Wireshark-bugs: [Wireshark-bugs] [Bug 7471] "Bad" UAT entry can't be edited/removed via the Wire
Date: Sun, 21 Apr 2013 15:24:07 +0000

Comment # 9 on bug 7471 from
(In reply to comment #8)
> (In reply to comment #7)
> > In general I like the approach. A few misc questions:
> 
> - any reason to use
> > g_array_append_vals instead of just g_array_append_val?
> 
> Copy/paste of how user_data array was handled.
> 
> - the use of
> > gboolean* valid is entirely redundant, I think
> 
> I guess you could revalidate when the UAT dialog comes up, but at the
> momemnt, validation is only checked on "startup" and when an individual
> entry is editted, so I stuck with that paradigm.  Not sure how many placed
> need to "reverify", which is why I cached it.

That's not quite what I meant: you append valid_rec to the valid_data array,
then take a pointer to the new entry and reassign that pointer the value of
valid_rec (which it is guaranteed to already have). There's no need to take the
pointer or reassign the value, and with both of those removed the variable is
unused (I think).

> - the call to
> > UAT_UPDATE(uat) got removed from uat_remove_record_idx, was this
> > intentional?
> 
> UAT_UPDATE is for user_data only.  uat_remove_record_idx is only called from
> the GUI, which only deals with raw_data.

Right, that makes sense.

> My gut feeling re. backporting is no, but if you want to land
> > this before 1.10 branches tomorrow that might be a good idea.
> 
> Do you think it's good enough to submit the functionality without the GUI
> element of highlighting "bad" entries?  Good enough just because of the ABI
> change and with 1.10 so close to release?  I took a few stabs at the GTK,
> but wasn't successful, which is why I didn't already commit the patch.

If you've tested it then it's probably worth landing. We can always back it out
if it causes problems and getting it in trunk will make it easier for somebody
else to fiddle with the GTK (which is getting deprecated after 1.10 anyways).

Either way the behaviour is a step up over current, which requires manually
fiddling with the text file in order for the GUI to be usable again.


You are receiving this mail because:
  • You are watching all bug changes.