Differences

This shows you the differences between two versions of the page.

--- 2013:dataflow_is_the_architecture_you_need [2012/10/24 07:34] – created jonjagger
+++ 2013:dataflow_is_the_architecture_you_need [2016/06/11 14:05] (current) – external edit 127.0.0.1
@@ Line 4: / Line 4: @@
 \\
 **Title**: Dataflow is the architecture you need\\
-**Proposer**: [[2013:russell_winder]]\\
+**Proposer**: [[2013:russel_winder]]\\
-**Type**: Workshop/Tutorial\\
+**Type**: Presentation (with interaction)\\
 **Duration**: 90 minutes\\
-**Description**:
+**Description**: Shared-memory multi-threading is all very well just as long as you don't do it for applications
+programming. Yet we need concurrency and parallelism in our applications, and threads are the current
+principal tool to provide this.\\
+\\
+Java championed shared-memory multi-threading, but Java 8 will change that game. Groovy and GPars are
+emphasizing message passing over shared-memory multi-threading. Go eschews shared memory in favour of
+message passing between processes (cf. CSP). Scala emphasizes the use of actors. The C++11 standard
+introduced a standard thread and memory model, apparently indicating that C++ promotes shared-memory
+multi-threading as a good thing.  Fortunately, asynchronous function call with futures made it into the
+standard giving folk a high-level concurrency and parallelism construct.\\
+\\
+There are implementation of actors, dataflow and CSP (communicating sequential processes) for Python,
+Groovy, Scala, Java, D, Go, C++.  In this session we will look at why the dataflow approach is an
+improvement for programmers, and why shared-memory multi-threading is the wrong tool for the job unless you
+are writing an operating system or something equivalent.\\
+\\
+Roger: Yes
+\\ Asti - yes \\