Wireshark-dev: Re: [Wireshark-dev] LLRP dissector support
From: John R. Hogerhuis <jhoger@xxxxxxxxx>
Date: Tue, 8 Apr 2008 05:50:04 +0000 (UTC)
Jaap Keuter <jaap.keuter@...> writes:

> 
> Hi John,
> 
> I've been looking at this submission from the start, and frankly I don't like 
> it. It is like Ronnie says in 
> http://bugs.wireshark.org/bugzilla/show_bug.cgi?id=1957#c4, this code is very 
> hard to read, let alone maintain.
> I don't want to sign off on that and burden myself and other with the 
> maintenance chores. So I left it alone for another core developer to 
> eventually pick it up. It seems none is confident enough to commit it.
> 

Well that's a clear statement of the problem, thanks for the reply.

It appears Matt is responding favorably to requests to make specific
improvements. General criticisms about hard to read/maintain and how he has
abstracted the message parsing are obviously harder to address. My understanding
is that parts of the code are generated based on XML descriptors of the binary
protocol available from http://sourceforge.net/projects/llrp-toolkit (I am a
developer for this project but not the LLRP dissector).

If the code could be simplified to avoid wrappers are there other issues for you
or Ronnie that would stand in the way of commit?

Thanks,

-- John.