And Vlad presenting on rules engine
I’d like to think i’m a fairly smart guy, I’ve read everything in that post and watched everything in these videos BUT I’m still al confused where Webcore is going and what’s coming next???
is it just me?
Let me try and explain what I understood, Ady has developed a new webCoRE using the new Rules API, called webCoRE 2.0, that will be the new official rules engine like Smart Lighting, but first the hub needs new firmware which will come around February, and there’s no migration tool, but it will run local. You can’t run both webCoRE’s at the same time, but there will be a 3 month grace period until webCoRE 1.0 and groovy will be turned off. but at the moment it only does 60% of what webCoRE 1.0 does and it’s up to people like @WCmore, @RobinWinbourne and @ipaterson to bring it up to speed with variables and expressions.
Does that mean we will be re-writing our pistons with the new Rules API?
That’s what I understood yes, re-writing our pistons from scratch, and re-adding our devices to the new webCoRE, but I’m sure @ady624 will be along soon with some clarification.
From the videos, it looks like we can run webCoRE 1.0 and 2.0 simultaneously… but (of course) if the logic in one conflicts with the logic in the other, there may be funky results.
I view ST wants to update the rules engine to support variables, etc. I expect there won’t be a full functioning webcore on the rules api until that happens.
At some point ST wants to disable running groovy code in their cloud - likely it is just too expensive. That said rules API is not up to the task yet with lack of variables, etc.
Some of the issues ST is grappling with is that they want everything distributed, or local, so in the end it has to be architected for cloud distributed. This is forcing the constant re-instanciation of groovy apps and the need for micro services.
A pure local run, does not require constant re-instanciation that Ady mentions. This is one of the many optimizations done for HE.
The current rules engine is much less functional than webcore requires, and it is likely less functional than rules machine on HE (which is less functional than webcore).
I did not hear them state dates (I may have missed it). If they really plan to shutdown in 6 months on ST, this will be a train wreck for ST users imho. Forcing everyone to convert to something that does not exist sounds a lot like Google…
I agree with @E_Sch…
I really hope the overlapping window is much longer than it was when they switched us over from WUnderground to TWCWeather…
If they follow that same pattern, our houses will be “dumb” for months while converting old pistons to the new API…
Perhaps I should apply for the job… God knows we need good documentation!
It’s because of you that we have what we have now. Which is awesome BTW.
I’ve got no problem with re-writing pistons if the payoff includes local execution.
I totally agree, @GRClark…
I just know that in my house,
variables are what holds most of my logic together. Without
variable capabilities in the API, my SmartHome will only do a small percentage of what it currently does now.
variables, I have no interest in the API. I may just “retire” for 6 months until they are added to the new code. (since I don’t want to invest time “dumbing it down”, just so I can “beef it up” a few months later)
I guess, for clarification, I am glad they are redesigning the architecture… I just don’t want to shift over until the core components are in place. (is a 12 month overlap asking to much?!?)
Can I throw a bone out?
Is there any plan for Alexa integration other than virtual switches?
Vlad seemed open and interested in Alexa and IFTTT integration, but it sounds like nothing is “officially” planed yet on that front.
I agree that more local execution would be nice. However…
THIS. More important than local execution, variables are a very large and important part of webCoRE’s power and flexibility. Every single one of my pistons use variables. A lot of them use many variables. Hopefully, the gap between the current version and version 2.0 isn’t very long (none would be best).
same here… (as you know Wcmore) 80% of my pistons use variables and expressions… Without those webcore is just a new app for my grandpa.
How about coding (writing) part…
I have a good grip on webcore but I’m not a coder…
Are we going to write everything with all comas, slashes etc by typing???