4 ms·
I am usually wary of new languages, because they come at a great cost. All the code I've written, libraries I've discovered, standard library I'm used to and ta
by shurcooL 13y ago
I am usually wary of new languages, because they come at a great cost. All the code I've written, libraries I've discovered, standard library I'm used to and take for granted are suddenly gone, and I have to start from scratch. On the plus side, the new language may have desirable aspects that are improved.
However, this is not that. Despite looking like a different language, in reality, it's just a different _interface_ to writing/reading Go. It's completely compatible with all the libraries, tools that work on Go code because it _is_ Go code, simply expressed via a different representation of the Go AST.
I think that is a really nice way to experiment with the positive aspects of new languages/syntaxes. Nicely done!
- sitkack 13y agohttp://en.wikipedia.org/wiki/Bidirectional_transformation http://en.wikipedia.org/wiki/Bidirectional_transformation It might be interesting to hook up different AST representations directly into an editor, in this case iGo would never exist on disk. Really all he has done is push what would be an editor completion into a new sytnax.