I'd say in the API we should be closer to the data and hence use 0-255, in the slider we convert it to 0-100%. We could possibly have a slider that allows for 255 positions but prints numbers from 0%-100%?
Are you aware of systems where there are more values than 8 bits for the alpha channel (for which we should prepare)? It wouldn't be a much bigger deal to go this road, but I think it shouldn't be purely prophylactic without real needs in mind. Matthias On 02/20/2017 03:19 PM, John Hawkinson wrote: > > An outstanding question is what range we should allow? Behind the > > scenes it's generally going to be converted to a 0-255 value, but for > > users a 0-100% may be more explanatory. Opinions? > > The tools I use again express opacity in 0-100%, and I think that's > the correct thing for the user interface. Why should users presume > the internals are going to be 8-bit unsigned? Why shouldn't it be > 0-65536? Or floating point? > > > The API is ALL OVER THE PLACE here, and should also be fixed. We have > > a mix of transparency/opacity/alpha and ranges of 0-1, 0-100, 0-255. > > It's confusing for developers too! > > So, I imagine at some point in the future someone is going to > decide 8-bit alpha channels are too limiting. The API should > probably be forward-looking with respect to that. I'm not > sure exactly what that means, but to me it suggests > not 0-255. Maybe 0-1 floating point. But I dunno. > > --jhawk > _______________________________________________ > Qgis-developer mailing list > [email protected] > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer > _______________________________________________ Qgis-developer mailing list [email protected] List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
