant with commercial software as well, and has a more catastrophic failure mode (e.g. with open source and the right license you can fork and continue work on your own, whereas dead commercial software might be truly dead).
It could be that dead open-source projects are more public. If a company goes down, it takes its servers with it. Dead open source projects stay on sourceforge or as a gzip on some .edu server.
It's exactly the opposite. With open source software there's no risk at all. You can always hire developer to support this unmaintained software, because sources are available, or just to export data. With proprietary software you just can't do it legally and if you want to reverse-engineer it (allowed in some countries), you'll need exceptional expertise.
It depends on the amount of money you can spend on it. For most companies, where money is not an issue, it is better to have commercial support for software.
> For open source, poor-quality software is a problem to be explained away or a reason to eschew the software altogether. For free software, it is a problem to be worked through.
is a big unexplained pile of garbage, and without that the whole thing falls.
If it respects your freedoms, then it's a viable replacement for an equivalent program, even if buggy or lacking in certain regards. If it's completely unusable, then obviously that's not practical, but if you reject proprietary software, then the alternative is simple: nothing.
A good example that I'm experiencing right now is Replicant. It lacks many features---not to any fault of its own---that require proprietary drivers. The front-facing camera, wifi, bluetooth, GPS, NFC, certain video drivers, etc simply don't work. Consequently, it can't take video or run certain programs like FF/IceCat that require special drivers for rendering.
But I can do everything I need to do with the phone except for take videos of my children.
So your problem is that some totally separate pieces of code are not open source. Yeah, well, that is not the fault of open source.
Whoever wrote Replicant could just as well have written it in GPL and it would have made absolutely no difference unless some of the driver makers chose to distribute Replicant. Open source is friendlier for gradual adoption of openness, and that can make the process a lot simpler. It also allows users to have mixed stacks which adds a whole new set of people that can contribute code.
> So your problem is that some totally separate pieces of code are not open source. Yeah, well, that is not the fault of open source.
No, I don't have a problem. It's unfortunate, but it is what it is until someone develops a viable replacement. My understanding is that the Android version in Replicant 6 will solve many outstanding issues.
In any case, it's the fault of the developers of those drivers.
All the product qualities have value only as long as they're important to the customer. "Respecting the users' freedom" referenced here is a thing that most users don't even understand, let alone value. In their eyes, free (or open source) software doesn't have this inherent advantage the author talks about — and for most professional users (as in those who make money using the software) the "free as a beer" isn't that important as well.
I use a lot of open source, but there are definitely times I prefer to write something myself because it's faster than learning how someone else implemented it, and needing to hack a lot for special customizations. My favorite open source project is probably three.js.
Honestly I grow tired of rms and his Ivory-Tower ideals. I'm sure that if I lived in my office, had no family, and someone paid me a salary to do nothing but give the same talk over and over, then I too would have lots of time to wax poetic about how freedom is more important than actually being able to use the software.
But I don't.
I have a job, and when I come home I have a family to pay attention to, so I don't have time to hack away at my desktop environment or perhaps my phone's OS. I'll just stick with Android, even though it is not strictly "free", because for me the choice is between nonfree and not having a phone.
...for me the choice is between nonfree and not having a phone.
Surely it would be better to have the free option too? That would just give you one more choice.
That's why RMS takes the idealistic position. If no one does it, the available choices inevitably get watered down to "nonfree or nothing at all", as you put it.
It's like human rights: the principled stance is important in itself and must avoid the appearance of corporate or government influence. You wouldn't want the American Civil Liberties Union to be controlled by Wal-Mart and ExxonMobil. For the same reason, we shouldn't want open source to be controlled by Google and Facebook.
Even while it's not good enough to be a viable alternative, it does act as competition - it provides a "floor" of price/performance that noone can afford to breach and acts as a (weak, but still) limitation against a particular proprietary product exploiting a de facto monopoly position.
Even if now the only real choice is one particular proprietary software package, then you probably can imagine some level of price increase and/or usage restriction where the currently poor free alternative would be worth considering, if you and others put some effort and/or money towards it instead of the proprietary licence - and the proprietary vendor can imagine this as well, so the existence of the free alternative by itself is putting a limit on how much they can exploit the market.
Perhaps in some markets, you may be right. In aerospace there are very few free alternatives and most of the clients are large enterprises who can afford multi-thousand dollar license fees.
> I'd love to live in the world rms lives in but most of the time free software just isn't good enough to be a viable option.
I've made this argument before myself, but I wonder how true it really is. The reality is, in 2016, that it's easy to assemble a free software environment that's far above and beyond anything I'd have had access to back when I started computing. It's true that that fit and finish of the best 2016 FOSS software doesn't match the fit and finish of the 2016 proprietary software, but in many ways, it's more than match for 1995 era software. While that may seem like faint praise, it was easily possible to do good, significant work in that kind of environment.
One other thing I'd add is that I also remembering struggling as a student to pay for compilers and other development tools. ($99 for Microsoft Quick C, etc.) These days, reasonable tools with far more diversity are available for free. Compared to the situation when I started, the current availability of development tools is nothing short of miraculous.
A few months after meeting me, my girlfriend wanted to try a Linux system and I was happy to help her install it. In the end she didn't see what I thought was better about it. I thought I never said it was better, but apparently I gave off that impression, leading to a false expectation.
And it's true, Linux desktops are still notorious despite being actually very stable these days. Open source programs like Firefox, Chromium and Telegram all work the same; and some others like the average file explorer is functionally equivalent too. It's just not better, at least not for most people. It might even eat more battery because of inferior drivers.
The inherent advantage of free software being free (as in freedom) is what the article is about, and that is a good point.