Erik Meijer: Duality and the End of Reactive [video]
channel9.msdn.com
Erik Meijer: Duality and the End of Reactive [video]
1–10 of 68 posts
Re: Erik Meijer: Duality and the End of Reactive [video]
#2The first six minutes alone are worth a listen if you're in a rabble-rousing mood.
Re: Erik Meijer: Duality and the End of Reactive [video]
#3i thought it was a good presentation, however, I recently watched a Netflix one, published by Matt Marenghi, which to me was infinitely better.
I think the description of events as just another collection, albeit a time based collection flow was nicer ( Observable === Collection + Time ), and also that this collection must have the ability to tell you that it has terminated.
Re: Erik Meijer: Duality and the End of Reactive [video]
#4(Shameless self-promotion) I just finished by thesis on a new programming language called SolScript, which is reactive and based on "math".
It's a hard real-time language for avionics software, BUT it has type inference and duck typing (everything is "symbolic"). SolScript is supposed to look familiar to anyone with a basic knowledge of mathematics. It's also a literate language, because every SolScript file is a Markdown file (and SolScript code is inside Markdown code blocks).
[link redacted]
Re: Erik Meijer: Duality and the End of Reactive [video]
#5The first six minutes alone are worth a listen if you're in a rabble-rousing mood.
I definitely wouldn't mind watching more of his presentations. He's pretty good.
Re: Erik Meijer: Duality and the End of Reactive [video]
#6Now I want to know who was in the audience (esp. the two 'math' guys off to the right).
Re: Erik Meijer: Duality and the End of Reactive [video]
#7The first six minutes alone are worth a listen if you're in a rabble-rousing mood.
Wow, I listened to the first six minutes on your recommendation and was pretty put off.
The speaker mocks his audience of Microsoft engineers (like not in fun, he is really saying their culture sucks).
He is dismissive of "reactive programming" as nonsense from architecture astronauts, without giving a remotely fair description of what it actually is. His straw-man of "var x = 10; println(x); x = 42; println(x)" is not reactive, because "x" is not a function of other mutable state in the model. I've never heard of reactive programming until five minutes ago, but I can tell this just by reading the Wikipedia excerpts from his own slides (and a little follow-up reading confirms this).
I'm all for rabble-rousing, but if you're throwing punches you should know what you're talking about.
At 17:30 he makes a joke about a Steve Jobs function that says "iPhone", "iPad", "iCloud", and then "terminates naturally". Holy bad taste.
EDIT: others have pointed out that with extra context this is more obviously a self-deprecating and good-natured shtick.
Re: Erik Meijer: Duality and the End of Reactive [video]
#8(Shameless self-promotion) I just finished by thesis on a new programming language called SolScript, which is reactive and based on "math". It's a hard real-time language for avionics software, BUT it has type inference and duck typing (everything is "symbolic"). SolScript is supposed to look familiar to anyone with a basic knowledge of mathematics. It's also a literate language, because every SolScript file is a Mar…
Just had a little glance (I'm on my way out the door), but if it's possible to construct arbitrarily large lists (as seems to be implied by the availability of e.g. "range N" for any N, how can you guarantee that e.g. "sum large_list" can actually terminate within the time bound set by the frequency?
Also there seems to be no actual proof of safety of the language's semantics (are the semantics formally specified?), but perhaps I missed it.
Re: Erik Meijer: Duality and the End of Reactive [video]
#9The first six minutes alone are worth a listen if you're in a rabble-rousing mood.
Wow, I listened to the first six minutes on your recommendation and was pretty put off. The speaker mocks his audience of Microsoft engineers (like not in fun, he is really saying their culture sucks). He is dismissive of "reactive programming" as nonsense from architecture astronauts, without giving a remotely fair description of what it actually is. His straw-man of "var x = 10; println(x); x = 42; println(x)" is n…
When you say that he's mocking Microsoft engineers, is that because of the Visual Basic comments? Because Erik actually enjoys VB and was responsible for adding some features to it. He's a former employee, so there's context...
And I don't think he's being dismissive of reactive programming, he's dismissive of that definition, which is too broad. He was one of the main proponents of Reactive programming inside Microsoft.
Re: Erik Meijer: Duality and the End of Reactive [video]
#10The first six minutes alone are worth a listen if you're in a rabble-rousing mood.
Wow, I listened to the first six minutes on your recommendation and was pretty put off. The speaker mocks his audience of Microsoft engineers (like not in fun, he is really saying their culture sucks). He is dismissive of "reactive programming" as nonsense from architecture astronauts, without giving a remotely fair description of what it actually is. His straw-man of "var x = 10; println(x); x = 42; println(x)" is n…
When Erik mocks reactive programming I believe he is reacting to the Reactive Manifesto (http://www.reactivemanifesto.org/) currently being pushed by Typesafe and others.
My reading and feelings concerning the Manifesto mirror Erik's so I enjoyed listening to his mockery of it. I didn't really notice the MS bashing.
The issue with the Reactive Manifesto is that it is largely waffle. There are styles of programming that account for time, such as functional reactive programming. They tend to have formal definitions, or at least concrete implementations, so one can make objective statements about them. The Manifesto on the other hand is vague and full of platitudes. It's great for the Pointy-haired Boss, but there is little of substance than one can discuss in any objective manner.
[Update] I can understand how it's off-putting to listen to the presentation without this background.