OPEN LETTER or PLEA - Please give me some guidance about investing time in Smartthings


(Tony) #1

Firstly a little background.

I’m a technically literate individual with over 20 years experience of IT, I run a number of different machines using various OSs.

I have a limited amount of time on my hands but am interested in Home Automation, partly because of it’s intellectual interest and partly for the benefits it can bring.

I have a Smartthings Hub V2 and have ‘played’ with some functionality over the last few weeks to familiarise myself with it’s capabilities and limitations.

I have read a number of different threads and bits and pieces and am slowly starting to come to the conclusion that Smartthings is not yet ready for primetime.

Please understand that I’m not looking for responses that are based on individuals own prejudice, just fairly factual and honest responses based on peoples experience.

My questions in no particular order of importance are…

  1. Why did someone decide that naming the ‘new’ Smartthings app ‘Smartthings’ and the ‘old’ app ‘Smartthings Classic’ was a good idea? Anyone with any experience of technical nomenclature would understand that by recycling the existing name you introduce a huge amount of confusion when users, particularly new users to the platform, do searches for information as any results found cannot be trusted as to which version they relate to.

  2. Is the frequency of the servers being off line as bad as it appears to be? Has no-one looked at something as obvious as different site redundancy to ensure stability of the platform? Having worked in Investment Banking IT I can honestly state that people, including senior IT managers and directors, would be fired for not designing and maintaining systems that had better then 99.9% uptime.

  3. Have I understood correctly that the API for the ‘new’ app does not support tiles on the app the way the ‘old’ classic app does? As an example I have some Model 75 versions of the Aeotec energy monitoring Plug/Switch. The tiles in the ‘old’ app are aesthetically pleasing and quickly convey the information I want. However In the ‘new’ version of the app I have to use the generic Z-Wave energy monitoring switch DH, that prompted me to look at writing (nearer migrating) the older version of the DH to work in the new environment wit the aesthetics I want. I understand from what I’ve read however, that the ‘new’ app doesn’t support tiles the same way (properly?) thus the functionality is lost. Is this correct?

  4. Why are Samsung not addressing these problems at a fundamental level? They have what appears to be the basis of a good platform here, but unless senior managers get a grip on this and fairly quickly, as this market starts to exponentially grow they’ll lose their lead and all of the many existing developers time will have been wasted. This is not the phone market where people are platform tied in, transparency of what’s happening and a clear plan are important for people to commit to investing their time. Marketing by confusion will not work in this space.

  5. What is the timescale for full transition to the ‘new’ app and when will the full range of devices be supported? This non answer “we’re working on it” is frankly amateur hour and unbecoming of an organisation with Samsungs resources. In reality, until named senior managers are willing to put their credibility and reputation on the line and give dates that then get delivered against this will continue to ‘float down river’ with no serious commitment to set and meet transparent goals. If you don’t believe that this could disappear surprisingly quickly do a little research on Borland.

The users deserve better, the front line staff dealing with stuff deserve better, when are management going to stand up and be counted?

Just my 0.02c, feel free to rip me a new one.


(Brian Harding) #2

I am not going to rip you a new one nor am I going to defend Samsung/Smartthings, but you will probably not get an answer to your questions from the staffs of either company and many of us would like answers to these questions. That being said my understanding is that Samsung had left Smartthings to their own devices for several years, meanwhile they had their own apps, etc. for their equipment, especially their phones. At some point Samsung decided to get involved and wanted Smartthings to combine all of their apps into one app and Smartthings having limited resources found this to be a huge undertaking especially without disrupting the customer bases too much. What the final product will become, nobody but Samsung and Smartthings know for sure.
Now, from a reliability standpoint, I have only had several outages in the past 3 years, some from their servers failing and some from drunk drivers knocking down the telephone poles in front of my house thus killing my power and internet. A couple have been caused by severe storms between my home and their servers. Other than that, it has been quite reliable for me and allowed me to add video monitoring that my paid, professional alarm system which did not do such a good job of doing. In other words it complimented my professional alarm system in a big way.

Brian


#3

The reliability question is objectively answered in the documented bug reports.

There is an official status page, but a lot of problems never get posted there even though support will tell you it’s a known problem. It’s hard to figure out exactly what does make the official status page. (for example, the official integration with Samsung smart televisions was broken in the fall of 2017 and was not fixed for several months, yet the problem never made the official status page.)

https://status.smartthings.com

The community – created wiki has a bug report page which has a lot more stuff. It’s worth looking at if you want to see what kinds of things go wrong when. Note that there has been a major outage at least once a month for the past 20 months or so, sometimes more than once, except for, I think, January 2017.

http://thingsthataresmart.wiki/index.php?title=Bug:_First_Reports

In Addition, smartthings can and does push out firmware updates which will take your hub off-line for anywhere from a few minutes to a few hours. These can be neither delayed nor denied. Usually we get a few days notice of a planned update, but not always, and then sometimes there will be additional unannounced updates over the next two or three days as they try to fix problems from the first one. Since these outages are planned as part of maintenance, they are not posted to the status page or the bug reports page. I’d say there’s typically one a month, but I don’t have data on that one. You could check the announcements category, I guess.

https://community.smartthings.com/c/announcements

Whether or not that pattern is acceptable to you depends very much on your own priorities and preferences. Most of the outages are short and most of the glitches are pretty minor and may just require popping the batteries on some devices or opening a page in the app and then re-saving it. Some glitches, however, persist for months.

This can’t just be solved with more redundancy, because I would say that over 80% of the problems are caused by platform changes intentionally introduced by staff.

What would be required would be a change in corporate commitment to a MFOP ( maintenance free operating period) more similar to that seen in home appliances like dishwashers or washing machines. Which is to say typically one year.

However, smartthings has been heavily rewarded by the technical press, The investment press, and the marketplace for instead rapidly introducing new features and support for new devices. that means a lot of change, and in turn a lot of instability. But they became a “top of mind” leader in IOT (not top of unit sales, which is almost certainly the boring cable company offerings) because they continually offer new integrations. But that has resulted in an MFOP of under 15 days if you count the planned outages.

Many people don’t even notice the outages because they aren’t home when they occur, or because most do seem short and minor. Others who are home most of the time will notice their frequency, but won’t care that much because the way they have it set up it just doesn’t make much of an impact if the system is glitching. Still others who have set reliability as a higher personal priority will quietly move on to one of the other competitors Who can deliver a longer MFOP, even if it does mean giving up A lot of the power and versatility of a smartthings-based system.

So the short answer is that smartthings delivers the MFOP that it does because it has been rewarded for doing so. Not in unit sales, but in industry buzz.

If you want a system with a longer MFOP you either have to pay a lot more money or switch to a system with many fewer devices supported and a much simpler rules engine.

So it all comes down to what you yourself need from your home automation system. Every system has pros and cons, and different things work for different people.


(Bob Kerr) #4

I am not even sure that money solves the MFOP problem. Simplicity does. You can go with less complex systems that offer increased reliability on one aspect of your home automation, for a limited number of devices. But there isn’t a system out there that integrates as many devices as SmartThings does, with an equivalent feature set. Or is it?

When in doubt about SmartThings, just go here:


(Jimmy) #5
  1. SmartThings has not released documentation yet on how to create your own UI in the new app for hub connected devices. So your assessment is correct. Using a generic UI with your custom handler is currently the only way to get custom handlers to work at all, and even that is more reverse engineering then officially published documentation. Maybe they will be ready with documentation for SDC in November.

  2. Who knows. Even before Samsung, SmartThings marched to the beat of its own drum. And frankly, for reasons JD touched on, its worked for them.

  3. They haven’t, and probably never will, release a timeline. With that said, the new app has already closed the gap in features from the Classic app, and in some cases implemented improvements that have been asked for in the Classic app for years. Still a ways to go, but the feature parity gap is shrinking.


#6

A full throated control4 implementation supports many more devices (Smartthings says “over 200 hundred” and control4 says “nearly 11,000”) than smartthings does , and with excellent reliability, but typically costs 10% of the value of the home plus an annual maintenance fee, and an additional programming fee when you want anything changed.

https://www.control4.com/for/connects-with-control4


(Tony) #7

With regard to any MFOP this again is a clear management issue, anyone with a reasonable amount of experience would be using some kind of version control with ‘branching’, the 'branching could be defined by the user to nominate themselves as a number of different ‘levels’ that then become the determining factor on when updates are rolled out to them. For example let’s just say there are 5 user levels,

Production (live box managing critical systems - no updates without user confirmation)
Live (live box managing systems - most updates except critical ones require user confirmation)
Beta Tester (semi live box - user is using it to manage systems but is willing for tested updates to be implemented with just notification)
Alpha Tester (semi live or pure development box - user gets all updates automatically installed on a pre scheduled cycle every say three to seven days)
Development (development box that is constantly updated in real time)

On top of the settings the user can also specify a set of update rules, such as all updates can only happen when I’m on the premises and after 1am (further rulesets can be introduced that include things like the likely disruption time from an update).

None of this is rocket science and is SOP in any business where IT based systems are a critical requirement for the core business, Samsung has the resources to be able to implement this kind of structured system.


(Bob Kerr) #8

(Tony) #9

While a system that can control 11,000 devices could be nice it’s also unrealistic for most users.

If you make a count of all the light fittings and all the power points in your property multiply it by between two and three and use that as a base you should be able to work out your likely top end for the number of devices you’ll want to manage.

I have a fairly large five bedroom house and am only just nudging the 200 device limit for everything I’d want to manage. If push came to shove I’d ‘split’ the house in two and have two ST hubs.


(Alex) #10

My questions in no particular order of importance are…

  1. Why did someone decide that naming the ‘new’ Smartthings app ‘Smartthings’ and the ‘old’ app ‘Smartthings Classic’ was a good idea? Anyone with any experience of technical nomenclature would understand that by recycling the existing name you introduce a huge amount of confusion when users, particularly new users to the platform, do searches for information as any results found cannot be trusted as to which version they relate to.

I’d say this is irrelevant to helping you decide whether to adopt the ST platform. I agree it might be confusing, even though it was pretty clear to me… however that may be because I stayed on top of it reading updates and so on.

  1. Is the frequency of the servers being off line as bad as it appears to be? Has no-one looked at something as obvious as different site redundancy to ensure stability of the platform? Having worked in Investment Banking IT I can honestly state that people, including senior IT managers and directors, would be fired for not designing and maintaining systems that had better then 99.9% uptime.

ST sells a sub $100 hub and does not charge for the service. Just this aspect should set expectations or at least justify some tolerance to outages. I have seen a few but they were not very disruptive. I had to turn on a few lights “the old way by pressing a button” or some lights stayed on all day. Nothing critical. How big of an issue an outage will be will largely depend on what you automate. I am of the opinion that overly complicated automations are likely to just cause problems. Unless a sequence of actions always, 100% of the time, happens in a specific sequence, I do not automate it. Humans are not robots and I do not want to force my family to do things in a very specific way just so the automation will work as expected. I see many examples of community members that in my humble opinion go way overboard… so I am not surprised when I see them complain for things going nuts and as harsh as it may sound, I believe they are victims of their own doing. My automations are generally simple and reliable. They are meant to assist my family eliminating repetitive tasks or ensuring that certain things get done (devices turned off, doors locked, etc). To give a practical example… While my doors lock automatically, I still expect my family to lock the door behind them when leaving the house. The automation serves as an assurance that we do not forget the door open. If I were to rely on it to always lock it for me, one failure would be too many…

  1. Have I understood correctly that the API for the ‘new’ app does not support tiles on the app the way the ‘old’ classic app does? As an example I have some Model 75 versions of the Aeotec energy monitoring Plug/Switch. The tiles in the ‘old’ app are aesthetically pleasing and quickly convey the information I want. However In the ‘new’ version of the app I have to use the generic Z-Wave energy monitoring switch DH, that prompted me to look at writing (nearer migrating) the older version of the DH to work in the new environment wit the aesthetics I want. I understand from what I’ve read however, that the ‘new’ app doesn’t support tiles the same way (properly?) thus the functionality is lost. Is this correct?

The GUI of the ‘old’ app is far from being user friendly or even aesthetically pleasing. I have run into lots of shortcomings. ActionTiles (@tgauchat) addresses many of those shortcomings so you might want to look at that as a way to control your home. The new app is “not ready” for use and ST says so as well. No point in trashing it at this point. I am expecting ST to deliver on same or better GUI once they are done. If not, then we do have a problem Houston! My biggest complaint is that both apps seem to be geared for small numbers of devices and they get really painful to use when you have a few hundred devices. I am counting on the new app to fix that. Once large numbers of devices become more popular… they will have no option but to do so. As a side note this applies to Alexa 100% as well.

  1. Why are Samsung not addressing these problems at a fundamental level? They have what appears to be the basis of a good platform here, but unless senior managers get a grip on this and fairly quickly, as this market starts to exponentially grow they’ll lose their lead and all of the many existing developers time will have been wasted. This is not the phone market where people are platform tied in, transparency of what’s happening and a clear plan are important for people to commit to investing their time. Marketing by confusion will not work in this space.

I give them the benefit of the doubt. I am not sure ST if generating massive revenue so they may not be of massive importance to Samsung in the big scheme of things. The biggest value is likely indirect revenue they get… I want to change my appliances so that I can monitor/control them via ST for example. Samsung will get more money from me that way… certainly not by selling me a few sensors and hubs.

  1. What is the timescale for full transition to the ‘new’ app and when will the full range of devices be supported? This non answer “we’re working on it” is frankly amateur hour and unbecoming of an organisation with Samsungs resources. In reality, until named senior managers are willing to put their credibility and reputation on the line and give dates that then get delivered against this will continue to ‘float down river’ with no serious commitment to set and meet transparent goals. If you don’t believe that this could disappear surprisingly quickly do a little research on Borland.

I can’t blame for saying zip on the timeline. They kept doing so in the past and got trashed each time they missed the timeline. It is never soon enough as I am hungry for new features/capabilities but I try to relax and enjoy what trickles down to us.

The users deserve better, the front line staff dealing with stuff deserve better, when are management going to stand up and be counted?

Just my 0.02c, feel free to rip me a new one.

You sound frustrated by ST before even starting so you might not want to adopt this platform (that I enjoy tremendously) for the sake of your stress levels. I am suggesting it sincerely and with no intent to rile you up.


(Bob Kerr) #11

Harmony Elite works with over 270,000 that doesn’t really mean anything. I agree that Control4 is a polished, more reliable system, but when you mix cloud and local processing the MFOP is just not there. The difference is that if you have Zigbee interference in your home, you don’t hassle yourself with it, you call a dealer to do it for you. That’s what the extra dollars get you, when you buy into Control4 universe.


(Tony) #12

Alex

Your post didn’t rile me up in the slightest, because it was thought out and articulate.

I would take issue with a couple of your comments/observations though.

While Samsung do not ‘charge’ for the service, they are doing this as they clearly see that if they can become the defacto standard for ‘simple’ home automation that gives them enormous leverage regarding the implementation of APIs going forward. There will be a huge revenues in charging Sony, LG, et al for a licence to use Samsung/Smartthings implementation code rather then having to have a their own department write and then taking it through acceptance testing.

I have no problem giving them the benefit of the doubt to a point, but the reality is that Alexa, Siri, GH and others are going to drive this market forward in leaps and bounds over the next five years. If they don’t have the correct management already in place to address the issues I’ve raised pretty much immediately there is a serious chance the opportunity will pass them by.

This is not small number were talking about, if you imagine that say ten years from now that every electrical item in your home has a level of automation control and integration think of what that market is worth globally, it’s in the hundreds of billions of dollars.

To be still making after this long, what I honestly have to describe as schoolboy errors, is a damning indictment in my opinion and needs to be addressed quickly.


(Tony) #13

Bob

Having designed and managed systems for Investment banks I’d be very surprised if their compensation package would meet my expectations.


(Alex) #14

@stilllearning - I agree with your latest post but adopting ST for the time being is not a lifelong decision. I am more than glad to jump boat if ST can’t keep up and something better comes along. I love technology/innovation and will not stick with a boat anchor, if ST were to turn into that. For now, I see great value in ST and I enjoy the openness and low cost. If I were looking for absolute reliability I would look into other options such as Control4 but I don’t want to pay 10k and above and have to rely on a 3rd party to automate my house for me. The fun and satisfaction I get when my automations work as desired makes it worth the occasional frustration when the cloud goes haywire. Don’t get me wrong, I am not fully excusing all the troubles but I am willing to consider them as part of the price of this low cost solution (in hope for things to get better over time).


(Tony) #15

I can’t blame for saying zip on the timeline. They kept doing so in the past and got trashed each time they missed the timeline. It is never soon enough as I am hungry for new features/capabilities but I try to relax and enjoy what trickles down to us.

Sorry I meant to comment on this point.

This is still a management issue.

People that can’t forecast and then implement to a timetable are frankly just not up to the ‘management’ part of the job, they may well be technically brilliant but not everyone has both skillsets.

Back in the mid nineties I was involved in building and implementing a migration of 2000 traders to a new custom built trading floor in a new building in Canary Wharf in London.

We had six months to plan, at the end of which we had a Gantt chart with slightly over 50,000 individual line items, the line item including responsibilities, dependencies, costings etc…it was pretty comprehensive and most important included who was the responsible owner for that line item.

As the plan began to roll out to a ninety day timetable there were daily reviews of the current state of the plan and weekly audits to ensure what was being reported was what was being met. If any line item started to slip for any reason a management decision was made to either commit more resources to bring it back on schedule (and the new resources would be fed back into the updated plan) or, if it was felt it was an inability of the line item owner to deliver they were quickly replaced. The project was delivered three days ahead of schedule and almost exactly on budget forecast.

Good management has vision AND makes things work, after all vision without execution are just stories.


#16

Sorry if I wasn’t clear. Those numbers aren’t the maximum number of devices in any one installation. They are the numbers of different third-party models that could be selected for someone’s account. The number of possible integration partners. That was the question that had been asked.


(ActionTiles.com co-founder Terry @ActionTiles; GitHub: @cosmicpuppy) #17

There are many businesses, projects, and “miracles of engineering” that absolutely would not exist without incredible project management and coordination. These range from relatively small, but critical (the management of a patient from entry to a hospital all the way through surgery and discharge), to the incomprehensible 400,000 parts that have to be coordinated from hundreds of vendors to build an passenger airplane, to the code that launches a satellite into orbit, or the timing and measurement tolerances required to erect a bridge or skyscraper…

In general, software tends to be an area with a lot more delays and mistakes. Some manufacturers do better than others. The software & systems at the bank I worked at were almost laughable in quality, but they managed sufficient reliability to operate and increase their bottom line by 10% annually,


But that’s irrelevant here. The reason Samsung doesn’t give out a timeline for new features, is that they have decided there is no net benefit to do so. And they aren’t alone in the tech industry: Apple doesn’t announce features of the next iPhone in advance of one official press release or reveal event (when the device is already complete and headed to mass production). Amazon doesn’t publish a timeline of the features they plan for the next update(s) to Alexa’s software. Most consumer product companies do not layout the future publicly - but some choose to.

If you want to know the roadmap or timeline of the product you are buying, just don’t buy a product which doesn’t offer you that - and best of luck that whatever product you find actually delivers on time and according to specifications.

It’s a free market and you’re a consumer. You can “plea” all you want - but we have observed how Samsung operates SmartThings. If that’s unsatisfactory to you, then you can choose to not be one of the millions of customers of SmartThings. They won’t miss you. Harsh? Nope, just reality - businesses exist to execute whatever strategy they want and hopefully maximize profit of their shareholders. If that leaves opportunities open for other companies, then other companies enter the free market and operate however they choose as well.

I’m just glad that SmartThings doesn’t make airplanes. That’s my only plea :relieved:.

In the meantime, Alex made a recommendation which I agree with completely:


(Bob Kerr) #18

Samsung can make phones that explode, if you cannot stomach that, one should not try SmartThings at home. :smile:


#19

It was the washing machines that exploded. The phones just caught fire. :stuck_out_tongue_winking_eye:


#20

I think the main point that counts here is that a number of community members have similar backgrounds to you, quite a few of us have worked for global 2000 companies (I myself worked at IBM), and many of us had very similar reactions to the one you are having now. You will find discussions of these, many making exactly the same points You’re making about software development, going back to 2015.

Company representatives including the SmartThings CEO, have come into the forums from time to time, always promising that they were going to make reliability their top priority. :innocent: And this has been going on for years.

Many of those community members have moved on to other platforms, but smartthings is still here, And still seems to have the same “run fast and break things” development philosophy.

I do think it’s important to note that smartthings has never had patents on anything. They’re not doing anything that anybody else couldn’t do. Samsung on the other hand has thousands of patents, and if there’s anything that they intend to be used by other companies, it’s what’s coming out of Artik, not out of the current hub-based SmartThings system.

After four years, my best guess, which I have mentioned before, is that Samsung bought SmartThings in 2015 for the name and the buzz. But they have their own long term IOT plans, and they aren’t based on this platform.

So I will say what I always say when this topic comes up, which it does about every six weeks: if you are a tinkerer and can tolerate the level of instability, SmartThings offers a tremendous amount of flexibility and power at a very low cost.

If you need a longer MFOP For whatever reason, there are a number of competing systems out there which may be a better match for you.

But there just isn’t much point in raising your stress level in trying to argue That SmartThings should have a different development philosophy. The one they have is working for them or they would’ve changed it two or three years ago.

If you enjoy the argument (some people do), feel free to carry-on. Or if you just want to vent. :rage: But realistically, SmartThings is unlikely to change because of anything said in this forum.

We hear you, we sympathize, many of us agree with what you’re saying, it just isn’t going to change anything.

FWIW