AF Presentations at NI Week 2013

[Elijah_K]

As David points out - the hands-on session isn't new (though I am updating it slightly this year), and it's very basic by comparison.  It's also being given twice, so you might still be able to make both.  If you can' t make the other one, you can go through the hands-on material at your own pace later - go to David and Daklu's session at NIWeek!

[Elijah_K]

For any of you who will be at Alliance Day, I will be giving a session I created for the European CLA Summit on creating Plugin Measurement Abstraction Layers using the Actor Framework.  The official title in the program is: "Designing a Plug-In Measurement Abstraction Layer Using the Actor Framework (10-10:55 AM)"

[SteenSchmidt]

Nancy_Hollenback wrote:

I just sat in on a dry-run of the Staab/Daklu presentation.  Don't miss this one!

Seriously bad news when I can't attend :-(( Purely selfish of course, as in every other way it's great news 🙂 NI Week is deterministically impossible for me to attend this year. Grim.

Hope you'll all enjoy it, and I'll be looking forward to the discussions in the OO/design patterns/AOD departement afterwards.

Cheers,

Steen

[komorbela]

If someone could record these sessions and share the video... That would be cool! I would really like to be there but cannot attend for sure

[Daklu]

(Cross-posted to the LV Architects Forum)

Michael Lacasse asked this question during the presentation:

"Question for Daklu (not urgent): Have you ever assembled UML diagrams that succinctly and sufficiently describe your architecture using Actors?  NIs AF is lacking satisfactory UML documentation in my opinion. This presents an obstacle to adoption."  (Emphasis added)

The short answer is no.  As always, I have a longer answer too. 

The long answer is, yes, I have used UML to describe an application's design I have implemented using actors.  I don't think that's what you're asking though--I think you're asking if I've documented the general patterns I use to build apps.  That answer is no.

First, my approach to actors isn't an architecture, a framework, or a design pattern.  It's more of a development methodology or style than anything else.  (I purposely haven't added an actor framework to LapDog because I haven't been able to invent one that is more effective than what I currently do.)  I don't know how to use UML to describe a methodology.

Second, although the UML class diagram doesn't necessarily mean "these are classes in my source code," that is usually how it is used and interpreted.  However, AoD and OOP are orthogonal ideas.  My goal, not only with this presentation but with all my posts on the subject, is to help people understand the fundamentals of AoD by removing all the complexities of OOP.  Once people understand the basic building blocks--actor, MHL, helper loop, msg transport, etc.--and how they fit together then it becomes very easy to implement those ideas however you want.  I believe using UML to describe the relationships would strengthen the connection between AoD and OOP in people's minds.

Third, I haven't been able figure out how to "succinctly and sufficiently" describe what I do.  UML diagrams won't be enough.  I've started white papers 3 or 4 times trying to get the information in my head out in a way that is understandable.  These slides are the closest I've been able to get to that goal.  I have an outline for a new paper (tentatively titled, "Agile Actors for LabVIEW Programmers") that formed the foundation of these slides.  We'll see if I get anywhere with it. 

[MarkBalla-Tecnova]

For thoes that can't be there in person.

I am planning to record the presentation providing the presenters give me permission.

Mark

[jdebuhr]

Thanks Mark,

I will want to see those since I cannot go this year.. but maybe now I can go next year

[Daklu]

I am planning to record the presentation providing the presenters give me permission.

It's okay with me.

[David_Staab]

MarkBalla wrote:

I am planning to record the presentation providing the presenters give me permission.

I hope to share the material in a "better" format on my own after the conference, so I prefer that you not publish a recording of my part of the session.

[AristosQueue (NI)]

I am somewhat amazed. There were six NIWeek sessions  on the AF and two hands-on sessions. Only the hands-on and two presentations were done by NI employees (and there were community folks helping out with the hands-on). I am delighted with a) how much the community values the AF and b) how much the community has picked up the work of documenting and educating other users about it. Thank you to all of you who have contributed to it.