Design practices from the CLD Exam Prep Guide

Question

    I have a question about design practices as stated in the CLD Exam Prep Guide found at   http://ftp.ni.com/evaluation/certification/cld/cld_exam_prep_guide_english.pdf
    Please refer to statements
2.e.1. Set the value or attributes of a control, statically using the property dialog box of an object, or dynamically using Property Nodes.
2.e.2. Initialize or set control values at application, load, start, and stop.
4.a.4. Use local variables to update controls.
    Because the words initialize and set and update all mean to change the value and two different methods are suggested to do so, I first saw this as contradictory.  Assuming there is no contradiction, it should be very clear when, where, and why each method is used.  
    From section 2.e I understand this to mean you should change (set) a control value at application, load, start, and stop by using Property Nodes.  Thus, application, load, start, and stop shall be clearly identified within your block diagram to avoid misunderstanding.  
    From section 4.a I understand this to mean you should change (update - as you might do repeatedly within a while loop) a control value by using local variables everywhere else in your block diagram.  If I understand this correctly, there is no contradiction.  Otherwise....
Please advise!  Scott Pierskalla
Certified LabVIEW Associate Developer

Answer 1

spiersk wrote:
    I have a question about design practices as stated in the CLD Exam Prep Guide found at   http://ftp.ni.com/evaluation/certification/cld/cld_exam_prep_guide_english.pdf
    Please refer to statements
2.e.1. Set the value or attributes of a control, statically using the property dialog box of an object, or dynamically using Property Nodes.
2.e.2. Initialize or set control values at application, load, start, and stop.
4.a.4. Use local variables to update controls.
    Because the words initialize and set and update all mean to change the value and two different methods are suggested to do so, I first saw this as contradictory.  Assuming there is no contradiction, it should be very clear when, where, and why each method is used.  
    From section 2.e I understand this to mean you should change (set) a control value at application, load, start, and stop by using Property Nodes.  Thus, application, load, start, and stop shall be clearly identified within your block diagram to avoid misunderstanding.  
    From section 4.a I understand this to mean you should change (update - as you might do repeatedly within a while loop) a control value by using local variables everywhere else in your block diagram.  If I understand this correctly, there is no contradiction.  Otherwise....
Please advise!  Scott Pierskalla
Certified LabVIEW Associate Developer
Scott,
Thank you for drawing my attention to this thread.  I understand your confussion but, there is no contradiction.  I will attempt to clairify the statements.
Section 2 addresses design of the GUI or, how to set up the user experience.  Several questions any developer needs to ask are:
Is the data represented correctly? (what data type is needed?)
How will the user change this data? (Enum / Ring / Combo box pull-down, maybe the increment / decrement buttons, Perhaps direct text entry from the keyboard? If so do we wish to update value while typing or end text entry with "Enter"?)
How can I make it "Pretty?"  (What colors are approachable, how can I present logical groupings of the data the user needs....?) 
Section 2 is all about the user.  Your application will be used by users and, they need to interact with your product.  Design the interface correctly and they will never notice, design it poorly and they will wish to break you nose because the software just does not "FEEL" right.
Section 4 addresses Programing practices.
Avoid Localitis
Avoid sequenceitis
don't ABUSE Globals
use common design patterns
Basically "Write Software" not "Functional spaghetti code"
Writting "Good Software" in LabVIEW is one of the things that CAN BE DONE with LabVIEW.  Unfortunatly, because LabVIEW is often approached by people with no understanding of Software Engineering principals, many LabVIEW applications fall into the "Functional Code" bucket and are poor "Software".
The CLD exam will require you to domenstrate that you can apply software engineering principals in the LabVIEW IDE.
Good Luck-  Sling some wires!
Jeff