[Richard van der Hoff <richardv@xxxxxxxxxxxxx>]

On Sat, Jun 26, 2004 at 07:29:25AM +1000, ronnie sahlberg wrote:
> There are some problems with the patch.
> 
> Cosmetic one:
> It replaces the explicit hf variables with an array,   that is
> inconsistent with all the other dissectors and might be confusing for
> someone browsing the source.

Mmm. Fair enough. I've added some comments (and done a little
refactoring) which I hope will make things clearer; however, I'm loth to
return to individual variables for each IE purely on the basis that very
few other dissectors use an array for their hfs, as I think the clarity
gained by making dissect_ies() orthogonal outweighs any potential
confusion elsewhere. Certainly, I would prefer to maintain the source
with an array than with (ultimately) hundreds of individual variables
and a huge switch statement.

> Serious one:
> It creates a lot of hf fields with the same abbrev    such as
> "iax2.iax.unknowndata"
> Ethereal doesnt really like to have multiple different hf fields with
> the same abbrev.
> This one must be fixed.

Well, that's not new to this patch, to be honest; and it only uses that
abbrev for IEs which aren't really understood (because I haven't been
able to investigate how they work); if people want to be able to
search/filter on such entries, they'll need to add proper dissection of
them anyway.

However, I've now changed it so that IEs of different format (byte,
word, string, etc) use a different abbrev. That will at least mean that
filtering has some chance of working if there is only one unknown IE (of
each format) per packet.

Cheers,

-- 
Richard van der Hoff <richardv@xxxxxxxxxxxxx>
Systems Analyst
Tel: +44 (0) 845 666 7778
http://www.mxtelecom.com

Attachment: iax2.patch.gz
Description: Binary data