Julia v1.7.0 has been released

[dmm]

julialang.org/blog/2021/11/julia-1.7-highlights/

Julia 1.6 is the new LTS (long-term support), instead of Julia 1.0 which is no longer maintained.

Comments

[dmm]

Switch to Julia 1.6.4 for LTS.

I hope that Diffractor.jl first release is not too far away: https://github.com/JuliaDiff/Diffractor.jl

[dmm]

This is still listed as experimental:

https://github.com/JuliaLang/julia/blob/master/base/opaque_closure.jl

Hopefully, this will all work Diffractor-wise...

[dmm]

OK, so there is a Hacker News Julia 1.7 highlights post: https://news.ycombinator.com/item?id=29399844

And there is a long write-up on Diffractor.jl by Keno Fisher in that post:

https://news.ycombinator.com/item?id=29401071

(Via https://twitter.com/KenoFischer/status/1465925184578363392 )

[dmm]

It's no longer enough to "add IJulia", one actually does need to "build IJulia" for the new version to be added to Jupyter.

Otherwise, everything is running very smoothly and rapidly so far.

[dmm]

Interesting drift in numerical things (pre-optimization).

May 2021:

> loss(m_im0, p_m_im0_warped_norm)

20.39865222030749

October:

> loss(m, target_value)

20.39928388822998

now, attempting to exactly reproduce the October run:

> loss(m, target_value)

20.400127548040803

Not much difference, 1.0000413573248672 is the ratio of the last two numbers, but still... uhmmm... what did they change?

[dmm]

Julia 1.6.5 exists now for LTS.

Julia 1.7.1 has been released today.