Closed johnnychen94 closed 2 years ago
Merging #77 (37d5ba9) into master (6e6ac5e) will decrease coverage by
0.05%
. The diff coverage isn/a
.
@@ Coverage Diff @@
## master #77 +/- ##
==========================================
- Coverage 97.16% 97.10% -0.06%
==========================================
Files 14 13 -1
Lines 247 242 -5
==========================================
- Hits 240 235 -5
Misses 7 7
Impacted Files | Coverage Δ | |
---|---|---|
src/api/color.jl | 90.00% <ø> (ø) |
|
src/clustering.jl | 100.00% <ø> (ø) |
|
src/DitherPunk.jl |
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 6e6ac5e...37d5ba9. Read the comment docs.
@johnnychen94 there are still some issues with Julia 1.0:
LoadError: iteration is deliberately unsupported for CartesianIndex. Use `I` rather than `I...`, or use `Tuple(I)...`
caused by src/error_diffusion.jl:57
:
CI = CartesianIndices(filter) .- CartesianIndex(1, offset)
Should I replace this by something like
CI = [CartesianIndex(I .- (1, offset)) for I in CartesianIndices(filter)]
or is there a more elegant way?
Thanks for all the benchmarks!
I see. That's probably the only way to work around (or one could probably use Compat.jl) the compatibility issue.
Maintaining 1.0 compatibility becomes more and more troublesome these days because many dependencies (e.g., UnicodePlots, ArrayInterface) require >=1.6. I propose to save ourselves some effort and just set the Julia lower bound to 1.6. How do you think?
Sounds good to me. I'll merge this and remove 1.0 in a separate PR to make it visible in the commit history.
Loading time is almost the same:
The benefits are that users don't need to battle with conditional dependencies.
Before:
After:
@adrhill can you help update the documentation accordingly?