Parallel section of the Julia Manual has errors

[Jim]

I think Julia is making great progress, I have been following it closely for the last 6 months and I still have to refer to the documentation. The great thing about Julia is that it can take advantage of multicore processors,but the documentation does not keep up with code updates. This is a big turn off for new comers. For example in the Parallel section, such as, drand apparently is not correct. I have tried to use the following: A=drand((10,10,10),3) and get the following error:

no method ref(Int64,Range1{Int64})

 in DArray at darray2.jl:31

 in DArray at darray2.jl:46

 in drand at darray2.jl:130

However if I use A=drand(10,10,10), I get a list of random numbers, which I expected.

Then if I try to execute the following example:

function comp(A::DArray)

    B = DArray(eltype(A), size(A), 3)

    for i = 1:size(A,3)

        @spawnat owner(B,i) B[:,:,i] = sqrt(A[:,:,i])

    end

    B

end

load("comp")

comp(A)

I get an error:

no method ref(Int64,Range1{Int64})

 in DArray at darray2.jl:31

 in DArray at darray2.jl:46

 in comp at /Users/jim/JuliaSrc/comp.jl:2

I'm using the Dec 13 version of Julia

I have an Imac I5 (4 cores) running OSX 10.8.2 with 16 GB of memory

Is there a location that I'm not aware of that contains updated docs? I hope somebody can give some help with this.

Jim

[Viral Shah]

Could you post this as two issues on github - one for drand, and one for the documentation?

The darray stuff has undergone an overhaul, and some old functionality has yet to be restored. In general, the documentation has not kept up with the pace of change.

One of my personal goals for the 0.1 release is to get some of the parallel stuff into good shape. Once we release 0.1, we will put in effort into ensuring that documentation is in sync with the releases.

-viral