Re: [PATCH] enable memory mapped access to ethernet

["Gianluca Varenni" <gianluca.varenni () cacetech com>]

----- Original Message ----- 
From: "Guy Harris" <guy () alum mit edu>
To: <tcpdump-workers () lists tcpdump org>
Sent: Monday, December 10, 2007 1:29 PM
Subject: Re: [tcpdump-workers] [PATCH] enable memory mapped access to 
ethernet


> Gianluca Varenni wrote:
>
> > why not using a different return value instead of a string message which 
> > is human readable but not easily computer readable?
>
> Probably because that's the way that other pcap APIs work.  That doesn't 
> mean it's the *right* way to work, it just means it fits in with the way 
> the rest of the library works.

I don't want to start a flame here, but we already have pcap APIs that 
return a set of different values, pcap_next_ex() (which can return 1, 
0, -1, -2).
In any case, at least I wouldn't call it errbuf. It's *not* an error.

Have a nice day
GV


> A "next-generation" pcap API should probably work differently, as
>
> 1) that allows the application to choose how to react to the error (the 
> reaction might be more than just "put up a dialog box");
>
> 2) that allows internationalization of the error.
>
> (That's why Wireshark's library for reading capture files works that way, 
> although, for some errors, it also provides an additional text string - 
> but that's primarily for the benefit of developers, to give details about 
> the error to help them diagnose the problem, not for the benefit of users, 
> for whom the additional information is probably meaningless.)
> -
> This is the tcpdump-workers list.
> Visit https://cod.sandelman.ca/ to unsubscribe. 

-
This is the tcpdump-workers list.
Visit https://cod.sandelman.ca/ to unsubscribe.