Announcement

Collapse
No announcement yet.

HS4 vs Hubitat

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #16
    There's also a big gulf between user expectations and actual device capabilities. That's bad enough, when people complain about what it "can't do" when that's not even a reasonable expectation. Pair that with the sort of failures you rightly point out and you've got a vicious circle of negative pressure on the market.

    Me, I'm grateful Homeseer has continued to plod forward through all this.

    Comment


      #17
      Originally posted by upstatemike View Post

      It occurs to me that the problem is not so much with Homeseer as it is with the Home Automation industry in general. If things worked the way they are supposed to then Homeseer could write a driver for a "Standardized" Z-Wave Switch or Zigbee Bulb and they would work with any Z-Wave switch or Zigbee bulb from any manufacturer. That is the world we were promised but it is not the reality we see. Every single device from every manufacturer seems to need an automation hub or software maker to write smething special to support their unique product. As the universe of automation products continues to expand I don't see that as a sustainable model for any automation controller company to follow. The focus of any company needs to be improving their core product and reacting to the latest revision of each product from every fly-by-night company who decides to dip their toe in the Home Automation craze. Hubitat supports new product drivers more quickly than Homeseer at the cost of less reliability and features and they will eventually hit a resource wall trying to keep up with new devices and new versions of existing devices.

      I think at some point there has to be a fundemental change in how standards for HA devices are defined and enforced because the model of writing a unique driver for every new device that comes along is simply not realistic.
      Zigbee seems to be better for basic functionality that zwave, though HS' Zwave implementation is about as good as anyone's. The new Zigbee spec driven by Google, Amazon and Apple should address a lot of this: https://www.apple.com/newsroom/2019/...vity-standard/ as these companies just want all the devices commoditized, and can drive things much more aggressively than HS, Hubitat etc... can.

      That said, the role of open source and user contributions here will always be important as innovative new devices that cant easily be supported in the normal frameworks will always be coming out, and AV integration is increasingly important, and we have very few good standards in that area. So I think things will get better, but integrators will always need to build frameworks that can allow 3rd parties to be easily integrated into their systems.

      Amazon, Google and Apple are growing capabilities significantly in this space, but they all seem to be focused on control and not real automation at this point. With more ubiquitous device support with the new standard, maybe that will change. But i really prefer more local control, and that will need HS and others to step up and take more initiative in working with these software intensive platforms to really deliver on that vision.

      Comment


        #18
        Originally posted by upstatemike View Post

        It occurs to me that the problem is not so much with Homeseer as it is with the Home Automation industry in general. If things worked the way they are supposed to then Homeseer could write a driver for a "Standardized" Z-Wave Switch or Zigbee Bulb and they would work with any Z-Wave switch or Zigbee bulb from any manufacturer. That is the world we were promised but it is not the reality we see. Every single device from every manufacturer seems to need an automation hub or software maker to write smething special to support their unique product. As the universe of automation products continues to expand I don't see that as a sustainable model for any automation controller company to follow. The focus of any company needs to be improving their core product and reacting to the latest revision of each product from every fly-by-night company who decides to dip their toe in the Home Automation craze. Hubitat supports new product drivers more quickly than Homeseer at the cost of less reliability and features and they will eventually hit a resource wall trying to keep up with new devices and new versions of existing devices.

        I think at some point there has to be a fundemental change in how standards for HA devices are defined and enforced because the model of writing a unique driver for every new device that comes along is simply not realistic.
        Personally I don't think HST is writing "Device Drivers" for every single device they support. I believe the standardized Command Classes are supported, they only thing you get personalized is parameter value descriptions for specific devices. There's a plugin that provides those in an xml file the way that Home Assistant does. From what I've seen recently, any changes to the Z-wave plugin are either supporting new Command Classes or fixing bugs/logic in old implementation (like the recent change where Rich admitted that he hard-coded for no more than 6 child devices). As new devices with more complex implementations are released those bugs are surfaced.

        There are probably some workarounds for devices that mix/match command classes in an unusual way, but those are fairly rare.

        Comment


          #19
          Originally posted by mterry63 View Post
          .. no more than 6 child devices). As new devices with more complex implementations are released those bugs are surfaced.
          Out of curiosity, what ends up with more than 6 child devices in a unit?

          Comment


            #20
            Originally posted by fresnoboy View Post
            Amazon, Google and Apple are growing capabilities significantly in this space, but they all seem to be focused on control and not real automation at this point.
            Given Google's demonstrated behavior of abandoning things I wouldn't hold much hope. Likewise Apple's "not invented here" syndrome is equally troubling, their support for 3rd parties is temporary, at best, and often ends up arrogantly crippling key 3rd party functionality long enough for the 3rd party solutions to falter... and then intro an apple-branded solution shortly after. Amazon is certainly no saint, but they don't screw around when it comes to keeping cloud services running.

            That said I too greatly prefer as much on-site integration as possible.

            Comment


              #21
              Originally posted by wkearney99 View Post

              Out of curiosity, what ends up with more than 6 child devices in a unit?
              I believe it was some Aeotec Smart Implant device that supports 7 temperature sensors along with 2 analog inputs and 2 switched outputs. Rich fixed this is Z-wave beta 3.0.4.0 (actually earlier, but 3.0.4.0 is the current beta).

              Comment


                #22
                To be honest, Hubitat is a hot mess of alpha junk. Nothing but hardware issues and bugs.

                I hate their system and the year and a half it robbed me of my free time. Every once and a while I will troll their forums and see the same old hub issues and rules running by themselves and their shitty *** rule machine app.

                The one thing I will say, they are very involved with the community... but what I would like to do is do a write up of their system to save people the hassle of using their system.

                Lastly, since I’m complaining. I find it humorous that people argue how expensive homeseer is but find in turn they buy multiple hubs so they can support their device and app load, why because their crap hub has constant issues!

                okay I’m done for now...

                Comment


                  #23
                  Originally posted by mik3 View Post
                  Lastly, since I’m complaining. I find it humorous that people argue how expensive homeseer is but find in turn they buy multiple hubs so they can support their device and app load, why because their crap hub has constant issues.
                  Their use of a low-end IOT-type gizmo concerned me from the outset. With Homeseer and other systems there's at least the option to spend as much money as I'd like to get performance handled. Yes, with well running software a LOT can be done with limited resources.

                  But if there are any software issues having more CPU and RAM often lets you hide from it. Memory leaks when you've got 8 times as much RAM as "needed" sometimes gives you better headroom to find them before they become performance issues. Likewise CPU, something that runs quickly enough doesn't end up running afoul of conflicting locks and such.

                  Granted, better software helps, that's obvious. It's better to find long-standing bugs that hide behind excess resource consumption because they're often the same kind that have security risks. And having to support a wider range of underlying hardware does complicate the development process. Support resources aren't unlimited, so a 'known' platform is one way to keep costs down.

                  But they are at least making a lot of interop possible. Yes, partially because limits in their system pretty much 'require it' but better that than crippled interop (or none).

                  Comment


                    #24
                    Originally posted by mik3 View Post

                    I hate their system and the year and a half it robbed me of my free time. Every once and a while I will troll their forums and see the same old hub issues and rules running by themselves and their shitty *** rule machine app.
                    I never expected HT to do the work of events (rule machine) and I only use mine for getting data from devices, like Zigbee. From there I pull those into HS with the mcsHubitat plugin. I haven't had any issues doing that. Quite awhile ago my HT hubs would stall once and awhile but I just created an event in HS that reboots the hub (s) twice a week in the middle of the night and no more issues. If I was running HS on a pi (which I don't) with the five plugin limit this is the way I'd go for sure. I uninstalled a lot of HS plugins which I own and use HT for that data acquisition with the mcsHubitat plugin, no worries. If you don't expect HT to do the heavy lifting of events you will be fine.

                    Comment


                      #25
                      Originally posted by Bigstevep View Post

                      I never expected HT to do the work of events (rule machine) and I only use mine for getting data from devices, like Zigbee. From there I pull those into HS with the mcsHubitat plugin. I haven't had any issues doing that. Quite awhile ago my HT hubs would stall once and awhile but I just created an event in HS that reboots the hub (s) twice a week in the middle of the night and no more issues. If I was running HS on a pi (which I don't) with the five plugin limit this is the way I'd go for sure. I uninstalled a lot of HS plugins which I own and use HT for that data acquisition with the mcsHubitat plugin, no worries. If you don't expect HT to do the heavy lifting of events you will be fine.
                      That’s a whole other question. It’s not clear where the line in the sand is when referring to heavy lifting and not heavy lifting. Nobody can clearly point that out.

                      When I was working with their system all I was trying to do was get lighting control to work. What a mess...

                      At this point in the game, you couldn’t even pay me to use their junk.

                      Comment


                        #26
                        Originally posted by mik3 View Post

                        It’s not clear where the line in the sand is when referring to heavy lifting and not heavy lifting. Nobody can clearly point that out.
                        And nobody ever will as it varies depending on the individual's skill / knowledge.

                        Comment


                          #27
                          Originally posted by mik3 View Post
                          At this point in the game, you couldn’t even pay me to use their junk.
                          Tell us how you REALLY feel Mike. heh.

                          Comment


                            #28
                            Originally posted by wkearney99 View Post
                            Tell us how you REALLY feel Mike. heh.
                            lol I’m a very happy homeseer user.

                            Comment


                              #29
                              Originally posted by fresnoboy View Post

                              . The new Zigbee spec driven by Google, Amazon and Apple should address a lot of this: https://www.apple.com/newsroom/2019/...vity-standard/ as these companies just want all the devices commoditized, and can drive things much more aggressively than HS, Hubitat etc... can..
                              Chip is its own spec. It is manage by Zigbee Alliance with the core classes and commands of zigbee (aka dotdot) . It is zigbee's corporate half-sibling.

                              IMHO, CHIP is just a way for all the wifi manufacturer's engineers skills to be portable and to reduce the cost of firmware.


                              There are so may intentionally vague parts of the spec, even this late in the process, that nothing to make me think a given SKU will actually be interoperable across google/amazon/apple. I mean, it could. But they won't.

                              The inclusion of Wyze and Tuya had me expecting cloud controls but the block chain system confirmed it. The block chain firmware tracker alone ensures the device is expected, if not required, to report back to a CHIP cloud and be uniquely identifiable. (Great idea for industrial applications, gives me the heebie jeebies as a consumer). With that mechanism, Apple can exclude any device whose serial number wasn't sold with homekit firmware based on the ledger history. Same goes for Amazon and google.

                              Having said that, I expect some companies, like Shelly, might release a CHIP device that can be run by any controller that wants to bother. It has an advantage over mqtt of having a complete dictionary of features and commands which, again, reduces the cost of manufacturing and software support. That's the goal.

                              Consumer benefit is likely only in reduced device cost (where it doesn't get turned into increased margins)

                              Comment

                              Working...
                              X