Quite right. My new designer is quick enough to understand that he can develop the entire skin in html, but for the dslocalizator skin object. I have no problem.
Well, I do have a problem. I have a stake in the success and survival of dslocalizator. Many of your current users must feel the same way.
First of all, there is no easy way to migrate from a dslocalizator solution to a core ML system. It will require in its current state a lot of work from the operators to migrate.
Secondly, for many users, dslocalizator will still be the easiest and most resource-friendly option. It seems the dnn development team seems to understand this dilemma and has vowed to keep dnn compatible with external ML solutions. That means, your users will continue to use dslocalizator and the use of dslocalizator remains meaningful.
Having said that, it is my interest that dslocalizator remains an attractive option.
New users will want to try building skins using using the html format first. Having a working sample skin in html format would make your solution much more accessible to potential users.
The localeselectorPA also provides a sample xml file but no information on how to implement that in the html skin (if that is what it is to be used for).
I will make you a deal. You explain me how to implement the new <object> syntax and I will upload you a new sample skin, 100% use of <div>s and dnn5 syntax.