I’d love to have a frank and open discussion about D&U Mode (that isn’t terminated abruptly with “we don’t support it”).
The move away from the “old” model of hard phones on desks is accelerating. Softphones, floating users and shared devices are the future. Frankly, I think that Sangoma/FreePBX are missing a huge opportunity to be at the forefront of this trend. (Of course there will always be a place for deskphones, particularly sexy new gear like the Sangoma devices and Ubiquiti’s Unifi phones.)
D&U Mode is a perfect fit for this future. I believe it is the only way to support both “one user, many devices” and “one device, many users” simultaneously. So why is FreePBX/Sangoma trying so damn hard to suppress it?
We are a nursing home (a 24/7/365 business) running FreePBX in D&U Mode on a Sangoma 60 box. We have (so far) 4 deskphones (fixed), 3 Android phones (fixed), 6 Android phones (ad hoc) and 13 iPads (ad hoc) - servicing 58 registered Users. D&U Mode works perfectly for us and is now “Business Critical” here. Are we a unique business? Are we even an unusual business? - certainly not! Will more and more businesses operate like this in the future? Yes!
Frankly, the argument that “D&U Mode breaks many modules” is just a cop-out. It could be made to work for the majority of modules - any that really can’t be supported could be marked as such - no harm, no foul.
It is not fair to say that D&U is being suppressed by Sangoma or anyone else. The stated policy of the project is that D&U is ‘unsupported’, but we are aware that some users rely on it. We are also aware that some users rely on it without realizing that it’s unsupported. Unsupported means that we are not actively adding features, nor are we necessarily testing new code in a D&U configuration. It is not at all correct to think ‘unsupported’ means that current features are being actively removed, nor do I think you will find any recent examples of this.
The devs strive not to break any D&U features, and there have been recent tickets proving exactly this. Any tickets indicating that new code breaks existing D&U functionality will get attention, and provided the fix is not too onerous, will probably be addressed by Sangoma. And of course, contributions from the community are always welcome.
So what does this mean for the D&U user? Not a whole lot other than to be aware that module upgrades might interfere with D&U, so you need to know how to rollback modules in Module Admin. Maybe you will want to ensure that you have a good backup prior to doing a mass upgrade so you have something to revert to. You will absolutely need to test major new versions of FreePBX well prior to putting into production.
The current policy is no more than recognizing that resources are limited. Some things get priority and some things don’t, and while we could assign a low priority to D&U internally without notifying users of this, obviously transparency is better for all, even it it is not what some users want to hear.
Thank you, Lorne - that’s a more positive statement than I’ve read anywhere else but only in terms of reassuring us that D&U is not going to disappear per se. Nevertheless, there are dangerous undertones in your statement, particularly:
(my added emphasis)
What I’d like to hear, though, is WHY this increasingly relevant and useful capability is being slowly, but surely, killed. Do the developers not agree that simultaneous one-to-many and many-to-one support IS an increasingly NECESSARY feature in a comms world where handheld devices are already dominant? I understand your point about resources. I’d stay quiet if the subject were a “legacy” feature in any way - for example, support for deprecated endpoint hardware - but it’s actually the precise opposite. BYOD and corporate-owned handheld devices are, quite obviously, THE hot topic of today. Supporting them fully (as D&U does) should be a priority for the project. “We’ll do our best not to break it” is not only a worrying response, it’s an illogical one!
Dangerous undertones? Your choice of words is incongruous with the topic being discussed.
The policy is not new, and to my knowledge is applied consistently. You bring up some noteworthy points, and if enough people agree with you, the policy may well will be revisited. This would be an excellent place for others to chime in with opinions.
Yes it does - but only that one scenario: most often (I’m guessing), a user with a company deskphone and a personal mobile. Critically, though, pjsip is still AFAIK marked “not ready for prime time”. Under D&U we use one-user-many-devices for two managers who each have a desk phone and a smartphone - it enables them to go out onto the “shop floor” (in our case a 38-bed nursing home but it could equally be a hotel, a warehouse, a factory or whatever) and remain “online”. Very useful.
But the majority (19) of our devices are in ad hoc mode under D&U - users (in our case nurses and carers but could equally be hotel staff, warehouse staff, factory staff or whatever) who grab a “device” (Android or iPad according to their role) from the charging rack at the start of their shift, sign in to it (*11…) and then spend their entire day on the “shop floor”. It has been a revolution in our business - no more “Has anyone seen Claire/Bob/Charlie?” games - a massive saving in man-hours spent simply walking around the 50-plus room premises looking for somebody or disturbing everybody (patients included) with tannoy announcements. EDIT: to reinforce this point: this is probably the single most effective efficiency improvement we’ve ever made in our twenty-year history - more effective inter-staff communication and less wasted time equals better care for our patients.
To emphasise the flexibility afforded us by D&U mode, one of the managers recently left their own smartphone at home by mistake. No problem - just pick up an Android from the rack and *11 into it with their extension number. This is exactly the kind of flexibility that users expect these days. D&U takes it all in its stride - it’s how a phone system should operate in 2016!
No offence intended, Lorne - “dangerous to me” is what I meant.
I believe you’ve hit the nail on the head. The policy has, indeed, been clearly stated for many years now. Given that the comms landscape HAS changed since that policy was established, I do believe it’s time to revisit it.
I’m not intending to argue with you, just present some alternate technologies or methods. I agree with your points.
Other methods of doing similar are hot-desking that re-provisions the phone on the fly (Sangoma phones) and WebRTC where the user logs in through the web browser.
We used to use D&U extensively - it is a major training issue (people do NOT want to learn feature codes - phones with buttons make this easier) but once the people are trained, it works quite well.
I thought this is what UCP with WebRTC was moving towards - especially with “Soft” devices, a UCP login works like a charm - couple UCP with iSymphony and you have a pretty functional “Roaming” user.
MANY years ago we used to sell Interactive Intelligence - a Windows-Based Call Center PBX. Their big claim to fame was that when you logged into a PC that was associated with an endpoint, you “Claimed” the phone - it was very natural and automatic, but it assumed fixed devices and roaming people.
I too would hate to see this functionality go away - especially now when phones with buttons are good and cheap (we used to use Polycom - VERY button-poor).
How can I help keep this functionality alive? (In other words, devs, what can I do as a user that will help you make sure it still works as it should? Testing? Bug Reports? Feature Requests?)
Equally, not arguing with either of you - just completing the picture for the sake of new readers…
I looked at Bria Stretto but (for us) it’s a costly solution - the per-user-per-year pricing model is punitive in our application. It’s aimed at corporate executive users as their licence allows (I think) five devices per user. Thus it really only suits the one-user-many-devices scenario. It also ties in to UCP, which we don’t need or use.
We are using Media5-fone MPS - their one-off charge per-device model is a much better fit for our many-users-one-device need. I had to twist their arm to persuade them to allow me access to the portal: their system is designed for much large user-bases than ours. I wanted the remote provisioning facility because I support our installation from 1000 miles away. D&U will, of course, work perfectly well with most of the free and paid SIP apps out there - CSIPsimple, Zoiper, vanilla Bria, etc.
The Sangoma phones are lovely to look at and natively support hot-desking (without reboot) but they are, of course, desk phones. If they ever come up with an Android-based handheld device, I’ll certainly be in the queue to buy. Digium Phone Apps are an improvement over most IP phones, some of which are laughable - the cheap Akuvox desk phones we use claim hot-desking as a feature but the user has to enter server address, SIP port, username, password, etc. - I can just see users typing in a 32-character alphanumeric-plus-symbols password on a telephone keypad every morning :).
The Unifi phones look to me like a step in the right direction. Never tried one but they seem aimed (like the top end Grandstreams and other devices) at desk-bound executives. Haven’t found any mention of their hot-desking support but they should work fine under D&U. Pity that they’ll never release their app as a standalone product.
Polycom offshoot Spectralink have their Pivot device - SIM-less Android, aimed at hospital, hotel and warehousing sectors but they’re very pricey.
Worth mentioning, too, that we have two Unicom ICW-1000G wifi phones - good wifi performance (better than most smartphones) but fragile construction and no viable support for a remote directory.
For anyone who says “device and user” mode is going away it shows they don’t know how any of this works. The “device and user” mode simply changes a couple menu items.
FreePBX, specifically the extensions are all device and user based. In extension mode, these items are managed as a set. When you turn on Device and user mode it hides the joint configuration and exposes the configurations as separate items. So what “device and user” mode does is it gives you the power to break things. Some people like the above understand the plumbing and do things exclusively in this mode. They basically handle what we automate. Many others switch it on and don’t get how it works and then nothing works.
All features we write are written with the assumption that we are handling both items. We test under this assumption and release under this assumption.
Unsupported means exactly that. You are running an un-verified and untested environment. We won’t put out “it works” stamp on it because we don’t know if it is or is not broken.
We have a solid team of community members such as Greg above that does well to let us know if something broke. Often fixes are trivial and will be fixed assuming the fix does not break things for people not in D&U mode. These fixes are not really a priority but they are often trivial enough that they just get done.
tl;dr.
FreePBX extensions are ALL a combination of Devices and Users.
Depreciation would be simply removing the menu item switch
There is no plan to remove that advanced setting at this time
It is broken because we don’t test it, not because it doesn’t work.
Trivial non-breaking fixes are typically implemented when a bug is filed.
No, that is it exactly - Back when I was using Polycom’s the current models were 335, 450, 550, 560 and 650 - The 335 had exactly 2 programmable buttons and you needed one of them for the line! A 450 had 3 buttons so you could have a logon and logoff button and a line! 550 had a whopping 4 buttons and the 650 had 6 - so unless you wanted to put 650’s on everyone’s desk, they had to learn feature codes!
Now that there are Sangoma. Aastra (Mitel), Yealink and Grandstream phones with lots of buttons it’s a non-issue. Just when I was using it it was a pain.
As for who was using them, several Real Estate Offices and a couple of clinics - MA’s would sign in (Doctors were too good to learn anything…) at the clinic they were working at that day. Hot-Desking at the re-estate office meant they didn’t have to have desks for all the agents they had.
With respect, you are simply reiterating the same message that has been trotted out for practically every question about D&U Mode for years. We get it. And the inner workings are adequately explained in the wiki - I’m certain anyone that has engaged D&U mode (itself not a trivial single-click task) has read that and understood what they are doing. Furthermore, it is perfectly possible for D&U Mode to “go away” - if any of the critical basic modules were to change such as to preclude the use of D&U Mode then a whole load of us would be left out in the cold (given the project’s current stance on the matter).
What surprises me is this repetitive “stone wall” response when numerous users have, over the years, declared their interest in, implementation of and business reliance on this mode. Not to put too fine a point on it, but shouldn’t you perhaps be listening more to users? (I feel able to say such a thing because I’m not enjoying a “free” product - I paid for a turnkey box with what I consider to be commercial software on it - on that basis I want my voice as a user to be heard.)
Fundamentally this can’t happen with how FreePBX was designed.
I fail to see the stone wall you speak of. I believe you are talking about commercial modules of which the stance is “unsupported” (this seems to be a word that really irks you). However we also do not support IE 9 but we do accept tickets and make every effort to make it work. The same applies to User and Device mode. You can see this from the tickets we’ve fixed.
The end result of what James said is basically we can stop with the FUD. U&D mode isn’t going away. If there are issues then report them. The developers (me included) don’t like the mode much, but why does that matter? We also love pjsip and that seems to have caused the same type of controversy recently (RE: Why is FreePBX going to remove chansip). But the fact is we will probably never remove chansip, same as why we will never remove U&D mode
I’ve seen two Sangoma employees come in here and give you their views on U&D mode, both with completely different wording and it just didn’t seem to cut it for you, so please detail what you feel we aren’t listening to our users about.
Perhaps you just don’t like the fact that we say it’s “unsupported”??
This is extremely unlikely to happen. As @jfinstrom said, D&U is the fundamental method of how FreePBX works. The ‘unsupported’ means exactly that. It doesn’t mean ‘will go away’, or ‘we hate it’, it means that we do not feel comfortable supporting it as it exposes a lot of things that make it extremely easy to break your machine.
I think a lot of the confusion here comes from people mis-interpreting that one word. Unsupported only means ‘not supported’, as in ‘we can not support it’.
Would you like us to lie? We’re just telling you the truth. 8-(
D&U is not going away, it’s extremely unlikely to go away, and we’re fixing bugs that crop up. It’s just not supported.
Does that clarify things a bit better for you?
Edit:
Three now. And if you look at how FreePBX works, you’ll understand why it’s almost impossible to remove D&U, because that was the fundamental design of FreePBX. That’s the entire point of the AMPUSER tree in astdb, and all of the macro-dial stuff is there, explicitly to handle it.
You may be confused. FreePBX, specifically the part that would use "device and user mode "(Core) is free as in freedom and beer. Open source, all that jazz.
Any commercial module you pay for would not use “device and user mode” they hook into FreePBX where required to do so but none are marketed to D&U mode.
If you are using a commercial solution from a 3rd party that does require D&U mode perhaps your outrage should be focused on them.
I have been working with the Asterisk community and the FreePBX project for 7 years both as an open source contributor, a vendor and now an internal member. Device and User mode has its uses but it is by no means a normal use case. In millions of installs, D&U mode users probably does not touch 1%
Okay, okay! So it’s not “going away”. As a business user, that’s reassuring. Nevertheless, as a business user, the “unsupported” position will always be a concern. It’s not a comfortable position for a business to be in. Just sayin’.
All four of you have avoided mentioning the request to revisit the policy - can we move back to that part of the discussion?
That’s surprising - perhaps you’re speaking from a coding viewpoint? From where I’m standing it’s a brilliantly elegant design - devices, users and a simple, editable link between them. It’s exactly what I was looking for, exactly what my business needed. And that’s the crux of it - as I said before, it’s how a modern telephone system should operate.
If James is correct that the usage rate is only 1%, I’d propose that it’s the “unsupported” policy that has held it that low. Business users will (like me) be extremely cautious of implementing an “unsupported” product. The larger the company, the less likely that such a move would ever be approved. I’d say, too, that the 1% figure is misleading. A huge proportion of FreePBX’s installs will be home users, tinkerers and hobbyists - a user-base that will intrinsically have less need for D&U mode.
D&U Mode “breaks things” because the FreePBX core and the commercial modules are not, by your choice, written to support it. But that, surely, doesn’t mean that full support of the mode is technically impossible? I guess I’m trying to sell the idea that fully embracing D&U Mode would align FreePBX with the needs and expectations of today’s users, making FreePBX an even greater product than it already is. That 1% would rocket to (amongst business users) 30%, 60%… who knows? I believe such a product is what today’s market wants for the new, mobile-heavy landscape.
