> The other question would be: assuming there's no replacement for some > direct field access, are people willing to add the needed functions not only > into 1.1 but also into 1.0.x, in order to avoid code duplication for end users? The idea of a 'shim' layer that helps bridge the gap between 1.1 new functions and 1.0.x exposed struct's seems like a great idea and hopefully not a lot of work.
