PG1204 commented on PR #8552:
URL: https://github.com/apache/texera/pull/8552#issuecomment-5734107989

   > Not a blocker for this PR and not in its diff, but since it touches the 
heat-map overlay: the operator fills read as quite saturated on the canvas.
   > 
   > The ramp in `heatmap-color.ts` is a reasonable choice in itself 
(ColorBrewer RdYlBu reversed, colorblind-safer than a rainbow). Two things push 
it brighter than the reference palette, though: the cold end sits at 59% 
saturation where ColorBrewer's own `#4575b4` is 45%, and the midpoint `#ffffbf` 
is pure yellow at 100%.
   > 
   > The larger factor is probably not the ramp but how it is applied. 
`joint-ui.service.ts:537` uses the color as an opaque fill across the whole 
operator box, while the same ramp in the thin legend bar does not read as harsh 
at all. Blending the fill toward white (roughly 35%) would keep the hue 
separation and the colorblind-safe property, take the edge off the canvas, and 
improve label contrast on top of the fill.
   > 
   > Entirely out of scope here, so no action needed on this PR. Happy to open 
a separate issue for it if that is preferred.
   
   
   Numbers hold: 59% against #4575b4's 45%, and 537 paints an opaque 
rect.body/fill.
   
   Two notes. #ffffbf is RdYlBu's own midpoint, so the pure yellow is 
inherited, not introduced here. The cold end is the only real departure, and 
it's deliberate: lightened for label readability, per the comment on the 
constant.
   
   On ~35% toward white, CIELAB ΔE end-to-end drops 87.5 -> 54.7, about 37% 
less discriminable range (spacing stays uniform, the ramp just shortens). Helps 
harshness and contrast, so a trade rather than a free win.
   
   Your legend point suggests a cheaper fix: the same ramp is fine in a thin 
bar and harsh on the box, so the variable is painted area. A strip inside the 
operator keeps full range. The border's taken - #6213 gave it to execution 
status.
   
   Out of scope here, agreed. I'll open a separate issue for this.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to