LabVIEW Idea Exchange

cancel
Showing results for 
Search instead for 
Did you mean: 
cy...

Shift Register - Retain value if unwired (right terminal)

Status: New

Good day forum

 

Proposal: add "retain value if unwired" option to shift registers. particularly useful to reduce wire clutter inside and outside of loops.LIX.png

Have a great day

 

 

CY (expired CLAD)
60 Comments
cy...
Active Participant

imagine the first special tunnel right side terminal unwired... that should be the case...

 

my paint.exe is giving me sharing violation messages and my new uploads for the png all cannot even make it to the preview after upload, except for the 2nd upload... after trying 3-4 times (changing file names), I too didn't want to spend time on something that might never happen.

CY (expired CLAD)
cy...
Active Participant

an additional note, I proposed the specialized tunnels to be separated by height in the BD, to improve readability of the code. having multiple in the same height or in close proximity may result in confusion in which terminal to which...

 

just like the "if(logic,if(logic,if(logic,true,if(logic,true,false)),if(logic(true,false)),false)" scenario

 

maybe colour coding (bad idea), or matching terminal names like the formula node to separate them? I don't know... what do you think?

CY (expired CLAD)
Intaris
Proven Zealot

"separated by height"..... separated from what? From other pairs or from each other?

 

I find the idea of inherent pairs being at different heights will not help readability at all.  Neatness maybe, but not readability.

cy...
Active Participant

hmm... readability in terms of screen real estate? or the inherent feed through pairs in the same height?

 

cos for the latter, unless we have unique identifiers for these pairs, the left and right terminals may not be easily distinguishable within the same height. if it is of different data type, color can be used as the identifier.

CY (expired CLAD)
Intaris
Proven Zealot

If you have 2-3 or more pairs, decoupling the vertical alignment for input and output makes a completely unreadable mess without interaction with the diagram (clicking on one tunnel autohighlights its pair for example).  IMHO, instant readability (no interaction required) trumps beats interactive readability every time.

cy...
Active Participant

point taken... may I suggest that perhaps the structure background colors to be customizable (low saturation, medium high brightness)? this can probably improve readability and distinguishability of pairs.

CY (expired CLAD)
Intaris
Proven Zealot

I think NI has been very hesitant to exploit colour as a differentiator in this regard as

 

1) colours have pre-defined meanings in LV and overloading the colour palette makes instant context detection dfifficult

2) colour blindness is real

3) It's straining on the eyes over time (yes, time catches up on us all and old eyes get tired more quickly then young eyes) and I've yet to meet a LV programmer who's getting younger

 

But if there would be a repeating list of, say 4 different brightnesses or saturations (or other variation?) of a given colour (so that datatype colour can be maintained) which could be used to differentiate multiple such items within a single structure, that might work.  But the number of colours I think must be fixed and few.

wiebe@CARYA
Knight of NI

>may I suggest that perhaps the structure background colors to be customizable (low saturation, medium high brightness)? this can probably improve readability and distinguishability of pairs.

 

Structure background is customizable as it is. Not a great feature, IMHO.

Structure BG.PNG

cy...
Active Participant

pardon me for I didn't know that is customizable, is it by structure type, or structure by structure basis? how to access this parameter anyways?

 

I think following options can be considered for differentiating the special tunnel in the same height:

1) by color, be it in the borders of the tunnels or structures, or border gradients of structures, switchable in options.

2) by text/symbol terminal identifiers, like terminals in formula nodes (but in pairs).

LIX3.png

 

 

example of structure border gradients

LIX4.png

 

 

color option to be based on depth of structure within the BD. For example:

1) standard producer-consumer loops, the first level timed loop and while loop will share the same color identifier, indicating its depth

2) standard state machines, the first level while loop will have a different color identifier than the case structure

 

Textual/symbolic identifiers on the other hand can be more readable, as it doesn't rely on the color, to be identified intuitively, and unique identifiers can even handle varied structure depths. for example: terminal X

  • left terminal in: declare a variable X and write a value to it (initialize) 
  • wired right terminal: action (overwrite new value)
  • un-wired right terminal: no action   
  • right terminal out: read initialized/overwriten value

*can call them "variable tunnels", for obvious reasons?

 

seemed like the textual/symbolic identifier may be better... what do you guys think?

CY (expired CLAD)
wiebe@CARYA
Knight of NI

>pardon me for I didn't know that is customizable, is it by structure type, or structure by structure basis? how to access this parameter anyways?

 

Simply select the coloring tool, and color each individual structure diagram or structure border as you would a control or decoration.

 

And (IMHO) please don't.