> Neko looks like a very promissing project for dynamic languages. > I'm impressed with the speed at which it is being developped, the size of > the vm and the (unoptimized) speed of execution.
Thanks > I just stumbled upon www.nekovm.org and a couple of questions came up. > > * Does neko (or nekoml) support generators like python > (http://www.python.org), Generators can be implemented using continuations. Actually with continuations you can do much more than generators only. Continuations are planned for Neko 1.3 > * Unicode: most current programming languages use a unicode string > representation. I'm not sure about this, but Neko seems to have a single byte > string representation, but with support for UTF8. Is it wise to stick to the > (old fashioned?) single byte representation? Neko source programs can be either UTF8 or ISO. Literal strings can then be either UTF8 or ISO, but identifiers are required to be ASCII. Neko strings are just byte buffers, they are then independant of any encoding. It then all depends of the API you're using to manipulate them. You can user either the String Builtins which are manipulating bytes (then suitable for ISO) or the UTF8 API which is manipulating UTF8 charcodes, whatever you like. > * Benchmarking. Since it is/will be an issue for the adoptation of neko, it is > nice to implement some of the benchmarks of the computer language shootout > (http://shootout.alioth.debian.org/) and apply for an entry in this shootout. I started doing it (see Neko CVS /src/benchs) but I don't have enough time to complete it. You're more than welcome to contribute additional programs. I guess we need a good part of them before applying for an entry. Best, Nicolas -- Neko : One VM to run them all (http://nekovm.org)
