My 12 grade math teacher would have said ‘Ee lad, you got verbal diarhea there’ 
, can you re-express in english?
2 Likes
This was an issue with HyperV VM’s only? We migrated several FreePBXhosting systems from 15 to 16 inline without issue (other than losing the “free” EPM module that FreePBXHosting had provided up through v15). Also no issue with onprem devices (we roll our own onprem devices as Sangoma’s are signifcantly overpriced for the same hardware).
Yes hyperV only as Centos had removed support in the kernel for it by default as far as I recall.
2 Likes
Also, can anyone at Sangoma explain why Debian became the default Linux version and not one of the Centos “forks” like Rocky or Fedora which it would seem would have been much easier to migrate from Centos to?
It does not appear that anyone at Sangoma is willing to provide any “clarification” on the issue of the FreePBXhosting “partnership” - or any other concerns raised in this thread.
@chuckjuhl I have raised this internally as a question for someone else to answer. I don’t know if we can or can not answer, and I also have no knowledge of it. As I stated in the other thread, getting answers to things can take time.
No offense intended, but honest and forthright answers as often as not do not come at all from Sangoma, much less “just takes time.” Recently I’ve been following Sangoma’s financial disclosures and earnings calls, and from the information gleaned from those sources, it certainly appears that Sangoma’s management’s direction is away from supporting the open source community and apps like asterisk and FreePBX that Sangoma’s commercial products are (rather ironically) built on.
Frankly, it may be time to fork FreePBX much like Centos forked from RHL amd SuiteCRM forked from SugarCrm. Bigger companies than Sangoma have tried heavy handed-control of open-source code - Oracle’s acquisition of OpenOffice from Sun Microsystems resulted in Libre Office and now OpenOffice is virtually dormant. Oracle’s Java (also aquired from Sun) is longer the tool of choice for a good many projects, including commercial vendors like Ubiquiti, which have moved to the Eclipse Foundation’s Temurin OpenJDK runtimes.
I’ve been in IT since 1974 (I’m old as dirt, I know). I started with Cobal and Fortran, moved to Microsoft’s BASIC interpreter (when it was a little shop in an Albuquerque strip mall Named Micro Soft) and coded over the years in pretty much everything from C to Pascal to ADA (does anyone remember Borland’s Turbo Pascal and Delphi IDE?) I had stopped hands on coding pretty much by the late 90’s and moved into “administration” after my MiSM, but recently with the newer AI tools I’ve started coding more again (or more accurately giving instructions to ChatGPT to code for me) instead of outsourcing projects. Certainly, the tools are here to fork FreePBX with minimal effort and expense. It may be time to divest from Sangoma if Sangoma continues to view the Asterisk and FreePBX open-source community like an unwanted stepchild.
I’m not holding my breath waiting for answers from Sangoma’s management, but I will continue raising those questions here and in related forums.
Personally my opinion is that people do what is right for them - and that includes forking things. From an open source project perspective though if it’s a relatively minor fork (or you just do it for yourself) and stays that way it can be fairly easy to keep up with things. If it grows and gains momentum as a fork, the investment can be substantial and underestimated. The expectations of people grow and there’s lots of ancillary things to actually facilitate a project and keep it going. If there’s also an established user base that may move over, that can put more pressure on it. This is based on my experience leading the Asterisk project and being part of it over so long.
On the topic of LLM based coding solutions, they’ve certainly improved but they still require oversight - particularly on established large code bases that may have varying quality of code.
I think you may be overstating the “challenges.” I’ve watched Libre Office and SuiteCrm start from virtually zero. And LLM coding tools are actually very good. There’s actually a number of studies showing LLM developed code is as good and even getting better than “hand coding” when properly prompted.In fact, Prompt Engineering looks to be the future of programming. A properly prompted LLM can generate code, and changes, in minutes where it would take weeks to develop the same code even in a decent IDE. I work with several universities, and Prompt Engineerig is being incorporated into Computer Science programs now. Hand coding JAVA, PHP, etc will become as “obsolete” as programming in machine language was back in the 70’s with first PC I owned (an Altair 8800). And “oversight” is far less resource intensive than building the actual code base, and we even have tools now to automate much of that “oversight” and QC. Heck, even the Linux kernel is entirely open source and several of the Linux Foundation’s projects are good examples of entirely community-driven development and support. Sangoma’s business model is basically repackaging open-source operating systems (Linux) with open-source applications (Asterisk, FreePBX) and acquired commercial modules using open-source development tools. By Sangoma’s own statements in its latest earning reports, it is only providing minimal resources to the open-source projects beyond what benefits Sangoma’s commercial products, while extracting the benefit of the Community contributions.
I didn’t give a complete list of challenges, I’m merely saying that you have to think about everything involved. Looking at it purely from a code perspective won’t give you the full picture of what is all involved to operating an active open source project. It is achievable, certainly, but it does take effort, time, and resources. The examples you’ve provided certainly show it can be done, but they had to put all of that in and continue to.
True, but just not nearly as much time, effort, and resources as 20 or even 10 years ago. Where it begins to get narly is when the stewards of the code attempt to monetize it.
Sangoma is turning out to be a textbook business model of what happens when putative open-source development becomes subservient to aspirations of commercial profits. Has Sangoma ever turned an actual profit other than EBITDA fiction? It has been profitable for CEO Charles Salameh - to the tune of over $3 million a year as I recall.
Saying something like this is like saying “Well it had 100 steps to follow but not it only has 90”. It doesn’t take nearly as much time or effort but ignores the true amount of time and effort of something.
If I was to fork all 70+ OSS models of FreePBX tonight what does that do? The project has been forked, you could get it from GitHub not from Sangoma. Does that do anything, really?
the new v17 ISO is potentially simpler than the old v16 ISO.
That particular installer boot option was designed for INTernal use on Sangoma appliances, but it may work for others.
In the context of installing operating systems, it usually means Fully Automated Installation.
No problema. 
It MAY POTENTIALLY work on others? Its BETA. I actually bill for beta testing for my commercial clients like Bausch and Lomb, eClincalworks, et. al.
I seriously doubt it can be simpler than the V 16 ISO, and definitely NOT simpler for migrating cloud installs on FreePBXhosting com.
Sangoma answers have become Orwellian frankly.
In any event, my unfamiliarity with your acronyms gives away my advanced age 
The clarifications are appreciated. “Int” in programming languages generally refers to integer, and FAI is taught in engineering and in manufacturing processes to mean a “First Article Inspection” - a production validation process and was pretty much a standard requirement in data hardware procurement from my D.O.D. days. And the only place I’ve heard the term “spice level” is at Taco Bell Did you make these terms up yourself?
Really? What did forking SuiteCRM from SugarCRM accomplish, or forking LibreOffice from OpenOffice, or forking Centos from RHL for that matter? It seems to me that the benefits are self-evident.
And its more like 100 steps have become 10.
I think you’re missing the bigger picture here. Just forking the code isn’t enough. There needs to be infrastructure, resources (time and money), management, support and a whole bunch of things in order to do what all those other projects you pointed out have done.
Not missing the “big picture” at all. This ain’t my first rodeo Actually, I’m exploring the idea of putting together those resources with the assistance of the Computer Science departments of CSU-LB and Golden West Community College - both of which I draw interns from. This could be a great project as a for-credit project and would be a significant feather in those institutions’ C.S. programs, especially Golden West CC. To be very upfront, if Sangoma doesn’t get its act together supporting the Open Source Community that its built on, I will be exploring the viability of a project to fork FreePBX following the model of SuiteCRM’s fork from sugarCRM and OPNsense fork from pfSense.
As I’ve mentioned previously, this ain’t my first rodeo Actually, I’m exploring the idea of putting together those resources with the assistance of the Computer Science departments of CSU-LB and Golden West Community College - both of which I draw interns from. This could be a great project as a for-credit project and would be a significant feather in those institutions’ C.S. programs, especially Golden West CC. To be very upfront, if Sangoma doesn’t get its act together supporting the Open Source Community that its built on, I will be exploring the viability of a project to fork FreePBX following the model of SuiteCRM’s fork from sugarCRM and OPNsense fork from pfSense.
I’ve gone ahead and purchased the domain names opnpbx.org and opnpbx.com. I’m having my ip attorney start the trademark registration process. I’m putting feelers out to the suitecrm project as well as the opnsense project to see if there is any interest in collaboration. The suitecrm project would benefit especially from a close collaboration with an open source UC application. This may turn into a viable opportunity
Well, you could try it out, on your test bench, side by side, and let us know what you think ?
This was covered a bit at AstriCon 2025 – some videos should be up shortly on YouTube.
Agreed, although there is a FOG spice level for cloud-friendlier setups, but more manageable migrations are on the list for v16 → v17.
They are a combination of acronyms and abbreviations – nothing too clever though. (INT = INTernal.)
Interesting, but not the intent here. The FAI Project is a good general place to look into ideas related to Fully Automated Installation of various Linux distros although not what was used in the case of this ISO.
Not all of them, no.