previous example with the peek operation: 1- The producer puts a value to a unbuffered (chan) by doing (>! c v) 2- The go-loop unparks from (peek<! c) without consuming the value, the producer keeps parked 3- The go-loop contacts the external-service 4-A If the external-service answer is ok, the go-loop consume (and discard) the value by doing a normal (<! c), and the producer unparks 4-B If the external-service answers it cannot process the value, the go-loop waits until a timeout to retry step 3
The producer only unparks when the value is effectively consumed by the external service. That's my objective. I think your pub proposal replaces the take-if proposal given before, but I think take-if (and pub) doesn't work for this scenario. Saludos, Nahuel Greco. On Sun, Oct 5, 2014 at 1:20 PM, <adrian.med...@mail.yu.edu> wrote: > Then how would peeking at the value help? > > On Sunday, October 5, 2014 12:14:32 PM UTC-4, Nahuel Greco wrote: >> >> Adrian: I don't see how a pub can help here, in the previous example to >> consume or not the value was decided not on some property intrinsic to the >> value (one you can create a topic from), but on the result of sending it to >> an external service. >> >> >> Saludos, >> Nahuel Greco. >> >> On Sun, Oct 5, 2014 at 12:59 PM, <adrian...@mail.yu.edu> wrote: >> >>> I think you can achieve an effect similar to what you want by using a >>> pub with an appropriate topic function that classifies the input in some >>> way, and then subscribing to the topic whose value you want to see. This >>> also has the benefit of automatically 'mult'ing the channel input, so you >>> can have multiple consumers looking for the same value. >>> >>> On Sunday, October 5, 2014 11:33:16 AM UTC-4, Nahuel Greco wrote: >>>> >>>> Picture the following: >>>> >>>> producer ---> go-loop ---> external service >>>> >>>> 1- The producer puts a value to a unbuffered (chan) by doing (>! c v) >>>> 2- The go-loop consumes the value with a take operation, **unblocking** >>>> the producer >>>> 3- The go-loop contacts the external-service but the external service >>>> answers it can't process the value yet >>>> 4- The go-loop waits some timeout to retry the request to the external >>>> service >>>> >>>> After step 2 the producer continues to compute (suppose an expensive >>>> computing) a new value but the previous one wasn't effectively consumed by >>>> the external service. >>>> I don't want that, I want to enforce an end-to-end flow-control setup >>>> where the producer blocks on (>! c v) (the step 1) until the value is >>>> consumed by all parties, >>>> >>>> Sure, this flow control can be solved adding an ack channel and sending >>>> an ack from the go-loop to the producer when the external service >>>> effectively consumes the value, previously blocking the producer after step >>>> 1 waiting that ack. >>>> But I think a peek operation in step 2 will be more elegant. Also, I >>>> was curious if the implementation of core.async channels limits in some way >>>> adding a peek operation. >>>> >>>> A take-if with a pure predicate can't solve this, because you need to >>>> contact the external service to decide to consume the value or not. >>>> >>>> >>>> Saludos, >>>> Nahuel Greco. >>>> >>>> On Sun, Oct 5, 2014 at 9:52 AM, Fluid Dynamics <a209...@trbvm.com> >>>> wrote: >>>> >>>>> On Sunday, October 5, 2014 12:51:04 AM UTC-4, Nahuel Greco wrote: >>>>>> >>>>>> I was thinking in a single-consumer scenario with a buffered chan, in >>>>>> which you want to check if you can consume the value before effectively >>>>>> consuming it. As you said, a peek operation has no sense if the channel >>>>>> has >>>>>> multiple consumers. >>>>>> >>>>> >>>>> And if you can't consume the value, then what? Nothing ever does, and >>>>> that channel becomes useless? >>>>> >>>>> Actually the only "peek" operation that to me makes much sense would >>>>> be a (take-if pred chan) or something similar, which atomically tests the >>>>> next value with pred and consumes it or not, so, it can't be consumed >>>>> elsewhere between the pred test and optional consumption here. And if not >>>>> consumed, two behaviors both occur to me as possible -- return nil or some >>>>> other sentinel value for "do not want" or block until the unwanted object >>>>> is consumed by someone else and then test the next item, etc.; at which >>>>> point you've got four versions of take-if you'd want, the inside-go and >>>>> outside-go versions cross product with the two when-not-wanted behaviors. >>>>> >>>>> At that point, you'd probably be better off just writing a consumer >>>>> that splits off the pred-matching items into one out channel and feeds >>>>> everything else into a second channel, with your original consumer taking >>>>> from the first of these and the others taking from the second. That gets >>>>> you the block until version of the behavior. The other version can be had >>>>> by making the pred-using consumer the sole consumer of the in channel, >>>>> which takes a value, applies pred, and branches, on the "want" branch >>>>> doing >>>>> whatever and on the "do not want" branch putting the value onto an out >>>>> channel that feeds the other consumers before taking its own "do not want" >>>>> actions. >>>>> >>>>> -- >>>>> You received this message because you are subscribed to the Google >>>>> Groups "Clojure" group. >>>>> To post to this group, send email to clo...@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+u...@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+u...@googlegroups.com. >>>>> For more options, visit https://groups.google.com/d/optout. >>>>> >>>> >>>> -- >>> You received this message because you are subscribed to the Google >>> Groups "Clojure" group. >>> To post to this group, send email to clo...@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+u...@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+u...@googlegroups.com. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- > 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. > -- 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.
