I have easily touched a thousand or so systems this past year. From support experience, James’ 1% is an over-estimate of D&U usage. By keeping tabs on discussions here, we know that D&U topics don’t come up very often, another data point that anecdotally suggests low usage. This thread has so far garnered only a single user (in addition to OP) with usage experience. Obviously there are selection biases in these cases, and your point about users being discouraged due to the policies in place probably dampen the numbers somewhat, but claims that a significant minority are using D&U look unsupportable from my perspective.
Users engage on the forums, but developers engage on the issue tracker, and it is probably this disconnect that fosters an impression that we are not listening. A forum thread about an issue with D&U will likely result in the user being informed the feature is unsupported; not particularly helpful but it is important they know that. For some reason (this thread included) “unsupported” gets interpreted as “deprecated”. However, that same user files a bug ticket properly describing the issue, it gets discussed internally and in all recent cases I’m aware of, a fix is coded.
I’m grateful for so many explanations of “unsupported”. But, Lorne, here’s your statement that concerns me:
(my emphasis added)
jfinstrom is even more definite. As a commercial user, I therefore face the possibility that any update could break my production system. My only option to remove that threat would be to maintain a testing platform, complete with dedicated endpoints, and run extensive tests for every update. Not a realistic proposition. I fully understand (and always have) the meaning of “unsupported” as well as the potential consequences. I didn’t start this thread to discuss semantics.
I am trying to start an open discussion about revisiting this policy, about moving FreePBX towards fully embracing and exploiting Asterisk’s underlying D&U structure. I can imagine a future version of FreePBX where Devices and Users can be freely mixed in one-to-one, many-to-one and one-to-many relationships. What a product that would be! As a former coder (and system architect) I know it would be do-able. As a business user I know it’s what I need and want - and I’m very sure that today’s market would widely appreciate such elegant flexibility.
So what does the team, and the community, think of that idea?
This is absolutely correct. If you want to ensure a use case that is not tested does not break production system you would have to do your own testing.
It seems from the constant back and forth that you don’t understand the true difference between “device and user” vs “extensions”. That is ok, we don’t really expect anyone to know because you shouldn’t have to. You mention home and hobby users. These are the people more likely to use D&U mode. They like to tinker, play, and take things beyond their normal uses. It is likely any tutorial you have read on device and user mode was from one of these home/hobby users. I would go easy on them. FreePBX is written with business use in mind. Most small to medium size businesses like to have their receptionist do their move,add,changes. We have non-technical people do usability testing. This is not meant to be rocket science.
FreePBX is used as everything from an overcomplicated answering machine in some guys house, to complex corporate PBX’s. It is used in major enterprises and in government agencies. It is used at airports across the country for emergency response.
FreePBX is not just for one market segment. All decisions made have to account for all market segments.
If you would like to contract and pay our QA team to test Device and User mode for every update I am sure our sales people can give you a quote on that.
At this point this post seems to be in some asinine loop So I assume all points of view have been expressed. I will flag this to the forum staff to consider closing.
No - please take a moment to re-read my opening post. Nobody has addressed the original question/issue! Almost the entire conversation has been hijacked to define the word “unsupported”, propose (partial) alternatives to D&U Mode and so on.
Look at my first sentence… prophetic, eh? As in every other thread concerning D&U, you hammer the “unsupported” thing and then want to close thread! Why won’t you guys actually discuss this? Somebody put the D&U switch in there in the first place… why? What use cases did you intend to cover at that time? Don’t you agree that the landscape has changed considerably since then? Do you agree or disagree that embracing D&U in the GUI could yield a more flexible, attractive product with even wider appeal? etc…etc…
I don’t believe I’ve been out-of-order in this thread, but the latent aggression and condescension from the other side has got me exasperated, I’ll admit. I just wanted a discussion but all I’ve got is pedantic lectures - a very poor customer experience indeed!
Jeremy,
Unfortunately, I disagree with you that this isn’t being discussed, as multiple members of the team have chimed in here. As we have pointed out, Device and User mode is at the core of FreePBX from it’s early inception, so without major rewrites its not going away. But as we also pointed out from the data we have (via bug report, support requests, forum posts, and conversations with community member) there aren’t a lot of people that use it. Now that doesn’t mean nobody uses it, or it’s going away, or that we don’t care. Instead it just means we rely more on the community to bring issues to our attention or submit patches to fix issues people may be seeing in that mode. You also have to remember that the issues people really needed Device and User mode for are handled much better when using PJSIP and that commercial modules have never supported Device and User mode (again based on data). You also have to remember that FreePBX is an open source project, so if anyone from the community wants to contribute tests for Device and User mode where they are generally seeing issues, we’d certainly welcome them as well any bug reports (remember we triage these weekly with the team) and pull requests. We also welcome our community members to openly discuss this and let their voices be heard by responding to this post, opening tickets, conversing with us in IRC, having conversations with us at an event like FreePBX World, etc.
Like Jeremy, I am also an “edge case” user, in that I use unsupported channel drivers to support the phones my customers prefer to use (DAHDI and Cisco Skinny).
Over the years, I’ve made some suggestions about additions to FreePBX that would make it easier for me to support them (and their receptionists). A few have been adopted, and few have been met with “send us code.”
Some of the changes that have been made have made it my responsibility to support the stuff I like to use (the SCCP Management module for pre-Version 12 FreePBX was an example). The “hook” model for the “Asterisk Info” module is an example of places where the team has made it so I need to pull myself up.
If my business is reliant on something, I make sure that I’m responsible for making sure it’s going to work before I drop it into production. If this means a “subset” testbed, then it means a testbed. If it means I try the changes out on my development server, then it means I test them out. Whatever the case, I wouldn’t rely on this band of scoundrels to make sure that ANY update is going to work flawlessly. It’s irresponsible and ill advised to trust anyone with mission-critical assets like this.
In any of my areas of responsibility, if what I want to do is not supported by some basic product I use, it’s up to me to make it happen. That’s the model in every industry I know of - food service, telemarketing, solar energy production, printing, networking, industrial control manufacturing, communications, etc. If my stuff fails because of something that is unsupported that my business relies on, then it is no one’s responsibility to ensure it works but me.
Keeping that in mind, though, don’t for a moment think that I feel like I’m out here on my own. I’ve worked with several of the developers to make things work better, to suggest changes that make my life less cumbersome, and that I’ve submitted code that was used as a suggestion for a future feature. I work with my suppliers (FreePBX being just one of them) all the time, making suggestions that make it easier for me to support my stuff easier.
So, Jeremy, I suggest that you might want to change your mindset. You have seen several people from the team (including managers) that D&U mode is “unsupported”. If you NEED D&U mode to work, you should establish an update schedule that works for you (no more than semi-annual for a production system) and Beta your software to make sure it meets your needs. Another approach might be to change the way you look at the product set and see if there might not be an update or new feature that support what you want done, rather than looking at your task as getting things done they way you want them done.
If you have suggestions, you don’t need to discuss them - submit them. Put in feature requests. Come up with some real, concrete examples of what you want and how you want it to work. Write some code that implements the simplest of your examples. Ultimately, this is a community supported project, so …
As far as I’m concerned, the answers to all of these questions is “who cares!” Whining about past decisions it clearly only useful if you are looking to lay blame. The decisions were made. What’s far more important is the route forward.
Whether people agree or disagree with your assertion isn’t really important or pertinent. If you have some recommendations about improving the GUI to support D&U mode, send them in as feature requests. If you can think of some way to improve support for D&U mode, then suggest it.
Finally:
Are you really a customer? Truly? You pay for every bit of support? You pay a retail value for every service you get from the project?
FreePBX (with the exception of the commercial modules) is community supported. Are you putting in as much as you are getting out? Are you really working toward what the rest of us are? Are you doing everything you can to make it better, or are you using the “millennial” approach, and sitting back waiting for someone to wait on you while you consume resources that could be put to better use somewhere else?
I’ll have a lot more support for your position when I see that you are actively working to make your edge case work as well as mine does.
Shared wifi-connected devices… Android/iOS smartphones and tablets. Very popular these days. Hotels, hospitals, care homes, warehouses, retail management, education, manufacturing plus, I’m sure, many other sectors. We’re certainly not talking niche or vertical here! Unlike the user-centric BYOD arena (nicely handled by pjsip in Extension Mode), this new mode of usage requires a device-centric model to utilise these shared devices for voice communications through a PBX. “Users” (who may or may not also have a deskphone) must be able to attach (temporarily) their unique identity (extension number and CID name) to a “Device”. Hot-desking but with handheld devices, not deskphones.
MDM (Mobile Device Management) vendors (MobileIron, for example) are already catering to this new business need. Polycom/Spectralink have launched devices specifically aimed at this market. It’s “hot”. I’m asking that FreePBX acknowledges this huge new field, at least to the extent of engaging in an open discussion that could possibly range across the requirements, the benefits, the pitfalls and possible timescales of re-thinking and re-working parts of the FreePBX GUI in order to embrace this significant trend. I’m not suggesting a straight-forward switch to full support of D&U Mode as is. I’m offering my own experience from my recent small but highly effective implementation of this model in my own business, as well as active participation in any activity that may come out of such discussion.
Side Note:
I’m very familiar with open source projects, having contributed (very) significantly to two others in the past. I try never to take without giving back: you may have noticed that, to balance the help I received here on a couple of small implementation issues, I did my best to help in some other threads.
We’ve had that open discussion, by the way. It’s this thread… Basically, it can be summarised ast this:
You want more time spent on Device and User Mode
Everyone else says that it’s not really useful, except in a small number of cases, and that PJSIP is much better for 99.9% of the times D&U would be used.
You disagreed
So, now that everyone knows where things stand, what is the way forward from here?
You should create Feature Requests
You should report bugs
You should write documentation
You should submit patches
If you think that D&U mode is such a critical thing, then I think the best thing you could do is to write documentation in the wiki, explaining how to do the things you think are cool.
If other people think it’s cool, they will start doing it too, and then more bugs and feature requests will get created, and it’ll get more visibility, and more people will start using it, which means more bugs and feature requests, and more users, and more documentation, etc…
You’re kidding, surely? All I got was tetchy dismissals, repetitive lectures and unwarranted attacks. Nobody acknowledged, let alone addressed, the real subject of the post.
I gotta say, this forum functions like a typical open source project - abundant over-sensitivity, some startling arrogance and generally poor reading, compression and communication skills. (Yourself excepted of course - you stepped in and helped me on another thread and I thank you sincerely for that.) If this was pre-Sangoma days and if I’d just downloaded the product for free to install on my own hardware… well, then I wouldn’t have a foot to stand on, naturally. But neither is the case. I bought a turnkey IP-PBX box (and some commercial modules), aware that the core software was OS, but with expectation of at least some of the advantages of a commercial product. Depressing, then, that I’m challenged to write documentation, submit code/patches and meekly ingratiate myself to “the community”. You’ve got a real identity problem here - Sangoma and FreePBX rather messily mixed together: Sangoma markets/sells complete IP-PBX systems direct to businesses but the FreePBX side still behaves like a hobby site.
“Cool”? Really? I’m obviously in the wrong forum - perhaps you could direct me to the Sangoma/FreePBX forum where I can discuss my business needs with the originators of the product I’ve purchased?
P.S. Please don’t anyone take any personal offence at this rant - I’m railing against the whole, not any individual.
I’m pretty sure that’s what I did? Am I missing something here?
No, I think you’re confused.
Sangoma pays my wages. They pay me to work on (amongst other things) FreePBX. FreePBX is an Open Source project. It also has commercial modules, which are commercially supported. You can also pay Sangoma to commercially support your Open Source FreePBX, with or without commercial modules. There is no difference in the FreePBX Open Source Project pre or post Sangoma. Anyone who says otherwise is fundamentally wrong, and doesn’t understand Open Source software.
This, the forums, is the community support if you do not want to pay for commercial support. The url is even ‘community.freepbx.org’.
However, even if you pay for commercial support, the answer will still be ‘Device and User mode is not supported’. We don’t support it because… well, look, I don’t need to repeat myself. You seem to be dismissing the reasons why we don’t support it as ‘tetchy dismissals, repetitive lectures and unwarranted attacks’.
If you’re going to continue to refuse to listen to us, then I can’t really see any reason to keep responding in this thread.
I have not criticised the free support offered here - it’s generally admirable. I’m more than aware of the support options open to me. I’m not asking for support in this instance, simply discussion - and on that I’ve been comprehensively refused.
The refusal to listen is coming from your side.
Lorne was the only person to acknowledge that there was possibly something worth discussing, but left it at that.
Presented below is everything that I wrote which remains, for the most part, without response:
@GSnover and @billsimon chipped in with support - thank you for that, guys. Anybody else that might have been of a mind to join in has, I’m sure, been dissuaded by the length and repetitive nature of this thread.
Now we really are going round in circles! You gave me a little speech above about planting little acorns to grow mighty oaks. Don’t you see that you guys have done the exact opposite? Every time D&U gets mentioned one (or more) of you chimes in with “Not supported!”. Of course you’ve got few commercial users of the mode - you’ve made it abundantly clear over the years that it’s “untouchable” for commercial use. You’re killing it by chopping off every shoot that tries to get out of the ground. Just ask Google.
- you stepped in and helped me on another thread and I thank you sincerely for that.) If this was pre-Sangoma days and if I’d just downloaded the product for free to install on my own hardware… well, then I wouldn’t have a foot to stand on, naturally. But neither is the case. I bought a turnkey IP-PBX box (and some commercial modules), aware that the core software was OS, but with expectation of at least some of the advantages of a commercial product. Depressing, then, that I’m challenged to write documentation, submit code/patches and meekly ingratiate myself to “the community”. You’ve got a real identity problem here - Sangoma and FreePBX rather messily mixed together: Sangoma markets/sells complete IP-PBX systems direct to businesses but the FreePBX side still behaves like a hobby site.