Maybe the solution is one appliance (like the existing system) and one virtual appliance offering that you could run on your own platform . . . with appropriate caveats for both scenarios.
1 Like
Yes, this would also (I think) satisfy people who want to tightly integrate it with lots of other things that they might host on the same platform . . .
I think sales would increase.
The problem would be on the end user support. I would imagine they would require a bigger support team to filter out end user errors vs software bugs.
But going this route, they could also dive into sensors, plugs, etc. Thus could potentially increase revenue...
1 Like
I'm actually surprised no one has written an open-source SmartThings-esque Groovy runtime environment for apps (and I guess LAN devices, which you'd need to actually make this usable)--basically what Hubitat did for both apps and devices/drivers to create their platform. This presumably took a year or two of effort, but I imagine a lot of that might have been Z-Wave and Zigbee development that could be avoided by focusing as much as possible on just apps. That would let people run what they think might be resource-intensive apps off in whatever environment they choose and make some people who are asking for this happy. 
You can pretty much already do this with MQTT and Home Assistant or NodeRed and other options people are using, except you have a different "app" paradigm.
Hubitat has stated in the slightly distant past--but nothing has changed recently as far as they've said--that they are not considering a VM-type install (or anything on other hardware) at this time.
1 Like
Personally, after doing some game development (which is highly resource constrained as well) I can say that I would prefer the priorities be better spent developing the underlying platform further to harden what is already there.
- Database: There appears to be weakness in the db tech, as corruption seems to occur rather easily, so if this could be improved (different tech?) then that would certainly help.
- Mesh Reporting: The zigbee routes page is useful, but should be integrated into a better monitoring/reporting interface (and something for z-wave?) so that users can identify troublesome devices more easily. I know the team don't want to fiddle with the mesh like other hubs do, but mesh weirdness seems to be common issue that impacts the feeling many have about the platform.
- Best Practices: Obviously developer documentation is still a work in progress, but without knowing what is ideal when it comes to development on the platform it's sort of a shot in the dark for many of the community devs. @bertabcd1234 Goes out of his way to code to what he sees as 'the hubitat way', which I respect. I just wish the there was more guidance. Mike and Bruce do provide it throughout the forum, but it would be nice to have it more codified.
- Dashboard: Not sure what further development is occuring in this space, but it goes without saying that there needs to be more here. Looking at the Home Assistant new interface skin just shows that without effort, Hubitat will fall behind here easily.
- Mobile App: I appreciate the app for what it is, but would love to see it do more. What about having the means for providing an answer back to a notification to trigger something else? a notification history?
I appreciate that the 'throw more hardware' at the problem solution seems appealing but that doesn't help you get an efficient reliable platform necessarily, it just kicks the issues down the road a bit. Hubitat architecture is still in its infancy, and it is amazing what has already been achieved but I'd like to see that improve over just additions of hardware.
Cheers,
Glenn.
30 Likes
Aside from rule execution performance and z-wave/zigbee constraints, I'm wondering about the significance of the Ethernet connection on local dashboards. On my system, local and remote dashboards are extremely slow.
Will Hubitat ever release a "performance" edition?
It seems to me that what you guys are missing is the size of the market for those who want a "performance" edition. It's tiny, compared to the market that Hubitat serves. There is not a positive business case for this concept. The suggestion that we offer our software to run on other hardware is a non-starter -- it also suffers from an untenable business case. So these two concepts that a small number of people clamor for aren't likely to happen, although we expect the clamoring to continue (it always has despite us saying these things numerous times before).
The suggestions made by @gbrown above, on the other hand, are ones that make sense, and which would benefit not only those who want a "performance" edition, but many other customers as well, and hence make sense for us to pursue.
No, it has a quad-core CPU, 1G memory, 8G flash.
19 Likes
7 Likes
Or:
2 Likes
Please note I am joking .. If it weren’t obvious
2 Likes
I actually rather like the concept. How much would you be willing to pay??
3 Likes
Well a decent supermicro based server.. I’d say about $3k
I have a server room with 3 racks of these
I'd pay that 
But since it looks like we're not getting any faster hardware, could we at least get a built-in unit-test toolkit with something like timeit (Python) and other helper methods to make it easier to do unit testing? The Metrics library is probably a bit too heavy and overkill, but something we can "easily" see when code-changes makes a difference in performance (or break things completely) would be nice. I realize this CAN be done with a "bit" of custom code, but it would be nice to have a "standardized" way to measure code performance on the platform while developing. These are just my 2 cents...
Doesn't matter. You already said it was a niche market and not worth it. 
I am new to HE (from ST), and i am disapointed with my HE experance to date. Ive keept a look out at HE wating for it to become more mainstream, but then at christmas ST cloud issues and the retiring of the old app pushed me to take the jump.
Yes the user interface is a bit clunky, but i can live with it. The only issue i have is z-wave locking,the fact that (as its local?) i cant monitor my own hub preformance to identfy issues/trouble shoot is realy frustrating.
Im looking for something rock solid that can run my 30 or so devices (on one hub) reliably.
And... how long has chromecaast been in beta???
what other hub fits that build/requirement ?
Indeed. If HE were to address the hub slow downs/DB corruptions it would make the Hubs run as fast as new out of the box - with ample performance to spare, which would make the case for a "pro" version somewhat moot . . .
4 Likes
But we are pushing half a year with these problems.
