Windows 10 support ends - Are you interested in Linux support by FAS?

[WhiteRoom]

It was all about whether FAS was familiar with Linux.

I sympathise with your frustration that you can't run Edit on your chosen OS, but it sounds as if FAS can't prioritise a Linux version for all the reasons other users have stated in this discussion.

What are the current issues running it in Bottles after your earlier attempts?

 

[l4mpi]

That's the important bit. The requirements are for the stated purpose. Nothing to do with the embedded OS...

The "UI Development for Embedded Linux using QT" part does not sound like it's about android development though, so it does beg the question why this is included and seems to imply that some of the devices indeed run embedded linux... which is interesting as I thought I read somewhere that they use the TI RTOS instead, but maybe that was about older hardware or I'm misremembering things.

 

[FPFL]

It was all about whether FAS was familiar with Linux.

No you, see it isn't. You are still oversimplifying. You know enough to be dangerous but aren't in software dev or support, or so it seems.

This next part isn't Fractal specific but it applies to them as well-

Writing for a specific embedded Linux device is not the same as supporting a large, professional user base on commercial versions of Linux based OS types. Linux is the kernel, but it's far from the only thing users and vendors need to run a full desktop client. Oldies like me remember when it was called GNU/Linux more commonly for this reason among others.

Paying professional users except real support and that means being specific with which distro and details to give them a specific experience. This is a kernel and a lot more.

The costly detail "its just Linux" people are talking past is the real work in support after the 1.0 install is built. More OS types to support costs them money as OS updates happen, there are more dev/test cycles every update, more customer help emails, phone calls, etc. All software vendors deal with this I assure you. We track it b/c it costs money. Support calls often drive those .1 feature releases.

Linux based OS types are also still just a tiny fraction of the global desktop/laptop market. Spend precious support money for a tiny user base, a diverse user base with the least predictable hardware and upgrade cycles? Not exciting.

IMHO a waste of precious resources for Fractal.

This is why vendors rarely support linux except where they have to, or get some competitive benefit out of it - like Steam OS now does for gaming by cutting out Windows OS and Store which is good competition for us all.

Be well.

 

[AlGrenadine]

The "UI Development for Embedded Linux using QT" part does not sound like it's about android development though, so it does beg the question why this is included and seems to imply that some of the devices indeed run embedded linux... which is interesting as I thought I read somewhere that they use the TI RTOS instead, but maybe that was about older hardware or I'm misremembering things.

smells like AxeFx IV
Or maybe the VP4 UI is made in Qt ?

 

[Rock Chalk]

Also, would just like to point out that whatever OS is running on Fractal devices (most likely proprietery given the very specific nature of the hardware) has precisely dick to do with what computer OS can run the editor.

So even if (IF... and that's a big if) Fractal devices were running some flavor of Linux under the hood of their hardware, it wouldn't somehow mean that they could easily port the editor over to it.

 

[dariosicily]

I bought a 100 Euro refurbished junky laptop just to use it with Fractal products, even on stage without caring too much about something precious..

It does suck, but don't care as its only purpose is to run the Editor.

I'd love to run it on Linux (I have a Zorin distro installed on my home PC)

 

[fuzzydude]

With 1 out of 3 of those polled saying they would like Linux support, I think this topic deserves more conversation than "FAS should just focus on what is already available and works right now." That might be a very short-sighted opinion that could one day be a missed business opportunity. Ultimately FAS will do whatever they think is in their best interests, but I'd like to offer an argument on how supporting open source in general could work for their benefit in the not-so-distant future.

I've organized my thoughts into 3 points. Feel free to skip to the one that interests you but I bet you'll want to read all of them.

Point 1) The best reasons for why FAS should NEVER support anything open source.
Point 2) What does FAS stand to lose?
Point 3) Why FAS should invest in open source and the inevitable.

Point 1

The best reasons for why FAS should NEVER support anything open source.

Even a tiny precompiled closed-source Linux library that would allow access to only switching presets or some other basic options with nothing proprietary would produce cheaper solutions that would compete with their foot controllers. Those basic options might be enough for some users to prevent them from even considering buying a FC6/FC12. I don't like it, but I understand the decision to protect their products from a business perspective. They also lose control over how their interface is designed for open source systems and in turn those solutions may offer a worse experience that reflects badly on FAS. These are the same type of arguments used to defend most closed-source systems.

Point 2

What does FAS stand to lose?

Every FAS firmware that has been released, and everything that has been encrypted including Bitcoin wallets, will be deciphered in as little as 5 years from now. Why? Quantum computing. All of our current in-use encryption was not designed to be hardened against quantum computing. I don't want to spend too much time describing what quantum computing is, but I'll say it's a new type of computer and it's been estimated to only need 4,099 qubits of processing power to guess an encryption key in about 10 seconds from what used to take 300 trillion years with a classic computer. In December of 2023, IBM made a quantum computer "Condor" with 1,121 qubits of processing power. That was more than twice the previous record holder "Osprey" which had 433 qubits at the end of 2022. You can see how quickly this is progressing. The event where all encrypted documents become easily deciphered is called the "quantum apocalypse" and if you don't know how a quantum computer will break encryption, then I suggest you inform yourself sooner rather than later. It's not a question of "if," it's a question of "when." If you happen to be an optimist and think we already have or will have a "post-quantum cryptography" (PQC) standards by that time, then I suggest you look at the probabilities. It's almost a guarantee that the first few versions of widely adopted "PQC" encryptions will be poorly implemented and have a weakness that will be exploitable, sometimes by design. Additionally, it is highly likely that the first few PQC encryptions will not be open source because of the incredible value they will have. If a software-based business (FAS) wants to guarantee its future firmware software encryption is not compromised and is willing to pay for a private-party closed-source encryption, then that software-based business (FAS) will be at the whim and wills of the company selling the encryption. Imagine how even today Microsoft and Apple can make FAS jump through hoops if they were compelled to by some new software policy that would cost FAS heavily. Now imagine adding a few exponents of power to that Microsoft and Apple already have if they were the arbiters of encryption. What does open source have to do with any of this? See Point 3.


Point 3

Why open source?

The long-term solutions for truly difficult closed-source problems such as PQC will always be found with true integrity in open source. It's free with little chance of controlling the future of your company. Closed source software needs open source solutions in order to exist, as has ALWAYS been the case. For FAS, or any closed source company to maintain some sovereignty over itself, it will need encryption software from the open source community rather than rely on closed source solutions. This isn't just true for FAS, but for every one of us as individuals. Even as we wax poetic over how perfect this or that closed-source operating system is, realize the only thing keeping those systems from being more oppressive is their single greatest existential threat, open-source alternatives. Closed source software will be with us for as long as I can imagine, even after a "quantum apocalypse" (see Point 2), but understanding and supporting open source is one of the best investments for the very-long-term that a long-lasting closed source software company can make. This is why Google, Apple, and Microsoft are so heavily invested in open source instead of fighting to destroy it. They get it.


I know FAS has the know-how on Linux already, but investing now for the future where additional knowledge and experience with open source will help FAS weather the "quantum apocalypse" future is the best way forward. 5 to 10 years happens in the blink of an eye and a lot will change. If FAS finds a way to leverage open source solutions to protect its own future closed source software, they might also consider giving back to open source by allowing their protected software to have some compatibility with the open source world that made its existence possible. I bet those Linux operating systems will be looked at much more favorably post-"quantum apocalypse."

My background: Run/own an Internet security business since 2010. Software developer by trade.

 

[TheRedDevil]

What are the current issues running it in Bottles after your earlier attempts?

FM3-Edit doesn't work, AXEEdit is a bit slow with RTA and often refreshes the rendering w/o any reason.

 

[TheRedDevil]

No you, see it isn't. You are still oversimplifying. You know enough to be dangerous but aren't in software dev or support, or so it seems.

This next part isn't Fractal specific but it applies to them as well-

Writing for a specific embedded Linux device is not the same as supporting a large, professional user base on commercial versions of Linux based OS types. Linux is the kernel, but it's far from the only thing users and vendors need to run a full desktop client. Oldies like me remember when it was called GNU/Linux more commonly for this reason among others.

Paying professional users except real support and that means being specific with which distro and details to give them a specific experience. This is a kernel and a lot more.

The costly detail "its just Linux" people are talking past is the real work in support after the 1.0 install is built. More OS types to support costs them money as OS updates happen, there are more dev/test cycles every update, more customer help emails, phone calls, etc. All software vendors deal with this I assure you. We track it b/c it costs money. Support calls often drive those .1 feature releases.

Linux based OS types are also still just a tiny fraction of the global desktop/laptop market. Spend precious support money for a tiny user base, a diverse user base with the least predictable hardware and upgrade cycles? Not exciting.

IMHO a waste of precious resources for Fractal.

This is why vendors rarely support linux except where they have to, or get some competitive benefit out of it - like Steam OS now does for gaming by cutting out Windows OS and Store which is good competition for us all.

Be well.

https://forum.fractalaudio.com/thre...d-in-linux-support-by-fas.210620/post-2634254

 

[WhiteRoom]

With 1 out of 3 of those polled saying they would like Linux support, I think this topic deserves more conversation than "FAS should just focus on what is already available and works right now." That might be a very short-sighted opinion that could one day be a missed business opportunity. Ultimately FAS will do whatever they think is in their best interests, but I'd like to offer an argument on how supporting open source in general could work for their benefit in the not-so-distant future.

I've organized my thoughts into 3 points. Feel free to skip to the one that interests you but I bet you'll want to read all of them.

Point 1) The best reasons for why FAS should NEVER support anything open source.
Point 2) What does FAS stand to lose?
Point 3) Why FAS should invest in open source and the inevitable.

Point 1

The best reasons for why FAS should NEVER support anything open source.

Even a tiny precompiled closed-source Linux library that would allow access to only switching presets or some other basic options with nothing proprietary would produce cheaper solutions that would compete with their foot controllers. Those basic options might be enough for some users to prevent them from even considering buying a FC6/FC12. I don't like it, but I understand the decision to protect their products from a business perspective. They also lose control over how their interface is designed for open source systems and in turn those solutions may offer a worse experience that reflects badly on FAS. These are the same type of arguments used to defend most closed-source systems.

Point 2

What does FAS stand to lose?

Every FAS firmware that has been released, and everything that has been encrypted including Bitcoin wallets, will be deciphered in as little as 5 years from now. Why? Quantum computing. All of our current in-use encryption was not designed to be hardened against quantum computing. I don't want to spend too much time describing what quantum computing is, but I'll say it's a new type of computer and it's been estimated to only need 4,099 qubits of processing power to guess an encryption key in about 10 seconds from what used to take 300 trillion years with a classic computer. In December of 2023, IBM made a quantum computer "Condor" with 1,121 qubits of processing power. That was more than twice the previous record holder "Osprey" which had 433 qubits at the end of 2022. You can see how quickly this is progressing. The event where all encrypted documents become easily deciphered is called the "quantum apocalypse" and if you don't know how a quantum computer will break encryption, then I suggest you inform yourself sooner rather than later. It's not a question of "if," it's a question of "when." If you happen to be an optimist and think we already have or will have a "post-quantum cryptography" (PQC) standards by that time, then I suggest you look at the probabilities. It's almost a guarantee that the first few versions of widely adopted "PQC" encryptions will be poorly implemented and have a weakness that will be exploitable, sometimes by design. Additionally, it is highly likely that the first few PQC encryptions will not be open source because of the incredible value they will have. If a software-based business (FAS) wants to guarantee its future firmware software encryption is not compromised and is willing to pay for a private-party closed-source encryption, then that software-based business (FAS) will be at the whim and wills of the company selling the encryption. Imagine how even today Microsoft and Apple can make FAS jump through hoops if they were compelled to by some new software policy that would cost FAS heavily. Now imagine adding a few exponents of power to that Microsoft and Apple already have if they were the arbiters of encryption. What does open source have to do with any of this? See Point 3.


Point 3

Why open source?

The long-term solutions for truly difficult closed-source problems such as PQC will always be found with true integrity in open source. It's free with little chance of controlling the future of your company. Closed source software needs open source solutions in order to exist, as has ALWAYS been the case. For FAS, or any closed source company to maintain some sovereignty over itself, it will need encryption software from the open source community rather than rely on closed source solutions. This isn't just true for FAS, but for every one of us as individuals. Even as we wax poetic over how perfect this or that closed-source operating system is, realize the only thing keeping those systems from being more oppressive is their single greatest existential threat, open-source alternatives. Closed source software will be with us for as long as I can imagine, even after a "quantum apocalypse" (see Point 2), but understanding and supporting open source is one of the best investments for the very-long-term that a long-lasting closed source software company can make. This is why Google, Apple, and Microsoft are so heavily invested in open source instead of fighting to destroy it. They get it.


I know FAS has the know-how on Linux already, but investing now for the future where additional knowledge and experience with open source will help FAS weather the "quantum apocalypse" future is the best way forward. 5 to 10 years happens in the blink of an eye and a lot will change. If FAS finds a way to leverage open source solutions to protect its own future closed source software, they might also consider giving back to open source by allowing their protected software to have some compatibility with the open source world that made its existence possible. I bet those Linux operating systems will be looked at much more favorably post-"quantum apocalypse."

My background: Run/own an Internet security business since 2010. Software developer by trade.

1. But who would own and utillise these quantum computers? Probably just security agencies, state actors and possibly the original designers/owners. Bear in mind that where and whenever national (and in this case global) security is at risk who actually designed or owns the technology is very much a secondary consideration. The entrenched power hierarchy and the machinations of those who seek absolute dominion over all things are what then determines the availability of something this groundbreaking. I'm no expert, I'm just reading from history.

We won't be worrying about open source vs. closed source guitar gear under those circumstances

2. You don't need to encrypt source code to protect it. Mostly (except for security services and security related industry afaik), source code is protected by everyday practical solutions like code obfuscation. Even when decrypted it still requires a huge amount, (almost equivalent to the already well versed process of reverse engineering binary instruction sets), of effort to re-engineer. So unless there is a systematic global cracking (by the limited number of quantum computers available) of all the public source code repositories by some organisation that specifically WANTS to destroy closed source, it seems unlikely to me that it will happen. Not overnight, and maybe not for decades (at least), after which time quantum encryption methods may well have caught up. I'm not saying these scenarios aren't possible, I just think they're unlikely, given how much vested interest is at stake here. Apocalypse, you said... they'll be working hard to prevent that

 

[unix-guy]

Point 1

The best reasons for why FAS should NEVER support anything open source.

Even a tiny precompiled closed-source Linux library that would allow access to only switching presets or some other basic options with nothing proprietary would produce cheaper solutions that would compete with their foot controllers. Those basic options might be enough for some users to prevent them from even considering buying a FC6/FC12. I don't like it, but I understand the decision to protect their products from a business perspective.

There's nothing magic about what you've proposed. There's already dozens of cheap midi controllers in existence that can do this and are actively being used with Fractal devices. There's no requirement for a library from Fractal...

Point 2

What does FAS stand to lose?

Every FAS firmware that has been released, and everything that has been encrypted including Bitcoin wallets, will be deciphered in as little as 5 years from now. Why? Quantum computing.

Fractal doesn't encrypt the firmware... Flawed argument.

Point 3

Why open source?

The long-term solutions for truly difficult closed-source problems such as PQC will always be found with true integrity in open source. It's free with little chance of controlling the future of your company.

Doesn't apply, see my previous point.

Closed source software needs open source solutions in order to exist, as has ALWAYS been the case. For FAS, or any closed source company to maintain some sovereignty over itself, it will need encryption software from the open source community rather than rely on closed source solutions.

"Always"? No...

 

[rossipedia]

It's been a while since I looked into the legal issues surrounding it, but is the GPL still a mine-field when dealing with building products on top of Linux these days? If so and I were FAS I wouldn't wanna get within a 100 miles of that mess.

 

[Geoff]

Always like to see options- so voted yes

Not sure i'd use it right away, since mostly use Mac for home purpose now. But an Ubuntu based recording studio might be nice. U-he makes wonderful synths that work on linux. Don't think they are officially supported, but they do work.

If your needs are not too cutting edge, another option is just unplug the network cable on your old Windows machine- make it special purpose recording studio only. That is what I do, using an old Imac with no internet. Turn on wifi for a few minutes when you need the new fractal firmware. It's a bit limiting, but I don't need the latest updates, just need to practice more.

 

[Ingo]

It's been a while since I looked into the legal issues surrounding it, but is the GPL still a mine-field when dealing with building products on top of Linux these days? If so and I were FAS I wouldn't wanna get within a 100 miles of that mess.

There's no "mess" and no "minefield". Building for the Linux kernel has never been a legal issue. And when using a library people need to check the library's license, GPL or otherwise, and behave accordingly. In fact, complying with the GPL is very simple, it's just not what many software vendors want to do.

 

[fuzzydude]

There's nothing magic about what you've proposed. There's already dozens of cheap midi controllers in existence that can do this and are actively being used with Fractal devices. There's no requirement for a library from Fractal...

Correct. Except I'm not talking about MIDI. I'm talking about using the same method Axe-Edit uses to communicate.

Fractal doesn't encrypt the firmware... Flawed argument.

Really? Tell me more. It's not encrypted at all? That's not very smart. Are you sure? If it hasn't been, the Line6 guys are pouring over it right now then.

Doesn't apply, see my previous point.

I guess I'll disagree with you here. Unless you're looking at that code from the firmware, how it's uploaded and how it's read, then I would say there is at least some obfuscation. If there isn't, that would be huge mistake on the part of FAS. Can you really speak for FAS about this detail with certainty?

"Always"? No...

I meant that in a general way, but I'd be curious to hear some successful examples.

 

[fuzzydude]

1. But who would own and utillise these quantum computers? Probably just security agencies, state actors and possibly the original designers/owners. Bear in mind that where and whenever national (and in this case global) security is at risk who actually designed or owns the technology is very much a secondary consideration. The entrenched power hierarchy and the machinations of those who seek absolute dominion over all things are what then determines the availability of something this groundbreaking. I'm no expert, I'm just reading from history.
We won't be worrying about open source vs. closed source guitar gear under those circumstances

I see what you're saying, but your assumptions are provably wrong. It costs $5000 to own a quantum computer right now at this early stage. So anyone can own one. They'll get much cheaper very quickly too. It's the same as with any computer or technology. If the "entrenched power hierarchy" are determining the availability of something this groundbreaking, they aren't doing a very good job of it.

2. You don't need to encrypt source code to protect it. Mostly (except for security services and security related industry afaik), source code is protected by everyday practical solutions like code obfuscation. Even when decrypted it still requires a huge amount, (almost equivalent to the already well versed process of reverse engineering binary instruction sets), of effort to re-engineer. So unless there is a systematic global cracking (by the limited number of quantum computers available) of all the public source code repositories by some organisation that specifically WANTS to destroy closed source, it seems unlikely to me that it will happen. Not overnight, and maybe not for decades (at least), after which time quantum encryption methods may well have caught up. I'm not saying these scenarios aren't possible, I just think they're unlikely, given how much vested interest is at stake here. Apocalypse, you said... they'll be working hard to prevent that

So let me know if I'm summarizing what you're saying correctly. You're saying we'll have a new computer that can take 300 trillion years of guessing and condense it into less than 10 seconds to get the correct answer, but we won't have a computer that can properly decode and decipher the intent of obfuscated source code? Those two don't exist in the same reality. It sounds like you've seen this obfuscated source code from decompiled code before, I know I have. That doesn't mean there aren't already systems that can further connect the obfuscated code into a more coherent software system. It's existed for the last few decades. Those systems need massive processing power to make them more effective. Which is what quantum computing brings.

 

[unix-guy]

Correct. Except I'm not talking about MIDI. I'm talking about using the same method Axe-Edit uses to communicate.

It uses midi... SysEx to be specific, but that's irrelevant because every item you talked about can already be done with generic midi PC and CC commands, and if you're clever you could sniff and reverse engineer the SysEx (for midi controllers that can send it) for any items not already documented by Fractal in their 3rd Party Midi Specifications.

Really? Tell me more. It's not encrypted at all? That's not very smart. Are you sure? If it hasn't been, the Line6 guys are pouring over it right now then.

I guess I'll disagree with you here. Unless you're looking at that code from the firmware, how it's uploaded and how it's read, then I would say there is at least some obfuscation. If there isn't, that would be huge mistake on the part of FAS. Can you really speak for FAS about this detail with certainty?

Assume it was... How would things work today? What would decrypt it to make it usable?

I meant that in a general way, but I'd be curious to hear some successful examples.

Open source software hasn't always existed. There's no example needed.

 

[fuzzydude]

It uses midi... SysEx to be specific, but that's irrelevant because every item you talked about can already be done with generic midi PC and CC commands, and if you're clever you could sniff and reverse engineer the SysEx (for midi controllers that can send it) for any items not already documented by Fractal in their 3rd Party Midi Specifications.

That's good to know. I'm surprised someone hasn't already made a cheaper foot controller for the Axe units and taken the market from FAS. I suppose in time it's going to happen if you what you say is true. To my point, that was the best I could come up with for why FAS shouldn't support open source. Your argument proves even that isn't a good reason.

Assume it was... How would things work today? What would decrypt it to make it usable?

Someone could take that code, rewrite it to avoid copyright infringement (or not), buy their own DSP chips and sell the same product for 1/10 the price. You know China does this all the time? Then there's every other competitor out there who wants a slice of the pie.

Open source software hasn't always existed. There's no example needed.

It actually has. At least longer than you're assuming. Mathematics is probably the original open source software. Without one group sharing the discoveries of basic mathematics, you couldn't do much. All we create we do while resting on the shoulders of giants who shared what they learned.

 

[unix-guy]

That's good to know. I'm surprised someone hasn't already made a cheaper foot controller for the Axe units and taken the market from FAS. I suppose in time it's going to happen if you what you say is true. To my point, that was the best I could come up with for why FAS shouldn't support open source. Your argument proves even that isn't a good reason.

There are already midi controllers that have Fractal "integration" on the market.

Also, there are people on this forum who are doing this type of thing for their own use and even sharing with others. See the Other Midi Controllers sub-section of the forum.

Someone could take that code, rewrite it to avoid copyright infringement (or not), buy their own DSP chips and sell the same product for 1/10 the price. You know China does this all the time? Then there's every other competitor out there who wants a slice of the pie.

It's not "code" - it's compiled and would require reverse engineering first... And it's useless without the matching hardware.

Yes, I know China does this. A company I worked for 25+ years ago would receive returns for networking hardware they created that were counterfeits.

If the current firmware was encrypted, it would require some form of key to decrypt it. With the current hardware, that would almost certainly be something like a built-in "dongle". And if you've recreated the hardware necessary to run the firmware, you'd have it... Or just buy the hardware and use it.

It actually has. At least longer than you're assuming. Mathematics is probably the original open source software. Without one group sharing the discoveries of basic mathematics, you couldn't do much. All we create we do while resting on the shoulders of giants who shared what they learned.

Mathematics is a language...

Most programming languages are open source these days but that wasn't always the case.

You're really stretching here.

In any case, I'm done debating on this topic. I really don't care that much and you're entitled to your opinions whether or not I (or anyone) thinks they are valid.

 

[Dave Merrill]

That's good to know. I'm surprised someone hasn't already made a cheaper foot controller for the Axe units and taken the market from FAS. I suppose in time it's going to happen if you what you say is true. To my point, that was the best I could come up with for why FAS shouldn't support open source. Your argument proves even that isn't a good reason.

Someone could take that code, rewrite it to avoid copyright infringement (or not), buy their own DSP chips and sell the same product for 1/10 the price. You know China does this all the time? Then there's every other competitor out there who wants a slice of the pie.

It actually has. At least longer than you're assuming. Mathematics is probably the original open source software. Without one group sharing the discoveries of basic mathematics, you couldn't do much. All we create we do while resting on the shoulders of giants who shared what they learned.

There are lots of generic midi controllers out there, which is a much larger market. IMO there's pretty skimpy motivation to reverse engineer Fractal's proprietary stuff, do a lot of super detailed development to replicate it, update it every time Fractal does, design and manufacturer the hardware at a scale where you won't lose your shirt, and support it.

Re the source code of the firmware, i bet all that's in the units is the compiled microprocessor and DSP chip bytecode. That's even less human readable than machine code for a single processor system, really useful only to someone running that exact hardware.

Re clones, I'm not aware of any.