Hey, thanks for responding. My responses are inlined.
On Thu, May 1, 2014 at 1:34 AM, Atamert Ölçgen <mu...@muhuk.com> wrote: > I am not an expert on Component. But AFAIK it is not for managing mutable > state but for assembling and configuring components, that might or might > not be mutable themselves, in an immutable fashion. > Agreed, this is what I've found. Component isn't for the use case of state changes between components. I'm actually looking at Datascript <https://github.com/tonsky/datascript>as a way of managing internal, runtime, changing RIA state. > However from what I can understand, your component-a has an atom, like: > > (->component-a (atom something)) > > Which should be OK. I mean it shouldn't matter from which path you are > accessing this atom. > > Can you share the constructor/definition of the components? > I can't really share actual client code. But the system definition in fig.1. pretty much describes how the real components are stitched together. Also, have you tried confirming that only one :a is instantiated? > That one *:a* is not the same instance throughout all the dependant components. Seems that it's the [*:core :a*] bit that's passed to all the components in the *'start* *'stop* functions. I say this, because when one of my dependant components updated an atom in it's own [*:local :a*], it was [*:core :a*] that reflected that change. On Wed, Apr 30, 2014 at 9:13 PM, Timothy Washington <twash...@gmail.com>wrote: > >> Hi all, >> >> I'm having a weird state problem with >> Component<https://github.com/stuartsierra/component>. >> Let's say I have a system component, like in fig.1. Starting / stopping and >> loading state is fine. >> However, let's say I have 2 other components (:updater , :reader) that >> use component :a. This is the problem that occurs. >> >> 1. When *:updater*, modifies an atom in *:a*, that change appears in >> path [*:core :a*], not path [*:updater :a*] or [*:a*]. >> 2. Because of the abouve, when *:reader* goes to it's local path [ >> *:reader :a*] to read that change, it doesn't see those >> modifications. >> 3. Using this scheme, *:a* is duplicated 4 times, in the system map. >> However, the modifications only appear in path [*:core :a*]. Thus >> :reader is unable to access it (it's local [*:a*] is unchanged). >> >> >> (def system-components [:a :updater :reader]) >> >> (defrecord Core [env] component/Lifecycle >> (start [this] ...) >> (stop [this] ...)) >> >> (defn component-core [env] >> >> (component/system-map >> :a (component/using >> (ca/component-a env) >> {}) >> :updater (component/using >> (cs/component-updater env) >> {:a :a}) >> :reader(component/using >> (cs/component-reader env) >> {:a :a}) >> :core (component/using >> (map->Foobar {:env env}) >> {:a :a >> >> :updater :updater >> >> :reader :reader }))) >> >> *fig.1 * >> >> >> I was hoping to use Component to manage all internal application state. >> But maybe it's not designed for this use case (state changes between >> components). I imagine that immutability is preventing all those duplicates >> from seeing the modifications. However, in this case I do need an update to >> :a in one component, to be accessed by another component. >> >> Any suggestions on patterns here? I'm also looking at >> component/update-system<https://github.com/stuartsierra/component/blob/master/src/com/stuartsierra/component.clj#L116>. >> But again, I don't have access to the core *system* var, from within my >> components. >> >> >> Any insights appreciated >> >> Tim Washington >> Interruptsoftware.com <http://interruptsoftware.com> >> >> > -- > Kind Regards, > Atamert Ölçgen > > > www.muhuk.com > > -- -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
