The Author Online Book Forums are Moving

The Author Online Book Forums will soon redirect to Manning's liveBook and liveVideo. All book forum content will migrate to liveBook's discussion forum and all video forum content will migrate to liveVideo. Log in to liveBook or liveVideo with your Manning credentials to join the discussion!

Thank you for your engagement in the AoF over the years! We look forward to offering you a more enhanced forum experience.

Paolo Patierno (13) [Avatar] Offline
#1
Hi,

the images which show the simple topology of the "Yelling" app have a "Source Processor" node and then a "To Processor" node at the end.
Maybe it should be "Sink Processor" instead of "To Processor" ?

Thanks,
Paolo.
Bill Bejeck (47) [Avatar] Offline
#2
I might change that, but my naming convention was to follow the high-level DSL.

Thanks,
Bill
Paolo Patierno (13) [Avatar] Offline
#3
Following the DSL means that you should use "Stream Processor" and not "Source Processor" for the first node because in order to have a KStream from the KStreamBuilder you have to use the "stream" method. At same time for having a KTable you use "table" method. In this way it seems to be difficult following DSL for the pictures in the node. In my opinion, using Source and Sink processor works for all of them smilie

Thanks,
Paolo.
Phil Derome (48) [Avatar] Offline
#4
This relates to the same discussion but more specifically, I don't follow the distinction between terminal nodes and sinks other than being told that there are terminal nodes that are not sinks. Well, what can they be then?
Phil Derome (48) [Avatar] Offline
#5
Phil Derome wrote:This relates to the same discussion but more specifically, I don't follow the distinction between terminal nodes and sinks other than being told that there are terminal nodes that are not sinks. Well, what can they be then?


The explanation comes much later in 3.3.1 near Figure 3.7 surrounding "BUILDING SECOND PROCESSOR" subsection although this later discussion fails to hammer the nail by avoiding the expression "terminal node" but the figure does speak about two nodes, one being the sink inside the "to" method and the other one presumably being "terminal" as the "patternKStream".

Figure 3.8 and Listing 3.11 but the confusion between Sink and Terminal Node remains.

There is some opportunity here to clarify the matter in editing these two discussions or tying them together explicitly.

The ah-ah moment comes in Listing 3.13 item 5, quite later than when the topic is introduced: a sink is just the last terminal node attached to the topology using the DSL. There can be only one sink per topology but the terminal nodes are the end points of the directed acyclic graph.
Bill Bejeck (47) [Avatar] Offline
#6
Hi Phil,

Thanks for the comments. We're close to the final production release, but I should be able to make some changes to make things more clear.

Thanks again,
Bill
Bill Bejeck (47) [Avatar] Offline
#7
Phil,

We're in the final stages before publishing and I've tried to clear things up.

Thanks,
Bill