Update to 0.7 and 1.0 #156
Conversation
| ind = inds[i] | ||
| push!(sigs, Signal(first(ind))) | ||
| push!(axs, Axis{n}(ind)) | ||
| end | ||
| end | ||
| Signal(zr), SliceData{dims[2] < dims[1]}((sigs...), (axs...)) | ||
| Signal(zr), SliceData{dims[2] < dims[1]}((sigs...,), (axs...,)) |
There was a problem hiding this comment.
is the trailing , here needed now?
There was a problem hiding this comment.
I think so, and in fact I believe this was added by Femtocleaner (which I ran locally).
julia> x = (1,2)
(1, 2)
julia> s = (x...)
┌ Warning: Deprecated syntax `(x...)` at REPL[2]:1.
│ Use `(x...,)` instead.
└ @ REPL[2]:1
(1, 2)
|
Waits on JuliaLang/METADATA.jl#17497 (which I had forgotten was justifiably blocked), but otherwise this is hopefully good to go. |
|
Seems blocked by JuliaIO/ImageMagick.jl#133 |
|
If you want the Travis tests to run, you need to force use of a newer |
|
Sorry for the misleading advice; I forgot that Pkg runs tests in a separate environment. I think the |
|
It's also weird that QuartzImageIO got installed. I will look into this later, gotta run now. |
|
So this was worth merging, but I worry that users of older Ubuntu versions will encounter the same problem and have to manually use the same workaround. Short of building the new version statically into ImageMagick, I'm not sure I see a fix. Do you? |
|
This is an example of a larger problem in the foreign library business. I got a bit carried away and raised an alarm about this on Discourse, where Stefan K made it clear that the Pkg crew hopes to improve the situation. In this instance, there has been talk of using BinaryBuilder for Gtk, which might fix it. (I'm skeptical of that project, given how deep the linking of Gtk goes into the system, but I'll be happy to be surprised.) In the interim, it would be good to provoke more useful error messages. That's a nice challenge. |
credit to @RalphAS for solution; see JuliaImages/ImageView.jl#156 and JuliaIO/ImageMagick.jl#133
No description provided.