I (like many of us) have been reading and following RMS for many years now. It is fascinating how things that I used to feel were impractical about his ideas ("if proprietary software gets the job done, then why shun it? We live in a world where we must be practical, etc.") now seem imperative.
This point struck home with his mention of education: how frustrating it is that so much of it lives in, say, Matlab, as compared to any other numerical package!
How frustrating that we all depend on Microsoft Excel (or Google Docs) to do pivot tables!
How frustrating that we depend on github to store our code.
How frustrating that we depend on AWS for our servers.
I spend a lot of my professional life migrating from one closed system (eg, deployments on Rackspace to AWS; cc processing from PayPal to Braintree; accounting from Quickbooks to Netsuite) to another.
I wonder if, in practice, so many of these frustrations would have been alleviated if we, as an industry, had adopted the "impractical" view of insisting on using only free software.
A fundamental weakness in the free software argument is that it lacks a solid, broadly applicable proposal for remunerating the effort of software engineers.
A similar argument could be made for the abolition of copyright -- with the same weakness. It helps explain why the free software community has not yet delivered things like Google Docs, Github and AWS.
This does not necessarily invalidate Stallman's argument about the benefits of free software. But without a clear mechanism for producing free software it's little more than a letter to santa claus.
Here's a thought -- what about government support? My impression is that FSF tends to focus on volunteerism and heroics. But government support has produced a lot of the free and open core science in high tech (e.g. DARPA), and via procurement, even supported bringing it to market.
Why not steer the FSF troops towards advocating direct government funding for free software projects and services like a public cloud?
Both copyright and government provisioning of the good are ways around the problem. Both are effective in some situations and of course have problems of their own.
I think the government has a role to play in software development in terms of 'basic research' types of stuff. The US government is responsible for the creation of the Internet, and CERN money created the web. But to leave all software development in the hands of the government does not sound like a great solution to me, either, to put it mildly.
Did you know Siri was developed with $150M of government funding?[1] The government sponsors a lot more than "basic" research. They also support the commercialization of technology through procurement, for example.
Why would government-sponsored software development be any worse than corporate-sponsored? Governments can sponsor multiple teams that compete with one another, after all. That's why "basic science" works even when the money comes from the government -- right?
I'd be interested to here from FSF folks on this. One of the biggest "problems" with our existing Silicon Valley model -- which despite what posturing Libertarians think is very, very heavily based on government support (DARPA, procurement, etc.) -- is basically that the public doesn't receive the full benefit of taxpayer-funded R&D. It's pretty much handed over to the private sector, who take all the profits.
This might possibly be a more profitable avenue for the Free Software movement to take. Advocate that taxpayer-funded research stay Free.
The current market has problems, but I just don't see any evidence that moving software production to the government would be much of an improvement, and it might be a whole lot worse.
I'm curious, how do these problems appear in the core sciences, if at all? Since that is almost entirely government funded.
The results are mostly free and open once published. It's incredibly nichey; many scientists dedicate their lives to a very narrow area of study. Nobody, DARPA and the NSF included, is great at predicating what is going to bear fruit years down the line.
And yet science happens.
It's not subject to a crippling "free rider problem", even though the whole world benefits. It's not massively corrupted by bribery.
Are these just what industry folks say when they want to be subsidized by taxpayers? It's ok for core R&D to be publicly funded, but all these problems will emerge suddenly if they bring it to market, so that last mile has to be left to us.. and the profits too?
Limited time to write, but I'm not sure you fully digested the 'economic calculation problem' entry, and perhaps some associated things.
How do you know how much of something to produce? Sure, science happens. Enough? It's very difficult to say.
With a market, outside of big market failures, you can be reasonably sure that, roughly, enough of the right kind of a given good is produced because of the feedback mechanisms involved: South Korea has abundant food - more than enough to feed everyone. North Korea does not.
If the government were the sole purveyor of software, sure, it would still get written, but I think that not enough would, and that it would turn into a big ugly mess.
Presumably you don't think modern science is a "big ugly mess", yet you've failed to explain why the planning problem hasn't caused that. We could go a step further and ask how science would fare under a market system.
Or for that matter, high tech. Oh, you thought high tech isn't a product of a planned state-controlled economy?
> Oh, you thought high tech isn't a product of a planned state-controlled economy?
It isn't. The government is one actor, and certainly has many contributions - the Internet is a big, obvious one - but there are certainly a lot of private actors that count for something too. Apple, IBM, and Microsoft come to mind. So "product of a planned state-controlled economy" is what I would call either "science fiction" or "wildly inaccurate", just as I would say that "high-tech is entirely the result of the free market" lunacy as well.
> Presumably you don't think modern science is a "big ugly mess"
Actually, my wife does research at a university, and I very much do think there is a lot of messiness in science, and think it likely that it's vastly underfunded - in part due to the fact that it's difficult to bring market mechanisms into play for things like basic research.
I think the ransom model can help a lot in cases like this. Develop closed until the software is good enough and then negotiate a pay-day based on the future discounted value of the software, upon receiving the cash the software gets open-sourced.
This could work for other creative efforts as well.
People tried this out years ago and it never really worked out very well, but perhaps with stuff like Kickstarter, things will change.
Part of the problem with it is that it involves a lot of "make me think" on behalf of the acquirer, rather than simply "ok, I need this, the price seems ok, I'll buy it".
Not that I advocate it, but just for a slightly clearer picture of what that world would look like: a situation where lack of copying restrictions leads to the government being the only one paying for development of software for distribution is a very different one than where "all software development [is left] in the hands of the government" - people would still build on what's there, and may hire people to adapt those solutions to particular needs. And of course that's ignoring the fact that you'd probably have a comparable amount of volunteer FLOSS development going on in general.
> A fundamental weakness in the free software argument is that it lacks a solid, broadly applicable proposal for remunerating the effort of software engineers.
This is the main reason why after a period on my life, where I was a strong FOSS advocate, nowadays I don't care that much what license a software has. I was anyway already living from coding, before discovering the FSF back in 1994.
It is easy to be a FOSS advocate when still living at the parent's home or as an university student. The problem starts when you want to make a living out of it.
That is why in the end, most desktop software is closed source and the majority of FOSS friendly companies that sell software, have moved it behind a SaaS paywall.
Still one can see how few people are willing to support FOSS developers, when looking at how much money they donate, or how quick someone comes up with a fork when money is asked for.
> It is easy to be a FOSS advocate when still living at the parent's home
Not this ridiculous, childish argument again. Your personal issues have nothing to do with FOSS advocacy.
There are many people who are paid exclusively (and well) to write FOSS. There are many more who contribute back to the products they use. The FOSS economy is vast, but difficult to see, largely because so few people pay directly for it, but it makes possible things like Google, Facebook, most of the Top 500 supercomputers, Android phones, most network appliances and so on. The list is much larger than would fit here.
Well, it was the only example that I could think of where a dev wanted to make a living from their software and offer it up as opensource.
So what is a viable model then? I build a kick-ass TODO list app or a casual game. I open source it. How do I make money from my efforts if I want to try and make this full time? This is not a service. I can't charge for support. Donations don't work. What else???
FOSS lacks one key aspect of the traditional software industry - it's never scarce. Any business model based on artificial scarcity will fail. If game development is what you are into and you want your game to be free you may charge for the game while releasing the code and retaining copyright of all non-code assets (visuals, music, sound effects etc).
>Still one can see how few people are willing to support FOSS developers, when looking at how much money they donate, or how quick someone comes up with a fork when money is asked for
Do you have any examples of forks being made due to someone asking for donations?
As for the problem of people getting paid to write open source software, I really don't think that the 'hiding the source code and selling binaries' method is the only viable option.
I also don't think people are unwilling to support FOSS developers, I think it's mainly a problem of convenience (as in payment) and exposure.
Things like Kickstarter helps a great deal here and I think this is a much untapped resource for funding FOSS development. One very successful venture was Openshot, an open source video editor (GPL licenced) which had a goal of $20.000 but which reached $45.000.
Granted we aren't in a situation where it's possible to make a full-time career out of developing open source software through Kickstarter style donations, but I think it's not such a big step as one might think.
I would happily pay a small sum each month for someone (or someone's) to work on FOSS software I use a lot, and I think a large number of people out there would do the same, atleast enough for someone to make a decent living out of it.
Again I believe the key aspect is to make the payment convenient/safe and of course have a way for projects to gain exposure.
Paying a small lump sum each month for continued free open source development of software you use/rely on for your work is in my book a much better long term solution than to pay a larger sum for the latest update of 'proprietary software X' where you are solely dependant on that provider and have to accept their conditions in terms of price/licencing not to mention the intrusive DRM mechanism which are a de facto part of proprietary software today.
Compare that to the FOSS situation where, given that the source code is open (and due to copyleft any forks will also remain open), you will never be locked-in to one provider and their whims even if you are totally dependant on that specific piece of software.
Anyway I'm not pretending that this is the 'perfect solution', but I dare say proprietary software certainly isn't (and I make a living off writing proprietary software and have for the past 8 years) and I think that a shift from relying on proprietary to instead relying on open source code is of great importance for many of the reasons RMS outlined. YMMV and all that.
>I would happily pay a small sum each month for someone (or someone's) to work on FOSS software I use a lot[...]
You're in luck. Gittip is exactly the model you are describing, and it's already active and supporting (in part) the works of hundreds of FOSS developers:
Thanks, yes I knew about this though last time I checked (back around when it was announced I think) it seemed pretty dead (that exposure thing again) but now it seems to be picking up.
I guess it's time to have a look around to see if there are any developers there working on things that I'm interested in/relying on.
I do hope that in the future we'll see more developers who will offer less abstract development plans even though I realise that gittip is not comparable to the project specific type funding as with Kickstarter etc.
> Do you have any examples of forks being made due to someone asking for donations?
CentOS and MariaDB come to mind as two possible examples, that started as a means not to pay for the original projects.
> Compare that to the FOSS situation where, given that the source code is open (and due to copyleft any forks will also remain open), you will never be locked-in to one provider and their whims even if you are totally dependant on that specific piece of software.
This is exactly the problem when trying to make a living out of it.
While it is very convenient for the users, it does not guarantee that you as the author will get paid to work on it.
Tomorrow another guy might come along, fork the project and start getting the money instead.
The Android market is a prime example of this, where lots of apps are just plain forks, with the original authors getting nothing in return.
"CentOS and MariaDB come to mind as two possible examples, that started as a means not to pay for the original projects."
It is a stretch to even call CentOS a fork, and it is even more of a stretch to claim that Red Hat asks for donations. Red Hat charges for services, which include things like:
1. Maintaining binary packages so you do not have to try to track down dependencies by hand or subscribe to ten thousand mailing lists just to make sure your software is up to date.
2. Taking action when you file bug reports (how much you pay determines how fast that action is taken).
3. Sending people to your data center to help you get everything configured.
4. Get official ratings e.g. EAL4+, for those people who need their software to have such ratings.
>CentOS and MariaDB come to mind as two possible examples, that started as a means not to pay for the original projects
Well neither CentOS or MariaDB came to be because of 'asking for donations'. CentOS was forked from RHEL to be a subscription-free enterprise alternative and IS actually being funded by donations, which they ask for.
MariaDB is a fork created by the same person who created and then sold MySQL to Sun for $1 billion (which he was able to do due to MySQL having copyright attribution from all contributors).
I think these were at best very poor examples to support your statement.
>While it is very convenient for the users, it does not guarantee that you as the author will get paid to work on it.
True, it will be a competitive situation, however it's not very different from proprietary commercial development as you will need to continue to add features and improve performance etc, else users will turn to another application which does.
In open source it's extremely easy to fork (ie base upon existing code) however due to copyleft it won't be a fork with exclusive lock-in features as anything can be ported back, so in essence the payment will go to the person/persons who the end users think is providing the best value for their payment, again not unlike proprietary commercial development.
And in the tradition of open source software, I see no direct reason as for why developers can't consider sharing development and thus also the donations.
>The Android market is a prime example of this, where lots of apps are just plain forks, with the original authors getting nothing in return.
Well I think that is primarily a problem of original developers not creating Android versions of their open source applications to begin with, as such we are seeing opportunistic third-party developers creating quick ports of their software.
But is that really a bad thing? If the original programmers aren't interested in supporting the Android platform and someone else provides the service then I fail to see the harm (assuming of course that the licence is abided).
And I do think that when official projects enter the market place they will quickly gain the upper hand over third-party ports as not only are they again the 'official' version, but also I trust that they as core developers can typically adapt the original code better for Android use.
I also think we are slowly seeing more official Android ports of well known open source desktop software, VLC comes to mind.
Looking past whats already been discussed I'm sure there are viable venues which haven't yet been thought of, or at least given a fair chance.
One thing I forgot to highlight was the Valve Steam Workshop donation program which just recently resulted in enough money for two 6 month development contracts for working on Blender.
Again things like this makes me believe that exposure and ease of payment is the big hurdle, not that there's not enough people willing to donate for open software development.
> CentOS was forked from RHEL to be a subscription-free enterprise alternative and IS actually being funded by donations, which they ask for.
By people that did not want to pay RedHat for support.
> MariaDB is a fork created by the same person who created and then sold MySQL to Sun for $1 billion (which he was able to do due to MySQL having copyright attribution from all contributors).
Because they didn't want to pay Oracle for support.
> In open source it's extremely easy to fork (ie base upon existing code) however due to copyleft ...
Not all licenses require copyleft. With BSD, for example, I can take your work, modify it and sell it without giving you anything back.
> But is that really a bad thing? If the original programmers aren't interested in supporting the Android platform and someone else provides the service then I fail to see the harm (assuming of course that the licence is abided).
The forked applications were already in the Android market, they were mostly repacked and resold.
> One thing I forgot to highlight was the Valve Steam Workshop donation program which just recently resulted in enough money for two 6 month development contracts for working on Blender.
How many developers and what will they do after 6 months?
> Again things like this makes me believe that exposure and ease of payment is the big hurdle, not that there's not enough people willing to donate for open software development.
For example, to make a comfortable living in Germany, those donations need to be at least 40K € per year.
>By people that did not want to pay RedHat for support.
>Because they didn't want to pay Oracle for support.
The reason MariaDB was started (according to the author, 'Monty') was because he didn't trust Oracle's intentions regarding the open source nature of MySQL once they bought it through their Sun acquisition.
This fear has been validated as Oracle is focusing on providing MySQL functionality through proprietary modules, with slow development of the open source part of Oracle's MySQL offering as a result.
Your original statement was about numerous forks made due to developers 'asking for donations'. Neither of these examples applies.
Not only that, but both your examples are actually funded through DONATIONS.
>Not all licenses require copyleft. With BSD, for example, I can take your work, modify it and sell it without giving you anything back.
Yes, the development model I described certainly lends itself best to copyleft style licencing.
>The forked applications were already in the Android market, they were mostly repacked and resold
Do you have any examples of Android ports done by the original developers which has then been repacked and resold, I'd like to compare the popularity between the official versions and the third-party versions.
>How many developers and what will they do after 6 months?
This is 2 developers, however their 6 month employment was entirely covered by the very first Valve donation payment so it's a good sign. Hopefully the donations will coninue which will allow for extension of the contracts and hopefully even more developers being contracted.
>For example, to make a comfortable living in Germany, those donations need to be at least 40K € per year
Again I'm not trying to portray this as a viable replacement for your typical commercial developer employment today, but as something which can hopefully be a viable way to fund full-time open source developers in a not too far off future.
I'm not blaming you for being negative, up until the rather recent explosion of crowd-funding through ventures like Kickstarter I was not particularly hopeful myself even though I've donated to successful projects like Blender which is entirely running on donations.
But now, slowly seeing more and more open source projects find funding on overall project or specific feature basis aswell as new venues of funding showing up like that of Valve's donation program and even smaller scope funding like gittip, the concept seems very promising to me.
I am a professional software engineer working full time and making quite a decent living at it. I believe quite a bit that the AGPL3 or other strong copyleft licenses are the moral high road and would be quite happy to work on products using AGPL3.
In terms of monetization, I believe hosted SaaS AGPL3 solutions where companies offer support and maintenance is a reasonably viable business model. I am reasonably sure that B2B "enterprise" sales could be AGPL3 without significantly impinging upon cash flow. Generally it seems that enterprises are buying business solutions, not software; they are paying for the assurance that stuff works.
Open source software alone can't spin up hardware. It makes sense to pay someone for providing a physical service.
Github is getting paid already for developing what will someday influence a definitive open-source interface to git and collaboration.
Photoshop is really nice software. The GIMP is too, and perpetually improving. Closed-source programs are sometimes the vanguard to show the world that a particular idea is important and useful. Any sufficiently useful closed-source program will be cloned into open source in today's environment.
15-20 years ago, the world had a plethora of closed-source options for a gui-based OS. These days, open-source offerings compete on near-equal footing.
It's almost as easy as typing:
apt-get install debian
on any keyboard anywhere to get access to a functional GNU/Linux system. A lot of closed-source programmers earned a lot of money showing us what features of that system have mass appeal.
Is a government-funded public cloud not beholden to the government which funds it?
I think you are overly optimistic, LibreOffice remains less functional than Office 97 and GIMP is a study in bad GUI design - not to mention the rest of the Creative Suite being missing in action. I was recently shocked to learn that I can't rotate text inside a table. Also LibreOffice at one point included the full source code of Thunderbird, some OSS is well written but I am not convinced that it is better at the architecture level when compared to Microsoft's offerings.
After 15-20 years the OpenSource user-land remains confusingly fragmented. Multiple monitors are a pain and applications don't have the understanding of full-screen. Working with icons on the desktops was hard, so now we have launchers, and now we are doing away with the desktop all together - my theory being that drag-able icon interaction was too hard to implement. Distro's are fragmented by their philosophies on software rather then their GUI's - resulting in horrible user experiences. Why the hell does Linux mint come in so many flavors?
Drivers remain a problem, I recently had to recompile modify the kernel because my elan mouse was being detected incorrectly. Ubuntu has made large leaps but I have noticed that their GUI burns power and shutters on systems that are about 5 years old. Also, remember when computers could hibernate?
But, I use Gentoo for work, 'cause kcachegrind is good. I think OSS has delivered something unique to its users but it is whole different from a general purpose operating system like Windows.
I think the comparison between Latex and Word is a good comparison between he OpenSource and propriety world.
When people speak to highlight the progress of OSS I am tempted to ask:
Whenever I've asked people to provide concrete examples of this, their issue has been that they expect the UI to be a 1:1 clone of whatever application they're used to (usually Photoshop) and so they don't know how to do some things, or can't operate smoothly enough as they haven't taken the time to grow muscle memory comparable to what they have with Photoshop.
I'm not a big fan of GIMP, but I want to scream of frustration on the occassion that I have to work with Photoshop (which I don't know well at all).
I don't like the ideas of floating control panels everywhere. They tend to distract, I find myself constantly needing to open and close them and they also occupy a lot of space. The issue has existed for dozens of years, finally the folks at GIMP appear to have sobered up and are working on a single-window mode http://www.gimp.org/release-notes/gimp-2.8.html
Yes, the SDI vs. MDI debate has been a popular and flamy one. It's also a rather subjective issue, and preference depends on what you're used to, what window manager you run (and how you're using it's features), etc. GIMP used to get bashed a lot for this UI choice, but the haters often forget that Photoshop (originally Mac software) defaults (or at least used to, and I believe it still does?) to SDI in the Apple land.
I'm glad that those who want MDI have finally stepped up and implemented it. Recently I've been using a tiling window manager that doesn't deal particularly well with floating windows. Now I have the choice.
"what window manager you run (and how you're using it's features)"
Right, exactly. I've found that starting GIMP in ratpoison is a little glitchy - two windows spend time competing for a single frame until I pull one of them into another frame - but once I've done that I quite like the experience of using GIMP in ratpoison.
In the context of this discussion, GIMP wins. You (or a group of users) are "free" to create an alternative interface for GIMP. The rest of the features of GIMP are readily available to be integrated.
The GP on the other hand is forced to use the exact same interface that Adobe prescribes for Photoshop. Moreover, if Adobe shuts down tomorrow, it is the end of the road for Photoshop development.
One that I've hit recently is that an open colour picker window prevents quit from doing anything. It also doesn't show the colour picker window, so I don't know why GIMP isn't quitting, just that pressing a common key combination (cmd+q) doesn't work for no good reason (colour picking is not an operation that needs to block quitting).
What an appalling Gish gallop. Well, as usual, the only thing to do is point out the first wrong thing. Conveniently, everything you said was wrong, so I can just pick the first clear statement you made: you can rotate text within a table.
--edit, not sure why links to the bugzilla describing the issue are giving me downvotes...
--edit 2 : I actually hit this bug when making a flyer, the bottom had little flaps with our groups email. It broke PowerPoint compatibility, and I had to reboot into Windows.
--edit 3: You made laugh when you mentioned that "You can't invert the text quite so easily, you have to draw a rectangle with text in which can then be rotated any arbitrary degree." That sums up whats wrong with many user side facing programs. Too bad LibreOffice doesn't support pipes :-)
Probably because both those links specifically say you can rotate text 90 and 270 degrees. You can't invert the text quite so easily, you have to draw a rectangle with text in which can then be rotated any arbitrary degree.
Next time you trawl bug reports looking for anti-Linux material, try reading them more carefully.
"I can't rotate text inside a table". My memory is perhaps wrong, but I thought it was possible earlier in openoffice than in word. For information, in the menu "Paragraph style", you have a tab "Position" where you can select the "Rotation / scaling" "90 degrees".
I think OSS has delivered a lot of excellent products but the issue that remains unsolved is the UX side. UX is critical in mobile devices. A bad designed interface makes you switch to another application beyond the product features included.
LaTex is awesome, period. But the learning curve is too high for common documents, which is fine because it was designed to be a great typesetting, not a word processor.
On the other hand, Word is great too, I wouldn't write a book on it but most of the time I am not writing books, as for free alternatives to Word, I find them sufficient but I would rather pay 10$ per month to use word and excel than work around libre office and its limitations.
> Any sufficiently useful closed-source program will be cloned into open source in today's environment.
That is a parasitic and unsustainable model. Even if you could find enough volunteers to copy the work of every bit of successful commercial software, that original design and development work still needs to be funded.
Other commenters have noted the quality situation of many of the cloner projects out there. Even cloning takes work and sustainable, quality work needs remuneration.
"A fundamental weakness in the free software argument is that it lacks a solid, broadly applicable proposal for remunerating the effort of software engineers."
Quick counterexample: the many thousands of programmers whose job is to adapt or maintain free software for the organizations that use it. That includes people working for small companies, large companies, the government (civilian and military), non-profits, political campaigns, etc., etc., etc. It is almost never the case that any software package meets all its intended users' needs, and someone needs to adapt the software when that happens.
"without a clear mechanism for producing free software it's little more than a letter to santa claus"
Clearly such a mechanism exists, as free software is very important and widely used.
"My impression is that FSF tends to focus on volunteerism and heroics"
That bit from GNU on charging whatever you want for distribution of free software is, frankly, a disingenuous argument.
Ask any book or music publisher: you can't expect to profit much from distribution if you can always get the same thing for free next door. If you get anything, it pretty much amounts to a donation.
I think we all know that here.
Yes, you're right that there are organizations that subsidize free software development. But this is not a broadly applicable mechanism, in the commercial sector at least. Not every commercial company can pursue a leveraged approach.
This is exactly why I suggested a government-funded direction.
Free software is certainly widely used, but it pales in comparison to non-free software. Hence the previous poster's complaint about the lack of a competitive free Google Docs, AWS, etc.
"Ask any book or music publisher: you can't expect to profit much from distribution if you can always get the same thing for free next door. If you get anything, it pretty much amounts to a donation."
Lies. Using music as an example, look at Spotify (or Itunes or whatever) as an example vs. downloading mp3s for free.
I'm aware this is anecdotal evidence, but I haven't downloaded any music for years due to Spotify being so convenient to use. Neither have any of the people I know.
I strongly believe that given a superior distribution model, money will follow.
> "It helps explain why the free software community has not yet delivered things like Google Docs, Github and AWS."
I'd bet that each of those has benefitted hugely due to FOSS. I might even go so far as to say they wouldn't exit at all without it (though I don't know what underlies Google Docs).
FOSS is great for things 'under the hood' but very few things in the mainstream are FOSS throughout. I find that interesting, as it's indicative of the different pressures in open-source vs closed-commercial environments (remuneration is one such issue).
One reason would be that (particularly in light of some govt agendas revealed recently) that govt funding would inevitably steer and even corrupt the development of FOSS.
There's even talk about how the NSA, who contributes to open cryptography standards, may have been deliberately placing vulnerabilities in said standards.
> It lacks a solid, broadly applicable proposal for remunerating the effort of software engineers.
How about this: abolish capitalism. I mean, sure, I agree with you: under capitalism, where people are denied access to the necessities of life (food and shelter) unless they sell their labour to capitalists in exchange for money, it's difficult for people to find the time to fully commit to developing free software. The solution to this is to abolish the system that denies people access to the necessities of life. And this is not crazy: people already live in houses (and for the ones that don't, there are tonnes of empty houses that nobody is using). Food literally grows on trees (unlike money). There's no good reason why everybody couldn't automatically have a place to live and enough food to eat without having to pay rent to some capitalist. I wish more free software advocates understood the more radical implications of the free software argument (and as you say, the same goes for anti-copyright advocates) rather than pretending that it's ultimately compatible with capitalism. It isn't, and we shouldn't want it to be either.
Why not steer the FSF troops towards advocating direct government funding for free software projects and services like a public cloud?
"Government funds" come out of people's paychecks. What you're asking for is that everyone be forced to contribute to projects that you approve of, regardless of whether those people value those projects or not. I can't see a valid reason why someone that doesn't know about, need or want "free" software should be obligated to pay for it. What about their own needs, desires, goals and concerns? People need to be left free to keep the money they earn so that they can better their lives and pursue the things that matter to them.
I really like sports cars - I like them a lot. I think they're inspiring to look at, fun to drive, and I could make the argument that there are a lot of people, ranging from children to the elderly, that feel the same way about them. Each of us would love to own, maintain, collect, restore and design new sport cars. Generally speaking, love of these things is connected with all sorts of other interests, including science, math and engineering, art (sculpting, photography, etc.) Just like software, albeit to a lesser extent, sports cars have have some arguable "public benefit" (i.e., a concrete benefit for a finite, delimited subset of individuals in society).
Now imagine that we came up with some kind of government program would shower sports cars on anyone that applied for them. Why should my aunt - who barely drives - be forced to subsidize my sports car hobby, at the expense of her own desire to travel on a cruise ship to Alaska while she still can? (She is 79 years old.)
At this point, maybe you'd say, "Well, software benefits everyone far more than cars do. After all, even people that don't use computers at all benefit from the improved efficiency (lower prices, increased quality, etc.) that computers bring to our economy. So all we're doing here is recouping costs from free riders."
I don't have the space to do much more than to hint at the answer. This is by no means exhaustive, but one fundamental issue is this:
When you say, "benefit," what do you mean? In my view, a thing is not ultimately beneficial to a person unless they come to see for themselves how it is a benefit to their lives, via their own independent judgement. Sure, identifying what's good for other people is fine. But then insisting that they agree with you if and when they do not actually agree subverts their independent judgement, which is the very thing required to produce and select the things needed to improve their own lives. No one should be compelled to eat a meal that wipes out their olfactory bulb and tastebuds.
> What you're asking for is that everyone be forced to contribute to projects that you approve of
That already happens. Such projects include the Internet, microprocessors, airplanes, voice recognition, robots, lasers, satellite view, highways, engines, the assembly line method of production, and many, many more. Except of course it's not projects I approve, it's projects the government approves.
The question is not whether the government should be funding things -- that's already the case.
The question is how much control should investors have over their investments. And in the case of our high tech economy, taxpayers are the biggest investors.
Absent copyright, developing things that are non-excludable is a collective action problem. Just about everyone agrees that government is an appropriate venue for some of these (frequently defense, policing, roads, schools). You can draw your particular line where you want, of course, but I reject the notion that there is a clear, morally motivated hard line that everyone should obviously subscribe to between any of these things and other projects. Government is how we coordinate certain kinds of efforts that won't permit individual opting out but which (hopefully) benefit us on the whole.
Our governments already pay for software. In fact, they already pay for custom software (you can imagine that large packages "off-the-shelf" need a lot of custom work.) Why couldn't a requirement be that that software be released as libre?
Could a government insist that, if they did fund a next-generation sports car (humvee, armored vehicle, sedan, motorcycle, bicycle) that the schematics and designs of that car be released in a libre fashion? Using libre formats?
They could certainly insist on that. The company would likely charge more in some cases because they'd know they could not recoup anything from selling whatever to others. With copyrights, sometimes you can say "ok, it's going to cost X", knowing that you can sell the exact same thing to others at X or close to it. Without that, you have to make all your money from the first sale.
Granted, this is not true in all cases, there are exceptions, and so on, but the logic of it makes sense to me.
> In my view, a thing is not ultimately beneficial to a person unless they come to see for themselves how it is a benefit to their lives, via their own independent judgement.
Vaccines.
Most people don't spend a lot of time thinking about how they don't have polio.
Not having polio benefits you even if you don't think about it. Your being vaccinated against it benefits me even if neither of us think about it.
I'd argue not having polio is "ultimately beneficial to a person" regardless of their beliefs about value.
Jeebus, I jumped the shark by upvoting you after reading only the first three paragraphs (internet attention span, I know), but the last two make me want to double-downvote you. You seriously want government to spend even more, and on funding some people's hobbies FFS? Why not give handouts to, let's say, cabinet makers, to make beautiful furniture for everybody who wants it but doesn't want to pay for it? Talking about cognitive dissonance...
"A fundamental weakness in the free software argument is that it lacks a solid, broadly applicable proposal for remunerating the effort of software engineers."
How god damned magical is it that not only do all of these things exist but that they are cheap and easy to access. The many, many, many orders of magnitude increase in productivity and possibilities enabled by software in the last 50 years is beyond astounding.
I don't believe he said that the services themselves are frustrating, but rather the fact that we rely on individual providers (with proprietary implementations) of these services so much.
False analogy. For the average person, these things add more freedom (freedom to tinker) while removing the freedom to look at the source. For example, most scientists don't care about the latter when it comes to using Matlab.
"I wonder if, in practice, so many of these frustrations would have been alleviated if we, as an industry, had adopted the "impractical" view of insisting on using only free software."
We'd live in a world with a lot less software and where computing was restricted to hobbyists. Propietary software exists for a reason, and its not because the world is run by the selfish and evil.
I love free software, too, but the thing to remember, and this is true for any discipline -- a massive amount of people will support something because they have the incentive to get money from it. A lot of people love programming, but a lot of people love money even more, and if the geniuses who went to Harvard and Stanford and UC Berkeley decided that they could make much more money in biotechnology or surgery, then that's one less Microsoft or Google in the world that's bad for us users. If these people loved programming more than money, their software wouldn't be proprietary. It's much much easier to make money off proprietary software than free software. Yes, there is dual-licensing and support and donations, but ultimately you cannot directly and legally charge users, your biggest market, to use your program; this logic is supported by some evidence: recently Red Hat reached a billion dollar valuation, making it the most valuable company to be majorly devoted to free software. That's amazing. Bill Gates on his own is worth $56B, though.
I think most people, and most people encompasses the people that make the proprietary software that we rely on most, look for money first and idealism next. We'd have RMS's work and some of the early MIT and UCB hacker's work without the incentive of money, but to get so far as multibillion dollar companies that produce programs 99% of us are reliant on? Impossible without the incentive of money.
Money isn't only incentive. It also measures the peoples needs. "Do my clients need XXX?" correlates with "Do my clients will pay $$$ for XXX?".
For organisation, replacing the need by the money allow to take explainable decision. Fame, power, influence or passion are far less easier to measure than money.
I think you are looking at it wrongly. Most would look at this and say "How wonderful it is that we can use these things to do X."
Honestly, no one put a gun to your head to use Matlab. No one prevented anyone from creating libre free Matlab.
For some reason I think someone was incentivized with something, not sure what it was, to create such a useful system for doing numerical analysis.
You say you've spent your "professional" life doing these migrations ... Do you regret the money you've made? =)
I mean working with proprietary formats has issues but I don't think it is always intentional vendor lockin.
As far as the integrating payment systems and github etc etc ... It makes wonder if you're under the impression that there are no integration issues with free software systems??
You're forgetting that it may seem like no one put a gun to our heads to use Matlab, if you're in school - you're required to use it. The same goes for MS Office, Visio, and hundreds of other programs.
Can you imagine how good Octave, LibreOffice or any other open source application would be, if instead of forcing users to buy proprietary versions of software schools and/or students were required to donate just 1% of the cost of proprietary software to open source alternative?
So, do you think if these proprietary systems didn't exist, you would be able to do more interesting work, or not have a job at all, or just spend your time on writing scripts that convert file formats between version2.forka of database software x and version48.forkb of software x?
If you find your work mind-numbing, maybe you should have a look at yourself and find other work, rather than blaming others for the problems in your life. (ooh those evil software companies, they make me work a job where I get to sit on my ass in an office all day and still get paid double (/triple/ten times, depending on where you live) the median salary).
This point struck home with his mention of education: how frustrating it is that so much of it lives in, say, Matlab, as compared to any other numerical package!
Check out GNU Octave:
www.octave.org
It's not Matlab (particularly toolboxes), but for day-to-day work, it's fine, and it will run a lot of *.m files. All of my thesis data analysis is done in Octave. I could choose Matlab (we have a site license), but I want anyone in the world to be able to check my work exactly.
"but I want anyone in the world to be able to check my work exactly."
The problem is not that they can't, the problem is they don't want to. Solve the real problems first before worrying about imaginary ones. (this is one of the main flaws in most of GNU rhetoric, btw, and I say this as someone who 10 years ago made the 'access to source code is a moral right' argument and actually believed it).
I've no illusion that most people will never be interested in my source. For the 2-10 people that are, being able to duplicate my working environment may be of great use.
I too believe that our civilization should be build on free software and I think the best existing financing solution is crowdfunding. The idea is based on free market capitalism and while it has been around for a while we now know that it works. Kickstarter has demonstrate that it can not only raise millions to finance hardware development and commercial games [2] but also substantial money for free software [3].
I truly believe that at this point, all we need is to propagate this idea and help FOSS developers to crowdfund their projects. Once a project has collected enough money for a first working version, subsequent development can be financed by donations, support or merchandizing. Major updates could again be financed over crowd funding.
It's hard to get cheapskates to pay $.99 for an app, let alone donate -- at least not at a level to compensate for a company to make any money or even pay the bills. When clothing stores start operating as donationwear then maybe donationwaring the software industry could follow.
The problem is that many people fail to see software as an actual product that needs to be produced. Hopefully Ducati can lead the way with a donation-sponsored business model. Maybe that could illustrate the free-rider principle more clearly.
> many people fail to see software as an actual product that needs to be produced
True, but many is not all, by far. Star Citizen (a software product) raised over 19 million dollars [1] in crowdfunding! Way more than .99$ per app and yes, it does make enough money to compensate for a company.
And FOSS is already developed today and will be in the future, based on donations in terms of money and time as well as other income sources like support and merchandise. Just like I stated in my post.
The difference between the traditional donation model and the newer crowd funding model, ala Kickstarter, is that there is a clear feedback to the supporter. To me this is a game changer. When I spend money on, say, git-annex (which raised over 25,000 USD!) I know that the project won't get build in the proposed form if the goal is not reached and there are clearly defined stretch goals as well. Depending on the amount I also get a corresponding rewards, e.g. my name in the credits or special merchandise articles. Its also great feedback for the developer. If they can't raise the money they know that there time might be better spend on something else.
I'm really surprised that seemingly every commercial business is getting crazy about Kickstarter now, but so few FOSS proponents see the appeal.
PS: What is Ducati? Google only shows the motorcycle brand.
Here's the thing: Kickstarter is one of those SaaSS systems that RMS was arguing against.
Idealism gives you a target but not a roadmap. As others here have said without Propietary software much of what we have in software just wouldn't exist. Kickstarter wouldn't exist. If the whole world were full of enough people who shared RMS's Ideals then his Ideals would be workable. It's a bootstrapping problem. And RMS spends so much time complaining about the bootstrapping that it sounds like he'd rather not get there all if that's what it takes.
Maybe that's not what he actually thinks but that's how he comes across.
Me? I do want to get there and I'm willing to bootstrap to do so.
> Kickstarter is one of those SaaSS systems that RMS was arguing against.
I upvoted your post, but this statement misses the point.
How does it matter if Kickstarter is SaaS? The hardware that RMS software runs on isn't free either, neither are his tea bags. Maybe, it would be good if these things were free as well, but I think thats not what we are discussing here.
> idealism gives you a target but not a roadmap
I thought I had spelled out quite a detailed and realistic road map in my last post, no?
> As others here have said without Propietary software much of what we have in software just wouldn't exist
True. Maybe it will always stay this way, maybe it would change if more FOSS was funded by 'crowds'. Git-annex for example is quite innovative I would say. But again, that misses the point a bit. We would like to have the software we rely on to be free. Where the original innovation came from is less important in my opinion - as long as it sustainable. Let me know, if one day innovation dies out because FOSS is eating the launch of commercial software. I can't imagine this ever happening.
I am not advocating that we should not pay for commercial software any more. Just saying we should do our best to fund more FOSS, especially for core technologies like kernels, developer tools, file sharing, communication etc.
So for instance, you could pay Dropbox for their service but also donate to something like git-annex. In the long term git-annex might be cheaper and better so it would pay off.
In my opinion, crowdfunding with a pay-upfront mechanism (ala Kickstarter) is THE answer to one of THE major problems of the 21st century:
How can we compensate creators for information that is no more under their control once released?^
This applies to software, hardware designs, movies, music records and books.
^ Note that an alternative solution that has been proposed is DRM and even police state like laws. That alone makes it a very important problem for everyone.
Nonsense, we don't depend on Excel for pivot tables, lots of other software can do that, likewise we don't depend on github, AWS or any of the other examples. This is total hogwash.
It's a service designed around a free and open protocol, the other examples are binary software packages. You can use any local git repository as a drop-in replacement. Github is neither a dependency nor a walled garden for most people, but the other software on that list definitely is. Git is the tool people use, Github is just a public repository. There is no question that Github is itself closed source, but it doesn't really share key attributes with the other examples mentioned either.
People are getting very dependent on Github workflows beyond just the act of being a git repository. If all you need is a remote git repo, use gitorious, it is open source.
No, people use github for the social features, the wiki, the issue tracker, comments, etc. These features are not portable, and as a result if github went down, your entire workflow suffers.
His point was that people could use bare git without github, but many use Github, which isn't free according to three of the four freedoms RMS lists in the essay [1], no matter how much we hackers might love it. Lots of people love Excel, too, but that doesn't make it free software.
[1] Github:
(0) freedom to run the program as you wish - YUP
(1) freedom to study & change the source - NOPE
(2) freedom to make and distribute exact copies - NOT SO MUCH
(3) freedom to make and distribute copies of your modified versions - NO
It's actually worse: you don't heave the freedom to use old versions when they push out an update, and you may lose access to it completely for any number of reasons, including 'we don't like you'. Granted, part of this lack of freedom is fair due to hosted services being different from just plain software, but the power over the users are there nonetheless.
Again, and I mean this in a nice way, how are your complaints remotely relevant? I'm really trying to understand...?
So if someone builds me a house and I pay them ... Is it immoral for them to have built the house without being some kind indentured servant to me for the rest of their lives doing whatever I want for whatever reason ...
That seems like a weird thing to say, but what you're saying is just as silly. Should I have access to unlimited resources from this builder and if they do not consent to my demands then they tyrants?
It's a service that I pay for and I have agreed to conditions of said service and I can choose to stop paying if I am not happy or I can sue if I feel the terms have not been met. I don't see the problem but I do see clearly the alternative is charity.
Someone else responded, a user jszz on this thread and said something crazy about Germany and how we need become a self-sufficient commune. I'm not sure what it meant and I'm not sure that anyone that holds this view knows what they mean. =(
I don't really know anything about you, but I couldn't really understand the free software movement until I got that we are not advancing software as a community.
We are more like thousands of egotistical entities, the great majority in a "consumer" behavior, using software that someone (the "producer") writes and only he/she has the right and the capabilities to read and change.
In your house analogy, it was like someone (let's call her "Sarah") built a house for you, but in a way that you could not build a new room, fix the plumbing or even get your stuff out of it by yourself. Instead, you had to call Sarah, who would then decide if it was worth her time, and what should be the price.
You could just move to a new house with just your personal items and furnish it all over again, but as we inhabit bigger and bigger digital mansions, we can't escape the fate of always losing lots of things when we want to change. This situation may even coerce us to never change at all, even though there are so many fucking problems with this damn house.
I feel this is getting too metaphorical, so I'll end up here. When Stallman's point gets into your head, you see that this is just an instance of a more abstract class of problems with our societies (he has a wide range of controversial opinions outside software world). It really is about being part of a community.
In this analogy, they don't give you the blueprints for the house, and it's not available upon request. Wanna drill a hole in the wall to hang a picture? Better hire the same contractor or risk drilling into pipes or wires. Is the contractor no longer in the market? Well, tough luck!
Github provides an amazing service, but it's closed source SaaS.
But unlike many SaaS source-control sites, all your data on github is fairly easily available for bulk download via their APIs... It would of course be annoying to move to a different site, simply because any change in familiar tooling can be annoying, but it's quite doable to automatically migrate everything over.
I think github seems to have been fairly careful toeing the line, and I don't get the sense of being locked in the way I do with some other sites. They appear to be content to compete on the basis of usability and ease of collaboration, not lockin.
Take a look at Newsblur: There's a free tier, or you can pay for hosting, but at the end of the day the sources[0] for the software you're using are freely available so if the service fails, or the author makes changes you don't like, or you find yourself needing customisations that aren't appropriate for the author to provide, you can keep using the software by hosting it yourself or finding someone else to host it for you.
You might also find the Franklin Street Statement[1] to be an interesting read -- it tries to set out a useful collection of ideas for services that respect the users' freedoms.
The alternative is for the whole US to strive for actually being the alternative; the self sufficient decentralized yet United States. If Germany, Singapore, China and Japan provided alternatives, while we continue to hold on to old ways, like: not investing as much into energy efficiency, or thinking outside the box like Germany -> we will lose to white labeled and foreign hosted OSS. To solve this we would need a social market economy [0] or at least a mandate to tax for fractionally financed civic OSS, open hardware (self sufficient micro servers), data, and R&D.
I knew I went too off topic...What I mean by Singapore is the mandated efficient renewable buildings, Germany: it's social market economy and how they are paying people for energy, China: it's investment in renewable energy, Japan: the same. The alternative of a centralized structure is a decentralized or hybrid one. The equipment for what I suggested in other terms would have to be treated as a public good, taxed by the services it provides or as a government subsidized one. Which as an efficient microserver: communication, entertainment, and the option to subscribe to private services. The ideology here is getting the most people to use a box by paying for them to maintain it. The box just happens to be energy efficient to the point where one can exchange selling their excess electricity or storage for private services / nodes / other microservers. Another form of continuing this route is to be able to use energy, storage, or cpu / gpu as a currency or collateral. Also if one decides to give up control of their servers for hosting, or scientific research one can. It's this idea we're trying to evangelize at r/Nucleus.
I believe freeing software will eventually free the regulation on transactions, energy, and communications. Something like a government subsidized team that controls a vagrant-mesos-docker for connecting CJDNS nodes (the SaaSS). Time Warner cable and all of these other companies who don't run OSS are stopping the digital divide and prohibiting peer communications. WebRTC and websockets are enough!
The topic should be on why BSD software is more important right now than GPL. Since we're talking about freedom.
Don't bother making sense of extremist blabbering. I used to try to reason with religious fanatics, people who believed in homeopathics or quackery like ear candles, socialists, etc., on the basis of 'everybody has their right to an opinion' and 'it might be true'. And those two are still true, but the right answer to those two assumptions is 'yes but not all of them are equal', and 'the onus of proof is on those with prima facie extraordinary claims'.
Don't get sucked into arguing on a rational basis with people whose sole method of conviction is appeal to emotion and looking profound through obscurity.
I'm here to disprove my convictions and find the correct answers if something is not possible. Not to boost my ego. You could have easily said don't waste your time: he doesn't know 100% what he's talking about.
It's a service, doesn't really fall into the same category as software you install on your machine (which you trust). You have no reason to trust their servers, but ultimately you're not giving them control over your hardware anyway.
I love github, but as it is not open source there's no point pretending it doesn't share some of the frustrations that closed-source software bring. Out of all the software I use day-to-day it is the only one that regularly has me thinking "I wish I could change this, or look for an issue or PR discussing changing it". I want mathjax support!
This point struck home with his mention of education: how frustrating it is that so much of it lives in, say, Matlab, as compared to any other numerical package!
How frustrating that we all depend on Microsoft Excel (or Google Docs) to do pivot tables!
How frustrating that we depend on github to store our code.
How frustrating that we depend on AWS for our servers.
I spend a lot of my professional life migrating from one closed system (eg, deployments on Rackspace to AWS; cc processing from PayPal to Braintree; accounting from Quickbooks to Netsuite) to another.
I wonder if, in practice, so many of these frustrations would have been alleviated if we, as an industry, had adopted the "impractical" view of insisting on using only free software.