[Jeff Morriss <morriss@xxxxxxxxx>]

Guy Harris wrote:

> Jeff Morriss said:
>
> > (Since I imagine there's a lot of gtk-1.2 users out
> > there--sunfreeware.com still hasn't released gtk-2.x packages!--would it
> > be worth putting in this optimization anyway?
>
>
> It would - *unless* we could achieve the same goal without changing the
> GTK+ code at all.
>
> It might be that the right way to do it would be to have the loop in
> file.c maintain the row number itself; I *think* it should be able to
> check the "flags.passed_dfilter" member of the frame_data structure to see
> whether a particular packet is being displayed or not, and assume that the
> Nth displayed packet has a row number of N.
>
> I tried that quickly, and it didn't look as if it set the time correctly;
> I didn't have time to look further into it.

Very good idea!

I suspect maybe you were off by one row because the attached patch seems 
to work very well for me (both with and without a display filter).

However, this doesn't work at all if the user decides to sort by some 
other column than frame number...  :-(

I guess we could find a way to check if the sort column is the frame 
number column and if not use the old method (or the old method with my 
previous patch)?  I'll look into it...

Attachment: file.c.patch.gz
Description: GNU Zip compressed data