Render updates 3

[hollasch]

See #179

• Resolves Update render 1.05 #547
• Resolves Update render 1.06 #548
• Resolves Update render 1.07 #549
• Resolves Update render 1.08 #550
• Resolves Update render 1.02 #551
• Resolves Update render 1.03 #552
• Resolves Update render 1.04 #553
• Resolves Update render 1.09 #554
• Resolves Update render 1.10 #555
• Resolves Update render 1.11 #556
• Resolves Update render 1.12 #557
• Resolves Update render 1.15 #560
• Resolves Update render 1.16 #561
• Resolves Update render 1.17 #562
• Resolves Update render 1.18 #563
• Resolves Update render 1.19 #564
• Resolves Update render 1.20 #565
• Resolves Update render 1.21 #566

Book 1 render updates: text changes

• Add extra blank line before code listing highlighting changes
• Add missing text change highlights
• Add missing header includes
• Organize main() chunks: image params, then world def, then camera def,
• then image render
• Images parameters updated for new renders
• Some code listings get more context
• Fix trivial code format issues
• Add missing code listings
• Side-by-side antialising illustration
• Fix source file name error for fig-1.09
• Change render image includes to use new names (coming in next commit)
• Improve code for distinct world material + geometry definitions
• Include render parameters for final scene

Book 1 render updates: code changes

• Add more double-line spacers
• Organize main() chunks: image params, then world definitions, then
• camera defs, then render
• Change class order in material.h to match book definition order
• Eliminate duplicated code in dielectric::scatter()
• Delete vestigal vec3::write_color() method (now in color.h)

[trevordblack]

img-1.14-glass-always-refracts.png is incorrect.
However. It is delightful.

[trevordblack]

After replacing that image, it's approved.

[hollasch]

The code is broken at this point. img-1.13 has no corresponding source (I don't know how it was rendered), and 1.14 is broken from the reflection/refract/front-face changes. Keeping it in for now, rather than suggesting that readers should expect some other result.

See #558, #559 -- these are not resolved in this merge.

[hollasch]

BTW, we can at least now use the progression branch to get to a good debugging point for issues like this.