4 ms·

We need a new philosophical movement for computer technology. The Stallman movement failed because it was not practical for business applications. At the end o

by max_ 3y ago

We need a new philosophical movement for computer technology. The Stallman movement failed because it was not practical for business applications. At the end of the day, coders & hardware manufacturers need to eat just like accountants, plumbers and doctors. Technology can't rely 100% on altruism. What we need is something close to the Ethical standards of the Stallman movement, but also compatable with business owners. These philosophical ethical standards should also be ingrained into VC culture & corporate culture. The reason why Apple can propose to have something like CSAM, or Google banning someone account forever (Gmail, YouTube et al) because they had images of their naked daughter suffering from a skin disease (to be sent to the doctor). The reason is that there is no ethical-philosophical compass guiding computer technology. Tech Corporate MBAs only care about "playing it safe". I find them disgustingly spinless. Until we have this ethical-philosophy developed, I only expect technology to used as a tool for manipulation and control.

dools 3y ago
> The Stallman movement failed Citation needed
- max_ 3y ago
  How many applications, phones, computers do people use that are Stallman compliant?
  - version_five 3y ago
    Ideological(ish) movements don't succeed by converting everyone absolutely, they do it by changing the narrative in their direction. Stallman has absolutely done that.
  - quickthrower2 3y ago
    The Linux Kernel. But answering the “it failed”. Not really. It exists and is a choice on most license dropdowns and is a famous one. People should really default to GPL rather than more permissive licenses and hot take… should also default to closed source. And make open source or non GPL a conscious choice. The reason is GPL gives you more ability to commercialise since it gives you the copyright holder more of an edge over let’s say Amazon Inc who has decided to accept your license to bundle it with a cloud service and extend it. They are obliged to release their source code to their customers. That Ballmer called it a cancer is a great accolade. Close source as a default because it is a one way door publishing the code on the open web. So you are protecting your own interests. You owe the “open source community” nothing and should exploit your IP as much as anyone else, and use GPL to allow people to “fix the printer” if they need to.
    - DaiPlusPlus 3y ago
      > The reason is GPL gives you more ability to commercialise since it gives you the copyright holder more of an edge over let’s say Amazon Inc who has decided to accept your license to bundle it with a cloud service and extend it. They are obliged to release their source code to their customers. Not if it's GPLv2 or non-AGPL - and even then what you're describing is the opposite of the underdog-moral-problem a lot of smaller cloud services or database/infrastructure providers are facing whereby Amazon (et al.) simply offers your product or service as their service, which is allowed by all versions of the GPL because Amazon doesn't care about the source-code: they care about charging a premium for access to a hot open-source project without contributing anything back to the actual project maintainers. That's (unfortunately) an argument for going closed-source, or at least having to reserve copyrights/licenses for operating the software as a service for another company. I imagine it's like being Sherlocked by Apple, except Apple was using your work all-along.
  - bheadmaster 3y ago
    More than none: https://www.fsf.org/resources/hw/systems https://www.fsf.org/resources/hw/systems https://ryf.fsf.org/ https://ryf.fsf.org/
  - abwizz 3y ago
    not much, but most build upon foss and some would not be viable without it.
- nvm0n2 3y ago
  It failed, yes. If you want to go wide you could argue that all open source is "the Stallman movement" but given Stallman's noted hostility to other approaches than his own, that seems too much. The peak of GPLd software was in the late 90s/early 2000s and since then the base of GPLd code has shrivelled. I can't remember the last time I encountered a library or program in my work that was even LGPL licensed, let alone GPLd, with the exception of some very old programs like the Linux kernel and a few bits of the userspace (plus Java's weird GPL+exceptions thing). These days almost everything new is Apache licensed or similar. Just browse around GitHub for a while and see what I mean. Also: 1. The Linux kernel, the last bastion of Stallmanist philosophy, in reality has succeeded by not actually enforcing the GPL thus treating it more like LGPL or Apache2 in practice. GPL violations are rampant and rarely does anything happen. They also very publicly refused to switch to GPL3. 2. The most popular Linux based OS is Android where Google rewrote the entire userspace specifically to avoid the GPL, and then also created their own driver layer, again, to avoid the GPL and the Linux approach to technical "enforcement" (constant API churn). 3. The most popular UNIX based desktop OS is built on FreeBSD despite losing technical features, specifically because Apple refused to countenance anything GPL or FSF connected. The few dependencies they started with (like gcc) were rewritten into new BSD-style licensed codebases like LLVM at vast expense. In other words the people who actually build operating systems for a living avoid like the plague anything connected to Stallman's vision or strategy. It seems likely that eventually the Linux kernel would also be replaced, certainly Google attempted this with Zircon, but whilst enforcement is so lax and technical workarounds so numerous there isn't much reason for doing so.
  - dTal 3y ago
    > I can't remember the last time I encountered a library or program in my work that was even LGPL licensed, let alone GPLd I don't know what line of work you're in but this seems quite unlikely. GTK and Qt are both LGPL (and GNOME and Plasma are GPL). WebKit's JavaScript engine is LGPL. MySQL is GPL. Blender is GPL. Arduino is AGPL. GCC is GPL. Bash is GPLv3. It's hard for me to believe you use the Linux kernel without touching any of these. GPL is wildly popular. It's losing ground, yes, and you correctly identify a primary reason why: commercial software companies correctly see it as a threat, and are fighting it in earnest now. It is a threat. It was designed to be a threat. The purpose of the GPL is to create a self-reinforcing free ecosystem that outcompetes the balkanized proprietary hellscape that preceded the free software movement. At this, it succeeded beyond any reasonable imagining. A lot of the revolutionary spirit has left the free software movement. We are currently in the "good times create weak men" stage of the cycle. People who grew up in the garden of Eden created by the free software movement, who take it for granted that the popular web browsers are open source and Microsoft releases cross-platform IDEs that run on Linux, don't appreciate that the current state of affairs was wrought over decades by the kind of people who spelled it Micro$oft and were willing to tolerate their computers only half working in order to avoid proprietary software of any kind, on principle. But make no mistake, we are still at war. The big corpos know it, even if they've managed to fling enough bread and circuses to fool some into thinking we're all one big happy family now. Giving up on the GPL is how they win, and we lose.
- jongjong 3y ago
  As a Kubuntu user, I have to disagree. From my perspective as a developer, the Linux PC experience has surpassed both Windows and Mac. I really couldn't care less what the other 95% of people are using. In that case, it's even more remarkable that even with just 5% market share, Linux can deliver such a cohesive, compatible experience. 10 years ago, I couldn't have done my job without Windows. Now I can do absolutely everything using Linux. IMO, people who are still on Windows or Mac don't know what they're doing. They've probably just been drifting through life, not really thinking.
mordae 3y ago
I am sorry, but you basically say that we need more altruistic capitalists. Won't happen. > These philosophical ethical standards should also be ingrained into VC culture & corporate culture. Won't happen. Money always wins and exploiting people on the side of selling (or giving for free) them some stuff is always going to bring in more cash than respecting their privacy.
- max_ 3y ago
  Their are business like Source Hut that are built on the basis that they will not do some of the things people like Github do. And I am glad that they are profitable.
  - wheelerof4te 3y ago
    They are profitable up to a point. Once that point is crossed, they will have to compensate and adapt or risk losing their entire business.
jacquesm 3y ago
> The Stallman movement failed The stuff you read on hn... Seriously, you think it failed? I think it succeeded wildly beyond anything anybody imagined at the time, to the point that very large entities threw everything and the kitchensink at it to try to make it fail. Which exposed them for the criminals they were. We don't need anything that is more compatible with 'business owners' than what we've got. What we need is education, so that people stop telling each other that the free software movement failed, but instead point out that without that free software movement the operating system, the browser and the applications on the computer that I'm writing this on would not exist.
- rapnie 3y ago
  Given your last paragraph, what is the net trend in society of that happening with enough speed to outpace other negative developments? Would it be more accurate to say that the Free Software movement is in need for an update, or making next steps, to provide a better outlook in this regard? Edit: Changed "Stallman movement" to "Free Software movement". I am a FOSS advocate myself, but not interested in deeply engrained notions around Stallman himself.
  - jacquesm 3y ago
    I think it's doing fine as it is within the limits of the law. If you want more you're going to have to either operate outside of the law or you're going to have to have the laws changed. But within the law the Free Software Movement used all of the available space, to the point that almost every other license out there is less strict than theirs. What needs to be updated is the the knowledge of the users and of the programmers. But we've been co-opted as an industry for the most part by very large bags of money and a service oriented eco-system is the capitalist response to the free software movement. Doing an end run around that is hard to enforce (see the various licenses that address this loophole directly) without having the backing of a regulator. I see this as a simple swing of the pendulum, which it has done several times in the past regarding centralization and decentralization of computers: single machines -> single machines with terminals and time sharing -> mini computers with terminals -> microcomputers -> networked micro computers -> networked micro computers using peer-to-peer services -> networked micro computers with access to services -> special purpose computers (terminals!) with access to services. So if we want to do away with these we'd need to focus on the technical implementation of federated services that do not require central services at all. One thing you could advocate for is to restore peer-to-peer capabilities of residential internet but the fear of the hacker card will be played to counter that one. But there is also still pressure to push the pendulum one more step to the right: identified access to the internet and without that no service. This may well happen, I would expect a country like the UK or Australia to be first to implement this. Also note that there is no actual technical driving force behind the pendulum, it doesn't have to swing at all, it could just as easily get stuck. The important common factor between all of the above was that someone (or lots of someones) made a lot of money by driving the swing.
verisimi 3y ago
> The reason is that there is no ethical-philosophical compass guiding computer technology. The reason is that there is no ethical-philosophical compass guiding people in the greater part, computer technologists included.
no_time 3y ago
>We need a new philosophical movement for computer technology. I'm hoping for Cory Doctorow's new book[0] to ignite this discussion. This is the first thing I've preordered in my life. And despite his marketing being a bit cheesy, I think he might be the right guy at the right time to plant the seeds for this movement. [0]: https://www.kickstarter.com/projects/doctorow/the-internet-con-how-to-seize-the-means-of-computation https://www.kickstarter.com/projects/doctorow/the-internet-c...
torginus 3y ago
I feel like we already have a battle-tested answer to that question - which is dual licensing. It worked great on software like Qt - just have an open source GPL(ish) licensed version for free as in speech users, and a proprietary do-whatever-you-want for cash money, so that developers can keep the lights on. What needs to change is the open-source community's perception of this coexistence as something mutually beneficial as opposed to betraying their own principles. This hostility manifests in two ways - one is the belief that not fully open software is somehow evil, and shouldn't be supported, and one should always use (often janky) alternatives that are fully open. The second is the constant churn without any thought given to API or ABI compatibility (never mind making the same binary run across versions, even source-level breakage is extremely common), leading to people wishing to ship commercial applications having to package half of the userland with them.

by max_ 3y ago

dools 3y ago

> The Stallman movement failed Citation needed

max_ 3y ago
How many applications, phones, computers do people use that are Stallman compliant?
- version_five 3y ago
  Ideological(ish) movements don't succeed by converting everyone absolutely, they do it by changing the narrative in their direction. Stallman has absolutely done that.
- quickthrower2 3y ago
  The Linux Kernel. But answering the “it failed”. Not really. It exists and is a choice on most license dropdowns and is a famous one. People should really default to GPL rather than more permissive licenses and hot take… should also default to closed source. And make open source or non GPL a conscious choice. The reason is GPL gives you more ability to commercialise since it gives you the copyright holder more of an edge over let’s say Amazon Inc who has decided to accept your license to bundle it with a cloud service and extend it. They are obliged to release their source code to their customers. That Ballmer called it a cancer is a great accolade. Close source as a default because it is a one way door publishing the code on the open web. So you are protecting your own interests. You owe the “open source community” nothing and should exploit your IP as much as anyone else, and use GPL to allow people to “fix the printer” if they need to.
  - DaiPlusPlus 3y ago
    > The reason is GPL gives you more ability to commercialise since it gives you the copyright holder more of an edge over let’s say Amazon Inc who has decided to accept your license to bundle it with a cloud service and extend it. They are obliged to release their source code to their customers. Not if it's GPLv2 or non-AGPL - and even then what you're describing is the opposite of the underdog-moral-problem a lot of smaller cloud services or database/infrastructure providers are facing whereby Amazon (et al.) simply offers your product or service as their service, which is allowed by all versions of the GPL because Amazon doesn't care about the source-code: they care about charging a premium for access to a hot open-source project without contributing anything back to the actual project maintainers. That's (unfortunately) an argument for going closed-source, or at least having to reserve copyrights/licenses for operating the software as a service for another company. I imagine it's like being Sherlocked by Apple, except Apple was using your work all-along.
- bheadmaster 3y ago
  More than none: https://www.fsf.org/resources/hw/systems https://www.fsf.org/resources/hw/systems https://ryf.fsf.org/ https://ryf.fsf.org/
- abwizz 3y ago
  not much, but most build upon foss and some would not be viable without it.
nvm0n2 3y ago
It failed, yes. If you want to go wide you could argue that all open source is "the Stallman movement" but given Stallman's noted hostility to other approaches than his own, that seems too much. The peak of GPLd software was in the late 90s/early 2000s and since then the base of GPLd code has shrivelled. I can't remember the last time I encountered a library or program in my work that was even LGPL licensed, let alone GPLd, with the exception of some very old programs like the Linux kernel and a few bits of the userspace (plus Java's weird GPL+exceptions thing). These days almost everything new is Apache licensed or similar. Just browse around GitHub for a while and see what I mean. Also: 1. The Linux kernel, the last bastion of Stallmanist philosophy, in reality has succeeded by not actually enforcing the GPL thus treating it more like LGPL or Apache2 in practice. GPL violations are rampant and rarely does anything happen. They also very publicly refused to switch to GPL3. 2. The most popular Linux based OS is Android where Google rewrote the entire userspace specifically to avoid the GPL, and then also created their own driver layer, again, to avoid the GPL and the Linux approach to technical "enforcement" (constant API churn). 3. The most popular UNIX based desktop OS is built on FreeBSD despite losing technical features, specifically because Apple refused to countenance anything GPL or FSF connected. The few dependencies they started with (like gcc) were rewritten into new BSD-style licensed codebases like LLVM at vast expense. In other words the people who actually build operating systems for a living avoid like the plague anything connected to Stallman's vision or strategy. It seems likely that eventually the Linux kernel would also be replaced, certainly Google attempted this with Zircon, but whilst enforcement is so lax and technical workarounds so numerous there isn't much reason for doing so.
- dTal 3y ago
  > I can't remember the last time I encountered a library or program in my work that was even LGPL licensed, let alone GPLd I don't know what line of work you're in but this seems quite unlikely. GTK and Qt are both LGPL (and GNOME and Plasma are GPL). WebKit's JavaScript engine is LGPL. MySQL is GPL. Blender is GPL. Arduino is AGPL. GCC is GPL. Bash is GPLv3. It's hard for me to believe you use the Linux kernel without touching any of these. GPL is wildly popular. It's losing ground, yes, and you correctly identify a primary reason why: commercial software companies correctly see it as a threat, and are fighting it in earnest now. It is a threat. It was designed to be a threat. The purpose of the GPL is to create a self-reinforcing free ecosystem that outcompetes the balkanized proprietary hellscape that preceded the free software movement. At this, it succeeded beyond any reasonable imagining. A lot of the revolutionary spirit has left the free software movement. We are currently in the "good times create weak men" stage of the cycle. People who grew up in the garden of Eden created by the free software movement, who take it for granted that the popular web browsers are open source and Microsoft releases cross-platform IDEs that run on Linux, don't appreciate that the current state of affairs was wrought over decades by the kind of people who spelled it Micro$oft and were willing to tolerate their computers only half working in order to avoid proprietary software of any kind, on principle. But make no mistake, we are still at war. The big corpos know it, even if they've managed to fling enough bread and circuses to fool some into thinking we're all one big happy family now. Giving up on the GPL is how they win, and we lose.
jongjong 3y ago
As a Kubuntu user, I have to disagree. From my perspective as a developer, the Linux PC experience has surpassed both Windows and Mac. I really couldn't care less what the other 95% of people are using. In that case, it's even more remarkable that even with just 5% market share, Linux can deliver such a cohesive, compatible experience. 10 years ago, I couldn't have done my job without Windows. Now I can do absolutely everything using Linux. IMO, people who are still on Windows or Mac don't know what they're doing. They've probably just been drifting through life, not really thinking.

mordae 3y ago

I am sorry, but you basically say that we need more altruistic capitalists. Won't happen. > These philosophical ethical standards should also be ingrained into VC culture & corporate culture. Won't happen. Money always wins and exploiting people on the side of selling (or giving for free) them some stuff is always going to bring in more cash than respecting their privacy.

max_ 3y ago
Their are business like Source Hut that are built on the basis that they will not do some of the things people like Github do. And I am glad that they are profitable.
- wheelerof4te 3y ago
  They are profitable up to a point. Once that point is crossed, they will have to compensate and adapt or risk losing their entire business.

jacquesm 3y ago

> The Stallman movement failed The stuff you read on hn... Seriously, you think it failed? I think it succeeded wildly beyond anything anybody imagined at the time, to the point that very large entities threw everything and the kitchensink at it to try to make it fail. Which exposed them for the criminals they were. We don't need anything that is more compatible with 'business owners' than what we've got. What we need is education, so that people stop telling each other that the free software movement failed, but instead point out that without that free software movement the operating system, the browser and the applications on the computer that I'm writing this on would not exist.

rapnie 3y ago
Given your last paragraph, what is the net trend in society of that happening with enough speed to outpace other negative developments? Would it be more accurate to say that the Free Software movement is in need for an update, or making next steps, to provide a better outlook in this regard? Edit: Changed "Stallman movement" to "Free Software movement". I am a FOSS advocate myself, but not interested in deeply engrained notions around Stallman himself.
- jacquesm 3y ago
  I think it's doing fine as it is within the limits of the law. If you want more you're going to have to either operate outside of the law or you're going to have to have the laws changed. But within the law the Free Software Movement used all of the available space, to the point that almost every other license out there is less strict than theirs. What needs to be updated is the the knowledge of the users and of the programmers. But we've been co-opted as an industry for the most part by very large bags of money and a service oriented eco-system is the capitalist response to the free software movement. Doing an end run around that is hard to enforce (see the various licenses that address this loophole directly) without having the backing of a regulator. I see this as a simple swing of the pendulum, which it has done several times in the past regarding centralization and decentralization of computers: single machines -> single machines with terminals and time sharing -> mini computers with terminals -> microcomputers -> networked micro computers -> networked micro computers using peer-to-peer services -> networked micro computers with access to services -> special purpose computers (terminals!) with access to services. So if we want to do away with these we'd need to focus on the technical implementation of federated services that do not require central services at all. One thing you could advocate for is to restore peer-to-peer capabilities of residential internet but the fear of the hacker card will be played to counter that one. But there is also still pressure to push the pendulum one more step to the right: identified access to the internet and without that no service. This may well happen, I would expect a country like the UK or Australia to be first to implement this. Also note that there is no actual technical driving force behind the pendulum, it doesn't have to swing at all, it could just as easily get stuck. The important common factor between all of the above was that someone (or lots of someones) made a lot of money by driving the swing.

verisimi 3y ago

> The reason is that there is no ethical-philosophical compass guiding computer technology. The reason is that there is no ethical-philosophical compass guiding people in the greater part, computer technologists included.

no_time 3y ago

>We need a new philosophical movement for computer technology. I'm hoping for Cory Doctorow's new book[0] to ignite this discussion. This is the first thing I've preordered in my life. And despite his marketing being a bit cheesy, I think he might be the right guy at the right time to plant the seeds for this movement. [0]: https://www.kickstarter.com/projects/doctorow/the-internet-con-how-to-seize-the-means-of-computation https://www.kickstarter.com/projects/doctorow/the-internet-c...

torginus 3y ago

I feel like we already have a battle-tested answer to that question - which is dual licensing. It worked great on software like Qt - just have an open source GPL(ish) licensed version for free as in speech users, and a proprietary do-whatever-you-want for cash money, so that developers can keep the lights on. What needs to change is the open-source community's perception of this coexistence as something mutually beneficial as opposed to betraying their own principles. This hostility manifests in two ways - one is the belief that not fully open software is somehow evil, and shouldn't be supported, and one should always use (often janky) alternatives that are fully open. The second is the constant churn without any thought given to API or ABI compatibility (never mind making the same binary run across versions, even source-level breakage is extremely common), leading to people wishing to ship commercial applications having to package half of the userland with them.