GtkAdjustment, float and double

[Paul Davis <pbd op net>]

there are a number of application areas where GtkAdjustment's use of a
float makes it less-than-ideal (my own issue is with audio files
larger than 5 minutes in length with a sampling rate of 48kHz: a float
cannot store a sample-accurate location in such files).

now, it would be trivial to create GtkDoubleAdjustment by just doing
an intelligent search-and-replace on "float" in the GtkAdjustment
source code. but such an object would not be compatible with anything
else in GTK+ that expects a GtkAdjustment, since they plan on fetching
and storing floats. C promotion rules would help with some instances,
but not all. this means that you could not use GtkDoubleAdjustment
freely with, say, a GtkHRange or GtkSpinbutton. i think.

are there any plans (or ideas) to address this issue? its seem
fundamentally limiting to have GtkAdjustment represent a value with a
range that doesn't match that of the native word size on the host CPU.

--p