#1679 NORM Trial-2: XO colors can be too close together to distinguish

[Zarro Boogs per Child]

#1679: XO colors can be too close together to distinguish
-------------------------------+--------------------------------------------
  Reporter:  sj                |       Owner:  Eben   
      Type:  defect            |      Status:  new    
  Priority:  normal            |   Milestone:  Trial-2
 Component:  interface-design  |     Version:         
Resolution:                    |    Keywords:         
  Verified:  0                 |  
-------------------------------+--------------------------------------------
Comment (by Eben):

 There are a couple facets to my response.  First, ideally the color picker
 in the intro sequence design would also have value sliders much like the
 javascript one.  They were only omitted from the first pass mockups in
 order to save time, with the assumption that we could add them in later.
 With those, there are many more color combinations that can be made.

 Second, Pentagram and I have been making some assumptions about the color
 pairs that will be allowed while making the mockups.  In particular, a
 number of the designs require that the fill color be sufficiently dark
 such that white text will show clearly against it, for all possible
 choices of colors.  Additionally, we want them to be distinct against both
 white and black backgrounds.  That said, there are a number of parameters
 to take into account:

  - fill color must be >2 munsell steps away from white (10)
  - fill & stroke colors must be >2 munsell steps away from both black and
 white (0, 10)
  - stroke color must be >1 munsell steps away from fill color

 This yeilds valid ranges of (3 - 7) for fill and stroke, where valid
 (fill, stroke) pairs are (3, 5|6|7), (4, 6|7), (5, 3|7), (6, 3|4), (7,
 3|4|5), or a total of 12 value pairings, for a total of 768 total color
 combinations.  Actually, we might be better off allowing 10 chromatic
 steps around the wheel instead of the current 8 (This fits more closely
 with the Munsell spec, which defines Red, Yellow, Green, Blue and Purple
 "primaries" with "secondaries" between them).  That yields a total of 1200
 combinations within our limited ranges, which seems more than reasonable.

 Walter, does the above logic make sense to you?  Glancing at the tables in
 the Colors section of the HIG, it seems 3 value steps is enough to clearly
 read text against a background.  If we want, we could push this to 4
 steps, limiting the fill color to a range of (3-6), and limiting the total
 combinations to 900.

-- 
Ticket URL: <http://dev.laptop.org/ticket/1679#comment:4>
One Laptop Per Child <http://laptop.org/>