Dale LaBossiere commented on EDGENT-179:

Any Edgent user can create their own oplets, as can new connector 
implementations, or any other enhancement to Edgent as/if needed.  

Seems like a better alternative would be to randomly allocate colors but keep 
track of what's been allocated and redo the random allocation until it's "not 
too close" to an in-use color.  There can still be pre-allocated ones with 
predictable/static colors for "known" oplets (which of course will become out 
of date over time).  You could also have different color allocation schemes for 
org.apache.edgent defined oplets vs non-edgent defined ones (if that might be 
helpful to users).

There's no single place where all of the Edgent defined oplets are located in 
the repository nor registry of them.
To identify Edgent's currently defined oplets, in Eclipse you can open the 
"Oplet" interface and select the "Quick Type Hierarchy" action on it.  The 
result will include abstract classes (nothing the console will encounter) and 
test specific ones but these are pretty easy to identify based on their full 
class name.

> console: oplet coloring gradients "too close"
> ---------------------------------------------
>                 Key: EDGENT-179
>                 URL: https://issues.apache.org/jira/browse/EDGENT-179
>             Project: Edgent
>          Issue Type: Bug
>          Components: Console
>            Reporter: Dale LaBossiere
>            Priority: Minor
>         Attachments: EDGENT-179_oplet_colors.png, Screen Shot 2016-09-19 at 
> 2.22.00 PM.png
> The oplet color generator ends up yielding two light purple-ish colors for 
> two different kinds of (adjacent) oplets and one can't tell them apart.  One 
> is a Sink the other is a subtype of Peek - a StreamScope.
> I was seeing this in a ParallelBalancedRecipe app of mine.  I can provide it 
> if really necessary to repro/improve this.

This message was sent by Atlassian JIRA

Reply via email to