To me, it's the same issue. The way I see it, "Right to Repair" is the hardware equivalent of Free Software movement.
(It's also widely misunderstood. It's not, as some claim, assuming everyone can be a software developer / electrical engineer. It assumes everyone can either be one, or find one in their neighborhood - being able to ask, and possibly pay, other people to fix or improve your software/hardware is the goal here. Free Software and Right to Repair want to let people do what people do best: self-help within their communities, be it personally or via local markets.)
Also somewhat related, "Motor Vehicle Owners' Right to Repair Act" covered similar ground for automobiles. This help discourage auto manufacturers from keeping the digital aspects of repair (error codes, etc.) from being too secret.
> While the most egregious violations of user freedom come from companies publishing proprietary software
I would say the actual most egregious violations of user freedom and privacy come from companies that are powered by open source infrastructure and that actually are big backers of open source.
"Software freedom" is the big enabler and partial driver of centralized computing. It is the big enabler in that if you were paying a per CPU license for Windows NT Server, you would go out of your way to make sure that as much as possible was being done on the users computer. In addition, the most widely used copyleft license, the GPL, actually penalizes you if you make a client program that uses GPL libraries(by requiring you be release the source code for your program if the user asks for it), but happily allows you to instead run the program in a datacenter and never have to share the code with the user. Thus you are incentivized to put as much of your program in the datacenter as possible and not on the user's computer. The thing is, if software is on the user's computer, whatever the license, they actually have far more de facto control. They can control what is getting shared. They can reverse engineer the files to allow perfect export and backups. They can own their data. They can also verify what the software is doing with the data (even if it might take a lot of efforts). Once computation happens in a data center, the user is at the complete mercy of the corporation. The corporation can make it impossible to get a true copy of their data, requiring use of legal means instead of technical means. The corporation can store and share and process the data in ways the user has no control or even awareness about.
Sometimes I wonder if we would have been better off from a society point of view and user freedom and privacy, if the old Microsoft vision of the PC as the hub of a users digital life, and users paying money for software they find useful, instead of what we have now where the cloud is the hub of a users digital life (even stuff that could work on a local network, will by default try to contact the cloud), and users are provided useful software in exchange for their engagement, privacy, and attention.
While you are making valid points here I think the primary driver in the push for cloud was the rise of the smartphone. Suddenly regular people wanted to have the same view into their data both on a phone and on a computer.
There are ways to do this peer to peer of course, but google and apple clouds became default.
Google already was a website so it only made sense for them to put their office suite next to the other web stuff as it was "their turf". Of course license cost would have become an issue here. But they could have developed their own backend as well.
I would rather see it as: there was demand and it was useful. Google knew they wouldn't beat microsoft on a Windows machine or Apple on ios/Mac. So they created their own plattform and made sure it is as much web as possible (so Andeoid was/is only a way of pushing their web plattforms).
I think cloud had little to do with smartphones per se. The transition started earlier, and I believe the main drivers were what pg made his money on and praised so much in his early essays - the ability to write your software in whatever stack you wish, bugfix and update it for all users instantly, without having to ship anything to anyone... and ultimately, having a justification for subscription revenue.
There were a lot of technical things that had to fell into place for the cloud to spread everywhere - web technologies, ubiquitous Internet access, and perhaps even smartphones - but ultimately, the driver was always the same: if you control access to something, you can charge rent for it.
I'm convinced that the biggest thing the Free Software movement could do, is build a convenient voluntary payment model (and software implementing it very conveniently).
Right now, by far the most convenient model is ads/tracking - it's literally on every platform that respectively can display images/text and doesn't have extensive privacy features, which is basically every platform on the planet. And it's so convenient it requires zero clicks to pay with it!
Ads/tracking cause the "you're not the user, you're the product" problem, which IMO is the reason why everything needs to be online - if you're offline, they can't send you ads or receive your data.
And a proper user-based funding mechanism would also make corporate funding less necessary, which will enable Free Software devs to prioritize consumer users instead of having e.g. Canonical's "but most of our money comes from servers" problem that ends up de-emphasizing the desktop despite their best intentions.
> I would say the actual most egregious violations of user freedom and privacy come from companies that are powered by open source infrastructure and that actually are big backers of open source.
they shy away from the Affero GPL, and their key elements are proprietary.
That's why AGPL software can be divided into small ideologically motivated camp (won't use it, but I respect the choice) and "due diligence mine" as part of dual-licensing AGPL/proprietary setup where the point is to trip people into becoming paying clients.
I'm not saying other licenses aren't. In fact, I'll repeat, I respect AGPL used for ideological reasons. But I find it rarer than cases where it's used as part of bait-and-switch tactic by startups preying on other startups (a bit of simplification, but that's non-trivial component of it).
What I find interesting is that the proprietary software tmost egregious in its impact to the customers product is device drivers, arguably where the free software began. Yet the devices drivers in and of themselves are not exactly worth anything, the company is selling a physical product after all.
I agree with the general sentiment, but the author gets a few things simply backwards:
Individual hobbyists care about tinkering with the software of their choice, emphasizing freely licensed source code. These concerns do not affect those who do not make a sport out of modifying code. There is nothing wrong with this, but it will never be a household issue.
This couldn't be further from the truth. Consider the simple example of tinkerers making a patch that disables DRM or telemetry, and sharing it with "those who do not make a sport out of modifying code". This is no hypothetical - she writes:
Consider third-party clients to centralized unencrypted chat systems. Although two users of such a client privately messaging one another are using only free software, if their messages are being data mined, there is still harm.
The Off-The-Record plugin (https://otr.cypherpunks.ca/) encrypts otherwise unencrypted centralized chat systems. It works even if you "do not make a sport out of modifying code".
> ...large corporations claim to love “open source”. No, they do not care about the social movement, only the cost reduction achieved by taking advantage of permissively licensed software.
My job has me convincing and coaching large companies through their first open source steps. In my experience, the cost reduction is a nuanced issue in the boardroom. Mainenance and support costs (and availability), trust of strangers, loss of control, and the clash of openness in a highly secretive world, are strong motivators against the relatively tiny (perceived) expense of building it yourself.
The argument that carries the most water in my experience, is the argument of interoperability. For reliability at modern scale, multi-cloud is an inevitable step. And open source technologies drastically reduce the complexity and cost. Why would you build on Azure's (excellent) CosmosDB and set yourself on a course for painful integration and migration, when you can build with Cassandra and avoid those issues?
To the author's list of important areas impacted by software freedom, I would add diversity and inclusion. It's nice that a bunch of white guys with West US culture build great solutions to their problems that makes sense in their cultural context. But those off-the-shelf products don't help people operating in a different cultural context. With open licenses, it's possible to modify software to fit your own needs, like the users who got braille support into MuseScore, or the Swiss companies who built privacy-friendly email services to compete with gmail on Ux.
However, I disagree with the overall article. These important issues (sustainability, privacy, copyright, outrage-for-profit, D&I, etc) are not components of freedom. They're areas that benefit from user freedom. Software can be "Free as in freedom", and still be Bad For You. It's OK to have open source gambling software, or video games, or bloated office suites. The point is that users get to choose.
This is a nice argument for some kind of "software for good" label, like the kosher K on your food. We could certify that this software:
- respects the 4 FLOSS Freedoms
- does not leak or store _any_ information remotely
- is (reasonably) free of unnecessary hardware requirements and bloat
- does not try to manipulate you for "engagement"
That's true about all freedom (not only software).
That's why we need to be able to hack/boot other OS on our hardware and that's why we need to be allowed to have guns.
Another thing is that corporations use the open source tools to shape industry practices, often in ways which result in net loss of productivity for themselves and their competitors.
Corporations in the software industry are often monopolies. They derive their success primarily from network effects - It has nothing to do with the quality of the software that they produce or their development efficiency; in fact, big corporations are the least efficient of all at building software. Even Google has admitted in the past to completely rewriting most projects every few years.
Clearly this is a sure sign that they're not developing software correctly... What normal company under competitive pressure can afford to constantly rewrite projects? Why then are indie developers and startups spending so much time and effort trying to emulate these big corporations and trying to use the same low quality tools that they use?
The open source tools produced by corporations are heavily marketed and seem to cause lower productivity almost by design. Perhaps they are designed to create some Kafkaesque anti-competitive moat around their products. Big corporations can afford to be inefficient. Maybe they they want to make sure that everyone else is inefficient too. They don't want people to come up with more elegant solutions, they want everyone to be brute-forcing software development by the numbers, throwing hundreds of developers at every problem; because that's the only game they know.
They use their 'open source' software along with their monopolistic network effects to try to draw everyone into that Kafkaesque game.
The best open source tools are outside the corporate sphere but most people don't even know they exist because they are heavily suppressed by the corporate monopolies. They don't want to give up power.
They want to control the landscape of software development; to keep it at a low level of productivity that they are comfortable with.
18 comments
[ 2.2 ms ] story [ 53.6 ms ] threadIt seemed that companies are trying to exert power over the horizon of user property rights.
It had gotten to the point that "right to repair" movement became a thing.
(It's also widely misunderstood. It's not, as some claim, assuming everyone can be a software developer / electrical engineer. It assumes everyone can either be one, or find one in their neighborhood - being able to ask, and possibly pay, other people to fix or improve your software/hardware is the goal here. Free Software and Right to Repair want to let people do what people do best: self-help within their communities, be it personally or via local markets.)
"Software freedom" is the big enabler and partial driver of centralized computing. It is the big enabler in that if you were paying a per CPU license for Windows NT Server, you would go out of your way to make sure that as much as possible was being done on the users computer. In addition, the most widely used copyleft license, the GPL, actually penalizes you if you make a client program that uses GPL libraries(by requiring you be release the source code for your program if the user asks for it), but happily allows you to instead run the program in a datacenter and never have to share the code with the user. Thus you are incentivized to put as much of your program in the datacenter as possible and not on the user's computer. The thing is, if software is on the user's computer, whatever the license, they actually have far more de facto control. They can control what is getting shared. They can reverse engineer the files to allow perfect export and backups. They can own their data. They can also verify what the software is doing with the data (even if it might take a lot of efforts). Once computation happens in a data center, the user is at the complete mercy of the corporation. The corporation can make it impossible to get a true copy of their data, requiring use of legal means instead of technical means. The corporation can store and share and process the data in ways the user has no control or even awareness about.
Sometimes I wonder if we would have been better off from a society point of view and user freedom and privacy, if the old Microsoft vision of the PC as the hub of a users digital life, and users paying money for software they find useful, instead of what we have now where the cloud is the hub of a users digital life (even stuff that could work on a local network, will by default try to contact the cloud), and users are provided useful software in exchange for their engagement, privacy, and attention.
There are ways to do this peer to peer of course, but google and apple clouds became default.
Google already was a website so it only made sense for them to put their office suite next to the other web stuff as it was "their turf". Of course license cost would have become an issue here. But they could have developed their own backend as well.
I would rather see it as: there was demand and it was useful. Google knew they wouldn't beat microsoft on a Windows machine or Apple on ios/Mac. So they created their own plattform and made sure it is as much web as possible (so Andeoid was/is only a way of pushing their web plattforms).
There were a lot of technical things that had to fell into place for the cloud to spread everywhere - web technologies, ubiquitous Internet access, and perhaps even smartphones - but ultimately, the driver was always the same: if you control access to something, you can charge rent for it.
Right now, by far the most convenient model is ads/tracking - it's literally on every platform that respectively can display images/text and doesn't have extensive privacy features, which is basically every platform on the planet. And it's so convenient it requires zero clicks to pay with it!
Ads/tracking cause the "you're not the user, you're the product" problem, which IMO is the reason why everything needs to be online - if you're offline, they can't send you ads or receive your data.
And a proper user-based funding mechanism would also make corporate funding less necessary, which will enable Free Software devs to prioritize consumer users instead of having e.g. Canonical's "but most of our money comes from servers" problem that ends up de-emphasizing the desktop despite their best intentions.
they shy away from the Affero GPL, and their key elements are proprietary.
The idea that GPL is ideological and other licenses aren't is false.
Allowing companies to proprietarize software and not caring about user's freedom is just a different set of values.
Individual hobbyists care about tinkering with the software of their choice, emphasizing freely licensed source code. These concerns do not affect those who do not make a sport out of modifying code. There is nothing wrong with this, but it will never be a household issue.
This couldn't be further from the truth. Consider the simple example of tinkerers making a patch that disables DRM or telemetry, and sharing it with "those who do not make a sport out of modifying code". This is no hypothetical - she writes:
Consider third-party clients to centralized unencrypted chat systems. Although two users of such a client privately messaging one another are using only free software, if their messages are being data mined, there is still harm.
The Off-The-Record plugin (https://otr.cypherpunks.ca/) encrypts otherwise unencrypted centralized chat systems. It works even if you "do not make a sport out of modifying code".
My job has me convincing and coaching large companies through their first open source steps. In my experience, the cost reduction is a nuanced issue in the boardroom. Mainenance and support costs (and availability), trust of strangers, loss of control, and the clash of openness in a highly secretive world, are strong motivators against the relatively tiny (perceived) expense of building it yourself.
The argument that carries the most water in my experience, is the argument of interoperability. For reliability at modern scale, multi-cloud is an inevitable step. And open source technologies drastically reduce the complexity and cost. Why would you build on Azure's (excellent) CosmosDB and set yourself on a course for painful integration and migration, when you can build with Cassandra and avoid those issues?
To the author's list of important areas impacted by software freedom, I would add diversity and inclusion. It's nice that a bunch of white guys with West US culture build great solutions to their problems that makes sense in their cultural context. But those off-the-shelf products don't help people operating in a different cultural context. With open licenses, it's possible to modify software to fit your own needs, like the users who got braille support into MuseScore, or the Swiss companies who built privacy-friendly email services to compete with gmail on Ux.
However, I disagree with the overall article. These important issues (sustainability, privacy, copyright, outrage-for-profit, D&I, etc) are not components of freedom. They're areas that benefit from user freedom. Software can be "Free as in freedom", and still be Bad For You. It's OK to have open source gambling software, or video games, or bloated office suites. The point is that users get to choose.
This is a nice argument for some kind of "software for good" label, like the kosher K on your food. We could certify that this software:
- respects the 4 FLOSS Freedoms - does not leak or store _any_ information remotely - is (reasonably) free of unnecessary hardware requirements and bloat - does not try to manipulate you for "engagement"
etc.
Corporations in the software industry are often monopolies. They derive their success primarily from network effects - It has nothing to do with the quality of the software that they produce or their development efficiency; in fact, big corporations are the least efficient of all at building software. Even Google has admitted in the past to completely rewriting most projects every few years.
Clearly this is a sure sign that they're not developing software correctly... What normal company under competitive pressure can afford to constantly rewrite projects? Why then are indie developers and startups spending so much time and effort trying to emulate these big corporations and trying to use the same low quality tools that they use?
The open source tools produced by corporations are heavily marketed and seem to cause lower productivity almost by design. Perhaps they are designed to create some Kafkaesque anti-competitive moat around their products. Big corporations can afford to be inefficient. Maybe they they want to make sure that everyone else is inefficient too. They don't want people to come up with more elegant solutions, they want everyone to be brute-forcing software development by the numbers, throwing hundreds of developers at every problem; because that's the only game they know.
They use their 'open source' software along with their monopolistic network effects to try to draw everyone into that Kafkaesque game.
The best open source tools are outside the corporate sphere but most people don't even know they exist because they are heavily suppressed by the corporate monopolies. They don't want to give up power.
They want to control the landscape of software development; to keep it at a low level of productivity that they are comfortable with.