On Oct 19, 2018, at 7:51 AM, Jeff Morriss <jeff.morriss.ws@xxxxxxxxx> wrote:
> Is it just me or is there no reason for ett[] arrays:
>
> /* Setup protocol subtree array */
> static gint *ett[] = {
> &ett_PROTOABBREV
>
> to be static?
Given that they represent persistent UI state - i.e., whether, when packet details are displayed, a particular type of subtree should be displayed as opened up or not - that would seem to indicate that they should be static (in the "in static storage" sense, not in the "with internal linkage" sense).
The ett_ variables themselves must be persistent. The arrays pointing
[for registration purposes]
to them? Not so much.