revert to using hex for fill colors instead of RGBA

[tlnagy]

2nd part of #322

[codecov-io]

Codecov Report

Merging #324 into master will decrease coverage by 48.13%.
The diff coverage is 71.42%.

@@            Coverage Diff             @@
##           master    #324       +/-   ##
==========================================
- Coverage   79.44%   31.3%   -48.14%     
==========================================
  Files          14      18        +4     
  Lines        1192    3031     +1839     
==========================================
+ Hits          947     949        +2     
- Misses        245    2082     +1837

Impacted Files	Coverage Δ
src/property.jl	20.25% <ø> (-79.75%)	⬇️
src/svg.jl	60.09% <71.42%> (-18.6%)	⬇️
src/Compose.jl	10.12% <0%> (-89.88%)	⬇️
src/table.jl	9.52% <0%> (-84.6%)	⬇️
src/list.jl	19.23% <0%> (-80.77%)	⬇️
src/misc.jl	29.62% <0%> (-54.59%)	⬇️
src/form.jl	44.18% <0%> (-52.76%)	⬇️
src/measure.jl	24.76% <0%> (-51.71%)	⬇️
src/cairo_backends.jl	36.25% <0%> (-46.02%)	⬇️
... and 10 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update afb9b07...391859c. Read the comment docs.

[bjarthur]

bump

[bjarthur]

is this ready to merge? or is it stale?

[Mattriks]

The function svgalphatest needs to be added to this PR. See this post and the subsequent posts. If you merge this PR, I can add svgalphatest in another PR?

[tlnagy]

As @Mattriks says, this PR needs to throw a warning to alert users to the change in behavior.

If you merge this PR, I can add svgalphatest in another PR?

This is fine, IMO

[bjarthur]

@Mattriks add the test and then we can tag

[Mattriks]

The above merge had a side-effect: the fill(nothing) statement in a parent context now clobbers the fill(color) statement in a nested context (for svg). See the 2 examples in the Transformations gallery, which both use nested contexts. I should be able to sort that out.