Steve, I find it interesting that in the same issue you mention VaraHF/VaraFM (in the 'EastNET Quarterly Videoconference on 2024-03-14' section): "the primary use of VARA HF was not for networking per se, but moving messages regionally between Bulletin Board Systems", and later on you write: "Novel (to someone) modulation methods are not…
Steve, I find it interesting that in the same issue you mention VaraHF/VaraFM (in the 'EastNET Quarterly Videoconference on 2024-03-14' section): "the primary use of VARA HF was not for networking per se, but moving messages regionally between Bulletin Board Systems", and later on you write: "Novel (to someone) modulation methods are not encryption, as long as you’re prepared to prove that the modulation’s intent is not “obfuscation” of the content, and that the content can be recovered using a typical example of the system (or published protocol, or algorithm)."
As far as I know Jose Alberto Nieto Ros, EA5HVK hasn't published a recent specification of VaraHF. The only one I was able to find is an 8-page word document from about 7 years ago, and it begins with this sentence "VARA HF Modem is a propietary (sic) system developed by Jose Alberto Nieto Ros EA5HVK and can be used under shareware license."
Perhaps in my search I missed something, but I don't think that that Word document qualifies as a 'published protocol', that would allow a third party to recreate a new implementation of VaraHF/VaraFM without reverse engineering EA5HVK's software.
Franco - First, apologies for my late reply. Thanks for your comment. I said:
At least in US Amateur Radio, use of a proprietary system isn't a disqualification for use in Amateur Radio as long as (quoting myself):
"... the content can be recovered using a typical example of the system (or published protocol, or algorithm)."
Note that DMR is widely in use in Amateur Radio, and it uses a proprietary CODEC from the company DVSI, the algorithm for which has not been publicly disclosed. Another example is PACTOR-4 from the company SCS, which is in wide in use in Amateur Radio but the algorithm for PACTOR-4 has not been publicly disclosed. Thus I think that EA5HVK's decision not to publicly disclose his algorithms for the various VARA software versions - HF, FM, and SAT is on par with DVSI and SCS.
In all of those cases mentioned above, a "typical example of the system" is easily obtainable - the DVSI AMBE CODEC chip built into a DMR system, a SCS PACTOR-4 modem (expensive, but available), and the VARA software is available for download and can be used to receive. You can obtain these, and run a recording of a transmission through these units and reconstruct the content sent via the original transmission.
With these systems, and others, the intent of "different" data communications systems isn't to obfuscate (encrypt), but rather make data communications more reliable, require less bandwidth, require less transmission time, etc. Thus, such systems ARE legal for use in US Amateur Radio. (Again, I'm only familiar with US Amateur Radio regulations.)
Like everyone in Amateur Radio, I'd love it if the DVSI CODEC were released into open source, or PACTOR-4 could also be made open source, and if EA5HVK would make VARA open source. But they haven't chosen to do so, and that's their right for their creations.
Hmmm... didn't realize there is a length limit for comments. I was afraid that I had lost the last paragraph (below), but apparently Substack simply truncates comments that are "too long" per its standards.
But with the example of HOW these systems work, and how WELL these systems work, there's nothing wrong with developing workalike systems that do equivalent things and have equivalent performance. Indeed, I think with the tools software developers have available now, I think it's ultimately inevitable that we will have such equivalents. Note that "equivalent" is not the same as "reverse engineered" and not the same as "interoperable". Disclaimer - I was on record early in my discovery process of VARA FM of advocating for an open source equivalent of VARA FM that would be interoperable with VARA FM. I was eventually persuaded that interoperability was a poor goal, and what should be worked on is a new open source system that works equally well or better, and has equal or better performance / reliability.
first of all thanks for your detailed reply and taking some of your busy time to write it.
I think it is important to have this discussion between the two of us, and in the larger ham radio community because of the impact I believe it has on the future of ham radio.
A few comments on what you wrote below.
- I think that the fact that this problem exists for other systems, methods, and protocols currently used in ham radio is beside my point, and doesn't make the specific case of Vara more 'right' or 'wrong'. It is like a defense attorney telling the judge that his client is not guilty (or less guilty) because others have done it too; it doesn't work this way. Thus I'll keep my notes below to just VAra and EA5HVK.
- If the intent of the Vara is not to obfuscate, then I think they can prove it by releasing the current (and future) specifications of their modem; note that I say the specifications of this digital mode, not the source code for it. Take the case of FT8/WSPR: if Joe Taylor K1JT, Steve Franke K9AN, Bill Somerville G4WJS had released just binaries for Windows, Linux, and Mac free of charge, and hadn't publish extensive documentation on how these protocols work (see here for instance: https://wsjt.sourceforge.io/refs.html), people would probably still be using them as much as they do now, but I would have exactly the same issues with them. I think their outstanding contribution of the world of ham radio communication is coming up with the idea/protocol and making it public. Others can "stand on the shoulders of giants" to learn from them and further advance our hobby.
- I feel that the point of your sentence "You can obtain these, and run a recording of a transmission through these units and reconstruct the content sent via the original transmission." could apply to pretty much to anything: from that point of view, even a black box that takes say a vector of 2^N inputs and produces another vector of 2^N outputs could be described that way. I could make it available to anyone for free, but it might take a third party several billion years to go through all the inputs and analyze their outputs to figure out how it works. That's pretty much what encryption does.
- Interoperability (or the possibility of) is at the core of ham radio 'values' and how we use our public bands; after all one of the goals of ham radio is to promote open communication between people and open experimentation, and interoperability is critical to achieve that in my personal opinion. Technically there are probably systems and methods that are 'better' (in some respect) than Vara, but I think that doesn't mean that interoperability with Vara is less important given how widespread Vara use is and the fact that it is used to say send emails and other messages. Also if Vara current specifications were available, other open source projects could learn a thing or two from Vara, and perhaps improve their own algorithms.
- Finally I am not advocating that EA5HVK make his software open source, or that he give it away for free, or that he make the software also available for Linux or Mac. I write software for living, and I think one should be paid for their work, especially for making it convenient to install and run, for designing a nice and user friendly UI, and so on.
What triggered my initial comment was on a different level, to express a concern that I have about VaraHF/VaraFM, and, as you pointed out, on a 'trend' in today's ham radio environment.
Franco - I think you'll agree that the US Amateur Radio regulations are agnostic about open versus non-open. Thus, the issue of open systems versus non-open systems is an issue that each Amateur Radio Operator decides on their own particular value system. The only issue that US Amateur Radio regulations address are deliberate obfuscation of transmissions with the intent to obscure the meaning of the information being transmitted.
I posit that VARA is not deliberate obfuscation because one can easily obtain a compatible system (download the software) and play a transmission back through that system and reassemble the information. You don't need any "private" knowledge to reassemble the information such as an encryption key. In my mind, requiring "private knowledge", or not, is THE essential difference between a novel / unique modulation method like VARA versus a system that encrypts information.
As to whether such non-open systems such as VARA are "good" for Amateur Radio or within the "spirit" of Amateur Radio... in the absence of an applicable regulation, the decision of whether to use a system, or not, is left to each individual Amateur Radio Operator to apply their value system.
As for me, I WISH Amateur Radio had a system that provides equivalent performance and reliability to VARA FM as an open source project that we could learn from and port easily to different operating systems like Raspberry Pi OS. There are several potential open source alternatives in progress such as the MMDVM-TNC potentially operating at 19.2 kbps with FEC, and a few others that I can't easily cite without doing some research.
But, at the moment, we don't have anything else for Amateur Radio that can use a 20 kHz VHF / UHF channel, with off the shelf radios and audio interfaces, that can provide 25 kbps. That VARA is proprietary and runs only on Windows is a non-issue for the majority of US Amateur Radio Operators, that are adopting it widely. In my reading, that's because VARA FM is pretty "plug and play" with their existing radios for the cost of a VARA license key and perhaps an audio interface, and boom - they're getting much better performance and much better reliability versus 1200 bps packet radio (and 300 baud HF packet radio). In good conscience, I can't advocate that folks shouldn't use VARA FM and instead wait, perhaps indefinitely, for a "better" system that no one seems committed to bring into reality. I wish there was such an alternative, and I am CONSTANTLY on the lookout for such an alternative. Again, I had high hopes for MMDVM-TNC as near-equivalent to the performance of VARA FM, but that project seems stalled with no observable progress towards a working system that can easily be reproduced.
To conclude, I'm very, very glad that EA5HVK created VARA FM (and VARA HF), if for nothing else than to demonstrate what IS possible within the constraints of
* 20 kHz VHF / UHF channel
* Amateur Radio VHF / UHF radio with "9600 connections" (such as the Kenwood TM-V71A)
* A wide bandwidth audio interface such as the Masters Communications DRA series products.
That, with that equipment, a modest Windows PC, and a $69 VARA license key, VARA FM can achieve 25 kbps with Forward Error Correction isn't something I previously thought possible. But EA5HVK proved it IS possible, so the challenge for others who believe in open source is to create a system that provides equivalent performance and reliability of communications... as an open source project. EA5HVK was (and is... improvements are ongoing) FOCUSED on bringing HIS system into reality.
The open source community, to date, hasn't demonstrated equivalent focus to create a system that's comparable in performance and reliability to VARA FM.
thanks for the interesting perspective about Vara.
The idea that "one can easily obtain a compatible system (download the software) and play a transmission back through that system and reassemble the information" can be a little naïve in 2024. For instance the mapping between OFDM subchannels and actual symbols could be dynamically changed during the transmission based on say a pseudo-random sequence to limit noise and inter symbol interference; the purpose of that would be to improve performance and therefore acceptable, but the 'scrambling' algorithm would be very close to what encryption does.
What I am trying to say is that these days there's a very fine line between reasonable methods to enhance performance and obfuscation. In my opinion, what it matters is the intent, and that can be clearly shown by publishing the current and future specifications of the modem. As they say, the devil is in the details.
As per what the US Amateur Radio regulations say, I am 100% in agreement with you, and people are free to do as they want within those boundaries. One observation though: if instead of EA5HVK, the Vara modem, program, and distribution to the public came from say B1CCP (I just made the callsign up, but you get the idea) under the same conditions (Windows binary only, no detailed protocol specification), we and the US ham radio community in general might have a different conversation now.
Finally my comments were not intended as a 'call for action' of any sort, except perhaps for EA5HVK to release the details about VaraHF/VaraFM. They are more a personal reflection on the (sad) state of things and just food for thought for an interesting discussion like the one we are having.
Steve, I find it interesting that in the same issue you mention VaraHF/VaraFM (in the 'EastNET Quarterly Videoconference on 2024-03-14' section): "the primary use of VARA HF was not for networking per se, but moving messages regionally between Bulletin Board Systems", and later on you write: "Novel (to someone) modulation methods are not encryption, as long as you’re prepared to prove that the modulation’s intent is not “obfuscation” of the content, and that the content can be recovered using a typical example of the system (or published protocol, or algorithm)."
As far as I know Jose Alberto Nieto Ros, EA5HVK hasn't published a recent specification of VaraHF. The only one I was able to find is an 8-page word document from about 7 years ago, and it begins with this sentence "VARA HF Modem is a propietary (sic) system developed by Jose Alberto Nieto Ros EA5HVK and can be used under shareware license."
Perhaps in my search I missed something, but I don't think that that Word document qualifies as a 'published protocol', that would allow a third party to recreate a new implementation of VaraHF/VaraFM without reverse engineering EA5HVK's software.
73,
Franco Venturi K4VZ
Franco - First, apologies for my late reply. Thanks for your comment. I said:
At least in US Amateur Radio, use of a proprietary system isn't a disqualification for use in Amateur Radio as long as (quoting myself):
"... the content can be recovered using a typical example of the system (or published protocol, or algorithm)."
Note that DMR is widely in use in Amateur Radio, and it uses a proprietary CODEC from the company DVSI, the algorithm for which has not been publicly disclosed. Another example is PACTOR-4 from the company SCS, which is in wide in use in Amateur Radio but the algorithm for PACTOR-4 has not been publicly disclosed. Thus I think that EA5HVK's decision not to publicly disclose his algorithms for the various VARA software versions - HF, FM, and SAT is on par with DVSI and SCS.
In all of those cases mentioned above, a "typical example of the system" is easily obtainable - the DVSI AMBE CODEC chip built into a DMR system, a SCS PACTOR-4 modem (expensive, but available), and the VARA software is available for download and can be used to receive. You can obtain these, and run a recording of a transmission through these units and reconstruct the content sent via the original transmission.
With these systems, and others, the intent of "different" data communications systems isn't to obfuscate (encrypt), but rather make data communications more reliable, require less bandwidth, require less transmission time, etc. Thus, such systems ARE legal for use in US Amateur Radio. (Again, I'm only familiar with US Amateur Radio regulations.)
Like everyone in Amateur Radio, I'd love it if the DVSI CODEC were released into open source, or PACTOR-4 could also be made open source, and if EA5HVK would make VARA open source. But they haven't chosen to do so, and that's their right for their creations.
Hmmm... didn't realize there is a length limit for comments. I was afraid that I had lost the last paragraph (below), but apparently Substack simply truncates comments that are "too long" per its standards.
But with the example of HOW these systems work, and how WELL these systems work, there's nothing wrong with developing workalike systems that do equivalent things and have equivalent performance. Indeed, I think with the tools software developers have available now, I think it's ultimately inevitable that we will have such equivalents. Note that "equivalent" is not the same as "reverse engineered" and not the same as "interoperable". Disclaimer - I was on record early in my discovery process of VARA FM of advocating for an open source equivalent of VARA FM that would be interoperable with VARA FM. I was eventually persuaded that interoperability was a poor goal, and what should be worked on is a new open source system that works equally well or better, and has equal or better performance / reliability.
Steve,
first of all thanks for your detailed reply and taking some of your busy time to write it.
I think it is important to have this discussion between the two of us, and in the larger ham radio community because of the impact I believe it has on the future of ham radio.
A few comments on what you wrote below.
- I think that the fact that this problem exists for other systems, methods, and protocols currently used in ham radio is beside my point, and doesn't make the specific case of Vara more 'right' or 'wrong'. It is like a defense attorney telling the judge that his client is not guilty (or less guilty) because others have done it too; it doesn't work this way. Thus I'll keep my notes below to just VAra and EA5HVK.
- If the intent of the Vara is not to obfuscate, then I think they can prove it by releasing the current (and future) specifications of their modem; note that I say the specifications of this digital mode, not the source code for it. Take the case of FT8/WSPR: if Joe Taylor K1JT, Steve Franke K9AN, Bill Somerville G4WJS had released just binaries for Windows, Linux, and Mac free of charge, and hadn't publish extensive documentation on how these protocols work (see here for instance: https://wsjt.sourceforge.io/refs.html), people would probably still be using them as much as they do now, but I would have exactly the same issues with them. I think their outstanding contribution of the world of ham radio communication is coming up with the idea/protocol and making it public. Others can "stand on the shoulders of giants" to learn from them and further advance our hobby.
- I feel that the point of your sentence "You can obtain these, and run a recording of a transmission through these units and reconstruct the content sent via the original transmission." could apply to pretty much to anything: from that point of view, even a black box that takes say a vector of 2^N inputs and produces another vector of 2^N outputs could be described that way. I could make it available to anyone for free, but it might take a third party several billion years to go through all the inputs and analyze their outputs to figure out how it works. That's pretty much what encryption does.
- Interoperability (or the possibility of) is at the core of ham radio 'values' and how we use our public bands; after all one of the goals of ham radio is to promote open communication between people and open experimentation, and interoperability is critical to achieve that in my personal opinion. Technically there are probably systems and methods that are 'better' (in some respect) than Vara, but I think that doesn't mean that interoperability with Vara is less important given how widespread Vara use is and the fact that it is used to say send emails and other messages. Also if Vara current specifications were available, other open source projects could learn a thing or two from Vara, and perhaps improve their own algorithms.
- Finally I am not advocating that EA5HVK make his software open source, or that he give it away for free, or that he make the software also available for Linux or Mac. I write software for living, and I think one should be paid for their work, especially for making it convenient to install and run, for designing a nice and user friendly UI, and so on.
What triggered my initial comment was on a different level, to express a concern that I have about VaraHF/VaraFM, and, as you pointed out, on a 'trend' in today's ham radio environment.
73,
Franco K4VZ
Franco - I think you'll agree that the US Amateur Radio regulations are agnostic about open versus non-open. Thus, the issue of open systems versus non-open systems is an issue that each Amateur Radio Operator decides on their own particular value system. The only issue that US Amateur Radio regulations address are deliberate obfuscation of transmissions with the intent to obscure the meaning of the information being transmitted.
I posit that VARA is not deliberate obfuscation because one can easily obtain a compatible system (download the software) and play a transmission back through that system and reassemble the information. You don't need any "private" knowledge to reassemble the information such as an encryption key. In my mind, requiring "private knowledge", or not, is THE essential difference between a novel / unique modulation method like VARA versus a system that encrypts information.
As to whether such non-open systems such as VARA are "good" for Amateur Radio or within the "spirit" of Amateur Radio... in the absence of an applicable regulation, the decision of whether to use a system, or not, is left to each individual Amateur Radio Operator to apply their value system.
As for me, I WISH Amateur Radio had a system that provides equivalent performance and reliability to VARA FM as an open source project that we could learn from and port easily to different operating systems like Raspberry Pi OS. There are several potential open source alternatives in progress such as the MMDVM-TNC potentially operating at 19.2 kbps with FEC, and a few others that I can't easily cite without doing some research.
But, at the moment, we don't have anything else for Amateur Radio that can use a 20 kHz VHF / UHF channel, with off the shelf radios and audio interfaces, that can provide 25 kbps. That VARA is proprietary and runs only on Windows is a non-issue for the majority of US Amateur Radio Operators, that are adopting it widely. In my reading, that's because VARA FM is pretty "plug and play" with their existing radios for the cost of a VARA license key and perhaps an audio interface, and boom - they're getting much better performance and much better reliability versus 1200 bps packet radio (and 300 baud HF packet radio). In good conscience, I can't advocate that folks shouldn't use VARA FM and instead wait, perhaps indefinitely, for a "better" system that no one seems committed to bring into reality. I wish there was such an alternative, and I am CONSTANTLY on the lookout for such an alternative. Again, I had high hopes for MMDVM-TNC as near-equivalent to the performance of VARA FM, but that project seems stalled with no observable progress towards a working system that can easily be reproduced.
To conclude, I'm very, very glad that EA5HVK created VARA FM (and VARA HF), if for nothing else than to demonstrate what IS possible within the constraints of
* 20 kHz VHF / UHF channel
* Amateur Radio VHF / UHF radio with "9600 connections" (such as the Kenwood TM-V71A)
* A wide bandwidth audio interface such as the Masters Communications DRA series products.
That, with that equipment, a modest Windows PC, and a $69 VARA license key, VARA FM can achieve 25 kbps with Forward Error Correction isn't something I previously thought possible. But EA5HVK proved it IS possible, so the challenge for others who believe in open source is to create a system that provides equivalent performance and reliability of communications... as an open source project. EA5HVK was (and is... improvements are ongoing) FOCUSED on bringing HIS system into reality.
The open source community, to date, hasn't demonstrated equivalent focus to create a system that's comparable in performance and reliability to VARA FM.
I'll let you have the last word in this exchange.
Steve,
thanks for the interesting perspective about Vara.
The idea that "one can easily obtain a compatible system (download the software) and play a transmission back through that system and reassemble the information" can be a little naïve in 2024. For instance the mapping between OFDM subchannels and actual symbols could be dynamically changed during the transmission based on say a pseudo-random sequence to limit noise and inter symbol interference; the purpose of that would be to improve performance and therefore acceptable, but the 'scrambling' algorithm would be very close to what encryption does.
What I am trying to say is that these days there's a very fine line between reasonable methods to enhance performance and obfuscation. In my opinion, what it matters is the intent, and that can be clearly shown by publishing the current and future specifications of the modem. As they say, the devil is in the details.
As per what the US Amateur Radio regulations say, I am 100% in agreement with you, and people are free to do as they want within those boundaries. One observation though: if instead of EA5HVK, the Vara modem, program, and distribution to the public came from say B1CCP (I just made the callsign up, but you get the idea) under the same conditions (Windows binary only, no detailed protocol specification), we and the US ham radio community in general might have a different conversation now.
Finally my comments were not intended as a 'call for action' of any sort, except perhaps for EA5HVK to release the details about VaraHF/VaraFM. They are more a personal reflection on the (sad) state of things and just food for thought for an interesting discussion like the one we are having.
73,
Franco K4VZ