Have a 2nd look at `DashBasePlotsExt`

[etpinard]

specifically its ::Plots.Plot{Plots.PlotlyBackend} dispatch

DashBase.jl/ext/DashBasePlotsExt.jl

Lines 7 to 16 in f3ea1ae

	function DashBase.to_dash(p::Plots.Plot{Plots.PlotlyBackend})
	return if haskey(Base.loaded_modules, Base.PkgId(Base.UUID("a03496cd-edff-5a9b-9e67-9cda94a718b5"), "PlotlyBase")) &&
	haskey(Base.loaded_modules, Base.PkgId(Base.UUID("f2990250-8cf9-495f-b13a-cce12b45703c"), "PlotlyKaleido"))
	# Note: technically it would be sufficient if PlotlyBase is loaded, but thats how it is currently handled by Plots.jl
	sp = Plots.plotlybase_syncplot(p)
	DashBase.to_dash(sp)
	else
	Dict(:data => Plots.plotly_series(p), :layout => Plots.plotly_layout(p))
	end
	end

which looks up if PlotlyKaleido is loaded to determine if DashBase.to_dash can use Plots.plotlybase_syncplot.

---

After a quick over in https://github.com/JuliaPlots/Plots.jl, I couldn't find haskey(Base.loaded_modules, #= =#) signatures mentioned in the comment of the above DashBasePlotsExt snippet.

Maybe the pattern in Plots.jl has changed since the DashBase v1 release (back in August 2023) ? 🤔

Having PlotlyKaleido in the DashBase [compat] list may cause issues with Julia 1.10 (ref JuliaPlots/PlotlyKaleido.jl#6)

We should take a 2nd look at this.

[BeastyBlacksmith]

Plots uses a different mechanism. It tries to load these packages in
https://github.com/JuliaPlots/Plots.jl/blob/0d956739d57d87b7557649aaf4948fcf24fb60d4/src/backends.jl#L586-L591
and registers an error hint, if that fails.
But in essence it still works the same.

[NicolasRiel]

Hi, I am not sure if this is related but I am having issues with precompilling DashBasePlotExt on Windows. It works very well on Linux and Mac, but on windows this part hangs forever.

The continuous integration also shows similar behavior for windows.

I wish I could get more information to send you on what is happening, but I don't know where to get it!

Cheers

[etpinard]

thanks @NicolasRiel for the report!

[NicolasRiel]

Hi, this has been fixed with the last version of PlotlyJS 0.18.13
There is still some issues but I think they are unrelated.

Nico