On 29.01.2016 00:16, Edinson E. Padrón U. wrote:
Hi, Guillaume. In my very humble opinion (and it should be noticed that I'm very far away to know the Groovy community and language internals as well as you do), the Python 2.x vs 3.x 'war' was due to mainly a very slow adoption of the 3.x branch from the different third-party libraries. Even though the 3.x branch is far better than its predecessor, the community stuck with the 2.x branch because of the incompatibility of the libraries their depended on.
I wish you had any idea about how many projects did still use Groovy 1.8 a year ago. It required a CVE for them to even consider changing.
[...]
Jigsaw is inevitable and that for itself require to break backward compatibility.
yes and no.. no, because this does not *require* a new MOP, which is all Groovy3 originally was about. Yes, there will be breaking changes... our extension methods will for example have to use proper service provider mechanism, our modules may have to move a few classes because of the almighty no same package for two modules paradigm - just to just name two random items. It would be a good chance though to introduce a new MOP... here I agree.
bye Jochen
