Given a choice between 1:10==[1:10] returning false or throwing an error, I'd vote for "error".
On Mon, Aug 25, 2014 at 12:06 PM, Stefan Karpinski <[email protected]> wrote: > If you (or anyone) can come up with a clever scheme for hashing arrays and > ranges so that 1:n and [1:n] hash the same but hash(1:n) isn't an O(n) > operations, I'd be thrilled to switch this back. I could not figure out a > good way to do this, however. > > > On Mon, Aug 25, 2014 at 11:47 AM, Jacob Quinn <[email protected]> > wrote: > >> See this issue: https://github.com/JuliaLang/julia/issues/7867 and the >> discussion in https://github.com/JuliaLang/julia/issues/5778 for >> information on the change. >> >> -Jacob >> >> >> On Sat, Aug 23, 2014 at 9:02 AM, Ed Scheinerman < >> [email protected]> wrote: >> >>> Is there a document describing new features and significant changes >>> between versions 0.2 and 0.3? >>> >>> One item I noticed is that in 0.2 the express 1:5 == [1:5] evaluated as >>> true, but in 0.3 it's false. >>> >> >> > -- Ed Scheinerman ([email protected])
