4 ms·
One argument I've heard for this model is that, by having the UI in a declarative language, the design team can update/experiment with the layout without having
by cglong 3y ago
One argument I've heard for this model is that, by having the UI in a declarative language, the design team can update/experiment with the layout without having to understand the backing logic. I've only been on one team that actually subscribed to this model though.
- jandrese 3y agoI've heard this, but in practice it is hard to fully decouple the interface from the logic so experimentation with the front end usually involves work on the back end. Even simple changes like "we want to change this drop down to a set of radio buttons" ends up requiring changes on the back end most of the time.
- cglong 3y agoI think good tooling support is a requirement for this to work; the one project I used this in (a Silverlight app) had first-class support for this. You could specify separate ViewModels for run time and design time, and the latter would just return static values that the designer could work against. I guess this would require some cursory understanding of C# syntax, but I think the ramp-up for adding/modifying a property would be minimal.